Title of Thesis: 
ABSTRACT 
UNDERSTANDING MATERNAL EFFECTS AS A 
RECRUITMENT MECHANISM IN LAKE MICHIGAN 
YELLOW PERCH (PERCA FLAVESCENS) 
Degree candidate: Christopher James Heyer 
Degree and year: Master of Science, 2000 
Thesis directed by: Assistant Professor Thomas J. Miller 
Marine - Estuarine - Environmental Sciences (MEES) Program 
Changes that have occurred in the abundance and trait distribution of adult Lake 
Michigan yellow perch (Percaflavescens) suggest that maternal effects on larval traits 
may be substantially influencing the recruitment of this heavily exploited species. 
Maternal effects on yellow perch larvae at hatching and through 32 days post hatch (dph) 
were investigated in ten maternal lines to test the null hypotheses of no effect of maternal 
condition on offspring condition at hatching, no persistence of maternal effects under 
conditions of starvation and high food availability, and no difference in offspring survival 
under conditions of starvation and high food availability. Maternal effects were 
detectable at hatching and likely result in differences among females in size, age, gonadal 
somatic index, and egg production. Maternal effects at hatching were expressed by 
differences in larval total length, yolk volume, dry weight, and DNA quantity. Maternal 
effects persisted under conditions of starvation to 6 dph, after which point virtually all 
larvae had perished. Maternal effects resulted in a twofold difference in resistance to 
starvation among the maternal lines. Larvae that exhibited the lowest resistance to 
starvation were long with small yolk volumes, while those exhibiting the highest 
resistance to starvation were short with large yolk volumes. Under high food availability 
maternal effects persisted to 32 dph, and resulted in threefold differences in survival 
among the maternal lines. No clear mechanism was identified to account for these 
survival differences. The observed maternal effects in Lake Michigan yellow perch may 
have substantial implications on recruitment. 
UNDERSTANDING MATERNAL EFFECTS AS A RECRUITMENT MECHANISM 
IN LAKE MICHIGAN YELLOW PERCH (PERCA FLAVESCENS) 
by 
Christopher James Heyer 
Thesis submitted to the Faculty of the Graduate School of the 
University of Maryland at College Park in partial fulfillment 
of the requirements for the degree of 
Master of Science 
2000 
Advisory Committee: 
Assistant Professor Thomas J. Miller, Chair 
Assistant Professor Mary L. Haasch 
Professor Edward D. Houde 
PREFACE 
This research reflects a portion of a combined multi-institutional, multi-agency, 
multi-state, collaborative effort to understand the mechanisms responsible for the 
recruitment failure and subsequent population decline observed throughout the last 
decade in Lake Michigan yellow perch Percaflavescens. While the shear size of this 
research required the assistance of many individuals, the results, thoughts, ideas and 
conclusions presented in this thesis are my own. Upon the successful defense of this 
research, my intention is to submit three of the five chapters of this thesis as co-authored 
manuscripts for review and hopefully eventual publication 
11 
Acknowledgements 
As with any large research effort this project could not have been completed 
without the help, input, guidance and advice of many individuals. I would first like to 
thank Dr. Stephen Tettelbach from Long Island University/Southampton College for 
opening my eyes to the world of fisheries science. Steve has continued to offer guidance 
even beyond my baccalaureate education. I would also like to thank Dr. Edward Houde 
for my first opportunity to be a part of the Chesapeake Biological Laboratory community. 
Ed has always shared his lifetime' s worth of knowledge with me and continually 
challenged me to be the best scientist that I can. Dr. Thomas Miller, my advisor and 
mentor, once said that graduate students are like little toys, you just wind them up, point 
them in a direction and off they go. I would like to thank Tom for winding me up and 
pointing me in this direction. Tom has continually broadened my horizons to levels I 
never imagined I could reach, and for that I am forever grateful. I would like to thank 
Fred Binkowski for his continual guidance and for all that he has taught me over the past 
three years. Fred has an amazing ability to get the job done no matter what the 
circumstances, and it is that ability that has allowed this research to happen. I also need 
to thank Chris Ney and Steve Yeo for all of their help and guidance in setting up and 
conducting this research. I thank Chris and Steve for imparting their knowledge and 
know-how onto me. I also want to thank Liz Trendel for everything she has done for me. 
Liz is the amazing woman behind the curtain that makes everything run smooth, and she 
does it extremely well, for that I am forever indebted. I would like to thank Dr. Mary 
Haasch for her support through the largest headache of this research, namely the analysis 
of nucleic acids. Mary was always there to sound ideas off of and to troubleshoot with. 
1ll 
Furthermore, Mary has an uncanny ability to simplify molecular biology to a point where 
even a fishhead like me can comprehend it, and I will always be thankful for that. I also 
need to thank Elaine Caldarone for providing me with not only a working nucleic acid 
extraction protocol, but for also furthering the knowledge I acquired through Mary. 
Elaine was instrumental in the final completion of this research, and for that I sincerely 
thank her. I would like to thank Bryce Brylawski, Veronica Caceres, Mary Chenery, and 
Mike Frisk (the Miller lab group) for all of the input, comments, suggestions, stress relief, 
laughs and good times that they provided throughout the course of this research. They 
are a tremendous group and I have immensely enjoyed working with them all. Last, but 
certainly not least, I would like to thank my wife Kristen for all of her support. Kristen 
has been amazing from day one of the pursuit of this degree and that support has 
continued uninterrupted to its completion, for that I am forever deeply thankful. 
lV 
TABLE OF CONTENTS 
List of Tables .... ...................... ...... ........................................ .. .................................. ........ vi 
List of Figures .... ... .................. .. .. ..... .. .... ... .. .. ................ .... .... ......................... ..... .. ....... .. .. . ix 
Chapter 1: Introduction ............. ... .... .. ... ..... ..... ...... ........... .... .. ..... ..... ............................. .... 1 
Thesis Structure .. ............. ............ .... ... .. .... ................. .............. ..... .. ... ...... .... ......... 1 O 
References ...... ...... ......................... .............................. ....... ... ... .... ......................... 12 
Chapter 2: I. Presence of Maternal Effects ......... .... ......................... .... ... ... ...................... 16 
Abstract .......... .... ..... ............ ............. .............. ......... ...... ..... .... .... ..................... ... ... 17 
Introduction ..... ......... .. ..... ...... .............. ... .... .......... ......... ............. .............. ...... .. ..... 18 
Materials and Methods ............. ........................... .... .. ... .................... .......... ........... 23 
Results ....... ........... ...... ........ ...... .......... .......... ............ .. ........................... ....... ... ...... 31 
Discussion ................ ............... ................................ .............................................. 34 
Acknowledgements ....................... ............... ............. ....... .. ... ................... ............. 40 
References ...... ............ ............................. ....... ... .... ....... ................................ .. ... .... 41 
Chapter 3: II. Resistance to Starvation .... .... .......... ..... ... .... ................ ... ........ .. ..... .......... .. 62 
Abstract ......... ..... ............. ..................... ..................................... ....................... ..... 63 
Introduction ........................... ...................................... ..... ....... .. ...... ............... .. ..... 64 
Materials and Methods .......... .... .............. ...................... ....... .. .......... ..................... 66 
Results ... ... .... .. ....... ...... .... .. .... .. ...................... ............... .......... .... .................... ....... 71 
Discussion ........... ....... ... ............. .................. ........................ ............................... .. 74 
Acknowledgements ......... .......... ................................ ... ...... .................................. . 80 
References ....... ............ ..... .............. ....... .. ..... ............. .......... ........ ..... ..... ... .... ... ...... 81 
Chapter 4: III. Implications for Recruitment .......... .................... ....... .. .... ....... ............ .. .102 
Abstract ..... ..... .... ... .... .. ..... ...... .... ................. ............... ............ .... .... ..... ..... .. .. ....... 103 
Introduction ... ... ............... ..... .... ..... ......... .. ............ ... ... .. ................................. ...... 104 
Materials and Methods ...................... .................... ...... ........................ ......... ... .... 106 
Results .... ........ ..... .. ........................ ..... .......................... ... ... ....................... .... ... .. . 11 l 
Discussion ....... .. .. .. ......... ......... ....... ...... .... ........................... ............................... . 114 
Acknowledgements ....................... .... ........ .......... .............. .................................. 119 
References .. ..... ....... ....... ... ... .. ....... ........ .......... ................... ........... ....................... 120 
Chapter 5: Synopsis ......... .... ..... ...... .... .... ............ ............. .......... ... ................................. 140 
Appendix I ....... .... ............... ......... ................................ ...... ... ....................... .... .......... ..... 148 
Appendix II ............................ .... ...... ..... ..... .. ... ......... .. .............. .... ... ........... ........ ............. 157 
V 
LIST OF TABLES 
Table 2.1. Abbreviations of maternal and larval traits measured .... .. ................ .......... ... .45 
Table 2.2. Phenotypic characteristics of the ten female yellow perch mothers .............. .46 
Table 2.3. Initial number of eggs and measures of hatching success in each 
maternal line ...... ... ......... ...... .... .. .................................... .... ... ........ ..... ............... 47 
Table 2.4. Hatchling phenotypic characteristics from each maternal line ....... ..... ..... ..... .48 
Table 2.5. Summary of analysis of variance for presence of maternal effects in 
larval traits ................................... .. .... .......... ....... ... ........ ..... ..... ..... .. .. ... ........ ... 49 
Table 2.6. Summary of canonical correlation analysis for female trait loadings ..... .. ... ... 50 
Table 2.7. Summary of canonical correlation analysis for larval canonical 
variables ..... ....... ...... .... ... ........... .. .... .. ................... .... .. ... .. .. ... ... ....... .... .......... ... 51 
Table 2.8. Summary of canonical correlation analysis for larval traits .... ..... ..... ............. 52 
Table 3 .1. Abbreviations of maternal and larval traits measured ...... .. ... ......... ...... .... .. .. .. 85 
Table 3 .2. Dilution factors used in fluorometric determination of RNA content 
and DNA content ... .. .... .............. ............. .. .............. ...... .... ........... .. ....... ........ .. 86 
Table 3 .3. Summary of multivariate analysis of variance for the detection of 
maternal effects through 6 days post hatch under conditions of food 
limitation ........ ...... .. .......................... ....... ................... ....... ..... ...... ...... .. ... .. ..... .. 87 
Table 3.4. Time to 50% mortality of the offspring of the ten maternal lines under 
conditions of food limitation .. .... .... ............. ......... ... ...... ...... .. ............... .. ... ...... 88 
Table 3.5. Summary ofrepeated measures analysis of variance for survival of the 
offspring of the maternal lines under conditions of food limitation ............... 89 
Vl 
Table 3 .6. Larval traits at time of 50% mortality of the offspring of the ten 
maternal lines under conditions of food limitation ........ ...... ... ......... ... ... .... ..... 90 
Table 3.7. Percent change in the larval traits of the offspring of the ten maternal 
lines under conditions of food limitation .................... .................... ................ 91 
Table 3.8. Summary ofrepeated measures analysis of variance for the larval traits 
of the offspring of the ten maternal lines under conditions of food 
limitation ................................ .. ................ ......... .. .............................. .. ............ 92 
Table 4.1. Abbreviations of maternal and larval traits measured ...... .. .... ..... .... .... ... ...... 124 
Table 4.2. Dilution factors used in fluorometric determination of RNA content 
and DNA content .. ...... .. ...... .. .......... ..... ...... ........ ... ....... ..... ..... ....... .... .. ....... ... 125 
Table 4.3. Summary of multivariate analysis of variance for the detection of 
maternal effects through 32 days post hatch under conditions of high 
food availability .... ...... .. .. .... ..... ...... ... .... .. ...... .. .. .. ......... ... ........ .... ...... .... ..... .. . 126 
Table 4.4. Time to 50% mortality of the offspring of the ten maternal lines under 
conditions of high food availability .......... ........ .... .......... ...... .............. .. ....... . 12 7 
Table 4.5. Survival at 32 days post hatch of the offspring of the ten maternal lines 
under conditions of high food availability ...... .. .. .... .. .. .... ......... ..... ... ......... .. .. 128 
Table 4.6. Summary of repeated measures analysis of variance for survival of the 
offspring of the maternal lines under conditions of high food 
availability ... ..... .. .... ... ...... .... ...... ... .......... .... .. ...... ........ ... .. ..... .... ..... ... .. .... ....... 129 
Table 4. 7. Summary of repeated measures analysis of variance for the larval traits 
of the offspring of the ten maternal lines under conditions of high food 
availability .... ........................ ..... .. .... ......... .... ....... ..... ... .... .... ..... .... .. ..... ... .. ..... 13 0 
Vll 
Table 4.8 . Percent change in the larval traits of the offspring of the ten maternal 
lines under conditions of high food availability .......... ...... .... ... .. .... .... .... ..... . 131 
Table 4.9. Correlations of total length between sampling days of the offspring of 
the ten maternal lines under conditions of high food availability ........ ......... 132 
Table 4 .10. Correlations of dry weight between sampling days of the offspring of 
the ten maternal lines under conditions of high food availability .......... .... 13 3 
Table. A.1.1. Two-dimensional body area and dry weight of 1 day post hatch and 
23 day post hatch yellow perch larvae .......... ....... ... ....... ... .... ... ......... .... ... 148 
Table A.2.1. Summary of analysis of variance for effect of post death condition 
and age in larval yellow perch ....... ... ....... ........ ... ........ ...... .... .. ....... ... ... ..... 169 
Table A.2.2. Summary of analysis of variance for effects of type of death in 
larval yellow perch .... ...... ..... ... .... ....................... ..... .. ... ........ ... ........... .. .... 170 
viii 
LIST OF FIGURES 
Figure 2.1. Annual yellow perch catch per unit effort from Green Can Reef in 
Lake Michigan between 1986 and 1998 .. ............... .... .... ........... ............. ... ... . 55 
Figure 2.2. Average yellow perch size at age in Illinois waters of Lake Michigan 
between 1983 and 1995 ................................ ...... ....... ............. .... ..... ..... ....... .. 56 
Figure 2.3. Bivariate correlations among yellow perch hatchlings from each 
maternal line .......................... ............. .... ..... ...... ...... .......... ..... ....................... 57 
Figure 2.4. Total length of hatchlings from each maternal line .. .. .. ... ........................ ... .. . 58 
Figure 2.5. Dry weight ofhatchlings from each maternal line ....... ...... ....... ...... ....... .... ... 59 
Figure 2.6. Yolk volume ofhatchlings from each maternal line .... ..... .. ........ ......... .. ....... 60 
Figure 2.7. DNA content of hatchlings from each maternal line ..... ................. ....... ... .. ... 61 
Figure 3 .1. Relationship between time to 50% mortality and larval total length, 
yolk volume and dry weight at hatching in the offspring of the ten 
maternal lines .......... ............. ... ............. ...... .............................. .... ..... ... .......... 95 
Figure 3.2. Total length of offspring from each maternal line at the time of 50% 
mortality under conditions of food limitation .................................... ... ........ 96 
Figure 3 .3. Yolk volume of offspring from each maternal line at the time of 50% 
mortality under conditions of food limitation .................... .... ......... ..... ....... .. 97 
Figure 3.4. Dry weight of offspring from each maternal line at the time of 50% 
mortality under conditions of food limitation ...... ... .... .. .............. ............ .... .. 98 
Figure 3.5. RNA content of offspring from each maternal line at the time of 50% 
mortality under conditions of food limitation ........... .................. ... ... .. ... ... .... 99 
lX 
Figure 3.6. RNA:DNA ratio of offspring from each maternal line at the time of 
50% mortality under conditions of food limitation ........ ....... .. ... ............... . .100 
Figure 3.7. Relationship between time to 50% mortality and larval total length, 
yolk volume and dry weight at the time of 50% mortality in the 
offspring of the ten maternal lines ............. .... .. ... ............. .... ......... ....... ....... .101 
Figure 4.1. Change in total length of the fed offspring of the ten maternal lines 
between hatching and 32 days post hatch ...... .. .......... ..... ...... ...... ................ .135 
Figure 4.2. Change in dry weight of the fed offspring of the ten maternal lines 
between hatching and 32 days post hatch .. ... ............ ...... ............. ..... ........... 136 
Figure 4.3. Change in the ratio of RNA:DNA of the fed offspring of the ten 
maternal lines between hatching and 32 days post hatch ........................... .137 
Figure 4.4. Daily growth rate measured in total length of the offspring of the ten 
maternal lines under conditions of high food availability ... .. .... ... ........ ....... 13 8 
Figure 4.5. Instantaneous growth rate (G), instantaneous mortality rate (M) and 
the ratio of MIG of the fed offspring of the ten maternal lines between 
hatching and 3 2 days post hatch ..... .... .... ... ..... .. ........... .... .... ... ... ....... .. .... ..... 13 9 
Figure A.1.1. Relationship between two-dimensional body area and dry weight in 
larval yellow perch ... .......... ................ ...................... .. .. .. ........... ... ...... ..... 156 
Figure A.2.1. Percent shrinkage on death of larval yellow perch in air as a 
function of age ... .. ....... ... .... .. ..... ...... .............. ..... ....... .. ... ... .... ... .. .. ....... .. .. 1 72 
Figure A.2.2 . Percent shrinkage on death of larval yellow perch on ice as a 
function of age ........ ........ ......... ......... ..... .... ....... ........ .. ..... ..... ... ........ ....... 173 
X 
Figure A.2.3 . Percent shrinkage on death of larval yellow perch in liquid nitrogen 
as a function of age ......... ... .... ....................... .... ............................... .... ... 174 
Figure A.2.4. Percent shrinkage on death of larval yellow perch by death 
treatment as a function of age ... ............ .. .. .. ..... ... .. .... ................ .. ..... ... .... 175 
Xl 
Chapter 1 
Introduction 
1 
"The opinion generally prevalent hitherto was that the renewal of the stock of fish 
took place, as in the case of the increase of any human population, by means of a 
more or less constant annual increment in the form of new individuals; the results 
here arrived at, however, indicate that this renewal, in the case of the species 
investigated, is of a highly irregular nature." 
Dr. Johan Hjort, 1914 
Throughout much of the past century fisheries scientists have sought to predict 
and understand the mechanisms that regulate recruitment of exploited species. Since 
Hjort's landmark publication in 1914, fisheries scientists have recognized that 
recruitment is highly variable from one year to the next, and a substantial amount of 
research over the past 85 years has focused on unraveling the mechanisms that control 
and regulate recruitment. It is generally believed that year-class strength in finfish is 
determined during the first few months oflife (Houde 1987; Pepin 1989; Leggett and 
Deblois 1994). Historically, much of the previous research conducted on recruitment has 
focused on population-level assessments. However, in recent years many recruitment 
mechanisms have been found to be size-dependent and hence affected by variation 
among individuals (Miller et al. 1988). Recognition of the potential importance of inter-
individual variability suggests that recruitment research may be more successful if it were 
to focus on individuals rather than populations (Crowder et al. 1989). 
Starvation induced mortality was the focal point of much of the past century's 
research since Hjort 's 'critical period' hypothesis in 1914. Hjort (1914) hypothesized 
that mass larval mortality, and thus poor recruitment, would ensue if food abundance 
were low during the larval transition phase from endogenous to exogenous food sources. 
Alternatively, if food availability were high during this critical period, larval mortality 
would be low and subsequent good recruitment would occur (Hjort 1914 ). Leggett and 
2 
Deblois (1994) reviewed the pertinent literature and concluded that there was insufficient 
evidence to support Hjort's 'critical period' hypothesis as a strong driving factor in 
recruitment variation. Revision and broader generalization of Hjort's hypothesis by 
Cushing in 1972 (reviewed in Leggett and Deblois 1994) yielded the 'match-mismatch' 
hypothesis. Cushing hypothesized that larval mortality, and thus recruitment, was 
regulated by the temporal overlap between larval hatching and the seasonal production of 
plankton (reviewed in Leggett and Deblois 1994). Again, Leggett and Deblois (1994) 
found a lack of evidence to support this hypothesis as a strong driving factor in 
recruitment variation. Furthermore, mortality due to starvation depends not only on the 
food availability, but also on the individual larva's resistance to starvation (Rice et al. 
1987). Resistance to starvation increases with hatching size (Hunter 1981; Miller et al. 
1988) and begins to increase dramatically as feeding begins and energy reserves begin to 
increase, and as increases in mouth gape and swimming ability provide a wider range of 
available food sizes to the larvae (Blaxter 1969; Hunter 1981; Miller et al. 1988). 
Larval mortality directly attributable to predation has yielded one hypothesis that 
has been under constant review, the 'bigger is better' hypothesis. The 'bigger is better' 
hypothesis contends that larger larvae are less susceptible to predation than smaller larvae 
(reviewed in Leggett and Deblois 1994). However, in mesocosm and microcosm studies 
Litvak and Leggett (1992) found selectivity for older larvae when size was held constant, 
no selectivity when age was held constant and size was variable, and selectivity for larger 
larvae when both size and age were variable as they would be within a cohort. These 
findings counter the 'bigger is better' hypothesis. Litvak and Leggett (1992) pointed out 
that larger larvae swim more than smaller larvae and therefore may encounter more 
3 
predators, thus the selection for larger larvae may be confounded with increased 
encounter probability. 
The lack of evidence supporting the ' critical period' and the 'match-mismatch' 
hypotheses suggests that while starvation experienced at hatching may alone result in 
some mortality, it is not likely a controlling factor of recruitment by itself. The evidence 
for the 'bigger is better' hypothesis suggests that while predation is obviously of great 
importance in controlling recruitment, a clear understanding of how this agent operates 
independently cannot be established. If neither starvation nor predation can be 
determined and defined conclusively as independent agents that control recruitment then 
what can? Much recent research has begun to examine starvation not as a stand-alone 
recruitment controller, but as a modifier to predation. Starvation can increase 
vulnerability to predation through reductions in avoidance capabilities. Starvation can 
also result in reduced growth rates, which can lengthen stage duration and thus increase 
vulnerability to predation. Such interactions are more likely to control recruitment than 
starvation or predation alone. 
Interest in the role of starvation-mediated larval mortality has prompted 
researchers to evaluate the condition or "health" of larvae in the field. Condition has 
been used as a measure that indexes past growth, as indicated by recent feeding history, 
and likely future survival based on selectivity. Poorly conditioned larvae, resulting from 
reduced feeding (i.e. starvation), have been shown to exhibit reduced growth rates 
(Buckley 1984). Starvation-induced reductions in growth rate can potentially lead to 
selective mortality (i.e. predation). Morphological, histological and biochemical indices 
have been historically used to assess the condition of larval fish (see review by Ferron 
4 
and Leggett 1994 ). Of these measures, biochemical indices of condition, particularly 
nucleic acid based indices, are the most sensitive indicators of recent feeding history, thus 
extensive research has been conducted on various methodologies (Buckley 1980; 1984; 
Ferron and Leggett 1994). 
Since 1970 when Sutcliffe (reviewed in Bergeron 1997) demonstrated that the 
growth of small organisms could be estimated from their ribonucleic acid (RNA) content, 
much attention and interest has been focused on such methodologies. A commonly 
employed biochemical method of measuring larval fish growth is the use of the ratio of 
RNA to deoxyribonucleic acid (DNA) (see reviews by Ferron and Leggett 1994; 
Bergeron 1997). The ratio of RNA:DNA has also been used as a measure of larval 
nutritional condition ( see reviews by Bergeron 1997; Ferron and Leggett 1994 ). The 
assumptions behind the RNA:DNA ratio are that the amount of DNA in the somatic cells 
of individual species is relatively constant (Buckley 1980, 1984), and that the amount of 
RNA in the cell changes as a function of protein synthesis (Bulow 1970). In more 
general terms, larvae of good condition will tend to have a higher RNA:DNA ratio than 
larvae of poorer condition. Improvements in our understanding of larval growth and 
mortality may be accomplished through the use of biochemical techniques such as the 
ratio of RNA:DNA. 
During early life, selective mortality may operate differentially on individual 
larval traits (Houde 1997; Miller 1997). Parental contributions to the phenotypic 
expression of larval traits can be either genetic or environmental (Chambers 1993). 
Environmental factors expressed via parental contributions are generally of a maternal 
origin and are expressed in the early life stages of the offspring (Chambers 1993). These 
maternal contributions are commonly referred to as maternal effects and can be defined 
as comprising "a class of phenotypic effects that parents have on phenotypes of their 
off spring that are unrelated to the off spring' s own genotype" (Bernardo 1996). Maternal 
effects have potential implications on the susceptibility of a cohort to starvation and 
predation. Heavy exploitation may pass a spawning population through a phenotypic 
bottleneck through the removal of adults from the population that exhibit desired traits 
(i.e. large fish). This selective removal of adults could select for larval traits that exhibit 
lowered resistance to starvation and increased vulnerability to predation, and could 
potentially result in a substantial impact on recruitment. 
Lake Michigan yellow perch (Percaflavescens) have experienced shifts in the 
distribution of the phenotypic traits of the spawning population through high levels of 
selective exploitation. Aµ overall reduction in the spawning stock biomass has shifted 
the age structure of the population from a broad distribution to one dominated by adults 
greater than 4 years of age (Wisconsin Department of Natural Resources Fish Unit 
(WDNRFU), 600 E. Greenfield Ave. , Milwaukee, WI 53204, USA personal 
communication). The commercial and recreational fisheries target the larger and faster 
growing female portion of the population. Subsequently, shifts in the population sex 
ratio have occurred over the past decade such that the population in 1996 was dominated 
by a male composition of 90% (WDNRFU personal communication). Furthermore, 
female growth rates have shifted dramatically since the 1980s. Females in the mid 1980s 
reached a size of about 250 mm at an age of 12 years, by the mid 1990s females were 
reaching the same size at an age of only 4 years (S.M. Shroyer, Department of Biology, 
Ball State University, Muncie, IN 47306-0440, USA personal communication). 
6 
Historically, yellow perch have provided one of the most important commercial 
and recreational fisheries in Lake Michigan (Rakoczy and Rogers 1987; GLFC 1995; 
Francis et al. 1996). However, the recent decline of yellow perch in Lake Michigan and 
the truncation of the trait distribution of the spawning population suggest that maternal 
effects might be affecting the recruitment potential of the population. Yell ow perch are 
the only native Great Lakes fish that has sustained substantial, uninterrupted commercial 
and recreational harvests throughout the past century. Between 1889, when Lake 
Michigan yellow perch catch records were first recorded, and 1960, annual landings 
averaged 1090 metric tons (Baldwin et al. 1979). After 1960, commercial and 
recreational harvests of yellow perch fluctuated in response to a fluctuating population 
level (Francis et al., 1996). Population booms of the non-native Alewife (Alosa 
pseudoharengus) in the 1960s caused large yellow perch declines (Crowder 1980). 
Rebounds in the yellow perch population began to occur by 1972 and by the early 1980s 
the yellow perch population was at a substantially high level (GLFC 1995; Francis et al. 
1996). However, the population declined dramatically in the early 1990s and failed to 
recover throughout the last decade. The Yellow Perch Task Group (YPTG), a 
subcommittee of the Lake Michigan Committee, was established by the Great Lakes 
Fishery Commission in 1994 to provide analysis, technical information and advice on 
management alternatives toward resolving the continuing and rapid decline of the yellow 
perch population in Lake Michigan. The YPTG identified poor reproduction and 
recruitment failure as primary reasons for the continued decline of the Lake Michigan 
yellow perch population. The population has experienced practically constant 
recruitment failure since 1990. The production of weak year-classes in 1996 (WNRB 
7 
1996) and 1997 (personal observation) was an ominous reminder of the continuing 
recruitment failure. The recruitment failure of the Lake Michigan yellow perch 
population has prompted an intense regional effort to understand what went wrong. 
Researchers from around Lake Michigan are investigating many aspects of recruitment 
processes. 
Yellow perch spawning takes place during the months of May and June in Lake 
Michigan when the shallow inshore waters warm to temperatures of about 1 O 0c. Adult 
yellow perch will move inshore from the deeper offshore waters and begin to congregate 
on spawning grounds around the lake. Spawning normally occurs at night and is marked 
by a unique courtship. A single female will lead two side by side lines of males around a 
circuitous course before she releases her accordion-like gelatinous egg skein (Werner 
1980). The males then fertilize the egg skein. The accordion-like architecture of yellow 
perch egg skeins coupled with their semi-buoyant nature allows the egg strands to 
slightly compress and expand with the water currents, creating a pumping action which 
moves a fresh supply of water across the eggs. Fecundity has been documented to range 
between 950 to 210,000 eggs per female, with an average fecundity between 20,000 and 
30,000 eggs per female (Collette et al. 1977; Werner 1980). Incubation of yellow perch 
eggs in Lake Michigan can range between 14 and 21 days depending on water 
temperature and the rate of warming (personal observation). After hatching, the larvae 
move up into the upper three meters of the water column where they begin feeding on 
zooplankton within the first couple days (Forney 1971; Mills and Forney 1981). At a size 
of approximately 25 to 30 mm, yellow perch begin to make a transition to demersal life 
(Forney 1971; Mills and Forney 1981). Yellow perch remain demersal in the inshore 
8 
waters until they reach a size of about 65 mm, at which point they begin to migrate 
offshore (Forney 1971). Yellow perch are recruited to the fishery between ages 2 and 3. 
Year class strength in both yellow perch (P. flavescens) and European perch (P. 
fluviatilis) are known to exhibit high inter-annual variation in many systems (Forney 
1971 ). The variation in year class strength has been attributed to predatory control in 
many systems. Yell ow perch year class strength has been shown to be controlled by 
walleye (Stizostedion vitreum) predation on age-0 yellow perch in Lake Oneida (Mills et 
al. 1987), by northern pike (Esox lucius) in Hemming Lake (Lawler 1965), and by 
muskellunge (Esox masquinongy) in lakes where it was stocked (Forney 1971). There 
has been some limited evidence for the role of predation in the Great Lakes systems. In 
an embayment off of Lake Ontario Mason and Brandt (1996) documented extensive 
predation by alewives on larval yellow perch. The evidence is less clear on a lake wide 
basis. For example, in Lake Michigan there is some evidence that yellow perch year 
class strength is affected by alewife (Alosa pseudoharengus) predation on larval yellow 
perch and competition between alewife and yellow perch in some areas of the lake 
(Shroyer and McComish 1998). However, by and large it is not believed that year class 
strength in Lake Michigan is controlled through predation as it is in other systems. 
9 
The truncation of the spawning stock traits in recent years coupled with the 
selective mortality that may operate differentially on individual larval traits have lead to 
the focus of this research on the role of maternal effects as a controlling factor of 
recruitment in Lake Michigan yellow perch. The goals of this research were as follows: 
1. To determine the presence of maternal effects in Lake Michigan yellow perch 
2. To determine the influence of maternal effects on the susceptibility oflarvae to 
starvation in Lake Michigan yellow perch, and; 
3. To determine the implications maternal effects have on the recruitment of yellow 
perch in Lake Michigan. 
Thesis Structure 
Chapter 2 is a determination of whether maternal effects are present in Lake 
Michigan yellow perch. Background information on changes that have occurred to the 
population structure over the past 15 years is presented to lay the foundation for 
examining maternal effects. The presence of maternal effects in newly hatched yellow 
perch larvae was detected. The correlation between the maternal traits and the traits of 
the offspring were investigated. Conclusions on which offspring traits can be expected 
given a set of maternal traits are drawn. Furthermore, the roles of trade-offs between 
traits that occur at both the maternal and offspring levels are examined. While the 
importance of maternal effects expressed as differences in hatching success among the 
maternal lines are examined, their further importance is merely suggested here and is not 
empirically evaluated. 
10 
Chapter 3 is the first of two chapters that empirically evaluate the importance of 
the maternal effects detected in Chapter 2. In this chapter, the susceptibility of larvae 
from different maternal lines to starvation is investigated. Differences in survival and 
differences in the changes that occur in individual larval traits are tracked. The 
importance of maternal effects and size-dependent recruitment mechanisms are evaluated 
in terms of resistance to starvation. 
Chapter 4 continues the evaluation of the importance of maternal effects detected 
in Chapter 2. In this chapter, the implications of persistent maternal effects through 
metamorphosis on the recruitment of yellow perch in Lake Michigan are investigated. 
Differences in not only survival, but also differences in the change of individual larval 
traits are tracked. The importance of maternal effects and size-dependent recruitment 
mechanisms are evaluated and discussed. 
Finally, in Chapter 5 the implications of the results on the rebuilding of yellow 
perch in Lake Michigan and on the long-term future management of the stock are 
discussed. Furthermore, the utility of this research and implications for understanding 
recruitment mechanisms in general are discussed. Advice to managers about the dangers 
of phenotypic bottlenecking caused by high exploitation and uncertain stock assessments 
are given. 
11 
References 
Baldwin, N.S., R.W. Saalfeld, M.A. Ross, and H.J. Buettner. 1979. Commercial fish 
production in the Great Lakes 1867-1977. Great Lakes Fishery Commission 
Tech. Rep. No. 3. 
Bergeron, J.P. 1997. Nucleic acids in icthyoplankton ecology: a review, with emphasis 
on recent advances for new perspectives. J. Fish Biol. 57(Supplement A): 284-
302. 
Bernardo, J. 1996. Maternal effects in animal ecology. Amer. Zool. 36: 83-105. 
Blaxter, J.H.S. 1969. Development: eggs and larvae, p. 177-252. In: W.S. Hoar and D.J. 
Randall (eds.) Fish Physiology. Vol. III. Reproduction and growth. Academic 
Press, New York, New York. 
Buckley, L.J. 1984. RNA-DNA ratio: an index oflarval fish growth in the sea. Mar. 
Biol. 80: 291-298. 
Buckley, L.J. 1980. Changes in the ribonucleic acid, deoxyribonucleic acid, and protein 
content during ontogenesis in winter flounder, Pseudopleuronectes americanus, 
and effect of starvation. Fish. Bull. 77: 703-708. 
Bulow, F.J. 1970. RNA-DNA ratios as indicators ofrecent growth rates of a fish. J. 
Fish. Res. Board Canada. 27: 2343-2349. 
Chambers, R.C. 1993. Phenotypic variability in fish populations and its representation 
in individual-based models. Trans. Am. Fish. Soc. 122 404-414. 
Collette, B.B., M.A. Ali, K.E.F. Hokanson, M. Nagiec, S.A. Smimov, J.E. Thorpe, A.H. 
Weatherly, and J. Willemsen. 1977. Biology of percids. J. Fish. Res. Board Can. 
34: 1890-1899. 
12 
Crowder, L.B. 1980. Alewife, smelt and native fishes in Lake Michigan: competition or 
predation? Env. Biol. Fish. 5: 225-233. 
Crowder, L.B., J.A. Rice, T.J. Miller, E.A. Marschall. 1989. Empirical and theoretical 
approaches to size-based interactions and recruitment variability in fishes, p. 237-
255. In: DeAngelis, D.L. and L.J. Gross (eds.) Individual-Based Models and 
Approaches in Ecology. Chapman and Hall, New York, New York. 
Ferron, A. and W.C. Leggett. 1994. An appraisal of condition measures for marine fish 
larvae. Adv. Mar. Biol. 30: 217-304. 
Forney, J.L. 1971. Development of dominant year classes in a yellow perch population. 
Trans. Am. Fish. Soc. 100(4): 739-749. 
Francis, J. T. , S.R. Robillard, and J.E. Marsden. 1996. Yellow perch management in 
Lake Michigan: a multi-jurisdictional challenge. Fisheries. 21(2): 18-20. 
Great Lakes Fishery Commission (GLFC). 1995. Fish community objectives for Lake 
Michigan. Great Lakes Fishery Commission Special Publication 95-3. 
Hjort, J. 1914. Fluctuations in the great fisheries of northern Europe viewed in the light 
of biological research. Rapp. P-V Reun. Cons. Int. Explor. Mer. 20: 1-228. 
Houde, E.D. 1997. Patterns and consequences of selective processes in teleost early life 
histories, pp 173-196. In: R.C. Chambers and E.A. Trippel ( eds.) Early Life 
History and Recruitment in Fish Populations. Chapman and Hall, New York, 
New York. 
Houde, E.D. 1987. Fish early life dynamics and recruitment variability. Amer. Fish. 
Soc. Symp. 2: 17-29. 
13 
Hunter, J.R. 1981. Feeding ecology and predation of marine fish larvae, p. 33-79. In: R. 
Lasker (ed.) Marine Fish Larvae: Morphology, Ecology, and Relation to 
Fisheries. University of Washington Press, Seattle, Washington. 
Lawler, G.W. 1965. Food of the pike, Esox lucius, in Heming Lake, Manitoba. J F. h 
. IS . 
Res. Board Canada. 22: 1357-1377. 
Leggett, W.C. and E. Deblois. 1994. Recruitment in marine fishes: is it regulated by 
starvation and predation in the egg and larval stages. Neth. J. Sea. Res. 32(2): 
119-134. 
Litvak, M.K. and W.C. Leggett. 1992. Age and size-selective predation on larval fishes : 
the bigger-is-better hypothesis revisited. Mar. Ecol. Prog. Ser. 81: 13-24. 
Mason, D.M. and S.B. Brandt. 1996. Effect of alewife predation on survival of larval 
yellow perch in an embayment of Lake Ontario. Can. J. Fish. Aquat. Sci. 53: 
1609-1617. 
Miller, T.J. 1997. The use of field studies to investigate selective processes in fish early 
life history. In R.C. Chambers and E.A. Trippel (eds.) Early Life History and 
Recruitment in Fish Populations. Chapman and Hall, New York, New York. 
Miller, T.J., L.B. Crowder, J.A. Rice, and E.A. Marschall. 1988. Larval size and 
recruitment mechanisms in fishes: toward a conceptual framework. Can. J. Fish. 
Aquat. Sci. 45: 1657-1670. 
Mills, E.L. and J.L. Forney. 1981. Energetics, food consumption, and growth of young 
yellow perch in Oneida Lake, New York. Trans. Am. Fish. Soc. 110: 479-488. 
Mills, E.L., J.L. Forney, and K.J. Wagner. 1987. Fish predation and its cascading effect 
on the Oneida Lake food chain. pp. 118-131. In: W.C. Kerfoot and A. Sih (eds.). 
14 
Predation: Direct and Indirect Impacts on Aquatic Communities. University Press 
of New England, Hanover, New Hampshire, USA. 
Pepin, P. 1989. Predation and starvation of larval fish: a numerical experiment of size-
and growth-dependent survival. Biol. Ocean. 6: 23-44. 
Rakoczy, G. and R. Rogers. 1987. Sport fishing catch and effort from Michigan waters 
of Lakes Michigan, Huron and Erie, and their important tributary streams. April 
1, 1986 - March 31 , 1987. Michigan Dept. Nat. Res., Fish. Div. , Fish. Tech. Rep. 
No. 87-6A. 58 pp. 
Rice, J.A. , L.B. Crowder, and F.P. Binkowski. 1987. Evaluating potential sources of 
mortality for larval bloater (Coregonus hoyi): starvation and vulnerability to 
predation. Can. J. Fish. Aquat. Sci. 44: 467-472. 
Shroyer, S.M. and T.S. McComish. 1998. Forcasting abundance of quality-size yellow 
perch in Indiana waters of Lake Michigan. N. Amer. J. Fish. Manag. 18: 19-24. 
Sutcliffe, W.H. , Jr. 1970. Relationship between growth rate and ribonucleic acid 
concentration in some invertebrates. J. Fish. Res. Board Can. 27: 606-609. 
Werner R.G. 1980. Freshwater Fishes of New York State. Syracuse University Press. 
' 
Syracuse, NY. 186 pp. 
Wisconsin Natural Resources Board (WNRB) Memorandum. 1996. File Ref. #3600. 
Wootton, R.J. 1990. Ecology ofTeleost Fishes. Chapman and Hall, New York, New 
York. 404 pp. 
15 
/ 
' 
Chapter 2 
Understanding Maternal Effects as a Recruitment Mechanism in 
Lake Michigan Yellow Perch (Perea flavescens) 
I. Presence of Maternal Effects 
16 
Abstract 
Changes that have occurred in the phenotypic distribution of adult yellow perch 
(Percajlavescens) in Lake Michigan suggest that maternal effects on larval traits may be 
substantially influencing the recruitment of this heavily exploited species. We 
investigated maternal effects on yellow perch larvae at hatching in ten maternal lines to 
test the null hypothesis of no effect of maternal condition on offspring condition. 
Analyses lead to a rejection of the null hypothesis and indicate that the observed maternal 
effects likely result from differences among females in size, age, gonadal somatic index, 
and egg production. The observed maternal effects are expressed in the offspring by 
differences in larval total length, yolk volume, dry weight, and DNA quantity. 
Correlation analyses indicated trade-offs within female traits between GSI and all other 
female traits measured, and within larval traits between total length and yolk-volum.e. 
The observed trade-offs have potentially substantial implications on the offspring 's 
vulnerability to starvation and predation. We conclude that yellow perch exhibit 
maternal effects that are expressed in the distribution of larval traits at hatching. 
17 
1 
Introduction 
Despite the large focus of energy that has been directed towards understanding 
and predicting recruitment, recruitment variability is the single least understood process 
in fisheries science (Houde 1987). The difficulty stems from the "synergism of variable 
processes" that regulate recruitment (Houde 1987). These processes include, but are not 
limited to, interactions between spawners and their characteristics, recruit characteristics 
' 
larval growth and mortality, and trophic interactions together with the effects of the 
physical environment in which both the spawners and recruits exist (Laurence 1988; 
Crowder et al. 1989). This complexity underlies the high inter-annual variability 
characteristic of recruitment. 
Recruitment variability is likely attributable to processes in early life history 
(Houde 1987; Pepin 1989; Leggett and Deblois 1994). The extreme variability in vital 
rates that characterizes early life stages produces larvae with a wide range of 
characteristics on which phenotypic selection may act (Rice et al. 1993). Recognition of 
the potential importance of inter-individual variability has led to suggestions that 
recruitment research may be more successful if it were to focus on individuals rather than 
populations (Crowder et al. 1989). Many recruitment mechanisms have been found to be 
size-dependent (Miller et al. 1988). Because of this size-dependence, small initial size 
differences within a population can result in very large survival and recruitment 
differences (Adams and DeAngelis 1987; Crowder et al. 1989; Rice et al. 1993). 
Previous research has shown that observed variation in the early life history traits 
among individuals of many finfish species has been linked to maternal influences 
(Chambers and Leggett 1992, 1996). Maternal effects can be defined as comprising "a 
18 
J 
class of phenotypic effects that parents have on phenotypes of their offspring that are 
unrelated to the offspring 's own genotype" (Bernardo 1996). Maternal influences result 
from differential reproductive investment by males and females. While both male and 
female gametes are required for successful fertilization, it is the female's contribution to 
the egg that provides the nutritional requirements essential for proper development. 
Subsequent to hatching, it is again the investment of the female that provides the 
nutritional sustenance that is required for larval survival until first feeding. These 
investments by the female can be measured by the morphological and biochemical 
condition of the subsequent offspring. In contrast, paternal investment and its impact on 
subsequent survival is believed to be small. Recent research on parental influences in 
Atlantic cod (Gadus morhua) did not show any consistent effects of male size on 
fertilization success (Rakitin et al. 1999). Furthermore, differences in hatching success 
and survival in Atlantic cod could not be explained by differences in male size nor sperm 
density (Rakitin et al. 1999). 
Much of the previous research on maternal effects has focused on detecting 
maternal effects in the phenotypic traits of eggs. Significant female effects on egg sizes 
have been observed in Atlantic salmon Salmo salar (Kazakov 1981 ), chum salmon 
Oncorhynchus keta (Beacham and Murray 1985), rainbow trout 0. mykiss (Gall 1974), 
cape1in Mallotus vi!losus (Chambers et al. 1989), Atlantic cod (Knutsen and Tilseth 1985; 
Chambers and Wai wood 1996), haddock Melanogrammus aeglefinus (Hislop 1988), 
Atlantic silverside Menidia menidia (Bengston et al. 1987), black porgy Acanthopargrus 
schlegeli (Huang et al. 1999), winter flounder Pseudoplueronectes americanus (Buckley 
et al. 1991 ), and ye11owtai1 flounder Pleuronectes ferrugineus (Benoit and Pepin 1999), 
19 
1 
among others. Chambers and Leggett (1996) also identified Atlantic herring Clupea 
harengus, walleye pollock Theragra chalcogramma, striped bass Morone saxatilis, turbot 
Scophthalmus maximus, and flathead sole Hippoglossoides elassodon, as species that 
exhibited significant maternal effects on egg size in their review of maternal effects. In 
contrast, research on maternal effects on the size, condition and growth rates of larvae has 
been limited (Benoit and Pepin 1999). Significant maternal effects in larval traits have 
been observed in a few species. The traits for which these effects have been documented 
are size at hatch ( chum salmon, Beacham and Murray 1985; Atlantic silversides, 
Bengston et al. 1987), yolk volume at hatching ( chum salmon, Beacham and Murray 
1995; capelin, Chambers et al. 1989), and size at metamorphosis (winter flounder, 
Chambers and Leggett 1992). Chambers and Leggett (1996) concluded, "body size 
during larval life was likely to be significantly influenced by maternal effects. " For most 
other species where significant maternal effects are reported in larval traits, such 
correlations are inappropriately extrapolated from significant maternal effects in 
phenotypic egg traits (Pepin and Miller 1993; Benoit and Pepin 1999). Of those studies 
in which maternal effects in larval traits have been documented, few have used 
biochemical traits such as the ratio of RNA to DNA to evaluate larval condition. 
Potentially, maternal effects can substantially impact the recruitment dynamics of 
heavily exploited populations. Since maternal effects are an expression of the phenotypic 
characteristics of the spawning population, overfishing can lead to phenotypic 
bottlenecking which can cause the population to shift from its fitness peak (Solemdal 
1997). Changes in the distributions of size, age and condition of the spawning population 
that result from exploitation can potentially cause changes in egg production, fertilization 
20 
success, and hatching success (Rjinsdorp 1989). Furthermore, changes in the size and 
condition distribution of newly hatched offspring of a population can potentially lead to 
changed susceptibilities to starvation and predation. In theory, such shifts can result in 
increased mortality, decreased recruitment, and ultimately recruitment failure. 
Historically, yellow perch (Percaflavescens) has provided one of the most 
important commercial and recreational fisheries in Lake Michigan (Francis et al. 1996). 
Substantial harvests have been ta.ken over the past century, even in the face of 
perturbations such as the alewife invasion and its subsequent population boom in the late 
1960s. However, the population declined dramatically in the early 1990s and has failed 
to recover throughout the last decade. The Lake Michigan Yellow Perch Task Group, 
established in 1994 by the Great Lakes Fishery Commission to provide analysis, 
technical information and advice on yellow perch management, identified poor 
reproduction and recruitment failure as primary reasons for the continued decline of the 
Lake Michigan yellow perch population. The population has experienced practically 
constant recruitment failure since 1990. The production of weak year classes in 1996 
(WNRB 1996) and 1997 (personal observation) was an ominous reminder of the 
continuing recruitment failure. 
In 1998, the Lake Michigan yellow perch population level was reported to be 20% 
of 1990 levels (Wisconsin Department ofNatural Resources Fish Unit (WDNRFU), 600 
E. Greenfield Ave. , Milwaukee, WI 53204, USA unpublished data). Winter gillnet 
surveys catch per unit efforts (CPUE) declined from around 4000 to 100 yellow perch per 
1000 ft of gillnet per night between 1986 and 1998 (Fig. 1 ). The age structure of the 
population changed over this period as well (WDNRFU personal communication). In the 
21 
mid and late 1980s, the population was dominated by yellow perch age 2-4 years with a 
relatively small portion of the total population exceeding 4 years of age. By 1994, the 
population had shifted to one composed of almost exclusively greater than 4-year-old 
yellow perch (Fig. 1). The same WDNRFU winter gillnet surveys have also shown a 
shift in the sex ratio of the population from a ratio of almost 50% males to 50% females 
in the mid 1980s to a ratio of 90% males to 10% females in the mid 1990s. In response 
to the population decline, the average growth rates of female yellow perch have increased 
substantially since 1983 (S.M. Shroyer, Department of Biology, Ball State University, 
Muncie, IN 4 7306-0440, USA personal communication). In the early and mid-1980s 
female yellow perch reached a size of about 250 mm at an age of 12 years, while in the 
mid-l 990s, females were reaching that same size at an age of only 4 years (Fig. 2). 
The changes in the population structure that have occurred over the past decade 
reflect the mark of an extremely size selective fishery, which targets the faster growing 
and larger female yellow perch. The high fishing mortality rate, the highly size-selective 
nature of the fishery and a lack of recruitment have caused the overall population decline 
and shifts in the sex ratios which further weakened reproduction and recruitment. The 
changes in the female portion of the Lake Michigan yellow perch population structure 
suggest that the distribution of female phenotypes has been significantly altered. Large 
females that are predominantly greater than 4 years of age, but that are historically young 
for their size, dominate the female portion of the population (Figs. 1 and 2). The lack of 
recruitment further fuels this phenotypic bottleneck. These changes suggest that maternal 
effects may play a dominant role in the recruitment dynamics of yellow perch in Lake 
Michigan. 
22 
Given the changes that have occurred in the Lake Michigan yellow perch 
population structure, this study aimed to investigate the role of maternal effects within the 
population. The objectives of this study were threefold; to determine whether maternal 
effects on larvae of Lake Michigan yellow perch were present; to determine which larval 
traits accounted for the maternal effects if found, and; to determine which female traits 
correlate with the larval traits that account for the observed maternal effects if found. 
Materials and Methods 
We investigated our objectives within a single experiment, which involved 
quantifying the relationship between ten different females and their offspring. All 
females were collected in Lake Michigan. Subsequent experimental work was conducted 
at the University of Wisconsin 's WATER Institute, Milwaukee, WI. The experiment was 
conducted as a randomized complete block design, consisting of three blocks of 20 tanks 
with two replicate tanks per female per block. 
Collection of Parental Brood Stock and Gametes 
Yellow perch were collected off Green Can Reef ( 42° 50' 00" N, 87° 50' OO"W), a 
historic yellow perch spawning ground offshore of Milwaukee, WI, using graded mesh 
gillnet (mesh from 2 ¼ to 2 ¾ bar) in June of 1998. Independent bottom sets were made 
between 10 _ 15 m. of water in the afternoon prior to retrieval. Nets soaked overnight. 
Upon retrieval, all fish were carefully removed from the gillnet and their sex was 
determined. Female yellow perch were determined to be green (not ready to spawn, eggs 
not fully hydrated), ripe ( eggs fully hydrated and easily expressed) or spent. Male yellow 
23 
perch were characterized as green (not ready to spawn, no spenn), ripe ( spenn readily 
expressed), or spent. A minimum of 25 ripe males was held in an onboard holding tank 
until ripe females were captured. As ripe female yellow perch were captured, their eggs 
were immediately expressed into separate large stainless steel mixing bowls. The 
separate egg strands were then carefully transferred to Ziploc bags and weighed. The 
eggs were transferred back to the separate stainless steel mixing bowls and fertilized with 
the milt from no less than three and no greater than eight males. The number of males 
used to fertilize the eggs was detennined based upon each male 's spenn production. 
After approximately one minute, fresh lake water was added to the mixing bowls to keep 
the eggs wet and at ambient temperature. After a minimum of three minutes, the water 
was drained from the bowls and the fertilized eggs were rinsed with fresh lake water. 
The fertilized egg skeins were then transferred to separate 2-gallon Ziploc bags and filled 
with fresh lake water. The bags were stored in large coolers containing ice in a separate 
compartment until transported back to the laboratory. The females and males used to 
produce each fertilized egg skein were sacrificed, and kept in separate Ziploc bags on ice 
for transportation back to the laboratory. 
Due to the population decline of yellow perch in Lake Michigan it was impossible 
to obtain more than five ripe female yellow perch on any one given day, despite the use 
of more than two miles of commercial gillnet. As a result, egg skeins from five female 
yellow perch were obtained on 2 June 1998; five additional egg skeins were obtained on 
4 June 1998. 
24 
Determination of Maternal Characteristics 
We quantified eight maternal traits (Table I). The total length and weight of each 
female were measured in the laboratory. Males were also measured, although these data 
are not reported. Fulton's condition factor (k) was calculated for each female ask= 
(weight/length3) x I 00,000. Additionally, the body depths at the anterior base of the first 
dorsal fin and the posterior base of the second dorsal fin were measured. The sagital 
otoliths of each female were removed for subsequent age analysis. Each female 's 
gonadal somatic index (GSI) was determined as: 
(Wet Weight of Eggs+ Wet Weight of Stripped Ovary) 
GSI= . (Wet Weight of Female+ Wet Weight of Eggs+ Wet Weight of Stripped Ovary) 
When the ovary from each female was removed there were some green eggs remaining in 
each ovary. These eggs were determined to be eggs that would not have been released 
during spawning and would most likely have been reabsorbed by the female; therefore 
our estimates of total egg production for each female are based solely on the ripe eggs 
expressed. Since the green eggs remaining in the ovary were assumed to be eggs that 
would have been reabsorbed, their weight was included in the weight of the stripped 
ovary. 
Egg Sampling 
The total volume of each female 's egg skein was determined in the laboratory. 
Independent 1-ml sub-samples (n=3) of each egg skein were taken and the eggs in each 
sub-sample were enumerated. The total egg production per female was defined as the 
product of the skein volume and the average egg concentration per unit volume. An 
additional sub-sample of each female's eggs was videotaped for subsequent analysis of 
25 
morphometry (n=15). For unknown reasons, the gelatinous matrix of Female2 did not 
hold together. While we were able to get egg production estimates for Female w d'd 2, e I 
not videotape her eggs to avoid risking loss of too many eggs. The egg skeins obtained 
on 2 June 1998 were kept separated, but held in a single mass rearing tank until 4 June 
1998 when the second set of egg skeins was obtained. Because of the fragile gelatinous 
matrix of the egg skein from Female2, her eggs were held separately in a 10-gallon flow-
through aquarium. 
Husbandry of Eggs 
Once egg skeins from ten females were available, the eggs were randomly 
assigned to and incubated in flow-through, temperature-controlled 10-gallon aquaria. 
Based upon the egg production of each female, and the previously determined number of 
eggs per milliliter ( checked again on 4 June 1998 to adjust for swelling associated with 
water hardening), a volume of skein equivalent to - 2000 eggs was placed into each tank 
assigned to that female. The egg skein for Females was crushed in transport and some 
eggs were damaged; the total number of usable eggs for Females was only 1400 per tank. 
Hatching began on 14 June 1998 in all tanks. Hatching was encouraged by gently 
mixing the water and the broken egg skeins to allow those larvae trapped within the 
gelatinous matrices to be freed. The broken egg skeins, and any dead eggs and larvae 
were removed. 
26 
Larval Sampling 
We sampled 15 larvae at hatch from each tank systematically from the first tank 
in the first block to the last tank in the last block. Larvae from each tank were sampled 
five at a time to prevent mortality during sample processing. The larvae were first 
anesthetized with Tricaine Methansulfonate (MS-222), and then videotaped for analysis 
of morphometry. We sampled 12 larval traits, of which 9 were morphometric measures 
and 3 were biochemical measures of condition (Table 1). The individual larvae were 
then placed into cryovials and flash-frozen in liquid nitrogen, then stored at - 80 °c for 
subsequent extraction of nucleic acids. 
Post-Sampling Processing of Eggs and Larvae 
Analysis of morphometric landmarks of the eggs and larvae was conducted with 
Optimas imaging analysis system (v 6.1 , Media Cybernetics, Takoma Park, MD) at the 
Chesapeake Biological Laboratory in Solomons, MD on videotaped images. Calibrations 
for imaging analysis were derived from videotaped images of micrometers taken before 
sampling of both the eggs and the larvae. The two-dimensional surface area ( calculated 
internally by Optimas from the outline of the egg or egg yolk) and three sides of a 
triangle inscribed within the outline of each female's individual eggs and egg yolks were 
measured. Total egg volume and egg yolk volumes were then determined based upon the 
following equations adapted from Miller et al. (1995) . 
27 
4 Volwne(mm3) = -Jr R 2 
3 
R= abc 
4K 
K = ,.js(s-aXs-bXs-c) 
a+b+c 
S=---
2 
Where R is the estimated radius of the sphere and a, b and c represent the three triangle 
sides inscribed within each egg and egg yolk. 
Total length, eye diameter, body depth at the insertion point of the pectoral fin, 
body depth at the insertion point of the first dorsal fin (inclusive of the yolk-sac), and 
body depth at the insertion point of the anus were measured for each individual larva 
sampled from each female. The two-dimensional surface area of each larva's yolk-sac 
( calculated internally by Optimas from the outline of each larva's yolk-sac) was 
measured along with the major and minor yolk axes. Total yolk-sac volume was then 
estimated as: 
Volume(mm3 ) = ~(Length of the Minor Yolk Axis) · (Surface Area) 
.., 
The two-dimensional body area of each larva ( calculated internally by Optimas from the 
outline of each larva) was also measured, and a calibration from other newly hatched 
yellow perch larvae between two-dimensional body area and dry weight allowed 
estimation of individual larval dry weight (Dry weight (µg) = [(2 x 10-6) x (Body Area)2 
+ (6 x 10-5) x (Body Area)+ 3 x 10-5] x 1000, n = 171 , r2 = 0.9836, p = 0.0001; Appendix 
I). 
28 
Nucleic acids were extracted from each individual larva Larvae we d 
· re remove 
from -80 °C storage and placed on ice. Whole larvae were placed in individual 
microcentrifuge tubes containing 150 µl of 1 % sarcosyl solution (N-lauroylsarcosine: 
Sigma L-5125). The microcentrifuge tubes were vortexed at high speed for 1 hour to 
fully break down the tissue and dissociate the nucleic acids. After 1 hour of vortexing, 
1.35 ml of Tris buffer (5mM Tris-HCL (Trizma Base: Sigma T-8524, HCL: Sigma H-
1758), 0.5 mM EDTA (Sigma E-5134), pH 7.5) was added to each microcentrifuge tube 
to dilute the total nucleic acid concentration. Tubes were mixed by inversion, and 
centrifuged at 14,000 g for 15 minutes to separate any remaining tissue structures. 75 µl 
Aliquots of the supernatant from each microtube were combined with 75 µl ofEthidium 
Bromide (EB) (ISC Bioexpress C-5515-10), a nucleic acid specific fluorophore, in 96-
well microplates. Dilution series of calfliver rRNA (Sigma R-0889) and calf thymus 
DNA (Sigma D-4 764) standards, blanks of 0.1 % sarcosyl, and a 9 dph yellow perch 
homogenate standard were also combined with EB (75 µ1 : 75 µ1). The microplate was 
shaken at Jow speed for 5 minutes to fully mix the sample and EB. The micropJate was 
read on a fluorescence plate reader (Ex= 545 nm, Em= 575 nm) (IDEXX FCA-VIP, 
JDEXX Corp., Portland, ME) to determine total nucleic acid (TNA) fluorescence in each 
sample. 7.5 µl of RNase A (Sigma R-6513) was then added to each well of the 
micropJate. The plate was shaken at low speed for 20 minutes to digest the RNA in each 
sample, followed by fluorescence determination of the DNA in each sample. The 
fluorescence due to RNA was then determined for each sample by subtraction of the 
DNA fluorescence from the 1NA fluorescence. Calibration curves for RNA and DNA 
were generated from the fluorescence values obtained for the appropriate dilution series. 
29 
The concentrations of RNA and DNA in each sample were then detennined from the 
standard calibration curves, and the ratio of total RNA (concentration* 1.5 ml) to total 
DNA ( concentration* 1.5 ml) in each sample was calculated. To reduce error associated 
with any one day's reading, only a few of the larvae from an individual tank we d re rea on 
a given day. 
Statistics 
Separate statistical analyses were conducted to address our three objectives. All 
statistical analyses were conducted using SAS (v 6.12, SAS Institute, Inc., Cary, NC). 
We used 12 larval and eight maternal traits in our analyses (Table 1). The presence of 
significant maternal effects at hatching in the offspring of the ten females was detennined 
by multivariate analysis of variance (MANOV A; SAS Proc GLM) on all larval traits at 
hatching. The MANOV A was based on the mean value of each larval trait within each 
tank, for each female nested within her date of fertilization (2 June or 4 June 1998). 
Significance of MANOV A was detennined based upon a Wilks Lambda statistic with 96 
and 259.507 degrees of freedom. 
The larval traits that accounted for the observed maternal effects at hatch were 
detennined by univariate analysis of variance (ANOVA; SAS, Proc Mixed) on 
biologically important larval traits within each tank, for each female nested within her 
date of fertilization. Significance of each ANOV A was determined based on an F 
statistic with 8 and 48 degrees of freedom. 
To determine which female traits correlated with the larval traits that accounted 
for the observed maternal effects at hatch we performed a canonical correlation analysis 
30 
on the female and larval data (SAS, Proc Cancorr). Significance of the overall canonical 
correlation analysis was based on a Wilks Lambda statistic with 80 and 274 95 d f 
. egrees o 
freedom. We report the trait loadings on significant canonical correlation vectors only. 
Results 
The females that we did catch had a fairly wide distribution of phenotypic traits. 
Female total length ranged from 216 mm to 287 mm and female weight ranged from 1 l3 
g to 261 g (Table 2). Females ranged in age from 2 years to 6 years (Table 2). The GSis 
ranged from 0.22 to 0.63, while egg production ranged from 11,400 to 36,720 eggs 
(Table 2). Correlation analysis indicated that female total length, female weight, Fulton 's 
condition factor k, female body depth, age, and egg production were all positively 
correlated with one another, and negatively correlated with GSI. This suggests a tradeoff 
between the GSI of the female and the female's size, age and egg production. 
Mean hatching success of the eggs was highly variable among the ten maternal 
lines. The lowest mean number of offspring observed was for Fema1e2, 89 ± 20, 
representing a 4.5 % (n;nitial = 1968 ± 124) hatching success (Table 3). Excluding 
Female2, given the poor quality of her egg skein, hatching success narrowed to a range of 
896 ± 181 offspring for Female3, representing 44.8 % (n;nitial = 2000 ± 324) hatching 
success, to 82.5 % (I 148 ± 198 (n;0 ;1;a1 = 1392 ± 156)) for Females (Table 3). 
At hatching, 90 larvae from each female were sampled. There was a loss of a 
portion of videotape record for each female due to a problem with the video recording 
device that was not evident until analysis of the videotape began. Means of individual 
larval traits are based on variable sample sizes ranging from 67 to 81 larvae per female 
31 
(Table 4). The means of individual larval morphometric traits varied among females by 
between 10.2 % (mean larval eye diameter, followed by 10.3 % variation in mean larval 
body depth at the insertion of the pectoral fin) and 29.3 % (mean RNA) (Table 4). Mean 
RNA ranged from 2.7 ± 0.63 µg to 3.8 ± 0.55 µg while DNA ranged from 1.7 ± o.21 µg 
to 2.3 ± 0.30 µg between the hatchlings (Table 4). Despite relatively large differences in 
the mean quantity of RNA and DNA in the larvae, the mean ratio ofRNA:DNA only 
varied between 1.5 ± 0.32 and 1.8 ± 0. 78 among the offspring of the ten maternal lines 
(Table 4). 
Correlation analysis of tank mean values for 10 larval traits showed a highly 
variable covariance structure (Fig. 3). Some trait pairs were highly correlated (e.g. larval 
. body depth at the insertion of the pectoral fin and larval body depth at the insertion of the 
anus (r = 0.8297, n = 60, p = 0.0001); larval total length and larval dry weight (r = 
0.7093, n = 60, p = 0.0001); RNA and RNA:DNA ratio (r = 0.7406, n = 60, p = 0.0001). 
However, other trait pairs exhibited no significant correlation ( e.g. larval total length and 
larval yolk volume (r = -0.1309, n = 60, p = 0.3188); larval yolk volume and larval dry 
weight (r = 0.0527, n = 60, p = 0.6893). Both negative and positive correlations were 
present in the morphometric data (Fig. 3). Nucleic acid-based traits showed strong 
correlations internally. However, with the exception of a correlation between DNA and 
larval dry weight, there were only weak correlations between nucleic acid-based traits 
and morphometric ones (Fig. 3). 
We were able to reject the null hypothesis of no overall maternal effect among the 
.c-.c-. • f h t maternal Jines for larval traits based upon MANOVA results (F96 ousprmg o t e en , 
_ 2 21 _ 0 OOOl) Univariate ANOVA indicated an overall significant difference 259.507 - , , p - · . 
32 
- ----·--- - -
between the offspring of the ten maternal lines for larval total length, larval yolk volume, 
larval dry weight, and DNA (Table 5; Figs. 4, 5, 6, 7). 
There was a significant canonical correlation between maternal and larval traits 
(F so, 274.95 = 4.51 , p = 0.0001 ). The canonical correlation analysis indicated that canonical 
correlation vectors 1, 2 and 3 were the only significant vectors, with vector 1 explaining 
54% of the variance, vector 2 explaining 33.6%, and vector 3 explaining 5.5% of the 
variance. Because canonical correlation vector 3 explains only 5.5% of the variance, we 
did not consider this vector further. The female traits, age, GSI and egg production were 
strongly associated with the Female 1 canonical variable. Female age and egg production 
drive canonical variable Female 1 in a negative direction while GSI drives it in a positive 
direction (Table 6). This provides further evidence for a potential trade-off between 
female GSI and other traits such as egg production and age. The canonical variable 
Female 2 is a strong positive descriptor of female size and age (Table 6). The larval 
traits, larval total length, larval yolk volume and larval body depth at the insertion of the 
first dorsal fin are strongly associated with the Larval 1 canonical variable. Larval total 
length drives the canonical variable Larval 1 in a positive direction while larval yolk 
volume and larval body depth at the insertion of the first dorsal fin drive it in a negative 
direction (Table 7). This suggests a potential trade-off between larval total length and 
larval yolk volume. The canonical variable Larval 2 is a strong positive descriptor of 
larval size and nucleic acid content (Table 7). Furthermore, 54% of the variation we 
observed among the larvae of the ten maternal lines was explained most by differences in 
larval total length, larval yolk volume and larval body depth at the insertion of the first 
dorsal fin, and was associated most with differences in female GSI, egg production and 
33 
age (Table 8). An additional 33.6% of the variation among the larvae of the ten maternal 
lines was expressed most in larval size (particularly larval dry weight) and nucleic acid 
content, and was most associated with differences in female size and age (Table 8). 
Discussion 
We detected significant maternal effects at hatch among the offspring of the ten 
yellow perch females. Analyses indicate that these effects likely result from differences 
among females in size, age, GSI, and egg production. Furthermore, larvae express the 
maternal effects by differences in larval total length, larval yolk volume, larval dry 
weight and DNA quantity. However, we could not find evidence of significant 
differences among the eggs of the different females. 
Previous research on maternal effects has focused largely on correlations between 
the phenotypic traits of the female and the phenotypic traits of her eggs. There has been 
substantially less focus on the relationship between female phenotypic traits and the 
phenotypic traits of her offspring, both immediately following hatching and throughout 
the larval stage. This focus is perhaps driven by the need of fisheries scientists to predict 
recruitment as early as possible in the fish's ontogeny and by the underlying hypotheses 
that egg size is predictive of larval size and subsequent performance (see reviews by 
Chambers and Leggett 1996; Chambers and Waiwood 1996). Also, propagule size is an 
obvious expression of maternal effects that should correlate with the female's phenotype 
(Bernardo 1996; Solemdal 1997). In light of this, our failure to detect maternal effects in 
the eggs of yellow perch is surprising. We believe that our lack of data on egg traits most 
likely accounts for the results observed. The information that we were able to collect, 
34 
egg volume and yolk volume, bore no relationship to either maternal phenotype or larval 
phenotype. Had we been able to collect more data on the female's eggs (e.g. dry weight 
and lipid content), maternal effects in the egg stage of yellow perch may have been 
expressed. 
Our inability to detect maternal effects in the egg stage of yellow perch was not a 
precursor to a lack of detectable maternal effects in the newly hatched offspring. 
Differences in the phenotypic traits of the hatchlings from the ten maternal lines were 
significantly explained by difference in the phenotypic traits of the females. Previous 
research in other species has shown that maternal effects were expressed in hatching 
length, size, yolk volume, and oil globule volume (Chambers et al. 1989; Buckley et al. 
1991; Chambers and Leggett 1996). Our findings suggest that in yellow perch maternal 
effects are expressed most in larval length, yolk volume, dry weight and DNA content 
(Figs. 4, 5, 6, 7) (Table 5). We can ignore the expression of maternal effects in larval 
body depth at the insertion of the first dorsal fin because this measure is inclusive of the 
yolk-sac and is likely to be driven by yolk volume. Since the quantity of DNA in a cell is 
relatively constant within a given species, we expect a higher number of cells in larger 
larvae than in smaller larvae, and thus a higher quantity of DNA. Therefore, there should 
be a correlation between larval size and DNA content. Thus, if maternal effects were 
expressed in one trait they would be expressed in the other. Near allometric relationships 
have been shown between larval length and total DNA for capelin, cunner Tautogolarbus 
adspersus, and radiated shanny Ulvaria subbifurcata (Pepin et al. 1999). Our results 
provide further evidence for this pattern. 
35 
A lack of significant maternal effects in the RNA content and RNA:DNA ratio of 
the hatchlings from the ten maternal lines indicates that condition of larvae was not 
significantly different at hatching (Table 5) suggesting that differences in protein 
synthesis may not be exhibited until larvae begin feeding on exogenous food sources, or 
perhaps until they have utilized much of their endogenous food source. Changes in 
condition are likely to occur at different rates once the larvae begin depleting their yolk-
sacs and body reserves, and once they begin first feeding. Such variability in RNA:DNA 
ratios of first feeding larvae has been shown for Atlantic herring (Clemmensen 1994), 
Iberian sardine Sardina pilchardus (Chicharo 1998) and red drum Sciaenops ocellatus 
(Rooker et al. 1997). 
The age distribution of females in this study was surprisingly wide, given the 
recent population structure changes that have occurred. Our initial expectations had been 
that all mature female yellow perch caught in 1998 would be greater than 4 years old. 
Female GSI, age, size and egg production are the key female phenotypic traits that 
explained differences in the size, yolk volume and DNA content of offspring (Table 8). 
There appears to be a potential trade-off between the GSI of the female and her age, size 
and egg production. It appears that old, large females with high egg production exhibit 
low GSI values, while younger, smaller females have high GSI values while producing 
fewer eggs. The strong association between larval total length and larval yolk volume 
and female GSI, egg production and age suggests that it is these female traits that 
potentially determine the length and amount of yolk a larva may have at hatching. 
Additionally, the association between larval size and nucleic acid content and female size 
and age suggests that it is these female traits that potentially determine the size and 
36 
amount of nucleic acids of a larva at hatching. These results suggest that the traits of 
larvae produced by inexperienced spawners (i.e. spawners of a young age) may be very 
different from the traits of larvae produced by experienced spawners. While previous 
studies of maternal effects have not shown trade-offs between larval characteristics that 
are driven by female characteristics, some correlations have been found. Larval size has 
been positively correlated with female size (Bernardo 1996), and female size and age 
have been positively correlated with fecundity (Hislop 1988; Buckley et al. 1991 ; 
Chambers and Waiwood 1996). 
The canonical correlation analysis clearly identifies a trade-off between hatchling 
total length and larval yolk volume within canonical vector 1. Furthermore, evidence for 
this trade-off is observed in the box plots for hatchling total length and yolk volume. 
Across all maternal lines, those hatchlings with the highest total lengths exhibit the 
lowest yolk volumes, and vise-versa. The non-significant correlation between larval total 
length and yolk volume is likely due to the variation of those larval traits within maternal 
lines caused by a few unique individuals (outliers). Despite these correlation results, 
there is a trend in the data suggesting that a negative relationship between larval total 
length and yolk volume does exist, and may have become more evident with a larger 
sample size. The trade-off between larval size and yolk volume has potentially 
substantial implications on the larvae' s vulnerability to starvation and predation. 
Larger larvae are thought to have an advantage over smaller larvae in that they 
have been shown to swim faster and thus avoid predators more easily, search greater 
distances for food, capture larger sizes and quantities of prey, and survive periods of food 
shortages longer (Knutsen and Tilseth 1985; Miller et al. 1988; Chambers et al. 1989; 
37 
Chambers and Leggett 1996;). However, this does not take into consideration the 
dangers of initial starvation, or the consequences of potential spatial and temporal overlap 
with prey. Should a larva be faced with a lack of food upon hatching, it will be solely 
dependent upon its yolk-sac and body reserves for survival. Our results indicate a trade-
off potentially exists between larval size and yolk volume, and that trade-off is driven by 
key characteristics of the female spawning stock. Under circumstances of food limitation 
immediately after hatching, we would expect smaller larvae with larger yolk-sacs to have 
a survival advantage over larger larvae with smaller yolks, in terms of survival. 
However, assuming that the food limitation is not indefinite, then we would begin to 
expect the larger larvae to again have a survival advantage over the smaller larvae as 
feeding and predator avoidance begin to become important. This suggests that 
potentially, bigger is not necessarily better in all cases. 
Heavily exploited fisheries, especially those which target faster growing, larger 
females will be much more susceptible to the phenotypic bottlenecking of the female 
population and may subsequently be at higher risk to recruitment failure when 
environmental conditions shift. Solemdal ( 1997) argues that overfishing of the female 
portion of the spawning stock will cause a population to shift from its fitness peak since 
maternal effects are phenotypic. We believe that phenotypic bottlenecking, resulting 
from overfishing of the female portion of the population, was a strong component of the 
causes of the recruitment failure of yellow perch in Lake Michigan. The Lake Michigan 
ecosystem has dramatically changed over the past decade and a half with the introduction 
of many new exotic species, particularly the zebra mussel (Dreissena polymorpha). It is 
likely that changes in the environment and high fishing pressure, have resulted in stresses 
38 
on the population. No single recruitment mechanism is likely to be responsible for 
recruitment variation and failure, rather a combination of many mechanisms is likely 
responsible (Leggett and Deblois 1994). However, Solemdal (1997) has argued that 
because of the complex nature of recruitment mechanisms, maternal effects may be the 
only recruitment mechanism that can be effectively managed in a fishery through 
regulation. 
39 
Acknowledgements 
We would like to thank the Wisconsin Department of Natural Resources 
Milwaukee Fish Unit and Dan Anderson and his family for all of their efforts in obtainino 
. b 
our yellow perch brood stock. We thank Heather Brussat, Richard and Lori Fulford 
' 
Emily Kircher, Chris Ney and Steve Yeo for their help in the collection and husbandry of 
the yellow perch. We would also like to thank Mary Chenery and Hind Eloukkal for 
their help in the videotape imaging analysis. Mary Haasch and Ed Houde for all of their 
support, enthusiasm, contributions and efforts throughout this research. We thank Roger 
Dawson for technical guidance in fluorometric analysis and Bob Anderson for the use of 
his fluorometer to analyze our nucleic acid samples. Funding for this research was 
provided by the University of Wisconsin Sea Grant College Program. 
40 
References 
Adams, S.M. and D.L. DeAngelis. 1987. Indirect effects of early bass-shad interactions 
on predator population structure and food web dynamics, p. 103-117. In: W.C. 
Kerfoot and A. Sih (eds.) Predation: Direct and Indirect Impacts on Aquatic 
Communities. University of New England Press, Hanover, N.H. 
Beacham, T.D. and C.B. Murray. 1985. Effect of female size, egg size, and water 
temperature on developmental biology of chum salmon ( Oncorhyncus keta) from 
the Nitinat River, British Columbia. Can. J. Fish. Aquat. Sci. 42: 1755-1765. 
Bengston, D.A., R.C. Barkman, and W.J. Berry. 1987. Relationships between maternal 
size, egg diameter, time of spawning season, temperature, and length at hatch of 
Atlantic silverside, Menidia menidia. J. Fish. Biol. 31: 697-704. 
Benoit, H.P. and P. Pepin. 1999. Interaction ofrearing temperature and maternal 
influence on egg development rates and larval size at hatch in yellowtail flounder 
(Pleuoronectes ferrugineus). Can. J. Fish. Aquat. Sci. 56: 785-794. 
Bernardo, J. 1996. Maternal effects in animal ecology. Amer. Zool. 36: 83-105. 
Buckley, L.J. , A.S. Smigielski, T.A. Halavik, E.M. Caldarone, B.R. Bums, and G.C. 
Laurence. 1991 . Winter flounder Pseudopleuornectes americanus reproductive 
success. II. Effects of spawning time and female size on size, composition and 
viability of eggs and larvae. Mar. Ecol. Prog. Ser. 74: 125-135. 
Chambers, R.C. and W.C. Leggett. 1996. Maternal influences on variation in egg sizes 
in temperate marine fishes . Amer. Zool. 36: 180-196. 
41 
Chambers, R.C. and W.C. Leggett. 1992. Possible causes and consequences of variation 
in age at size and metamorphosis in flatfishes (Plueronectiformes): An analysis at 
the individual, population, and species levels. Neth. J. Sea. Res. 29: 7-24. 
Chambers, R.C. and K.G. Waiwood. 1996. Maternal and seasonal differences in egg 
sizes and spawning characteristics of captive Atlantic cod, Gadus morhua. Can. 
J. Fish. Aquat. Sci. 53: 1986-2003. 
Chambers, R.C. , W.C. Leggett, I.A. Brown. 1989. Egg size, female effects, and the 
correlations between early life history traits of Capelin, Mallotus villosus: An 
appraisal at the individual level. Fish. Bull. 87: 515-523. 
Chicharo, M.A. 1998. Nutritional condition and starvation in Sardina pilchardus (L.) 
larvae off southern Portugal compared with some environmental factors. J. Exp. 
Mar. Biol. Ecol. 225: 123-137. 
Clemmensen, C. 1994. The effect of food availability, age or size on the RNA/DNA 
ratio of individually measured herring larvae: laboratory calibration. Mar. Biol. 
118: 377-382. 
Crowder, L.B. , J.A. Rice, T.J. Miller, E.A. Marschall. I 989. Empirical and theoretical 
approaches to size-based interactions and recruitment variability in fishes, p. 237-
255. In: DeAngelis, D.L. and L.J. Gross (eds.) Individual-Based Models and 
Approaches in Ecology. Chapman and Hall, New York, New York. 
Francis, J.T. , s.R. Robillard, and J.E. Marsden. 1996. Yellow perch management in 
Lake Michigan: a multi-jurisdictional challenge. Fisheries. 21(2): I 8-20. 
Gall, G.A.E. 1974. Influence of size of eggs and age of female on hatchability and 
growth in rainbow trout. Calif. Fish and Game. 00(1): 26-35. 
42 
Hislop, J.R.G. 1988. The influence of maternal length and age on the size and weight of 
the eggs and the relative fecundity of the haddock, Melanogrammus aeg!ejinus, in 
British waters. J. Fish. Biol. 32: 923-930. 
Houde, E.D. l 987. Fish early life dynamics and recruitment variability. Amer. Fish. 
Soc. Symp. 2: 17-29. 
Huang, W, T. Chui, and C. Shih. 1999. Effects of maternal conditions on early life 
history traits of black porgy Acanthopagrus schlegeli. I. Appl. lchthyol. 15: 87-
92. 
Kazakov, R. V. l 981. The effect of the size of Atlantic salmon, Sa/mo salar L., eggs on 
embryos and alevins. J. Fish. Biol. 19: 353-360. 
Knutsen, G.M. and S. Tilseth. l 985. Growth, development, and feeding success of 
Atlantic cod larvae Gadus morhua related to egg size. Trans. Am. Fish. Soc. 
114: 507-51 I. 
Laurence G.C. 1988. Toward a consensus paradigm of early life stage survival of fishes 
and recruitment-research in the face of variability and uncertainty. ICES ELH 
No. l I. 
Leggett, W. c. and E. Deblois. l 994. Recruitment in marine fishes: is it regulated by 
starvation and predation in the egg and larval stages. Neth. J. Sea. Res. 32(2): 
119-134. 
Miller, T.J., T. Herra, and W.C. Leggett. 1995. An individual-based analysis of the 
variability of eggs and their newly hatched larvae of Atlantic cod (Gadus morhua) 
on the Scotian Shelf. Can. J. Fish. Aquat. Sci. 52: 1083-1093. 
43 
Miller, T.J. , L.B. Crowder, I.A. Rice, and E.A. Marschall. 1988. Larval size and 
recruitment mechanisms in fishes: toward a conceptual framework. c J F" h an . . IS . 
Aquat. Sci. 45: 1657-1670. 
Pepin, P. 1989. Predation and starvation of larval fish: a numerical experiment of size-
and growth-dependent survival. Biol. Ocean. 6: 23-44. 
Pepin, P. and T.J. Miller. 1993. Potential use and abuse of general empirical models of 
early life history processes in fish. Can. J. Fish. Aquat. Sci. 50: 1343-1345. 
Pepin, P. , G.T. Evans, and T.H. Shears. 1999. Patterns of RNA/DNA ratios in larval fish 
and their relationship to survival in the field. ICES J. Mar. Sci. 56: 697-706. 
Rakitin, A. , M.M. Ferguson, and E.A. Trippel. 1999. Sperm competition and 
fertilization success in Atlantic cod (Gadus morhua): effect of sire size and 
condition factor on gamete quality. Can. J. Fish. Aquat. Sci. 56: 2315-2323. 
Rice, I.A., T.J. Miller, K.A. Rose, L.B. Crowder, E.A. Marschall, A.S. Trebitz, D.L. 
DeAngelis. 1993. Growth rate variation and larval survival: inferences from an 
individual-based size-dependent predation model. Can. J. Fish. Aquat. Sci. 50: 
133-142. 
Rijnsdorp, A.D. 1989. Maturation of male and female North Sea plaice (Pleuronectes 
platessa L.). J. Cons. Int. Explor. Mer. 46: 35-51. 
Rooker, J.R. , G.J. Holt, and S.A. Holt. 1997. Condition of larval and juvenile red drum 
(Sciaenops ocellatus) from estuarine nursery habitats. Mar. Biol. 127: 387-394. 
Solemdal, P. 1997. Maternal effects - a link between the past and the future. J. Sea. 
Res. 37: 213-227. 
Wisconsin Natural Resources Board (WNRB) Memorandum. 1996. File Ref. #3600. 
44 
~ 
v-, 
Table 1. Maternal and larval traits measured in each of the yellow perch (Percajlavescens) females and sampled offspring of the ten maternal lines. 
Total Length (mm) 
Weight (g) 
Maternal Traits 
Uody Depth at the Anterior Dase of the First Dorsal Fin (mm) (FDD 1) 
Body Depth at the Posterior Uase of the Second Dorsal Fin (mm) (Fl3D2) 
Age (years) 
Egg Production (# Eggs) 
Fulton's Condition Factor k (g/mm3) 
Gonda! Somatic Index (GS!) 
Linear measures 
Biochemical measures 
Volume measures 
Derived measures 
Total Length (mm) 
Eye Diameter (mm) 
Minor Yolk Axis (mm) 
Major Yolk Axis (mm) 
Larval Traits 
Uody Depth at the Insertion Point of the Pectoral Fin (mm) (ODIP) 
Uody Depth at the Insertion Point of the First Dorsal Fin (mm) (DDID) 
Uody Depth at the Insertion Point of the Anus (mm) (13DIA) 
RNA Content (pg) (RNA) 
DNA Content (~1g) (DNA) 
Yolk Volume (mm3) 
Dry Weight (pg or mg) 
RNA:DNA Ratio 
Table 2. Phenotypic characteristics of the ten female yellow perch (Percaflavescens) caught off Green Can Reef 42° 50' 00" N, 87° 50' 00" W), Lake Michigan. 
Females 1-5 were collected on 2 June 1998, and Females 6-10 were collected on 4 June 1998. All trait abbreviations are as defined in Table I. 
Female Total Length Weight (g) Fulton's Condition Age FBD1 FBD2 GSI Egg (mm) Factor (k) (Years) Production 
1 250 145.3 0.929856 4 53 47 0.3589 22560 
2 227 124.7 1.066078 4 46 44 0.6284 11400 
3 280 255 .6 1.164222 4 72 57 0.5676 36075 
4 249 151.3 0.980228 2 54 43 0.6170 23560 
5 256 169.9 l.012683 3 55 46 0.3824 11600 
..i:,.. 6 285 261.4 1.29244 5 65 61 0.2869 36720 0\ 
7 287 255.7 l.081730 6 65 62 0.2168 29440 
8 216 113 .1 l.122479 3 47 40 0.4064 30600 
9 222 115.7 1.057760 3 45 41 0.2875 21970 
10 242 115.4 0.814465 4 49 46 0.3989 32760 
.i::,. 
-...l 
Table 3. Mean initial number of eggs(± SD) in each female's replicate tank (n = 6), mean number of offspring at hatching(± SD), and percent hatching success 
in the ten yellow perch maternal lines. 
Female Mean Initial Number of Eggs (n=3) Mean Initial Number of Offspring Hatching Success (%) (Hatching Success) (n=6) 
1 2002 ± 14 1615±289 80.7 
2 1968 ± ll 89 ±20 4.5 
3 2000 ± 18 896 ± 181 44.8 
4 2024 ± 15 1538 ± 133 76.0 
5 1392 ± 12 1148 ± 198 82.5 
6 2040 ± 12 1269 ± 228 62.2 
7 2020 ± 11 1610 ± 290 79.7 
8 2002 ± 6 1061 ± 381 53.0 
9 2032 ± 14 1379 ± 164 67.9 
10 1965 ± 17 1287 ± 188 65.5 
Table 4. Mean phenotypic characteristics(± SD) of the yellow perch (Percajlavescens) offspring of the ten maternal lines sampled at hatching on 14 June, 
1998. All trait abbreviations are as defined in Table I. 
Mean Mean 
Total Mean Eye Yolk-sac Mean Mean Mean Mean Dry Mean Mean Mean 
Length Diameter Volume BDIP BOID BDIA Weight RNA DNA RNA:DNA 
Female N (mm) (mm) (mm3) (mm) (mm) (mm) (mg) (µg) (µg) Ratio 
1 80 6.1 ± 0.47 0.38 ± 0.05 ± 0.7 ± 0.63 ± 0.56 ± 0.22 ± 0.023 3.4 ± 2.1 ± 1.7 ± 0.69 0.044 0.015 0.091 0.087 0.083 0.98 0.34 
2 70 6.2 ± 0.39 0.39 ± 0.04 ± 0.68 ± 0.64 ± 0.54 ± 0.22 ±0.019 3.2 ± 2.2 ± 1.5 ± 0.45 0.042 0.014 0.093 0.094 0.021 0.93 0.30 
3 78 6.1 ± 0.47 0.39 ± 0.04 ± 0.68 ± 0.66 ± 0.55 ± 0.22 ± 0.022 3.4 ± 2.1 ± 1.7 ± 0.47 0.05 0.012 0.086 0.091 0.087 0.5 0.37 
.j:::.. 0.38 ± 0.05 ± 0.73 ± 0.67 ± 0.59 ± 3.4 ± 2.2 ± 00 4 81 6.2 ± 0.46 0.23 ± 0.022 1.6 ± 0.43 0.044 0.015 0.088 0.083 0.079 0.66 0.29 
5 76 6.1 ±0.44 0.39 ± 0.05 ± 0.69 ± 0.64 ± 0.56 ± 0.22 ± 0.017 3.2 ± 2.1 ± 1.5 ± 0.32 0.043 0.015 0.086 0.087 0.081 0.6 0.25 
6 77 5.5 ± 0.48 0.42 ± 0.06 ± 0.67 ± 0.75 ± 0.53 ± 0.20 ± 0.018 3± 1.9 ± 1.6 ± 0.71 0.047 0.012 0.087 0.085 0.069 0.61 0.31 
7 67 6.1 ± 0.41 0.39 ± 0.07 ± 0.72 ± 0.8 ± 0.57 ± 0.24 ± 0.024 3.8 ± 2.3 ± 1.6 ± 0.29 0.045 0.017 0.077 0.11 0.088 0.55 0.30 
8 75 5.6 ± 0.4 0.38 ± 0.06 ± 0.69 ± 0.8 ± 0.51 ± 0.2 ± 0.019 2.9 ± 1.8 ± 1.7 ± 0.67 0.048 0.015 0.091 0.14 0.078 0.68 0.38 
9 70 5.5 ± 0.43 0.35 ± 0.05 ± 0.65 ± 0.73 ± 0.51 ± 0.19 ± 0.016 2.7 ± 1.7 ± 1.6 ± 0.41 0.04 0.01 l 0.081 0.096 0.075 0.63 0.21 
10 78 5.7 ± 0.42 0.42 ± 0.07 ± 0.69 ± 0.76 ± 
0.51 ± 0.21 ± 0.018 3.3 ± 1.9 ± 1.8 ± 0.78 0.036 0.014 0.096 0.112 0.081 0.53 0.42 
Table 5. Univariate ANOVA results for tests of significant maternal effects, for individual larval traits, among the offspring of the ten maternal lines (n = 752). 
All trait abbreviations are as defined in Table I. Univariate tests were performed with 8 and 48 degrees of freedom . 
Larval Trait F Pr>F 
Mean Total Length 5.86 0.0001 
Mean Eye Diameter 0.55 0.8096 
Mean Yolk Volume 3.45 0.0032 
Mean BDIP 1.69 0.1242 
MeanBDID 2.04 0.0617 
Mean BDIA 1.62 0.1437 
Mean Dry Weight 10.78 0.0001 
Mean RNA 1.68 0.1284 
.i:,. 
Mean DNA 7.31 0.0001 \0 
Mean RNA:DNA Ratio 0.00 1.0000 
~"'-~"'-~~'\..'\.."-~~ ~,~'s;:,...'\.~~~ 
Table 6. Canonical correlation analysis: female loading results for the eight female phenotypic traits of the ten female yellow perch (Percajlavescens). All trait 
abbreviations as defined in Table 1. 
Canonical Variables 
Female Traits Female 1 Female 2 
Total Length 0.0140 0.5477 
Weight -0.0476 0.4185 
Fulton's Condition Factor (k) -0.0474 -0.1657 
Age -0.4114 0.4898 
GSI 0.6980 -0.0551 
Egg Production -0.5160 0.0808 
v, FBD1 0.0800 0.4096 
0 
FBD2 -0.1976 0.4985 
Table 7. Canonical correlation analysis: larval loading results for the ten larval phenotypic traits from the ten yellow perch (Percajlavescens) maternal lines. 
All trait abbreviations as defined in Table I. 
Canonical Variables 
Larval Traits Larval 1 Larval 2 
Total Length 0.6226 0.4855 
Eye Diameter 0.4139 0.2733 
Yolk Volume -0.6858 0.4244 
BDIP 0.2217 0.4417 
BOID -0.7681 0.2451 
l3DIA 0.3847 0.3566 
V, Dry Weight 0.3887 0.7954 
...... 
RNA 0.1192 0.5172 
DNA 0.4416 0.7385 
RNA:DNA Ratio -0.1574 0.0429 
v-, 
N 
Table 8. Canonical correlation analysis: results for the eight female phenotypic traits and the 12 larval phenotypic traits of the ten maternal lines. Canonical 
correlation vector I explains 54% of the variation while canonical correlation vector 2 explains 33 .6% of the variation . Canonical correlation vector I 
is positively driven by female GSI and negatively driven by egg production. Canonical correlation vector 2 is positively driven by female size and age. 
All traits abbreviations as defined in Table I. 
Larval Trait 
Proportion of Variance 
Total Length 
Eye Diameter 
Yolk Volume 
BDIP 
BDID 
BDIA 
Dry Weight 
RNA 
DNA 
RNA:DNA Ratio 
Vector 1 
(Driven by Female GSI (+) and 
Female Egg Production (-) 
54% 
0.5852 
0.3891 
-0.6446 
0.2804 
-0.7220 
0.3616 
0.3653 
0.1120 
0.4151 
-0.1480 
Vector 2 
(Driven by Female Size and Age)(+) 
33.6% 
0.4410 
0.2483 
0.3856 
0.4013 
0.2226 
0.3240 
0.7226 
0.4698 
0.6710 
0.0390 
Figure Legends 
Figure 1. Annual yellow perch (Percajl.avescens) catch per unit of effort (number of fish 
per 1000 ft. of gillnet per night) from the WDNRFU winter gillnet survey over 
Green Can Reef (42° 50' 00" N, 87° 50' 00" W) (a historic yellow perch 
spawning ground) in Lake Michigan from 1986 to 1998. 
Figure 2. Average size at age for Lake Michigan female yellow perch (Percajl.avescens) 
collected from fishery independent surveys conducted by Ball State University 
in Illinois waters from 1983 to 1995 (S.M. Shroyer, Department of Biology, 
Ball State University, Muncie, IN 47306-0440, USA "personal 
communication"). 
Figure 3. Bivariate correlations among IO larval traits of yellow perch (Perea 
jl.avescens) . Data plotted are mean values of each trait based on analysis of 
between 67-81 larvae in each of 60 tanks that comprised the experiment. 
Figure 4. Box and whiskers plot of total length of yellow perch (Percaflavescens) 
offspring from ten maternal lines at hatching. The upper and lower limits of 
the box represent the 3rd (Q3) and 1st (Q.1) quartiles, respectively. The solid 
horizontal line within the box represents the median value of the data. The 
dashed horizontal line within the box represents the mean value of the data. 
The upper and lower extremes of the whiskers represent (Q3+ l.5*(Q3-Q1)) and 
(Q1-l.5*(Q3-Q 1)) , respectively. Data outside of the whiskers are considered 
outliers, and are represented by open circles. Closed circles represent the 
minimum and maximum values of the data. 
53 
Figure 5. Box and whiskers plot of dry weight of yellow perch (Percaflavescens) 
offspring from ten maternal lines at hatching. Box-whisker antennae as 
defined in Fig. 4. 
Figure 6. Box and whiskers plot of yolk volume of yellow perch (Percaflavescens) 
offspring from ten maternal lines at hatching. Box-whisker antennae as 
defined in Fig. 4 . 
Figure 7. Box and whiskers plot of DNA content of yellow perch (Percaflavescens) 
offspring from ten maternal lines at hatching. Box-whisker antennae as 
defined in Fig. 4. 
54 
U) 
U) 
8000 
-
..... ! 7000 I ~ ---- -\ {\ DAges >4 •Ages 2- 4 
-----
(1) 
C 6000 -
-
·-C, 
It- 5000 · 0 
wi 
::>. 0 
D. 0 4000 -
0 0 
~ 
-.c 3000 -U) 
·-It-
It-
0 2000 -
'-Q) 
.c 
E 1000 -::::, 
C: 
-
0 
1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 
Year 
V) 
0\ 
-.. 
E 
350 -
300 -
250 .. 
E 200 -
-
.c 
...., 
g> 150 
Q) 
..J 
100 
50 ·-
0 -,-
D / A 
/ tr' / / / . .. 
) 21 ,1:( . / ~ 13/ / -- / .-,-,/~ · --- • - -- -+--------• ---·- --• -----
/ 
ft _,,/ . - _,,,t--::,;;.,-- ----- • 
/ ,,,.,• / / 
f 
/ / ? A.fr __ .... - · ---· ----· -----· 
11 v ---• .---- • ----•----
~?~; ... . --·--- ·-
~ / 
····--,- ·-·--, . -·· ·--· , - · -T- --· 1 ··-· ··- ·1-
--- ------
·- ··• - 19a3 
- • - 1985 
--A- 1987 
- • - 1989 
···-A· 1991 
- G- 1993 
- 0 -- 1995 
·---------
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 
·-- -r·-·-i---·--r··--· r--
r-·----i 
Age (years) 
Vl 
-..J 
5 
TL 
BDIP 
YVol 
0.9 
BOID 
BDIA 
Q.4 
0.26 
RNA 
3 
Ratio 
DNA 
TL= Total Length (mm) 
ED = Eye Diameter (mm) 
YVol = Yolk Volume (mm3) 
BDIP = Body Depth at the Insertion of the Pectoral Fin (mm) 
BOID= Body Depth at the Insertion of the Dorsal Fin (mm) 
BDIA = Body Depth at the Insertion of the Anus (mm) 
DWt = Dry Weight (mg) 
RNA= RNA Content (~tg) 
DNA = DNA Content (µg) 
Ratio = Ratio of RNA:DNA 
D = Significant Correlation (n = 60) (SAS, Proc Corr) 
The numbers in the diagonal boxes represent the upper and lower ranges of each trait. 
8 · 
7.5 - · 
• T I T T 
• 
7 -
6.5 · 
-E 6 · E 
-
.c 
-g> 5.5 
Cl) 
_J 
V1 
00 ! 51 0 
• 4.5 
I 
• ..L 0 
4 I • 
• • 
3.5 J 
3 -•- -· ·· ·- -· --- ------ · ·-- -· ·- ---· 
1 2 3 4 5 6 7 8 9 10 
Maternal Line 
v-, 
\0 
0.00035 .. ,. 
0.0003 ., 
ai 0.00025 1. 
-
i 
..., 
J: 
C) 
~ 
~ 
C 0.0002 --
0.00015 + 
0.0001 ., ... · -- -
1 
t 
J 
2 
l t T 
l l @ 
-L • • 
• • 
--- - -- ·- ···---· - -·-·-·-- ·-
3 4 5 6 
Maternal Line 
• 
0 
T • 
l 
@ 
• 
..L Tr 0 
• i i 
---··· - - -----. ----- -- - ---· 
7 8 9 10 
0\ 
0 
-M E 
E 
-
0.14 ~-
0.12 -,--
0.1 -•·· 
c» 0 .08 ,_ 
E 
::s 
0 
> 
~ 0.06 -•-~ 
0.04 -
0.02 ,_ 
• 
• • 
0 
• 
0 ·'- -·-·· - ----- .. --- - 0 - ... - ---- . 
1 2 3 
• 
• 
• 
0 
8 
0 
• 
• 
-·- · ---- - ---- -- - - ------ -- ----- - ---- · -
4 5 6 7 8 9 10 
Maternal Line 
0\ 
....... 
en 
:::::, 
-,IJ 
C 
Cl) 
3.5 · 
3 -0-
2.5 --
1: 2 
0 
0 
<( 
z 
C 
1.5 · 
1 
0.5 -- -
• 
• 
1 
• 
0 
• 
2 
I 
B 
0 
• 
3 
0 
• 
4 
(:} 
0 
• 
5 
• 
0 
0 
0 
0 
• 
6 
Maternal Line 
0 
• 
7 
8 
e 
• 
8 
8 
• 
9 10 
- === 
Chapter 3 
Understanding Maternal Effects as a Recruitment Mechanism in 
Lake Michigan Yellow Perch (Percajlavescens) 
II. Resistance to Starvation 
62 
- ===-=== 
Abstract 
Maternal effects have been detected previously in Lake Michigan yellow perch 
(Perea flavescens). The persistence and potential importance of maternal effects under 
conditions of food limitation were examined between hatching and the time of 100% 
mortality in ten maternal lines. Furthermore, the importance of maternal effects were 
evaluated under such conditions to test the null hypothesis of no difference among the 
maternal lines in their resistance to starvation. By 4 days post hatch (dph) nearly all 
larvae had fully absorbed their yolk-sacs. Maternal effects were detectable at hatching 
and at four and six dph, but not at two dph. Analyses lead to a rejection of the null 
hypothesis and indicated that the detected maternal effects resulted in differences of a 
factor of two among the maternal lines in resistance to starvation. Furthermore, larvae 
that exhibited the lowest resistance to starvation were long and had small yolk volumes 
' 
while larvae that exhibited the highest resistance to starvation were characterized as short 
with and large yolk volumes. Rates of mortality differed among the yellow perch 
maternal lines. There was evidence for a mortality threshold based on an RNA:DNA 
ratio of about 1.15. The observed differences in resistance to starvation can have 
substantial implications for recruitment success. Maternal effects in yellow perch result 
in differential mortality of larval progeny under conditions of starvation. 
63 
Introduction 
Recruitment in fishes can vary substantially from year to year, a fact that has been 
recognized since 1914 when Johan Hjort first addressed it. For over 80 years fisheries 
scientists have sought to understand what mechanisms may control recruitment and cause 
the high inter-annual variability observed in many exploited fish species. One potential 
controlling mechanism of recruitment that received substantial attention during the early 
1900s and throughout most of the 20th century was starvation. 
The magnitude of mortality directly attributable to starvation, and thus the 
magnitude of starvation effects on recruitment remain unclear at the present time (Leggett 
and Deblois 1994). Starvation-induced mortality depends not only on food availability 
but also on the individual larva's resistance to starvation (Rice et al. 1987). Research has 
recognized that starvation does not affect all individuals equally. Resistance to starvation 
increases with larval size (Hunter 1981 ; Miller et al. 1988), and increased resistance to 
starvation affords greater flexibility in encountering food and thus increases the chance 
for survival (Chambers et al. 1989). 
Recently, researchers have recognized that even if the importance of food 
limitation as a direct recruitment controller may be small, its importance as a modifier to 
predation may be quite large. Food limitation can reduce escape swimming speeds 
(Elliot and Leggett 1997; Litvak and Leggett 1992) thus increasing a larva's vulnerability 
to predation. Furthermore, food limitation can lead to reduced growth rates (Buckley 
1984), which can result in lengthened stage durations and thus increased exposure to 
predation (Houde 1987; Bertram et al. 1997). Furthermore, food limitation reduces larval 
body tissues and reserves that are essential for survival. Such starvation-induced changes 
64 
in larval traits may result in increased vulnerability to predation. Larvae that are more 
resistant to starvation will exhibit reduced rates of change in larval size and thus, are 
likely to be less vulnerable to predation than larvae that are less resistant to starvation. 
Larvae that exhibit high resistance to starvation are more likely to exhibit greater 
swimming ability that will allow them to better avoid and elude predators (Rice et al. 
1987) and search longer and further for suitable prey (Miller et al. 1988) than larvae with 
low resistance to starvation. 
Maternal effects can control larval size at hatching (Beacham and Murray 1985; 
Bengston et al. 1987; Chambers et al. 1989; Chambers and Leggett 1992, and this thesis -
Chapter 2). Additionally, research on maternal effects has shown that larger females 
typically produce larger larvae that are more resistant to starvation (Huang et al 1999). 
Previously, we have shown that maternal effects were present at hatching in ten maternal 
lines of Lake Michigan yellow perch (Percaflavescens) (Chapter 2). These maternal 
effects were primarily driven by female size and age and were expressed in the hatchlings 
by differences in larval total length, yolk volume, dry weight and DNA content (Chapter 
2). 
Research has shown that yellow perch larvae can exhibit high sensitivity to food 
limitation. Yell ow perch larvae, fed on exogenous food sources until they reached a 
length of 10 mm, experienced 50% mortality in an average of 6 days after exogenous 
feeding was experimentally limited (Letcher et al. 1996). Letcher et al. (1996) also 
measured time to 50% inactivity, the time at which 50% of the larvae within a tank were 
laying inactive on the bottom. Time to 50% inactivity in yellow perch occurred, on 
average, 2.6 days before the onset of 50% mortality (Letcher et al. 1996). 
65 
' ~ 
These findings, when coupled with previous findings of maternal effects in yellow 
perch, lead us to evaluate the importance of maternal effects on differences in resistance 
to starvation among hatchlings of different maternal lines. As recruitment mechanisms 
such as starvation have been found to be size-dependent (Miller et al. 1988), we 
hypothesized that the offspring of the ten maternal lines would have varying 
vulnerabilities to starvation due to the initial differences in larval size and yolk volume 
that were observed at hatching (Chapter 2). This study aimed to investigate the 
importance of maternal effects on the resistance to starvation in Lake Michigan yellow 
perch larvae. We sought to determine if differences in larval traits observed at hatching 
resulted in differences in survival, expressed as resistance to continuous starvation from 
hatching. Additionally, we wished to quantify which larval traits most accounted for 
differences in resistance to starvation if found, and whether the maternal effects 
expressed at hatching persisted throughout the duration of starvation. 
Materials and Methods 
We investigated our objectives within a single experiment, which involved 
monitoring and quantifying several larval traits and survival of larval yellow perch when 
starved from hatch. All larvae used were the offspring of the ten maternal lines described 
in Chapter 2. Experimental work was conducted at the University of Wisconsin' s 
WATER Institute, Milwaukee, WI. The experiment was conducted as a randomized 
complete block design, consisting of three blocks of 20 10-gallon flow-through aquaria 
with two replicates per maternal line per block. 
66 
Full details of the collection and husbandry of eggs and larvae are given in 
Chapter 2, and are only summarized here. Male and female yellow perch were collected 
on 2 June and 4 June 1998 in the vicinity of Green Can Reef (42° 50' 00" N, 87° 50' 00" 
W) off Milwaukee, WI. Eggs were fertilized in the field and returned to the lab. Egg 
volumes, sufficient to produce approximately 2000 larvae, were incubated in each of 60 
10-gallon flow-through, temperature-controlled aquaria. 
Husbandry of Larvae 
Hatching began on 14 June 1998 in all tanks. After the removal of the broken egg 
skeins and any dead eggs and larvae, the larvae in all 60 tanks were sampled for the 
presence of maternal effects at hatch. After sampling at hatch, one of two treatments, fed 
or starved, was randomly applied to one replicate tank of each maternal line within each 
block. The 30 starved treatment tanks were not given any exogenous food throughout the 
duration of the experiment. The flow rate and temperature in the tanks were consistent 
with successful yellow perch rearing conditions established at the WATER Institute from 
previous research. Mortalities in each tank were removed daily and enumerated. 
Larval Sampling 
At hatching, 15 larvae were sampled from each tank. Subsequently, five larvae 
were sampled every two days post hatch (dph) until 100% mortality occurred in all tanks. 
We quantified 12 larval traits (9 morphometric and 3 biochemical measures of condition_ 
Table 1) for each larva sampled from each tank on each sampling day. Larvae were 
anesthetized with Tricaine Methansulfonate (MS-222), and then videotaped for analysis 
67 
of morphometry. Individual larvae were placed into cryovials, flash-frozen in liquid 
nitrogen, and then stored at -80 °C for subsequent extraction of nucleic acids. 
Post-Sampling Processing of Larvae 
Morphometric analysis of the larvae was conducted on videotape images of larvae 
using Optimas (v 6.1, Media Cybernetics, Takoma Park, MD) as defined in Chapter 2. 
Larval total length, eye diameter, two-dimensional yolk surface area ( calculated 
internally by Optimas from the outline of the yolk), the major and minor axes of the yolk, 
body depth at the insertion point of the pectoral fin, body depth at the insertion point of 
the first dorsal fin (inclusive of yolk-sac), body depth at the insertion point of the anus, 
and body area ( calculated internally by Optimas from the outline of the larvae) were 
measured for each individual larva sampled (Table 1). From these primary 
measurements yolk-sac volume and dry weight were estimated using methods defined in 
Chapter 2. Extraction and determination of nucleic acids was also performed as 
described in Chapter 2, with two exceptions. Microplates were read on a fluorescence 
plate reader at Ex = 590 nm, Em = 625 nm (Spectra Max Gemini, Molecular Devices 
Corp. , Sunnyvale, CA). Furthermore, growth and development of the larvae following 
absorption of their yolk-sacs required an additional dilution of the extracted nucleic acid 
supernatant (Table 2). This additional dilution factor was used in the conversion of RNA 
and DNA concentration to RNA and DNA content (concentration*l.5 ml*additional 
dilution factor). 
68 
Statistics 
Separate statistical analyses were conducted to address each of our objectives. 
All statistical analyses were conducted using SAS (v 6.12, SAS Institute, Inc. , Cary, NC). 
For all traits except larval yolk volume, the analyses were conducted from hatching 
through six dph. Virtually I 00¾ mortality had occurred by seven dph in all tanks. The 
persistence of significant maternal effects throughout the duration of the experiment 
(time at 100% mortality in all tanks) was determined by multivariate analysis of variance 
(MANOV A; SAS Proc GLM) on all larval traits, except survival, for each sampling day. 
The MANOVA was based on the mean value of each larval trait within each starved tank 
' 
for each maternal line nested within the date of fertilization (2 June or 4 June 1998). 
Significance ofMANOVA was determined based upon a Wilks' Lambda statistic with 
varying degrees of freedom. 
Variation in survival over time and among the maternal lines was evaluated by 
repeated measures analysis of variance (MANOVA; SAS Proc GLM). The proportion of 
larvae surviving on each day was arcsine transformed before analysis. The repeated 
measures analysis was based on the mean value of survival within each tank, for each 
maternal line on each sampling day. Repeated measures analysis was conducted o, 2, 4 
and 6 dph. Significance of the test of the null hypothesis of no significant difference in 
survival over time (i.e. time effect) was determined based upon a Wilks' Lambda statistic 
with 3 and 16 degrees of freedom. Significance of repeated measures analysis for the test 
of the null hypothesis of no difference in survival among the offspring of the ten maternal 
lines over time (i.e. interaction between mother and time) was determined based upon a 
69 
Wilks' Lambda statistic with 27 and 47.37 degrees of freedom. This test of significance 
examines whether the offspring of the maternal lines were dying at different rates. 
The repeated measures analysis was conducted on the sampling dates indicated 
above for all larval traits. The analysis was based on the mean value of each larval trait 
within each tank, for each maternal line on each sampling day. However, by four dph, 
full yolk-sac absorption had been observed in all but two of 150 sampled larvae. 
Therefore, repeated measures analysis for larval yolk volume was only conducted from 
hatching through two dph. Significance of repeated measures analysis for the test of the 
null hypothesis of no variation in the larval traits (Table 1) over time was determined 
based upon a Wilks' Lambda statistic with 3 and 16 degrees of freedom for the 
morphometric traits, with 1 and 20 degrees of freedom for larval yolk volume and with 3 
and 15 degrees of freedom for the biochemical traits. Differences among the starved 
offspring of the maternal lines over time were determined by repeated measures analysis 
on each larval trait at each sampling day. Significance of repeated measures analysis for 
the test of the null hypothesis of no difference among the offspring of each maternal line 
for the larval traits over time was determined based upon a Wilks ' Lambda statistic with 
27 and 47.37 degrees of freedom for the morphometric traits, with 9 and 20 degrees of 
freedom for larval yolk volume, and with 27-and 44.45 degrees of freedom for the 
biochemical traits. These tests of significance examine whether the ten maternal lines 
have different rates of change for a given larval trait over time. Significance of repeated 
measures analysis was determined based upon an F statistic with 9 denominator degrees 
of freedom for the test of the null hypothesis of no difference among the offspring of each 
maternal line for a given larval trait (i.e. mother effect). This test of significance is 
70 
analogous to evaluating differences in the absolute magnitude of a given larval trait 
among the offspring of the ten maternal lines. 
Results 
Significant maternal effects were detected at hatching and on two of the three post 
hatch sampling days. At two dph we were unable to reject the null hypothesis of no 
overall maternal effect among the offspring of the ten maternal lines for the larval traits 
based upon MANOVA results (Fso, 65.65 = 1.33, p = 0.1143). However, at four dph and 
again on six dph we were able to detect a significant overall maternal effect among the 
offspring of the ten maternal lines based upon MANOVA results (Table 3). 
Mortality rates differed among the offspring of the ten maternal lines by a factor 
of two. Time to 50% mortality ranged from two to six dph among the offspring of the ten 
maternal lines. The offspring of Maternal Line3 were the least resistant to starvation, 
reaching 50% mortality by 2.7 ± 0.58 days. The offspring of Maternal Line6 were the 
most resistant to starvation, reaching 50% mortality by 6.3 ± 0.58 days (Table 4). At 6 
dph the proportion oflarvae surviving varied from 10 ± 9.1 % to 54 ± 13.0 (Table 4). 
Repeated measures analysis showed that survivorship varied significantly over time 
(Table 5). We were able to reject the null hypothesis of no difference in the survivorship 
of the offspring of the ten maternal lines over time (Table 5). The rate of mortality was 
significantly different among the offspring of the ten maternal lines (F27, 47_37 = 3.097, p === 
0.0003). 
A positive relationship between mean larval yolk volume at hatching and mean 
time to 50% mortality was observed among the maternal lines (Fig. 1). However, no 
71 
strong relationships were observed between any of the other larval traits measured at 
hatching and time to 50% mortality. While there is a positive relationship between larval 
yolk volume at hatching and time to 50% mortality, there is a negative relationship 
between larval total length at hatching and time to 50% mortality (Fig. 1). 
We quantified the trait variability at 50% mortality for the offspring of the ten 
maternal lines. Mean larval total length at the time of 50% mortality ranged from 5. 7 ± 
0.17 mm (Maternal Line9) to 6.28 ± 0.096 mm (Maternal Line1) (Fig. 2, Table 6). Larval 
yolk volume at the time of 50% mortality varied by a factor of two among the offspring 
of the ten maternal lines, ranging from 0.0373 ± 0.0014 mm3 (Maternal Line2) to 0.0689 
± 0.0066 mm3 (Maternal Line1) (Fig. 3, Table 6). Body depth at the insertion of the first 
dorsal fin ranged between 0.58 ± 0.042 mm (Maternal Line2) and 0.71 ± 0.032 mm 
(Maternal Line7). Larval dry weight ranged between 0.199 ± 0.0034 mg (Maternal Lines) 
and 0.237 ± 0.0068 µg (Maternal Line1) (Fig. 4, Table 6). RNA content at the time of 
50% mortality ranged from 2.4 ± 0.18 µg (Maternal Lines) to 3.5 ± 0.53 µg (Maternal 
Line7) (Fig. 5), while DNA content did not exhibit large variation, ranging between 2.2 ± 
0.19 µg (Maternal Lines) and 2.9 ± 0.31 µg (Maternal Line1) (Table 6). Condition, the 
ratio of RNA:DNA, at the time of 50% mortality also exhibited low variation, ranging 
between 1.1 ± 0.37 (Maternal Line1) and 1.26 ± 0.074 (Maternal Line7) (Fig. 6, Table 6). 
A strong positive relationship was observed among the maternal lines between 
mean larval yolk volume at the time of 50% mortality and mean time to 50% mortality 
(Fig. 7). However, no strong relationship was observed between other larval traits 
measured at time of 50% mortality and time to 50% mortality. While there is a positive 
relationship between larval yolk volume at time of 50% mortality and time to 50% 
72 
mortality, there is a negative relationship between larval total length at time of 50% 
mortality and time to 50% mortality (Fig. 7). 
Starvation induced changes in the traits of the offspring from the ten maternal 
lines were observed between hatching and the time of 100% mortality for all maternal 
lines. Between hatching and two dph, yolk absorption was highly variable among the 
offspring of the ten maternal lines. However, by four dph virtually all larvae had fully 
absorbed their yolk-sacs. Between hatching and two dph, larval yolk volume decreased 
by as much as 40 ± 17.7 % (Maternal Line4) and by as little as 1 ± 29.7 % (Maternal 
Line10). Some offspring of Maternal Line9 had increased yolk volumes by two dph, thus 
the average change in yolk volume for Maternal Line9 was negative. By six dph, large 
variation in body depth at the insertion of the first dorsal fin was observed in all maternal 
lines. Between hatching and six dph, body depth at the insertion of the first dorsal fin 
decreased by as much as 40 ± 3.1 % (Maternal Line10) and by as little as 9 ± 38.7 % 
(Maternal Line7) (Table 7). Body depth at the insertion of the first dorsal fin increased in 
some offspring of Maternal Line 1, as indicated by a mean change in body depth at the 
insertion of the first dorsal fin of -1 ± 52.2 %. Variation in the dry weight of the 
offspring from the ten maternal lines was much lower by six dph than for other traits. 
Larval dry weight decreased by six dph by as much as 12 ± 6.6 % (Maternal Line1) and 
by as little as 1 ± 3.9 % (Maternal Line9). Offspring from Maternal Line6, Maternal Line7 
and Maternal Line10 exhibited increases in larval dry weight over the six day period 
(Table 7). RNA content and the ratio of RNA:DNA decreased in all maternal lines by six 
dph while DNA increased. Decreases in the RNA content of the offspring from the ten 
maternal lines ranged from 47.3 ± 38.15 % (Maternal Line1) to 69 ± 14.8 % (Maternal 
73 
Line4) (Table 7), while decreases in the ratio of RNA:DNA ranged from 53 ± 42.1 % 
(Maternal Line1) to 72.4 ± 10.46 % (Maternal Line4) (Table 7). DNA content increased 
by as much as 42 ± 6.1 % (Maternal Line2) by six dph, and by as little as 8 ± 12.5 % 
(Maternal Line4) (Table 7). 
All larval traits except body depth at the insertion of the pectoral fin varied 
significantly over time (Table 8). We were unable to reject the null hypothesis of no 
significant difference in the larval traits of the offspring of the ten maternal lines over 
time (Time*Mother Interaction in Table 8). This indicates that the change in larval traits 
observed over time was not significantly different among the offspring of the ten 
maternal lines. Furthermore, repeated measures ANOV A allowed us to reject the null 
hypothesis of no overall difference in larval total length, larval yolk volume, body depth 
at the insertion of the first dorsal fin and larval dry weight of the offspring of the ten 
maternal lines (Table 8). No significant differences in body depth at the insertion of the 
pectoral fin, body depth at the insertion of the anus, RNA content, and the ratio of 
RNA:DNA among the ten maternal lines were detected through ·repeated measures 
ANOV A (Table 8). 
Discussion 
We detected significant maternal effects at hatching and at 4 and 6 dph. There 
were no detectable maternal effects at two dph. There were significant differences in the 
survival of offspring from the ten maternal lines. Time to 50% mortality and the larval 
traits at time of 50% mortality varied greatly among the ten maternal lines. Furthermore, 
while the rate of mortality was significantly different among the offspring of the ten 
74 
maternal lines, the rates of change of aII the larval traits during the 6 day period were not 
significantly different from one another. 
It is not clear why we were unable to detect maternal effects at two dph. One 
possible explanation is that larval yolk volume was a strong driving factor in the analysis. 
Differential yolk absorption rates may have resulted in increased within maternal line 
variation in yolk-sac volume, resulting in an apparent reduction in between maternal line 
variation and thus a lack of detectable maternal effects. By four dph, virtuaIIy all larvae 
had fully absorbed their yolk-sacs and larval yolk volume was eliminated as a trait in the 
analysis for the detection of maternal effects at four dph. The absence of larval yolk 
volume at four and six dph in the MANOV A may have resulted in the significant 
detection of maternal effects among the offspring of the ten maternal lines for the 
remaining traits. 
The rate of mortality differed significantly among the offspring of the ten 
maternal lines. Raising the question of which traits could have accounted for the 
differences. We found significant differences in larval size and yolk volume among the 
offspring of the ten maternal lines. While the absolute measures of each of these traits 
were significantly different among the offspring of the ten maternal lines, the rate of 
change of these traits were not significantly different among the ten maternal lines. 
Furthermore, the non-significant differences in the rates of change of the larval traits 
among the offspring of the ten maternal lines indicates that differences in survival can 
only be attributed to differences in larval size and yolk volume at hatch. It is probable 
that the absolute magnitude oflarval size and yolk volume would result in survival 
75 
differences of the offspring due to size-dependent susceptibilities to starvation (Hunter 
1981; Miller et al. 1988). 
A two-fold difference in the time to 50% mortality was observed among the 
offspring of the ten maternal lines. Maternal Line2 and Maternal Line3 exhibited the 
lowest resistance to starvation. The offspring of Maternal Line2 and Maternal Line
3 
were 
characterized by the lowest hatching successes, long total length and small yolk volume 
(Chapter 2). The offspring of Maternal Line6 exhibited the highest resistance to 
starvation and were characterized by moderate hatching success, short total length and 
large yolk volume (Chapter 2). This suggests that larval total length and larval yolk 
volume are important factors in determining a larva's resistance to starvation. 
Furthermore, it appears as if the trade-off between larval size and larval yolk volume 
observed at hatching is more than an anomaly (Chapter 2). While previous research has 
demonstrated that larger larvae typically have larger yolk-sacs and are more resistant to 
starvation than smaller larvae (Hunter 1981, ~iller et al. 1988), these results suggest that 
in yellow perch a trade-off between larval total length and larval yolk volume exists and 
that it is not the size of the larva that determines resistance to starvation, but rather the 
relative size of the yolk-volume. 
When we examine the individual larval traits at the time of 50% mortality we find 
that differences in larval yolk volume and body depth at the insertion of the first dorsal 
fin, which is inclusive of larval yolk volume, characterize differences observed in the 
time to 50% mortality. Furthermore, we find that larvae with large yolk volumes at 
hatching exhibit longer times to 50% mortality. This provides further support to the 
conclusion regarding the importance of larval yolk volume at hatch. Offspring of 
76 
Maternal Line2 and Maternal Line3 had small yolk volumes and shallow measures of 
body depth at the insertion of the first dorsal fin while the offspring of Maternal Line
6 
had large yolk volumes and relatively deep measures of body depth at the insertion of the 
first dorsal fin. This suggests that the yolk volume and the body depth at the first dorsal 
fin in newly hatched larvae are important predictors of resistance to starvation, as 
measured by time to 50% mortality. 
The lack of variability in the ratio of RNA:DNA among the ten maternal lines at 
the time of 50% mortality is striking. The consistency among maternal lines suggests 
that there may be a minimum threshold in the ratio ofRNA:DNA that is necessary for 
survival. Our results indicate that an estimate for this mortality threshold in yellow perch 
based on an RNA:DNA ratio is about 1.15. This estimate is similar to values reported for 
other species under conditions of starvation. Reported values for RNA:DNA ratios at or 
near the 'point of no return ' have been reported for Atlantic herring Clupea harengus 
larvae as I. 1 and 1.2 (Clemmesen 1987; Ueberschar and Clemmesen 1992, respectively), 
for striped bass Marone saxatilis larvae as 1.2 (Wright and Martin 1985), for North Sea 
plaice Pleuronectes platessa larvae as 1.0 (Hovenkamp 1990), and for turbot 
Scophthalmus maximus larvae as 1.3 (Clemmesen 1987). 
The changes observed in the larval traits over the duration of starvation were by 
and large expected, and indicate which body morphometrics are first reduced when food 
availability is low. The yolk is the only food source that larvae have during periods of 
starvation, so naturally larval yolk volume would decrease. Decreases in dry weight of 
larvae also were expected due to the depletion of reserves and metabolic expenditures. 
Starvation-induced reductions in dry weight of 10-25 mm larval yellow perch have been 
77 
reported as high as 53% reduction from original mass (Letcher et al. 1996). We observed 
only 1-12% reductions in dry weight as a consequences of starvation. The differences in 
dry weight reductions of starving yellow perch may be attributable to age differences. 
Decreases in the RNA content and RNA:DNA ratio of the larvae throughout starvation 
also were expected. The condition of larvae decreases in the absence of food. The 
decline of RNA and the ratio ofRNA:DNA during starvation has been observed for many 
species in the field and in the laboratory (Buckley 1980; 1984; Clemmesen 1987· 
' 
Westerman and Holt 1994; Rooker and Holt 1996; Elliot and Leggett 1998). 
With the exception of the observed increase in DNA content, the observed 
increases in some larval traits were striking. Previous research has shown that increases 
in DNA content during periods of low food availability have been shown to be indicative 
of starvation when coupled with decreases in RNA content (Buckley 1979; 1981; 
Clemmesen 1994). Given that the quantity of DNA per cell is relatively constant within 
somatic tissues of a species (Clemmesen 1994), we would conclude that cell proliferation 
is continuing even though food is limited. The observed increases in some maternal lines 
in mean larval total length, body depth at the insertion point of the pectoral fin, body 
depth at the insertion point of the first dorsal fin, body depth at the insertion point of the 
anus, and dry weight may be indicative of within maternal line differences in resistance 
to starvation. 
Much of the previous research on maternal effects has neglected to focus on the 
expression of maternal effects in larval traits, and even less research has focused on the 
empirical evaluation of maternal effects on the survival of offspring. Many previous 
studies have only assumed that maternally influenced differences in larval traits at 
78 
hatching, primarily larval size, result in differences in survival (Chambers and Leggett 
1996, Solemdal 1997). These assumptions have been based upon findings that larval 
survival is size-dependent (Houde 1987; Hunter 1981; Miller et al. 1988). One example 
of direct empirical evidence of significant maternal effects on the survival of offspring 
under conditions of starvation has been shown for black porgy Acanthopagrus schlegeli 
(Huang et al. 1999). Huang et al. (1999) observed differences in resistance to starvation 
among the offspring of 30 maternal lines. They found that larger females produced larger 
offspring that were more resistant to starvation than smaller offspring from smaller 
females (Huang et al. 1999). Our findings contradict those found in black porgy. We 
have found that small females produce short larvae with low dry weight that have large 
yolk-sacs and that are more resistant to starvation than the longer larvae with high dry 
weight and small yolk-sacs produced by large females. This contradiction in findings 
suggests that maternal effects may operate differently across species. 
Understanding differences in the survival of offspring among different maternal 
lines under conditions of starvation is important to the understanding of what drives 
recruitment. If females with favorable phenotypic traits produce larvae that have a high 
resistance to starvation, than a portion of those females should be managed to conserve 
their traits to ensure against environmental uncertainty. Many changes have occurred in 
Lake Michigan over the past decade that have lead to changes in plankton abundance, 
consequently, it is important to protect yellow perch females with a wide distribution of 
traits that will produce larvae with a wide distribution of traits, in order to protect the 
population from environmental uncertainty and continued recruitment failures. 
79 
Acknowledgements 
We would like to thank Fred Binkowski, Heather Brussat, Richard and Lori 
Fulford, Emily Kircher, Chris Ney, Jim Rice and Steve Yeo for their help in the 
husbandry and maintenance of the yellow perch larvae. We would also like to thank 
Mary Chenery for her help in the videotape imaging analysis, and Hind Eloukkal for her 
help with the enumeration of daily mortalities. We thank Vanessa Javor for her help in 
sample preparation for nucleic acid extractions. We would also like to thank Elaine 
Caldarone for her continued guidance and support in the fluorometric analysis of our 
nucleic acid samples. Funding for this research was provided by the University of 
Wisconsin Sea Grant College Program. 
80 
References 
Beacham, T.D. and C.B. Murray. 1985. Effect of female size, egg size, and water 
temperature on developmental biology of chum salmon ( Oncorhyncus keta) from 
Nitinat River, British Columbia. Can. J. Fish. Aquat. Sci. 42: 1755-1765. 
Bengston, D.A., R.C. Barkman, and W.J. Berry. 1987. Relationships between maternal 
size, egg diameter, time of spawning season, temperature, and length at hatch of 
Atlantic silverside, Menidia menidia. J. Fish. Biol. 31: 697-704. 
Bertram, D.F, T.J. Miller, and W.C. Leggett. 1997. Individual variation in growth and 
development during early life stages of winter flounder Pleuronectes americanus. 
Fish. Bull. 95: 1-10. 
Buckley, L.J. 1979. Relationship between RNA-DNA ratio, prey density and growth 
rate in Atlantic cod (Gadus morhua) larvae. J. Fish. Res. Board Can. 36: 1497-
1502. 
Buckley, L.J. 1980. Changes in the ribonucleic acid, deoxyribonucleic acid, and protein 
content during ontogenesis in winter flounder, Pseudopleuronectes americanus, 
and effect of starvation. Fish. Bull. 77: 703-708. 
Buckley; L.J. 1981. Biochemical changes during ontogenesis of cod (Gadus morhua L.) 
and winter flounder (Pseudopleuronectes americanus) larvae. Rapp P-v Reun 
Cons. perm. int. Explor. Mer. 178: 547-552. 
Buckley, L.J. 1984. RNA-DNA ratio: an index of larval fish growth in the sea. Mar. 
Biol. 80: 291-298. 
Chambers, R.C. and W.C. Leggett. 1996. Maternal influences on variation in egg sizes 
in temperate marine fishes. Amer. Zool. 36: 180-196. 
81 
Chambers, R.C. and W.C. Leggett. 1992. Possible causes and consequences of variation 
in age at size and metamorphosis in flatfishes (Plueronectiformes) : An analysis at 
the individual, population, and species levels. Neth. J. Sea. Res. 29: 7-24. 
Chambers, R.C., W.C. Leggett, and J.A. Brown. 1989. Egg size, female effects, and the 
correlations between early life history traits of capelin, Mallotus villosus: an 
appraisal at the individual level. Fish. Bull. 87: 515-523 . 
Clemmesen, C.M. 1987. Laboratory studies on RNA/DNA ratios of starved and fed 
herring (Clupea harengus) and turbot (Scophthalmus maximus) larvae. J. Cons. 
Int. Explor. Mer. 43: 122-128. 
Clemmesen, C.M. 1994. The effect of food availability, age or size on the RNA/DNA 
ratio of individually measured herring larvae: laboratory calibration. Mar. Biol. 
188: 377-382. 
Elliott, J.K. and W.C. Leggett. 1998. Larval condition and vulnerability to predation: 
and analysis based on mixed-prey experiments. Can. J. Fish. Aquat. Sci. 55: 626-
630. 
Houde, E.D. 1987. Fish early life dynamics and recruitment variability. Amer. Fish. 
Soc. Symp. 2: 17-29. 
Hovenkamp, F. 1990. Growth differences in larval plaice Pleuronectes platessa in the 
Southern Bight of the North Sea as indicated by otolith increments and 
RNA/DNA ratios. Mar. Ecol. Prog. Ser. 58: 205-215. 
Huang, w., T. Chui, and C. Shih. 1999. Effects of maternal conditions on early life 
history traits of black porgy Acanthopagrus schlegeli. I. Appl. Ichthyol. 15: 87-
92. 
82 
Hunter, J.R. 1981. Feeding ecology and predation of marine fish larvae, p. 33-79. In: 
Lasker, R. ( ed.) Marine Fish Larvae: Morphology, Ecology, and Relation to 
Fisheries. University of Washington Press, Seattle, Washington. 
Leggett, W.C. and E. Deblois. 1994. Recruitment in marine fishes: is it regulated by 
starvation and predation in the egg and larval stages. Neth. J. Sea. Res. 32(2): 
119-134. 
Letcher, B.H., I.A. Rice, and L.B. Crowder. 1996. Size-dependent effects of continuous 
and intermittent feeding on starvation time and mass loss in starving yellow perch 
larvae and juveniles. Trans. Am. Fish. Soc. 125: 14-26. 
Litvak, M.K. and W.C. Leggett. 1992. Age and size-selective predation on larval fishes: 
the bigger-is-better hypothesis revisited. Mar. Ecol. Prog. Ser. 81 : 13-24. 
May, R.C. 1974. Larval mortality in marine fishes and the critical period concept. In: 
J.H.S. Blaxter (ed.). The Early Life History of Fish. Springer-Verlag, Berlin. pp. 
3-19. 
Miller, T.J., L.B. Crowder, I.A. Rice, and E.A. Marschall. 1988. Larval size and 
recruitment mechanisms in fishes: toward a conceptual framework. Can J. Fish. 
Aquat. Sci. 45: 1657-1670. 
Rice, J.A. , L.B. Crowder, and F.P. Binkowski. 1987. Evaluating potential sources of 
mortality for larval bloater (Coregonus hoyi): starvation and vulnerability to 
predation. Can. J. Fish. Aquat. Sci. 44: 467-472. 
Rooker, J.R. and G.J. Holt. 1996. Application of RNA:DNA ratios to evaluate the 
condition and growth oflarval and juvenile red drum (Sciaenops ocellatus). Mar. 
Freshwater Res. 47: 283-290. 
83 
Solemdal, P. 1997. Maternal effects - a link between the past and the future. J. Sea. 
Res. 37: 213-227. 
Ueberschar, B. and C. Clemmesen. 1992. A comparison of the nutritional condition of 
herring larvae as determined by two biochemical methods - tryptic enzyme 
activity and RNA/DNA ratio measurements. ICES J. mar. Sci. 49: 245-249. 
Westerman, M. and G.J. Holt. 1994. RNA:DNA ratio during the critical period and 
early larval growth of the red drum Sciaenops ocellatus. Mar. Biol. 121: 1-9. 
Wright, D.A. and F.D. Martin. 1985. The effect of starvation on RNNDNA ratios and 
growth in larval striped bass, Marone saxatilis. J. Fish. Biol. 27: 479-485. 
84 
00 
Vl 
Table 1. Larval traits measured in each of the yellow perch (Percaflavescens) offspring of the ten maternal lines. 
Linear measures 
Total Length (mm) 
Eye Diameter (mm) 
Body Depth at the Insertion Point of the Pectoral Fin (mm) (BDIP) 
Body Depth at the Insertion Point of the First Dorsal Fin (mm) (BDID) 
Body Depth at the Insertion Point of the Anus (mm) (BDIA) 
Biochemical measures 
RNA Content (~lg) (RNA) 
DNA Content (~lg) (DNA) 
Volume measures 
Yolk Volume (mm3) 
Derived measures 
Dry Weight (~lg or mg) 
RNA:DNA Ratio 
00 
0\ 
Table 2. Dilution factors used in the fluorometric determination of RNA content and DNA content in the offspring of the ten yellow perch (Percajlavescens) 
maternal lines. 
Age (dph) Mean Total Length (mm) Dilution Factor Supernatant (µI) 0.1 % Sarcosyl (µI) 
0 5.9 0 1000 0 
2 6.2 0 1000 0 
4 7.1 2 500 500 
6 6.1 2 500 500 
4'-'t'..'- \'\.1 '--.. \..--\~ '3 , , _ ,..~,~• "'--"It\., .. , , , \. 
00 
-..) 
Table 3. MA NOV A results for tests of significance for all morphometric and biochemical larval traits, among the offspring of the ten yellow perch (Perea 
jlavescens) maternal lines. 
Day df F Pr>F 
0 80, 255 .92 2.67 0.0001 
2 80, 65.65 1.33 0.1143 
4 72, 68.41 1.86 0.0053 
6 72,50 .16 1.66 0.0298 
.. '~'"'''--'\J -, ~ .. ',_ 1..\.,.~ ... ,._'C \ ,,,,,,' 
00 
00 
Table 4. Mean time to 50% mortality and percent survival at 6 dph (± SD) (n = 3) of the offspring of the ten yellow perch (Percajlavescens) maternal lines over 
time, based upon the mean value for each tank per female on each sampling day. 
Maternal Linc Minimum (days post hatch) Maximum (days post hatch) Mean (days post hatch) % Survival at 6 dph 
4 5 4.3 ± 0.58 17 ± 9.2 
2 2 4 3 ± 1.2 28 ± 6.8 
3 2 3 2.7 ± 0.58 10 ± 9.1 
4 4 6 5 ± 1.2 38 ± 8.5 
5 5 5 5 ± 0 28 ± 10.8 
6 6 7 6.3 ± 0.58 44 ± 7.6 
7 5 7 6 ±1 49 ± 3.1 
8 3 6 5 ± 1.2 36 ± 12.8 
9 5 7 6 ± 1.2 54± 13.0 
10 3 6 4 ± 1.5 42 ± 6.9 
' '\'-''"'\'--- '\.....;- ,;. -- ,, -1..;,.t ... , ,., .. " ,, .. , .... ,. 
00 
\0 
Table 5. Repeated measures A NOVA results for tests of significance among the yellow perch (Percajlavescens) offspring of the ten maternal lines (n = 118). 
Tests of significant differences among the offspring of the ten maternal lines were performed with 9 degrees of freedom (Mother). Tests of significant 
time effects were performed with 3 and 16 degrees of freedom (Time). Tests of significant differences among the offspring of the ten maternal lines 
over time (Time*Mother) were performed with 27 and 47 .37 degrees of freedom. 
Mother at Age (dph) 
0 2 4 6 Mother Time Time*Mother 
Survival ** ** ** ** *** *** 
Table 6. Mean larval traits at 50% mortality(± SD) (n = 3) of the offspring of the ten yellow perch (Percaflavescens) maternal lines over time, based upon the 
mean value for each tank per female on each sampling day. All trait abbreviations are as defined in Table I. 
Maternal Line 
Larval l 2 3 4 5 6 7 8 9 10 Trait 
- -
Total 5.92± 
Length 6.2 ± 0.15 6.25 ± 0.057 6.1 ±0.19 6.27 ± 0.097 6.12 ± 0.076 5.9 ± 0.23 6.28 ± 0.096 5.75 ± 0.082 5.7 ± 0.17 0.017 (mm) 
Yolk 0.04756 ± 0.0373 ± 0.0379 ± 0.0466 ± 0.0477 ± 0.0576 ± 0.0689 ± 0.0593 ± 0.0550 ± 0.0630 ± Volume 
(mm3) 0.00077 0.0014 0.0023 0.0049 0.0093 0.0084 0.0066 0.0021 0.0023 0.0061 
'° BDIP (mm) 0.73 ±0.092 0.68 ± 0.030 0.67 ± 0.010 0.72 ± 0.023 
0.680± 0.73 ± 0.088 0.78 ± 0.054 0.70 ± 0.012 0.66 ± 0.024 0.70 ± 0 0.0095 0.026 
BDID (mm) 0.60 ± 0.058 0.58 ± 0.042 0.59 ± 0.020 0.061 ± 0.058 ± 0.69 ± 0.032 0.71 ± 0.032 0.63 ± 0.035 0.63 ± 0.063 0.66 ± 0.043 0.027 0.049 
BDIA (mm) 0.55 ± 0.062 0.52 ± 0.024 0.52 ± 0.01 I 0.054 ± 0.51 ± 0.015 0.55 ± 0.053 0.59 ± 0.027 0.49 ± 0.019 0.50 ± 0.034 0.51 ± 0.035 0.028 
Dry Weight 0.213 ± 0.213 ± 0.208 ± 0.224 ± 0.2105 ± 0.213 ± 0.237 ± 0.199± 0.201 ± 0.21 3 ± 
(mg) 0.0065 0.0048 0.0073 0.0078 0.00065 0.0019 0.0068 0.0034 0.0031 0.0061 
H.NA (µg) 3 ± 1.2 3 ± 1.0 3.1 ± 0.60 2.9 ± 0.80 2.7 ± 0.41 2.5 ± 0.21 3.5 ± 0.53 2.4 ± 0.18 3 ± 1.2 2.7 ± 0.11 
DNA (µg) 2.6 ± 0.37 2.7 ± 0.31 2.6 ± 0.29 2.7 ± 0.37 2.5 ± 0.16 2.3 ± 0.21 2.9 ± 0.31 2.2 ± 0.19 2.5 ± 0.73 2.4 ± 0.15 
RNA:DNA I. I ± 0.37 I.I ± 0.23 1.2 ± 0.16 I. I ± 0.11 I. 12 ± 0.093 
1.116 ± 1.26 ± 0.074 1.2 ± 0.29 1.1 ± 0.15 1.21 ± 
Ratio 0.0082 0.079 
Table 7. Mean percent decrease(± SD) (n = 3) in the larval traits of the yellow perch (Percaflavescens) offspring of the ten maternal lines sampled over six 
days, based upon the mean value for each tank per female on each sampling day. Negative percentages indicate an increase in the value of the larval 
trait. All trait abbreviations are as defined in Table I. 
Maternal Line 
Larval 1 2 3 4 5 6 7 8 9 10 Trait 
- -
Total 3 ± 16.7 0.9 ± 4.3 0.2± 5.4 -2± 3.8 0.4±5.\ -9 ± 5.6 -4 ± 2.5 -6 ± 1.3 -5 ± 3.2 -9 ± 4.1 Length 
BDIP -32 ± 67.4 8 ± 14. l -5 ± 6.8 16 ± 7.8 10 ± 10.3 -32 ± 65.3 -41 ± 61.8 -2 ± 5.1 6 ± 5.1 3 ± 11.9 
\0 BOID -1 ± 52.2 28 :f: 20.9 24± 15.5 32 ± 4.2 30 ± 7.2 14 ± 43.0 9 ± 38.7 34 ± 2.5 38 ± 2.3 40 ± 3.1 
..... 
BDIA -6±51.6 I 1 ± 12.2 9 ± 17.0 31 ± 6.4 22 ± 12.6 -8 ± 48.9 -19±59.5 12 ± 7.8 15 ± 8.5 14 ± 12.6 
Dry 12 ± 6.6 5 ± 9.8 9 ± 9.3 11 ± 9.9 8.3 ± 3.1 -2 ± 3.4 -I ± 10.3 4 ± 3.4 1 ± 3. 9 -3 ± 4.0 Weight 
RNA 47 ± 38.2 58 ± 6.6 56± 19.6 69 ± 14.8 61 ± 17.8 62 ± 19.2 52 ± 3.5 53 ± 23.9 52 ± 9. 5 59 ± 16.0 
DNA -22 ± 14.4 -42 ± 6.1 -28 ± 18.3 -8± 12.5 -15 ± 18.8 -17 ± 9.9 -37 ± 22.3 -37 ± 15.2 -28 ± 13.3 -28 ± 2.3 
RNA:DNA 53 ± 42.1 54 ± 30.4 65 ± 17.2 72 ± 10.5 Ratio 
67 ± 11.6 68 ± 15.6 64 ± 1.6 67 ± 19.2 62 ± 8.9 69 ± 12.3 
\0 
N 
Table 8. Repeated measures ANOV A results for tests of significance among the yellow perch (Percaf/avescens) offspring of the ten maternal lines. Tests of 
significant differences among the offspring of the ten maternal lines were performed with 9 degrees of freedom (Mother). Tests of significant time 
effects (Time) were performed with 3 and 16 degrees of freedom for all morphometric larval traits (n = 118), except yolk volume, which was 
performed with I and 20 degrees of freedom (n = 6 I). Tests of significant time effects (Time) were performed with 3 and 15 degrees of freedom for 
all biochemical larval traits (n = 117). Tests of significant differences among the offspring of the ten maternal lines over time (Time*Mother) were 
performed with 27 and 47 .37 degrees of freedom for all morphometric larval traits, except yolk volume, which was performed with 9 and 20 degrees 
of freedom. Tests of significant differences among the offspring of the ten maternal lines over time (Time*Mother) were performed with 27 and 44.45 
degrees of freedom for all biochemical larval traits. All trait abbreviations are as defined in Table I. 
Larval Total Length 
Larval Yolk Volume 
BDIP 
BOID 
BDIA 
Larval Dry Weight 
RNA 
DNA 
Ratio 
0 
** 
*** 
* 
*** 
** 
Mother at Age (dph) 
2 4 
* 
*** 
*** 
*** 
6 Mother Time Timc*Mother 
*** *** 
*** *** 
** *** 
** 
** *** ** 
*** 
*** 
*** 
Figure Legend 
Figure 1. Relationship between mean time to 50% mortality (n = 3) and mean larval total 
length (r2 = 0.2462) , yolk volume (r2 = 0.3434) and dry weight (r2 = 0.0348) at 
hatching (n = 751) for each maternal line (n=IO) of yellow perch (Perea 
flavescens) . 
Figure 2. Box and whiskers plot of total length of yellow perch (Percajlavescens) 
offspring from the ten maternal lines at the time of 50% mortality (n = 3). The 
upper and lower limits of the box represent the 3rd (Q3) and 1st (Qi) quartiles, 
respectively. The solid horizontal line within the box represents the median 
value of the data. The dashed horizontal line within the box represents the 
mean value of the data. The upper and lower extremes of the whiskers 
represent (Q3+ 1.5*(Q3-Q1)) and (Q1-1.5*(Q3-Q1)), respectively. Data outside 
of the whiskers are considered outliers, and are represented by open circles. 
Closed circles represent the minimum and the maximum values of the data. 
Figure 3. Box and whiskers plot of yolk volume of yellow perch (Percajlavescens) 
offspring from the ten maternal lines at the time of 50% mortality (n = 3). 
Box-whisker antennae as defined in Fig. 1. 
Figure 4. Box and whiskers plot of dry weight of yellow perch (Percajlavescens) 
offspring from the ten maternal lines at the time of 50% mortality (n = 3). 
Box-whisker antennae as defined in Fig. 1. 
Figure 5. Box and whiskers plot of RNA content of yellow perch (Percajlavescens) 
offspring from the ten maternal lines at the time of 50% mortality (n = 3). 
Box-whisker antennae as defined in Fig. 1. 
93 
Figure 6. Box and whiskers plot of RNA:DNA ratio of yellow perch (Percaflavescens) 
offspring from the ten maternal lines at the time of 50% mortality (n = 3). 
Box-whisker antennae as defined in Fig. 1. 
Figure 7. Relationship between mean time 10 50% mortality (n = 3) and mean larval total 
length (r2 = 0.0797), yolk volume (r2 = 0.491) and dry weight (r2 = 0.0354) at 
time of 50% mortality (n = 15) for each maternal line (n = 10) of yellow perch 
(Perea flavescens) . 
94 
:-~ • • • • • 
5 ~ • 
4 j • 
3 -\ 
• 
• 
2 1 
I 
1 l 
0 
5.4 5.5 
5.6 5.7 
5.8 5.9 
6 6.1 6.2 
6.3 
Mean Total Length at Hatch (mm) 
• 
-
'E. 7 -\ 
~ ~ 61 
i :1 
3 \ 
: 
2 ~ 
1 -\ 0 ..:...- -_...,..----,-- -......,.----.--- --,-- - --r-----c---- -
0.04 0.05 
0.06 0.07 0 .08 
0.02 0.03 
Mean Yolk Volume at Hatch (mm
3
) 
7 
6 
0 0.01 
• • 
• 
• 
• 
• 
1 ~ 
i 
0 j_, ___ --, _____ ,.----____ .,--_ ___ -,------ -
0 .2 0.25 0.15 
0.05 
0.1 
Mean ory Weight at Hatch (mg) 0 
95 
6.6 ·,-
- 6.4 -,-
E 
i62 $ + :E 
';fl 6 . 0 
II) 
.... 
0 
G) I 1 · - -, \0 ~ 5.a I. T I + 0\ 
-cu 
.c 
-0) 5; 5.6 . 
..J 
s 
~ 5.4 .. 
5.2 
1 2 ·------- .. - ·-- ·- -. . -----3 4 5 6 7 8 ·· -- ----9 10 
Maternal Line 
-- - .. . -...... 
O') 
0 
ci 
I • []-•--
• 
I • 
+ 
+ 
CO ~ CD LO V M C\I 
O O O O O O 0 
O O O O O O 0 
<cww) ~!feJ.'OW %OS JO 8W!.1 Je awnfOA )ffOA 
97 
...... 
0 0 
0 
0 
...... 
co 
CD 
~ 
cu 
LO E 
Cl) 
u. 
C\I 
...... 
0:, 
co 
--·-rn-·--i 
+ co 
+ 
1----·-DJ-·--
1--...... --CD ..... ·-----I 
N 
..-
co L{) v C'? N ..- N 0:, co 
"' N N N N N N 0 ..- ..- ..-
0 0 0 0 0 0 0 0 0 
(6w) ~!f~Oll\f %OS JO awu. :ie :J46!aM A.IQ 
98 
+ 0 ""t--
f- • ~ -/ 
+ co 
•[O • I " 
' 
+ 
i 
I 
I 
i CD (1) 
.E 
[]• -I -• C'IS 
lO C: ... 
(1) 
-C'IS 
I ,~ ~ • I "'-t" 
• 
I : l ] • I I C") 
• 
I 
. l I • I I N 
• : l • 
""t--
CD LO v M N ""t-- 0 
(6n) ~!f~OW %OS JO aWJl :ie :JU8:JUO:) 'tfN~ 
99 
~--DJ---· --1 
.___...---11 : I --1 ·------t 
+ 
i----e-e [J • 
• ~" '---•----1 
• • 
N co 
100 
N 
0 
i 
0 
co 
L!) 
~ 
N 
Cl) 
C: 
..J 
ns 
C: 
I-
Cl) 
-ns :=: 
7 
6 
5 
4 -
3 
] 
5.7 
~ 71 6 
~ 5 
~ :1 
E 2 ~ 
• 
• 
5.8 
• • 
• • 
• 
• 
• 
5.9 6 6.1 
6 .2 6 .3 6.4 
Mean Total Length at Time of 50% Mortality (mm) 
i= 1 . 
m ol------..,------,-----r----c----.----,-----
0.05 0 .06 0 .07 0 .08 
0.02 
0.04 0 .03 
~ 0 
7 1 
6 ~ 
I 
5 -1 
4 ~ 
i 
3 -1 
i 
2 J 
I 
1 -i 
0 J_ 
0.19 
0.01 
Mean Yolk Volume at Time of 50% Mortality (mm
3
) 
• • 
• • 
-
• 
• 
-
\ 
• 
• 
________,.---------,-------,-------,-----
0.21 0.22 
0.23 0.24 
0 .2 
Mean Dry Weight at Time of 50% Mortality (mg) 
101 
Chapter 4 
Understanding Maternal Effects as a Recruitment Mechanism in 
Lake Michigan Yell ow Perch (Perea flavescens) 
Ill. Implications for Recruitment 
102 
Abstract 
Maternal effects in Lake Michigan yellow perch <Perea fiavescens) have been 
detected, and the importance of maternal effects in resistance to starvation has been 
demonstrated. The goal of the present s:udy was to determine the importance of maternal 
effects when food was not limited. We ::xarnined the persistence of maternal effects 
between hatching and 32 days post hatch (dph) in ten maternal lines and tested the null 
hypothesis of no difference in survival clllong maternal lines. Maternal effects were 
consistently detectable between hatching and four dph and between 16 and 32 dph. 
Detected maternal effects resulted in threefold differences in overall survival at 32 dph, 
and rejection of the null hypothesis. Furthermore, a three to four fold difference in the 
rate of mortality, measured as time to 50% mortality, was observed among the maternal 
lines. However, no clear mechanism was identified to account for these differences in 
survival. We conclude that maternal effects were detectable and present at least to 32 
dph and resulted in differences in survival among maternal lines. 
103 
..., 
Introduction 
A major goal of fisheries science and management is to predict the level of 
recruitment of a cohort in a given year, and to do so as early in the life history of the 
species as possible. However, despite over three quarters of a century of research, 
recruitment variability remains the single least understood process in fisheries science 
(Houde 1987). Our lack of understanding is attributable to high inter-annual variability 
in recruitment of fish populations. Attempts to explain this high variability have invoked 
the influence of abiotic ( e.g. temperature, wind, transport mechanisms, and retention 
mechanisms) and biotic (e.g. starvation, cannibalism, and predation) controls. Recently, 
research has focused on identifying characteristics of recruits as a guide to determine 
sources of mortality that may control survival (Crowder et al. 1989; Miller 1997). 
Many important mortality sources controlling recruitment are size-dependent 
(Hunter 1981; Houde 1987; MiUer et al. 1988; Bailey and Houde 1989). Larger and 
older larvae are commonly viewed as being less vulnerable to predation than their smaller 
counterparts (Bailey and Batty I 984; Blaxter I 986; Houde 1987; Hunter I 981; Miller et 
al. 1988; Pepin 1993). YeIIow perch (percaflavescens) is one species in which this 
view, commonly referred to as the 'bigger is better' hypothesis, has received support. 
Nielson (1980) observed that larger yeIIow perch individuals exhibited lower mortality to 
walleye Cstizostedion vitreum) predation than smaller individuals. Starvation can also be 
size-dependent (Hunter I 98 I; Miller et al. I 988). Post and Evans (1989) demonstrated 
that overwinter starvation mortality in yeIIow perch was also strongly size-dependent. 
We have shown previously that newly hatched yellow perch larvae with large yolk-sacs 
104 
--
are more resistant to starvation than their larger conspecifics that have smaller yolks 
(Chapter 3). 
While differences in larval size at a given point in time may account for a 
considerable amount of mortality observed in fishes, recruitment is a dynamic process. 
Thus, differences in individual larval growth rates need also be considered. Differences 
in larval growth rates influence the duration of the larval stage and the time it takes larvae 
to pass through stages of high predation mortality (Houde 1987; Bertram et al. 1997). 
Faster growth rates can reduce the time that larvae spend in vulnerable size classes 
(Bailey and Houde 1989; Rice et al. 1993): Larvae exhibiting faster growth rates may 
experience a lower probability of mortality than slower growing larvae (Meekan and 
Fortier 1996). Furthermore, size-dependent predation mortality may select for survival of 
faster growing larvae to be favored and selected for over slower growing larvae 
(Rosenberg and Haugen 1982; Rice et al. 1987). 
Differences in the size and growth rates observed among individuals within a 
cohort can vary substantially (Rosenberg and Haugen 1982; Meekan and Fortier 1996). 
One potential explanation for the high intra-cohort variability observed in larval size and 
growth rates is maternal effects. Previous research has shown that intra-cohort 
differences in growth rates can in part be attributed to maternal effects. Gall (1974) 
demonstrated that older rainbow trout Oncorhynchus mykiss females produced offspring 
that exhibited faster growth rates than larvae produced by younger females. Larger and 
older female striped bass Morone saxatilis produced larvae that exhibited faster growth 
rates than larvae produced by smaller and younger females (Monteleone and Houde 
1990). Other studies have demonstrated similar findings for other species, including 
105 
-chum salmon Oncorhynchus keta (Beachc.m and Murray 1985), Atlantic silversides 
Menidia menidia (Bengston et al. 1987) and winter flounder Psuedopleuronectes 
americanus (Buckley et al. 1991 ), among others. 
In yellow perch, maternal effects were detected at hatching and were expressed 
most by differences in larval size and yok volume (Chapter 2). Furthermore, these 
maternal effects observed at hatching cortributed to significant differences in resistance 
to starvation (Chapter 3). However, the influence of maternal effects may be dampened 
during larval ontogeny as the influence individual experience becomes more important 
(Chambers and Leggett 1996). This raises the question of whether maternal effects can 
be of sufficient duration to influence mortality later in larval life. This study aimed to 
determine if the significant maternal effects observed at hatching in yellow perch were 
present throughout the first month of life. We tested whether maternal effects lead to 
significant differences in growth rate and survival of offspring under conditions of high 
food availability. 
Materials and Methods 
We investigated our objectives within a single experiment, which involved 
monitoring and quantifying several larval traits, and the survival of larval yellow perch to 
32 days post hatch (dph). All larvae used were the offspring of the ten maternal lines 
described in Chapter 2. Experimental work was conducted at the University of 
Wisconsin's WATER Institute, Milwaukee, WI. The experiment was conducted as a 
randomized complete block design, consisting of three blocks of 20 10-gallon flow-
through aquaria with two replicates per maternal line per block. 
106 
Full details of the collection and husbandry of eggs and larvae are given in 
Chapter 2, and are only summarized here. Male and female yellow perch were collected 
on 2 June and 4 June 1998 in the vicinity of Green Can Reef ( 42° 50' 00" N, 87° 50' 00" 
W) off Milwaukee, WI. Eggs were fertiliz~d in the field and brought to the lab. Volumes 
of eggs sufficient to produce approximately 2000 larvae were incubated in each of 60 10-
gallon flow-through, temperature-controlled aquaria. 
Husbandry of Larvae 
Hatching began on 14 June 1998 in all tanks. After the removal of the broken egg 
skeins and any dead eggs and larvae, the larvae in all 60 tanks were sampled for the 
presence of maternal effects at hatch. After sampling at hatch, one of two treatments, fed 
or starved, was randomly applied to one replicate tank of each maternal line within each 
block. The 30 fed treatment tanks were fed ab /ibitum throughout the duration of the 
experiment. The newly hatched yellow perch larvae were fed 'green tank water', a mixed 
batch culture of small-sized crustaceans, rotifers and protozoans. By two days post hatch 
( dph), larvae were introduced to Artemia 1auplii in combination with continued feeding 
of green tank water, and by five dph their diet consisted solely of Artemia nauplii. A 
mixture of ground beef and liver was add~d to the larval diet by eight dph. By 12 dph, a 
commercially available dry pellet fish food was added to the diet. The combined diet of 
Artemia nauplii, beef liver mixture, and commercial fish food was maintained throughout 
the duration of the experiment. The food type given, the flow rate and temperature in the 
tanks were consistent with the laboratory's successfully established yellow perch rearing 
conditions. Mortalities in each tank were removed daily and enumerated. 
107 
Larval Sampling 
At hatching, 15 larvae were sampled from each tank. Subsequently, five larvae 
were sampled every two days through 12 dph, and then every 4 days until 32 dph. 
However, in order to maintain some level of equal spacing of samples, only samples from 
0, 2, 4, 6, 8, 10, 16, 24 and 32 dph were processed and used in the subsequent analyses. 
We quantified 9 larval traits ( 6 morphometric and 3 biochemical measures of condition -
Table 1 ), for each larva from each tank on each sampling day. Larvae were first 
anesthetized with Tricaine Methansulfonate (MS-222), and then videotaped for analysis 
of morphometry. Individual larvae were placed into cryovials, flash-frozen in liquid 
nitrogen, and then stored at -80 °C for subsequent extraction of nucleic acids. 
Post-Sampling Processing of Larvae 
Morphometric analysis was conducted on videotaped images of larvae using 
Optimas (v 6.1, Media Cybernetics, Takoma Park, MD) as defined in Chapter 2. Larval 
total length, eye diameter, body depth at the insertion point of the pectoral fin, body depth 
at the insertion point of the first dorsal fin (inclusive of yolk-sac), body depth at the 
insertion point of the anus, and body area ( calculated internally by Optimas from the 
outline of the larva) were measured for each individual larva sampled. From these 
primary measurements we also estimated dry weight using methods described in Chapter 
2. We also determined yolk volume by methods described in Chapter 2. However, due 
to absorption of yolk very early on in the experiment it was omitted from these analyses. 
Extraction and determination of nucleic acids also was performed as described in Chapter 
2, with two exceptions. Microplates were read on a fluorescence plate reader at Ex= 590 
108 
run, Em = 625 run (Spectra Max Gemini, Molecular Devices Corp., Sunnyvale, CA). 
Furthermore, larvae > 7 rrun total length required an additional dilution of the extracted 
nucleic acid supernatant (Table 2). This additional dilution factor was used in the 
conversion of RNA and DNA concentralion to RNA and DNA content 
(concentration* 1.5 ml*additional dilution factor) . 
Statistics 
Separate statistical analyses were conducted to assess each of our objectives. All 
statistical analyses were conducted using SAS (v 6.12, SAS Institute, Inc., Cary, NC). 
The persistence of significant maternal effects throughout the duration of the experiment 
(32 dph) was determined by multivariate analysis of variance (MANOVA; SAS Proc 
GLM) on all larval traits, except survival, for each sampling day. The MANOV A was 
based on the mean value of each larval trait within each fed tank, for each maternal line 
nested within the date of fertilization (2 June or 4 June 1998). Significance of MANOV A 
was determined based upon a Wilks' Lambda statistic with varying degrees of freedom. 
Variation in survival over time and among maternal lines was evaluated by 
repeated measures analysis of variance (MANOV A; SAS Proc GLM). The proportion of 
larvae surviving on each day was arcsine transformed before analysis. The repeated 
measures analysis was based on the mean value of survival within each tank, for each 
maternal line on each sampling day. Repeated measures analysis was conducted on the 
sampling dates indicated above. Significance of the test of the null hypothesis of no 
difference in survival over time (i.e. time effect) was determined based upon a Wilks ' 
Lambda statistic with 4 and 15 degrees of freedom. Significance of repeated measures 
109 
analysis for the test of the null hypothesis of no difference in survival among the 
offspring of the ten maternal lines over time (i.e. interaction between mother and time) 
was determined based upon a Wilks' Lambda statistic with 32 and 59.91 degrees of 
freedom. This test of significance exan1ines whether the offspring of the maternal lines 
were dying at different rates. 
Differences among the fed offspring of the maternal lines over time were 
determined by repeated measures analysis on each larval trait at each sampling day. The 
repeated measures analysis was based on the mean value of each larval trait within each 
tank, for each maternal line on each sampling day. Significance of repeated measures 
analysis for the test of the null hypothesis of no variation in the larval traits (Table 1) 
over time was determined based upon a Wtlks ' Lambda statistic with 8 and 11 degrees of 
freedom for the morphometric and biochemical traits. For all traits the repeated measures 
analysis was conducted on the sampling dates indicated above. Significance of repeated 
measures analysis for the test of the null hypothesis of no difference among the offspring 
of each maternal line for each larval trait over time was determined based upon a Wilks ' 
Lambda statistic with 64 and 69.94 degrees of freedom for the morphometric and 
biochemical traits. These tests of significance examine whether the ten maternal lines 
have different rates of change for a given larval trait over time. Significance of repeated 
measures analysis was determined based upon an F statistic with 8 denominator degrees 
of freedom for the test of the null hypothesis of no difference among the offspring of each 
maternal line for a given larval trait (i.e. mother effect). This test of significance is 
analogous to differences in the absolute magnitude of a given larval trait among the 
offspring of the ten maternal lines. 
110 
Results 
Maternal effects were detectable on six of the nine sampling days. Maternal 
effects were detected at hatching and at two and four dph (Table 3). At six dph, we were 
no longer able to detect significant maternal effects among the offspring of the nine 
maternal lines based upon MANOV A results (F 72, 68.4 1 = 1.15, p = 0.2825) (Table 3). 
Tests for the presence of maternal effects were marginally non-significant on eight and 
ten dph. However, at 16 dph and through 32 dph significant maternal effects were once 
again detectable (Table 3). 
Maternal Line2 was the only maternal line that had no survival at the end of the 
experiment at 32 dph. Maternal Line2 experienced 100% mortality, as a result of loss due 
to sampling in addition to daily mortality, by 13 dph in two replicates, and by 24 dph in 
the third replicate. Repeated measures analysis requires a balanced data set. Therefore, 
we analyzed differences in the change of larval traits over 32 days for only nine maternal 
lines. The results of the analyses presented here do not include those for Maternal Line 
2' 
however, where appropriate, the results are presented in the tables. 
Mortality rates and overall survival differed among the offspring of the maternal 
lines. Time to 50% mortality ranged from 2 to 21 dph among the offspring of the 
maternal lines (Table 4). The offspring of Maternal Line3 exhibited the highest 
vulnerability to mortality, reaching 50% mean mortality by 4 ± 2.6 days, while offspring 
of Maternal Line exhibited the lowest vulnerability to mortality, reaching 50% mean 
7 
mortality by 15 + 1.5 days (Table 4). 
The lowest survival of 7 ± 1.2 % of the initial number of offspring surviving to 32 
dph was exhibited by Maternal Line6, while Maternal Line7 had the highest survival with 
111 
23 ± 3 % of the initial number of offspring surviving to 32 dph (Table 5). We were able 
to reject the null hypothesis of no overall difference in the survivorship of the offspring 
of the nine maternal lines over time based upon repeated measures ANOVA results (F32, 
59
_
91 
= 2.196, p = 0.0047) (Table 6). This indicates that the rate of mortality was 
significantly different among the offspring of the maternal lines. 
All larval traits varied significantly over time (Table 7). Furthermore, we were 
able to reject the null hypothesis of no overall significant difference in the DNA content 
of the offspring of the nine maternal lines over time based upon repeated measures 
ANOVA (F 
994 
= 1.508, p = 0.0470) (Table 7). This indicates that DNA content was 
64, 6 . 
significantly different among the offspring of the nine maternal lines over time. We also 
detected significant differences among the offspring of the nine maternal lines in body 
depth at the insertion of the pectoral fin, RNA content and the ratio of RNA:DNA, but 
not among the maternal lines over time (Table 7). No significant differences were 
detected among the offspring of the mat~rnal lines for larval total length, body depth at 
the insertion of the first dorsal fin, body depth at the insertion of the anus, or larval dry 
weight (Table 7). 
Changes in the traits of the offspring from the maternal lines observed between 
hatching and the end of the experiment were highly variable. Between hatching and 32 
dph, larval total length increased by as little as 25 ± 7 .1 % (Maternal Line3) and by as 
much as 38 ± 1.8 % (Maternal Line5) (Fig. 1, Table 8). Changes in body depth at the 
insertion of the pectoral fin among the maternal lines were less variable than those 
observed for larval total length. Body depth at the insertion of the pectoral fin varied 
over the duration of the experiment by as little as 15 ± 2.9 % (Maternal Line6) and by as 
112 
much as 19 ± 2.4 % (Maternal Line5) (Table 8). Changes in body depth at the insertion 
of the first dorsal fin and body depth at the irisertion of the anus were higher among the 
offspring of the maternal lines than those observed in body depth at the insertion of the 
pectoral fin. Body depth at the insertion point of the first dorsal fin varied by as little as 
14 ± 5.4 % (Maternal Line3) and by as much as 23 ± 3.3 % (Maternal Line10) while body 
depth at the insertion of the anus varied by as little as 16 ± 5.0 % (Maternal Line8) and by 
as much as 24 ± 4.7 % (Maternal Line4) among the offspring of the maternal lines (Table 
8). A four-fold variation in the change of dry weight among the nine maternal lines was 
observed over the duration of the experiment with Maternal Line3 exhibiting the smallest 
change (1 ± 1.3 %) and Maternal Line10 exhibiting the largest change (4 ±1.3 %) (Fig. 2, 
Table 8). Large variation among the mater:1al lines in the change of RNA content, DNA 
content and the ratio of RNA:DNA were observed over the duration of the experiment. 
The quantity of RNA varied by as little as 2 ± 1.5 % (Maternal Line3) and by as much as 
5 ± 1.2 % (Maternal Line7) among the maternal lines, while DNA content varied by as 
little as 5 + 1.1 % (Maternal Line _) and by as much as 11 + 4.8 % (Maternal Line~) 
- ) - ., 
(Table 8). The ratio of RNA:DNA exhibit~d over five-fold variation in percent change 
among the maternal lines over the duratior. of the experiment with Maternal Line8 
exhibiting the smallest change (19 ± 5.7 ¾) and Maternal Line 10 exhibiting the largest 
change (120 ± 72.8 %) (Fig. 3, Table 8). Correlation analysis of larval total length and 
dry weight over time indicates that size at hatching becomes less important to size at later 
points in time after eight dph in totai leng1h and 10 dph in dry weight (Table 9, 10). 
Differences in larval total length, yolk volume and dry weight among the maternal 
lines begin to increase beyond 16 dph (Figs. 1, 2 and 3). Daily growth rate as measured 
113 
in total length (mm/day) increased in all maternal lines between hatching and 16 dph 
(Fig. 4). Daily growth rates declined between 16 and 24 dph in all maternal lines except 
Maternal Line2 and Maternal Line3, after 24 dph daily growth rates again increased. 
Beyond 24 dph, differences in daily growth rate increase (Fig. 4). 
All maternal lines exhibited increased instantaneous growth rates (G) between 
hatching and 6 to 8 dph, followed by declines in G (Fig. 5). Beyond 16 dph G either 
increased or remained relatively stable in all maternal lines except Maternal Line 1. 
Maternal Line1 exhibited increased G to 24 dph, followed by decreased G. Instantaneous 
natural mortality rates (M) decreased between hatching and 4 dph, followed by increases 
at 6 and 8 dph in all maternal lines (Fig. 5). Beyond 10 dph M remained relatively 
constant for all maternal lines. 
Discussion 
Significant maternal effects were present among the ten maternal lines in six of 
the nine samples between hatching and 32 dph. We were able to detect maternal effects 
consistently between hatching and 4 dph and between 16 and 32 dph. Tests conducted on 
samples collected on 8 and 10 dph were only marginally non-significant, suggesting the 
potential for maternal effects at these times also. However, analysis of samples collected 
at 6 dph did not indicate significant maternal effects. Overall, we rejected the null 
hypothesis that maternal effects were not detectable up to 32 dph, and we conclude that 
maternal effects are likely widespread through early life history. Furthermore, we 
conclude that the maternal effects observed at hatching and to 32 dph in yellow perch 
larvae result in differences in survival. 
' 
114 
The loss of all offspring of Maternal Line2 by 13 dph in two tanks and by 24 dph 
in the third tank may be a direct result of the poor quality of the egg skein of Female2 
(Chapter 2). For whatever reasons, be they handling of the egg skein or perhaps 
premature expression of the eggs, the egg skein ofFemale2 did not maintain its matrix-
like gelatinous structure. Undoubtedly, the reduced quality of the egg skein is directly 
responsible for the low hatching success observed (Chapter 2) and the dramatically 
reduced survival of the subsequent offspring. However, even when larvae from this 
female were removed from the analysis, we were still able to detect significant maternal 
effects in survival, both expressed as the time to 50% mortality and in overall survival to 
32 dph. 
Why were maternal effects not detectable at 6 dph in the offspring of the ten 
maternal lines? An initial evaluation is that differences among the ten maternal lines in 
the slopes of the larval traits results in an intersection at some point between 6 and 10 dph 
and a divergence at 16, 24 and 32 dph. However, significant interactions between mother 
and time in the repeated measures analysis indicates that the slopes of the larval traits are 
only significantly different among the offspring of the ten maternal lines for DNA 
content. A more parsimonious explana,ion may be that the loss of detectable maternal 
effects is due to increased variability among larval traits within each maternal line 
associated with yolk absorption and the onset of feeding. This increase in variability may 
be due to differential increases in growth associated with both the onset of feeding and 
the first initial days of feeding. Such dtfferences in feeding ability would account for 
differences in growth within a maternal line, and would thus increase size variability that 
could potentially mask previously detected maternally influenced differences in size. 
115 
Subsequently, we might expect to see growth compensation by, or selective mortality on 
smaller individuals, which may then reduce variability within maternal lines (Rice et al. 
1993). As the variability within maternal lines is reduced, we would expect to again see 
significant differences among maternal lines as maximum size achieved within maternal 
lines may vary among maternal lines. Growth compensation has been demonstrated for 
other species used in maternal effects studies. Compensatory growth in rainbow trout 
was demonstrated to occur by 4 weeks post hatching (Springate and Bromage 1985) and 
after 25 dph in channel catfish Jctalurus punctatus (Reagan and Conley 1977). 
The correlation analysis for larval total length and dry weight supports increased 
within maternal line size variation as a potential cause of the disappearance and 
reappearance of detectable maternal effects. Size, as measured by total length and dry 
weight, at later sampling days (i.e. after IO dph) was not correlated with size at earlier, 
adjacent sampling days . Furthermore, correlation analysis between paired larval traits 
over time showed that correlations between morphometric measures strengthened as the 
offspring of the ten maternal lines aged. This suggests that reductions in the variation 
within larval traits occurred as the larvae approached 32 dph. 
The large changes observed in the larval traits throughout the duration of the 
experiment were expected due to the high abundance of food given to the larvae on a 
daily basis . However, the variation in those changes among the ten maternal lines 
suggests that mechanisms aside from food availability and quality may be influencing the 
growth of individuals. We note that while all maternal lines were given identical 
quantities and quality of food, DNA content was significantly different among the 
offspring of the maternal lines over time. Given that quantity of DNA per cell is 
116 
relatively constant within somatic tissues of a species (Clemmesen 1994), we would 
conclude that the offspring of the different maternal lines were developing, at least at the 
cellular level, at different rates. 
The abrupt changes in larval total length, dry weight, and RNA:DNA ratio 
observed beyond 16 dph are a little surprising. If growth rates vary among maternal 
lines, then differences in size and condition would be expected to be extenuated over 
time. The projections of each maternal line for each of these traits over time would be 
expected to gradually diverge from one another if growth rates differed. However, the 
projections of each maternal line for larval total length, dry weight and the ratio of 
RNA:DNA over time do not gradually diverge from one another, rather there are sudden 
divergences after 24 dph. This pattern carries over to the daily growth rate measured in 
total length where large divergences in daily growth rate are observed beyond 24 dph. It 
is not clear what may explain this sudden divergence in growth rate, size and condition 
among the maternal lines. One speculation would be that the observed divergences are 
attributable to the density dependent effects that are a consequence of mortality 
differences within each maternal line. 
The expected change in instantaneous growth rate would have been a slightly 
sigmoidal response over time (i.e. with age) while instantaneous mortality would have 
been expected to decline in a curvilinear manner (Houde 1997). Furthermore, the ratio of 
MIG would be expected to decrease through the point at which M = G (MIG = 1) as 
mortality decreases and growth increases, then eventually begin to increase again as 
mortality increases and growth decreases (Houde 1997). In the yellow perch maternal 
lines, instantaneous growth rate declined in all maternal lines until about 10 dph, beyond 
117 
which point it slowly began to increase or in some instances remain stable. It is not clear 
why this response was observed. Instantaneous mortality rate of the maternal lines did 
behave in the expected manner. The decline in the ratio of MIG observed among the 
maternal lines indicates that biomass is declining over the period from hatching to 32 
dph, and this is a combination of natural mortality and sampling mortality. 
In the laboratory, in the absence of predators and in the presence of ad libitum 
food, one would expect larvae to have a relatively equal probability of survival. yet, 
even under these conditions maternal effects on survival were detected. In the field 
' 
sources of selective mortality are prevalent (Miller 1997); thus, it is likely that maternal 
effects on larval survival may be even more important in the field. Some offspring may 
survive not because they are lucky (Pepin et al. 1999), but because their mother was more 
fit. While these results clearly support the significance of maternal effects on survival, 
the results regarding the significance of maternal effects on individual morphometric and 
biochemical traits do not clearly suggest a mechanism. Furthermore, no particular 
combination of the maternal characteristics that we measured best explains the 
differences observed in offspring survival. While the mechanisms for the differential 
survival observed among the offspring of the yellow perch maternal lines are not 
understood the end result is clear. Differences among mothers, most likely in traits that , 
were not measured in this study, account for differences in the survival of their offspring. 
Much future work on this topic is still needed, particularly research that can evaluate 
differences between mothers in more detail than we have here, perhaps even at the 
cellular level. 
118 
,, 
,, 
o! 
' 
.. 
,;, 
,,, 
.. ,. 
:, 
. , 
.'1 
.. 
. , 
• 
' 
Acknowledgements 
We would like to thank Fred Binkowski, Heather Brussat, Richard and Lori 
Fulford, Emily Kircher, Chris Ney, Jim Rice and Steve Yeo for their help in the 
husbandry and maintenance of the yellow perch larvae. We would also like to thank 
Mary Chenery for her help in the videotape imaging analysis, and Hind Eloukkal for her 
help with the enumeration of daily mortalities. We thank Vanessa Javor for her help in 
sample preparation for nucleic acid extractions. We would also like to thank Elaine 
Caldarone for her continued guidance and support in the fluorometric analysis of our 
nucleic acid samples. Funding for this research was provided by the University of 
Wisconsin Sea Grant College Program. 
119 
' • 
, . 
.. 
'. ,,,
' * 
·-;~ 
.'i 
.. 
, 
' . , 
References 
Bailey, K.M. and R.S. Batty. 1984. Laboratory study of predation by Aurelia aurita on 
larvae of cod, flounder, plaice and herring: development and vulnerability to 
capture. Mar. Biol. g3: 287-291 . 
Bailey, K.M. and E.D. Houde. 1989. Predation on egg and larvae of marine fishes and 
the recruitment problem. Adv. Mar. Biol. 25: 1-83. 
Beacham, T.D. and C.B. Murray. 1985. Effect of female size, egg size, and water 
temperature on developmental biobgy of chum salmon (Oncorhyncus keta) from 
the Nitinat River, British Columbia. Can. J. Fish. Aquat. Sci. 42: 1755-1765 . 
Bengston, D.A. , R.C. Barkman, and W.J . Berry. 1987. Relationships between maternal 
size, egg diameter, time of spawning season, temperature, and length at hatch of 
Atlantic silverside, J\,fenidia menidia· J. Fish. Biol. 31 : 697-704. 
Bertram, D.F., T.J. Miller and W.C. Leggett. 1997. Individual variation in growth and 
development during early life stages of winter flounder Pleuronectes americanus· 
Fish. Bull. 95: 1-10. 
Blaxter, J.H.S. 1986. Development of sense organs and behaviour of teleost larvae with 
special reference to feeding and predator avoidance. Trans. Am. Fish. Soc. 115: 
96-114. 
Blaxter, J.H.S. and G. Hempel. 1963. The influence of egg size on herring larvae 
(Clupea harengus L.). Journal du Conseil, Conseil International pour 
l' Exploration de la Mer. 2s: 211-240. 
120 
Buckley, L.J., A.S. Smigielski, T.A. Halavik,E.M. Caldarone, B.R. Burns, and G.C. 
Laurence. 1991. Winter flounder Pseudoplueronectes americanus reproductive 
success. II. Effects of spawning time and female size on size, composition and 
viability of eggs and larvae. Mar. Ecol. Prog. Ser. 74: 125-135. 
Chambers, R.C. and W.C. Leggett. 1996. Maternal influences on variation in egg sizes 
in temperate marine fishes. Amer. Zool. 36: 180-196. 
Clemmesen, C. 1994. The effect of food availability, age or size on RNA/DNA ratio of 
individually measured herring larvae: laboratory calibration. Mar. Biol. us: 
377-382. 
Crowder, L.B., J.A. Rice, T.J. Miller, and EA. Marschall. 1989. Empirical and 
theoretical approaches to size-based :nteractions and recruitment variability in 
fishes, p. 237-255. In: DeAngelis, D.L. and L.J. Gross (eds.). Individual-Based 
Models and Approaches in Ecology. Chapman and Hall. New York, New York. 
Gall, G.A.E. 1974. Influence of size of eggs and age of female on the hatchability and 
growth in rainbow trout. Calif. Fish. and Game. OO(lf 26-35. 
Houde, E.D. 1987. Fish early life dynamics and recruitment variability. Amer. Fish. 
Soc. Symp. z: 17-29. 
Houde, E.D. 1997. Patterns and consequences of selective processes in teleost early life 
histories. pp. 173-196. In: R.C. Ch:1II1bers and E.A. Trippel ( eds.) Early Life 
History and Recruitment in Fish Populations. Chapman and Hall, New York, 
New York. 
121 
Hunter, J.R. 1981. Feeding ecology and predation of marine fish larvae, p. 33-79. In. 
R. Lasker (ed.) Marine Fish Larvae: Morphology, Ecology, and Relation to 
Fisheries. University of Washington Press, Seattle, Washington. 
Meekan, M.G. and L. Fortier. 1996. Selection for fast growth during the larval life of 
Atlantic cod (Gadus morhua) on the Scotian Shelf. Mar. Ecol. Prog. Ser. 137: 
25-37. 
Miller, T.J. 1997. The use of field studies to investigate selective processes in fish early 
life history. In: R.C. Chambers and E.A. Trippel (eds.). Early Life History and 
Recruitment in Fish Populations. Chapman and Hall. New York, New York. 
Miller, T.J., L.B. Crowder, J.A. Rice, E.A. Marschall. 1988. Larval size and recruitment 
mechanisms in fishes: toward a conceptual framework. Can. J. Fish. Aquat. Sci. 
45: 1657-1670. 
Monteleone, D.M. and E.D. Houde. 1990. Influence of maternal size on survival and 
growth of striped bas Morone saxatilis Walbaum eggs and larvae. J. Exp. Mar. 
Biol. Ecol. 140: 1-11. 
Nielson, L.A. 1980. Effect of walleye Cstizostedion vitreum vitreum) predation on 
juvenile mortality and recruitment of yellow perch CPerca flavescens) in Oneida 
Lake, New York. Can. J. Fish. Aquat. Sci. 3T 11-19. 
Pepin, P. 1993. An appraisal of the size-dependent mortality hypothesis for larval fish: 
comparison of a multispecies study with an empirical review. Can. J. Fish. Aquat. 
Sci. so: 2166-2174. 
Pepin, P., G.T. Evans, and T.H. Shears. 1999. Patterns of RNA/DNA ratios in larval fish 
and their relationship to survival in the field. ICES J. Mar. Sci. 56: 697-706. 
122 
Post, J.R. and D.O. Evans. 1989. Size-dependent overwinter mortality ofyoung-of-the-
year yellow perch CPerca flavescens): laboratory, in situ enclosure, and field 
experiments. Can. J. Fish. Aquat. Sci. 46(11r 1958-1968. 
Reagan, R.E. Jr. and C.M. Conley. 1977. Effect of egg diameter on growth of channel 
catfish. Prog. Fish. Cult. 39: 133-134. 
Rice, J.A., L.B. Crowder, and F.P. Binkowski. 1987. Exploration of mechanisms 
regulating survival in Lake Michigan bloater: a recruitment analysis based on 
characteristics of individual larvae. Trans. Am. Fish. Soc. 116: 703-718. 
Rice, J.A., T.J. Miller, K.A. Rose, L.B. Crowder, E.A. Marschall, A.S. Trebitz, and D.L. 
DeAngelis. 1993. Growth rate variation and larval survival: inferences from an 
individual-based size-dependent predation model. Can. J. Fish. Aquat. Sci. so: 
133-142. 
Rosenberg, A.A. and A.S. Haugen. 1982. Individual growth and size-selective mortality 
oflarval turbot reared in enclosures. Mar. Biol. 72: 73-77. 
Springate, J.R.C. and N.R. Bromage. 1985. Effects of egg size on early growth and 
survival in rainbow trout (Salmo gairdneri Richardson). Aquaculture. 4r 163-
172. 
123 
....... 
N 
.i::,. 
Table 1. Larval traits measured in each of the yellow perch (Percajlavescens) offspring of the ten maternal lines. 
Linear measures 
Total Length (mm) 
Eye Diameter (mm) 
Body Depth at the Insertion Point of the Pectoral Fin (mm) (BDIP) 
Body Depth at the Insertion Point of the First Dorsal Fin (mm) (BDID) 
Body Depth at the Insertion Point of the Anus (mm) (BDIA) 
Biochemical measures 
RNA Content (~tg) (RNA) 
DNA Content (~tg) (DNA) 
Volume measures 
Yolk Volume (mm3) 
Derived measures 
Dry Weight (µg or mg) 
RNA:DNA Ratio (Ratio) 
-.....1!.',t\';:..ff ... ~-..., .. ~ ~-
Table 2. Dilution factors used in the fluorometric determination of RNA content and DNA content in the offspring of the ten yellow perch (Percaflavescens) 
maternal lines. 
Age (dph) Mean Total Length (mm) Dilution Factor Supernatant (µI) 0.1 % Sarcosyl (µI) 
0 5.9 0 1000 0 
2 6.1 0 1000 0 
4 7.8 2 500 500 
6 7.4 2 500 500 
8 8.2 4 250 750 
....... 
10 
Vl 10 8.9 4 250 750 
16 11.4 8 125 875 
24 12.7 12.5 80 920 
32 18.5 16 62.5 937.5 
Table 3. MANOV A results for tests of significance for all morphometric and biochemical larval traits among the offspring of the ten yellow perch (Perea 
flavescens) maternal lines. 
Day Of F Pr>F 
0 80,255.92 2.67 0.0001 
2 80, 65.60 1.69 0.0142 
4 72, 68.41 1.90 0.0040 
6 72, 68.41 1.15 0.2825 
8 72, 68.41 1.43 0.0678 
IO 72, 68.41 1.37 0.0979 
...... 
N 
16 72, 56.24 2.1 I 0.0021 
°' 24 72, 56.24 1.58 0.0372 
32 63, 51.16 1.79 0.0164 
,_. 
N 
--:i 
Table 4. Mean time to 50% mortality(± SD) (n = 3) of the offspring of the ten yellow perch (Percajlavescens) maternal lines over time, based upon the mean 
value for each tank per female on each sampling day. 
Maternal Line Minimum (days post hatch) Maximum ( days post hatch) Mean (days post hatch) 
1 4 13 9 ± 4.7 
2 12.5 1 23.52 16 ± 6.4 
3 2 7 4±2.6 
4 10 15 12 ± 2.6 
5 12 18 15 ± 3.1 
6 6 15 12 ± 4.9 
7 13 16 15 ± 1.5 
8 3 21 12 ± 9.0 
9 13 17 15 ± 2.1 
IO 14 16 15 ± 1 
1The offspring had not yet reached 50% mortality by 12 dph, however, by 13 dph 100% mortality had ensued. 
2The offspring had not yet reached 50% mortality by 23 dph, however, by 24 dph 100% mortality had ensued. 
Table 5. Mean survival(± SD) (n=3) to 32 dph in the yellow perch (Percajlavescens) offspring of the ten maternal lines sampled over nine days, based upon the 
mean value for each tank per female on each sampling day. 
Maternal Line Initial Number (0 dph) Surviving Number (32 dph) % Survival 
1 1566 ± 433 167 ± 58 11 ± 2.7 
2 84 ±23 0 0 
3 875±214 91 ± 30 10.3 ± 0.87 
4 1485 ± 112 168 ± 37 11 ± 2.3 
5 1133±156 159 ± 43 14 ± 3.0 
...... 
N 
00 6 1175±265 76 ±6 7 ± 1.2 
7 1468 ± 344 331 ± 81 23 ± 3 
8 896 ± 529 206 ± 185 22 ± 8.6 
9 1378 ± 208 201 ± 24 14.7 ± 0.56 
10 1283 ± 202 115 ± 26 9 ±3.7 
,_. 
N 
\0 
Table 6. Repeated measures ANOVA results for tests of significance among the yellow perch (Percajlavescens) offspring of the ten maternal lines (n = 263). 
Tests of significant differences among the offspring of the ten maternal lines were performed with 8 degrees of freedom (Mother). Tests of significant 
time effects were performed with 4 and IS degrees of freedom (Time). Tests of significant differences among the offspring of the ten maternal lines 
over time (Time*Mother) were performed with 32 and 59.91 degrees of freedom. 
Mother at Age (dph) 
0 2 4 6 8 10 16 24 32 Mother Time Time*Mother 
Survival ** * *** ** 
..-
I..,.) 
0 
Table 7. Repeated measures ANOVA results for tests of significance among the yellow perch (Percaflavescens) offspring of the ten maternal lines. Tests of 
significant differences among the offspring of the ten maternal lines were conducted with 8 degrees of freedom (Mother). Tests of significant time 
effects (Time) were conducted with 8 and 11 degrees of freedom for all morphometric and biochemical larval traits (n = 263). Tests of significant 
differences among the offspring of the ten maternal lines over time (Time*Mother) were conducted with 64 and 69.94 degrees of freedom for all 
morphometric and biochemical larval traits . All trait abbreviations arc as defined in Table I. 
Mother at Age (dph) 
0 2 4 6 8 10 16 24 32 Mother Time Time*Mother 
Larval 
Total * *** ** *** 
Length 
BDIP * ** ** * *** 
BDID ** ** * * *** 
BDIA * * *** 
Larval 
Dry *** ** * ** *** 
Weight 
RNA * * * * *** 
DNA *** * ** ** *** * 
Ratio * * *** 
Table 8. Mean percent change(± SD) (n = 3) in the larval traits of the yellow perch (Percajlavescens) offspring of the ten maternal lines sampled between 0 
and 32 dph", based upon the mean value for each tank per female on each sampling day. All trait abbreviations are as defined in Table I. 
Maternal Linc 
Larval 1 2" 3 4 5 6 7 8 9 10 Trait 
--
Total 35 ± 2.0 52 ± 15.0 25 ± 7.1 37 ± 5.9 38 ± 1.8 27 ± 5.0 35.0 ± 0.66 29 ± 8.5 36 ± 3.4 36 ± 3.6 Length 
BDIP 18.0 ± 0.60 34 ± 13.5 15 ± 3.2 19 ± 2.7 19 ± 2.4 15 ±2.9 18 ± 1.9 15 ±4.0 17.9 ± 0.29 20 ± 2.0 
BOID 16 ± 1.3 35 ± 14.3 14 ± 5.4 18 ±2.7 18 ± 3.4 17 ± 5.0 21±3.4 18 ± 5.7 20.0 ± 0.56 23 ± 3.3 
...... BDIA 21.5 ± 0.67 31 ± 9.9 17 ± 5.6 24 ± 4.7 22 ± 3.0 17 ± 3.1 20 ± 2.2 16 ± 5.0 20.9 ± 0.50 22 ± 3.0 vJ 
...... 
Dry Weight 3.2 ± 0.14 16± 11.6 I± 1.3 4 ± 1.8 3.9 ± 0.83 2 ± 1.5 3.4 ± 0.41 2 ± 1.8 3.9 ± 0.76 4 ± 1.3 
RNA 4.4 ± 0.28 IO± 6.5 2 ± 1.5 4.5 ± 0.93 4 ± 1.0 3 ± 1.8 5 ± 1.2 3 ± 1.5 4.4 ± 0.59 5± 1.4 
DNA 5.6 ± 0.88 18 ± 8.2 11 ± 4.8 8 ± 1.8 5 ± I. I 8 ± 2.0 5.8 ± 0.74 7 ± 1.2 5 ± 1.6 4.2 ± 0.84 
RNA:DNA 63 ± 50.4 23 ± 27.3 39 ± 28.3 69 ± 45.1 34 ± 32.3 76±13.4 19 ± 5.7 66 ± 46.2 120 ± Ratio 42 ± 7.9 72.8 
"Changes in the larval traits of Maternal Line2 occurred only over 5 and 7 sampling days due to the occurrence of I 00% mortality in all tanks by those days. 
Table 9. Correlations of larval total length between sampling days for the offspring of the ten maternal lines of yellow perch (Percaflavescens). Shaded 
correlation values represent significant correlations. 
0 dph 2 dph 4 dph 6 dph 8 dph 10 dph 16 dph 24 dph 32 clph 
0 dph 1 0.6485 0.7697 0.6242 0.7212 0.2970 0.2727 0.3697 -0.0333 
2 dph I 0.7212 . 0.6606 0.5152 0.3091 0.3576 0.3939 0.0667 
4 dph 1 0.6242 0.6970 0.3939 0.2364 0.2849 -0.1333 
6 dph I 0.8182 0.806) 0.8061 0.7939 0.4167 
8 dph 1 0.5515 0.5152 0.5152 -0.0333 
..... 
w 1 0.9394 0.6000 0.4833 N 10 dph 
16 dph I 0.6121 0.5833 
24 dph 1 0.4667 
32 dph 
Table 10. Correlations of larval dry weight between sampling days for the offspring of the ten maternal lines of yellow perch (Perea jlavescens). Shaded 
correlation values represent significant correlations. 
0 dph 2 dph 4 dph 6 dph 8 dph 10 dph 16 dph 24 dph 32 dph 
0 dph 1 0.8667 0.7333 0.6910 0.7455 0.6606 0.4182 0.4667 0.0667 
2 dph 1 0.6970 0.3818 0.5636 0.3939 0.2606 0.1152 -0.4667 
4 dph 1 0.6242 0.6849 0.6485 0.6000 0.2000 0.0667 
6 dph 1 0.8788 0.9273 0.7333 0.7697 0.6333 
8dph 1 0.9394 0.7455 0.5515 0.2667 
..... 
w 10 dph I 0.8788 0.7333 0.5500 vJ 
16 dph 1 0.5879 0.5500 
24 dph 1 0.6667 
32 dph 1 
Figure Legend 
Figure 1. Change in the total length of the fed offspring of the yellow perch (p 
erca 
flavescens) ten maternal lines between hatching and 32 days post hatch (dph). 
Values on each sampling day for each maternal line are based upon the mean 
value of total length from a total of 15 larvae ( 45 larvae on O dph) from 3 
replicate tanks. 
Figure 2. Change in the dry weight of the fed offspring of the yellow perch (p 
erca 
jlavescens) ten maternal lines between hatching and 32 days post hatch (dph). 
Values on each sampling day for each maternal line are based upon the mean 
value of dry weight from a total of 15 larvae ( 45 larvae on O dph) from 3 
replicate tanks. 
Figure 3. Change in the RNA-DNA ratio of the fed offspring of the yellow perch (p 
· erca 
flavescens) ten maternal lines between hatching and 32 days post hatch ( dph). 
Values on each sampling day for each maternal line are based upon the mean 
value of RNA:DNA ratio from a total of 15 larvae ( 45 larvae on O dph) from 3 
replicate tanks. 
Figure 4. Daily growth rate measured in total length (mm/day) of the offspring of the ten 
maternal lines of yellow perch CPercajlavescens) from hatching to 32 days post 
hatch. 
Figure 5. Instantaneous growth rate (G), instantaneous mortality rate (M) and the ratio of 
MIG for ten yellow perch CPercajlavescens) maternal lines from hatching to 32 
days post hatch. 
134 
Maternal Line 
1
-- -------- -- -- -- - - -- - ------- --------- -- ------- - -- - - - ------- - -- --------- ] 
--e- 1 ~ 2 -A-- 3 ~; 4 --)~ 5 --a- 6 -+-- 7 - ,\ - 8 · ,_, -9 · ->!(·- 10 
-- · --- ···--· -- ··- -·-- - - -- --- - - -- -··------- -- --· - - ·- --~-- - - --· --- - . 
30 -
25 -
-E 
E 20 -
-
_. 
.c 
l,.) ... 
v-, C> 
C 15 -
QI 
..J 
~ 
0 10 -
t- ~ -~~ ~-
5 ]!'~r:- ,.. ., ---.:. . 
0 . - -- - ----- - - -· ·1 - - -
0 
-- ·· 1 --
5 10 
- ---- -i- - - -- -- --- ·-,- - - ···1 
15 20 25 30 35 
Age (dph) 
,_.. 
I.,.) 
°' 
30 
+' 
.c 
25 · 
m 15 · 
1 
~ 10 · Q 
5 
Maternal Line 
\ 
- - · -· · - ···--· . . .... ... . - ·· ·· - - ·· ------·· ·- · · .. . .. . . . . . . .. --- ... --_ ----··--·- ···-· ·-- · ·------] 
--e- 1 ~ 2 -ti:-- 3 '' -4 - -)(--· 5 ---- 6 -+- 7 h .. 8 · 9 · Y 10 
--- ---- --- -- --· --· - - - · - ---- .... ... ------------ --- --------- -- ----- - ·------------ -------- -
A f ' 
0 ~ .. j - - ---- r:: -=::-:-. -r:------· r-----':-:----1.~~ I ·-·, ·· ·-· ---·1 ··- ---- - -- 1 
0 5 10 15 20 25 30 35 
Age (dph) 
-l..,l 
--.l 
18 
16 
14 -
o 12 · 
+l 
cu 
~ 10 -
< z 
C 8 -
.. 
< z 6 ~ 
4 -
Maternal Line 
I ---- -------- -- --- --- ·-... ---·--·---------- --··· . ----- -- . ---- ----· ··-··-] -Er- 1 ~ 2 -1:r-- 3 , ; 4 - -)(---· 5 -II- 6 -+- 7 --- ,\ ·· 8 - ., --9 ---?K··· 10 
. --· ··- -·- -- .... ------ -- - - -- -- ··· ·· ··- --- -----· - . -- -· ---- - - -· -- - ----- -- -·- ·-- - - ---
2 t'::~ 
C~ '~ 0 ""<!t·~~--w··•. -,~- / -----· .. -- T .. .. · _ . . <F•'" -~~·=c - 1 - - .. - ---1 . ·--- ,... ·-· ·· ··-· 
0 5 10 15 20 25 
Age (dph) 
30 35 
Q) 
C: 
::J 
ro 
C: 
.... 
2 
ro 
~ 
: o l I..- i 
I 
I ! 
i I 
' .. 
i I 
i~ 
I 
I [t 
I I 
! 00 
~ 
I CD . 
I I 
' 
: • I 
1 
LO I 
! 
0 
i x . 
. ' 
' I 
N i 
i • I 
• D 
I 
I 
; ...... I 
; ' 
: ' 
, ~ I 
' I I 
I ' I 
-
<l • 
·~' N\ 
- 0 ('t') 
138 
' '\ • I \ 
~\} : .• ,~ 
D ·,' ·· .. * Lf/ ~ !Pl 
:: 1:PI 
: ; :lli 
: /, .. ~ 1/. 
:Fl 
', .: 1 :' ,4:i' 
·.;: I : ,::: 
· ...· l/r.: 
• t , 11 /.)'f 
.... ~ 
: \ t :. f , ·, ', 
. ,. 11., , . 
.\ \\i'f. ·. '. 
: : · \ ; · ·. , 
:· . '.' 
. .,\ ., 
/( /~\ . ··. •. 
. : , \~: I . ·. : / i\ l ·: . 
!.. 0 N 
LC) 
' 
...... 
. :/I\ \ ..  
o ... • "-~~ - ~ 
·-~ 
/1 \ 
' \ \ ~~ \.':>:fl~ \ ~.,~"':; 
.i?\ 
~-. I ~~ 
0 C1J CD .:;f" N 
0 0 0 0 
LC) 
0 
-.i= 
C. 
"C 
-Cl) 
C) 
.c,:: 
-':-
>. 
ns 
I 
0.3 
~ 0.25 
~ 
ns 
n:: 
.c: 0.2 
j 
~ 0.15 
U) 
::, g 0.1 
C: 
.5 
c: 0.05 
.5 
U) 
.!: 0 
1 0.35 1 
- 03 ~ . 
ns 
n:: 0.25 
~ 
~ 0.2 
0 
~ 0.15 
U) 
:, g 0.1 
C: 
.5 
c: 0.05 
ns 
-
.E O j 
-- -~-- 1 
• 6 
0 5 
····0 ··· -2 
_ __ 7 
Maternal Line 
···· -&· ··- 3 
~•-·a 
,...., 
, ,c,:J • .. 
·· ··X·· ·-4 
-+- 9 ~~o I 
·~~~ ... ~ - T • 
15 20 25 30 35 
10 
Age (dph) 
139 
Chapter 5 
Synopsis 
140 
Maternal effects likely exist to some degree in all species due to the female's 
energetic investment in the production of eggs. It is the female that provides the initial 
' survival kit' that a larva has to survive on until it can locate and begin feeding on 
exogenous food sources. This initial 'survival kit' includes not only the quantity of yolk 
in the yolk-sac, but also the quality of that yolk. Differences among individuals at 
hatching would be expected to result in differences in survival under sub-optimal 
environmental conditions. Beyond hatching, maternally derived differences in size and 
growth may remain important among larvae. If differences in hatchling traits are directly 
attributable to maternal contributions, the question remains, "For how long are such 
maternal effects important in early life history?" 
Maternal effects were detected in Lake Michigan yellow perch (percaflavescens) 
among the offspring of ten maternal lines at hatching. The differences among the 
offspring resulted from differences among their mothers with respect to size, age, gonadal 
somatic index (GSI), and egg production (Chapter 2). These maternal effects were 
expressed most by differences among the hatchlings of the maternal lines in larval total 
length, yolk volume, dry weight and the quantity of DNA. Furthermore, trade-offs in 
both the maternal and hatchling traits were observed. A trade-off between female GSI 
and all other female traits indicated that old, large females were characterized by high 
egg production and low GSI, while young, small females were characterized by low egg 
production and high GSI. In the hatchlings, a trade-off between larval total length and 
yolk-volume indicated that long larvae were characterized by small yolk-sacs, and short 
larvae were characterized by large yolk-sacs. The differences in larval traits of the 
hatchlings and the observed trade-offs suggest that the contribution of different females 
141 
to Year class strength of yellow perch in Lake Michigan will vary with changes in either 
the biotic b. . 
or a 1ot1c environment. 
For example, maternal effects were found to persist through six dph under 
conditions f fc . . . 0 
ood-hmitat10n (Chapter 3). The maternal effects resulted in differences 
among mat · 
ernal lmes in their resistance to starvation by a factor of two. The offspring 
that exh·b· 1 Ited the lowest resistance to starvation were characterized by high total length 
anct small Yolk-sacs at hatching. The offspring that were most resistant to starvation were 
charact · 
enzed by low total length and large yolk-sacs at hatching. This suggests that it is 
not the . . . 
size of the larvae that determines starvat10n resistance, but rather the amount of 
Yolk the larvae were provided with by their respective mothers. The characteristics of the 
least and . . . 
most starvat10n resistant larvae provide further evidence for a trade-off between 
larval total length and yolk-volume, and insights into the importance of such a trade-off 
Intenn 
s of starvation resistance. The observed differences in resistance to starvation 
among the maternal lines suggest that maternal effects do influence survival under 
conct· . 
ltions of food-limitation. 
Moreover, maternal effects were found to persist through 32 dph when food was 
not a r · fc 1 . d"ffc imiting factor (Chapter 4). The maternal efi ects resu t m i erences among 
Il1aterna1 lines in overall survival by a factor of three. Three to four fold differences 
among the maternal lines in the rate of mortality, as measured by time to 50% mortality, 
'Were also attributed to maternal effects. However, neither differences in the maternal 
characteristics nor the larval characteristics account for the observed differences in 
survival among the maternal lines. While differences in survival are attributable to 
Illaternal effects, the mechanisms responsible for those differences are not clear, and may 
142 
be due to differences in maternal and larval characteristics not measured in this study, 
such as feeding ability, sensory ability or genetics. 
Clearly, maternal effects in yellow perch result in morphometric differences 
among hatchlings. Furthermore, those differences in morphometry result in distinct 
differences in resistance to starvation. And, even though the mechanisms are not clear, 
maternal effects also result in differences in survival among maternal lines when food is 
not limited. While these maternal effects would be expected to have substantial 
implications on recruitment under environmental uncertainty, they may only be important 
when high fishing pressure has reduced the variation in maternal characteristics in the 
spawning population (through selective harvest) at low or declining population levels. 
The larval traits that confer a survival advantage likely vary from year to year according 
to the biotic and abiotic conditions in the lake. When the distribution of spawning stock 
traits is wide, a wide distribution of larval traits would be produced within a cohort. Thus 
it is probable that a portion of the larval cohort possess the appropriate characteristics. 
Thus, the importance of maternal effects would be small. However, when high levels of 
selective fishing remove a portion of the spawning stock and alter the distribution of 
spawning stock traits, a narrow or even skewed distribution of larval traits would be 
produced within a cohort. Under this scenario it is less likely that larvae will possess the 
appropriate characteristics to ensure strong recruitment. Under environmental 
uncertainty, the distribution oflarval traits artificially selected for through high fishing 
pressure may not be sufficient for adequate survival and successful recruitment. 
The changes that have occurred in the Lake Michigan yellow perch population 
indicate that there are fewer yellow perch than in past decades, and moreover, lower 
143 
proportions of females than in the past. Females that are mostly greater than four years 
of age, and that are historically large for their age characterize the female population that 
does exist. These changes in the female portion of the population suggest that the 
importance of maternal effects may have increased during the population decline in the 
early 1990s, and may be most important now as the population attempts to recover. The 
collapse of the Lake Michigan yellow perch fishery that occurred in the last decade of the 
20th century is likely not directly attributable to changes in the characteristics of the 
female spawning stock nor maternal effects. In fact, the initial cause or causes of the 
population decline may not ever be fully understood. However, the role maternal effects 
play in the rebuilding of the population may be very important. 
The results of this study suggest that the large females currently dominating the 
spawning population are likely producing a substantial portion of offspring that would be 
characterized as relatively long and heavy with small yolk-sacs. If food were limited, 
these larvae would be expected to exhibit a lowered resistance to starvation than their 
smaller conspecifics with large yolk-sacs. In the past decade, Lake Michigan has 
undergone dramatic changes as a result of the introduction and rapid population booms of 
exotic species, particularly the zebra mussel Cnreissena polymorpha). The introduction 
of this species has potentially resulted in a degradation of not only the food quantity, but 
also quality, that yellow perch larvae depend on as first feeding begins. 
If food were not limited, than the larvae being produced by the current spawning 
population would be expected to have a lowered vulnerability to predation than their 
smaller conspecifics. The opposed potential outcomes of being long with a small yolk 
versus being short with a large yolk, coupled with the environmental uncertainty that 
144 
exists in nature offer an indication of the difficulty the yellow perch population will face 
as it continues to recover. Several consecutive years of environmental conditions that are 
unfavorable to the types of offspring produced each year could result in high levels of 
embryonic and larval mortality that would substantially inhibit recovery. The Lake 
Michigan yellow perch population will remain at the mercy of the environment as it 
continues to rebuild itself, until it reaches a point where a wide distribution of traits are 
once again being produced each year. 
As the population recovers, years when the environmental conditions favor 
recruitment of several phenotypes will result in the replacement of "missing" young 
yellow perch with new recruits that exhibit a wide distribution of traits. If enough 
favorable years occur consecutively, the age and size distribution of the population will 
begin to widen. Eventually the spawning population will include relatively high numbers 
of young, small spawning females in addition to the high numbers of old, large spawning 
females that are currently present. The subsequent diversity in the age and size structure 
of the population will result in the production of diverse year classes. The production of 
offspring that display a wide distribution of traits ensures that in the face of 
environmental uncertainty and size-selective mortality, some larvae will survive and 
recruit, and the population will endure. 
High exploitation of a species in a size-selective manner will, over time, force 
shifts and truncation on the distribution of traits within that population. If we are 
constantly removing individuals from a population that exhibit some set of desired traits, 
we can naturally expect to see fewer and fewer individuals that exhibit those desired traits 
within the population. This presents a substantial risk if maternal effects within that 
145 
species have effects on the survival of offspring. Dramatic shifts or truncations in 
spawning population trait distributions may precipitate severe population declines and 
potential recruitment failures. The best protection against the inherent environmental 
uncertainty that exists in nature, and to ensure an adequate level of recruitment each year, 
is to ensure a wide distribution of characteristics within the spawning population. This 
will result in the production of offspring that exhibit a wide distribution of traits and will 
ensure some level of larvae survive, no matter what nature throws at them. 
The results of this research identify several issues that need to be addressed by the 
managers of Lake Michigan yellow perch in the near future. The commercial fishery was 
completely shut down in 1997. There is considerable public pressure for the fishery to be 
reopened. Several key policies should be implemented before reopening occurs. 
Previously, the entire month of June was closed to harvest to protect the perch during 
spawning. However, in recent years the spawning season has actually begun in May as a 
result of earlier warming of the lake each year. Therefore, managers may better protect 
the spawning population through extending this closure to include the month of May. 
Furthermore, protection of female yellow perch through the rest of the year may best be 
accomplished through the use of formal, quantitative stock-assessments to determine the 
sustainable amount of yellow perch that can be taken each year. Before the closure of the 
Lake Michigan yellow perch fishery, detailed stock-assessments had not been used by the 
management agencies, and undoubtedly, the lack of assessment lead to excessive harvest, 
an overall reduction in the population level and a truncation of the spawning population's 
trait distribution. If managers chose to explore stock assessments as a management 
146 
option, they will want to ensure that a substantial monitoring effort is put in place to track 
the trait distribution of the spawning population throughout the year. 
Future Directions 
Phenotypes are the expression of genotypes and the interaction between those 
genotypes and the environment. This definition of phenotype strongly suggests that 
research on maternal effects need focus some effort on genetic components. Secondly, 
many teleost species are known to exhibit a high phenotypic plasticity. This allows 
individuals to quickly adapt to change. For example, density-dependent factors often 
bring about rapid changes in growth rate and age at maturity. It is conceivable that 
maternal effects may also influence an individual's scope for plasticity, which could have 
implications on survival and reproductive success. Bearing these two points in mind, 
future maternal effects research should begin to explore such areas. Are there maternal 
influences on resistance to starvation and survival that are not detectable at the 
phenotypic level? Are there maternal influences on phenotypic plasticity that result in 
differential adaptability and survival under changing conditions? These deeper insights 
will require research that employs genetic, physiological, biochemical, and ecological 
approaches simultaneously to determine fully the role of maternal effects on the 
dynamics of fish populations. 
147 
Appendix I 
Relationship Between Larval Yellow Perch (Percaflavescens) 
Two-dimensional Body Area and Dry Weight 
148 
Overview 
In addition to the ratio of RNA:DNA, the dry weight of a larva is a measure that yields 
insight into that individual ' s overall health and condition. However, in order to measure 
the dry weight of a larva and then extract nucleic acids from that larva for the 
determination of the ratio of RNA:DNA, one must freeze-dry the larva over night. If the 
drying process were performed in a drying oven, the high heat and duration of the drying 
process would increase enzyme activity substantially, resulting in the breakdown of RNA 
and DNA. Unfortunately, we were not privy to a freeze-dryer at the time of 
experimentation, therefore we opted to develop a relationship between the two-
dimensional body area of a larva and the dry weight of that larvae. 
Materials and Methods 
Larval yellow perch CPerca jl.avescens) were sampled from a mass-rearing tank for each 
sampling age. Eighty-nine, one day post hatch (dph) yellow perch larvae were sampled 
from a mass rearing tank on 15 June 1998, and Eighty-one, 23 dph yellow perch larvae 
were sampled from a mass rearing tank on 7 July 1998. On each sampling day, the larvae 
were sampled five at a time to prevent mortality during sample processing. The larvae 
were first anesthetized with Tricaine Methansulfonate (MS-222), and then videotaped for 
analysis of morphometry. The larvae were then placed into individual pre-weighed 
aluminum foil envelopes and dried in a drying oven at 60° C to a constant weight 
(reached after about 60 hours). After drying, the individual larvae in the pre-weighed 
envelopes were re-weighed. The dry weight of each larva was determined as the 
149 
difference between the final weight of the larva in the aluminum envelope after drying 
minus the original weight of the aluminum envelope. 
Imaging analysis on the videotaped larval images was performed using Optimas imaging 
analysis system (v 6.1, Media Cybernetics, Takoma Park, MD) at the Chesapeake 
Biological Laboratory, Solomons, MD. The two-dimensional body area of each larva 
( calculated internally by Optimas from the outline of each larva) was measured. 
Results 
The relationship between two-dimensional body area and dry weight was determined to 
be curvilinear and can be described by the equation (Dry weight (µg) = [(2 x 10-6) x 
(Body Area)2 + (6 x 10-5) x (Body Area)+ 3 x 10-5] x 1000, n = 171, r2 = 0.9836, p = 
0.0001). 
Discussion 
The calibration between larval two-dimensional body area and dry weight accurately 
allows the researcher to determine dry weight from videotaped images of larvae. 
Furthermore, this technique allows the researcher to obtain accurate dry weights of larvae 
without the increased risk of RNA and DNA digestion that occurs when larvae are 
handled for long periods of time after death, before nucleic acid extraction. This 
technique also reduces the amount of processing time required for each individual larva. 
The time required to analyze videotape images for a group of larvae is far less than the 
time required to obtain their dry weights. 
150 
Table I. Two-dimensional body area and dry weights of individual yellow perch (Percajlavescens) 
larvae sampled at.I day post hatch (dph) and 23 dph. Consists of 4 pages. 
Age (dph) Sample Number Body Area (mm2) Dry Weight (g) 
1 2.06 0.0002 
2 2.16 0.00013 
3 2.21 0.00009 
4 2.22 0.00013 
5 2.22 0.00013 
6 2.22 0.00034 
7 2.28 0.00006 
8 2.29 0.00015 
9 2.34 0.00013 
10 2.40 0.00017 
11 2.40 0.00024 
12 2.42 0.00017 
13 2.45 0.00033 
14 2.45 0.00022 
15 2.48 0.00025 
16 2.49 0.00018 
17 2.51 0.00006 
18 2.52 0.0003 
19 2.53 0.00025 
20 2.54 0.00023 
21 2.56 0.00012 
22 2.56 0.00019 
23 2.61 0.00017 
24 2.61 0.00024 
25 2.62 0.00125 
26 2.62 0.0002 
27 2.63 -0.00013 
28 2.63 0.00018 
29 2.63 0.00018 
30 2.64 0.00005 
31 2.65 0.00017 
32 2.66 0.0003 
33 2.74 0.0002 
34 2.75 0.00035 
35 2.75 0.00038 
36 2.76 0.0004 
37 2.76 0.00006 
38 2.78 0.00031 
39 2.78 0.00038 
40 2.79 0.00031 
41 2.81 0.0003 
42 2.82 0.00029 
43 2.82 0.0003 
44 2.83 0.00025 
45 2.84 0.00016 
46 2.84 0.00013 
47 2.84 0.00001 
48 2.85 0.00036 
49 2.87 0.0003 
1 50 2.89 0.00013 
1 51 2.89 0.00016 
151 
Age (dph) Sample Number Body Area (mm2) Dry Weight (g) 
52 2.89 0.00024 
53 2.90 0.00023 
54 2.90 0.00002 
55 2.91 0.00012 
56 2.91 0.00026 
57 2.93 0.00025 
58 2.95 0.00007 
59 2.96 0.0003 
60 2.96 0.00031 
61 2.97 0.0003 
62 3.00 0.00034 
63 3.03 0.00029 
64 3.03 0.00023 
65 3.03 0.00016 
66 3.04 0.00026 
67 3.07 0.00006 
68 3.07 0.00022 
69 3.07 0.00026 
70 3.07 0.00025 
71 3.07 0.00028 
72 3.11 0.00026 
73 3.12 0.00025 
74 3.18 0.00022 
75 3.21 0.00028 
76 3.21 0.00032 
77 3.22 0.0003 
78 3.30 0.00031 
79 3.30 0.00031 
80 3.31 0.0003 
81 3.32 0.00036 
82 3.33 0.00039 
1 83 3.39 0.00022 
1 84 3.48 0.00025 
85 3.55 0.00034 
86 3.58 0.00032 
87 3.61 0.00026 
88 3.62 0.00025 
89 3.66 0.00067 
90 3.66 0.00133 
91 3.97 0.00034 
23 92 14.66 0.00113 
23 93 16.24 0.00133 
23 94 16.53 0.00126 
23 95 18.04 0.0021 
23 96 18.96 0.00185 
23 97 20.12 0.00202 
23 98 20.36 0.002 
23 99 20.80 0.00229 
23 100 20.80 0.0032 
23 101 21 .15 0.00223 
23 102 21 .30 0.00213 
23 103 21.60 0.00221 
23 104 21 .65 0.00214 
23 105 21.82 0.0023 
23 106 21.96 0.00237 
152 
Age (dph) Sample Number Body Area (mm2) Dry Weight (g) 
23 107 21 .98 0.0022 
23 108 22.13 0.00252 
23 109 22.21 0.00217 
23 110 22.39 0.00243 
23 111 22.69 0.00228 
23 112 22.80 0.00266 
23 113 23.18 0.00221 
23 114 23.26 0.00246 
23 115 23.53 0.00243 
23 116 24.02 0.00322 
23 117 24.39 0.00291 
23 118 24.49 0.00267 
23 119 24.54 0.00245 
23 120 24.84 0.00283 
23 121 24.86 0.00241 
23 122 25.27 0.00298 
23 123 25.39 0.00283 
23 124 25.76 0.00268 
23 125 25.80 0.0031 
23 126 25.88 0.0027 
23 127 25.91 0.00306 
23 128 26.03 0.00274 
23 129 26.13 0.00302 
23 130 26.37 0.00303 
23 131 26.42 0.003 
23 132 26.50 0.00288 
23 133 26.68 0.00299 
23 134 26.72 0.00287 
23 135 26.94 0.00362 
23 136 27.86 0.00332 
23 137 28.01 0.003 
23 138 28.17 0.00323 
23 139 28.62 0.00369 
23 140 28.64 0.00345 
23 141 28.96 0.00361 
23 142 28.97 0.00355 
23 143 29.20 0.00352 
23 144 29.41 0.00318 
23 145 29 .78 0.00352 
23 146 29.79 0.00333 
23 147 29.87 0.00352 
23 148 29.99 0.00335 
23 149 30.32 0.00367 
23 150 30.47 0.00391 
23 151 32.34 0.00392 
23 152 32.35 0.00387 
23 153 32.52 0.00407 
23 154 32.54 0.00349 
23 155 32.76 0.00424 
23 156 33.18 0.00503 
23 157 33.87 0.00489 
23 158 34.03 0.00539 
23 159 35.89 0.0044 
23 160 36.29 0.0047 
23 161 36.77 0.00498 
153 
Age (dph) Sample Number Body Area (mm2) Dry Weight (g) 
23 162 37.61 0.00514 
23 163 38.06 0.00497 
23 164 38.57 0.00609 
23 165 40.75 0.00626 
23 166 42.02 0.00547 
23 167 44.03 0.00638 
23 168 45.12 o.oo6n 
23 169 45.49 0.00665 
23 170 48.89 0.0078 
23 171 58.96 0.01019 
154 
Figure Legend 
Figure 1. Nonlinear relationship between larval yellow perch CPercaflavescens) two-
dimensional body area and dry weight for 1 day post hatch (dph) and 23 dph 
larvae. 
155 
0.012 
0.01 
0.008 
-tn 
-.... 
.c. 
tn 
'a, 0.006 
~ 
~ 
a 
0 
y = 2E-06x2 + 6E-05x + 6E-05 
R2 = 0.9836 
- -- 1 - ·-· I - - -·-1 - · -
10 20 30 
1 -- - ·-·- - - --- -, - -·-· -· -- · -- -,----- ---- ---·- ·· 1 
40 50 60 70 
Body Area (mm2) 
Appendix II 
Determination of the Effect of Post Death Condition, Preservation, and Age 
on the Shrinkage of Larval Yellow Perch (Perea jlavescens) 
Emily A. Kircher and Christopher J. Heyer 
--
- ---
The following paper represents work that was conducted in the summer of 1998 
by Emil K· . . . 
Y 1rcher, a student from Marquette Umvers1ty working at the WATER Institute 
~~~R . . 
esearch Experience for Undergraduates program. Emily conducted this 
research . 
under my direct supervision. This research will serve as a future reference for 
total length J d · 1· "d · 
corrections of field-captured yellow perch arvae preserve m 1qm mtrogen 
for subsequent 1 . "d . d 1 . nuc e1c ac1 extract10ns an ana ys1s. 
Abstract 
Many studies have assessed the performance of larval fish as a function of size. 
However, it has been demonstrated that larval fish shrink upon death and preservation. If 
this shrinkage is not considered it is possible to overestimate or underestimate the 
perfonnance and condition of larvae. The amount of shrinkage due to the effects of post 
deatb condition, preservation, and age was determined in yellow perch larvae (perca 
flavescens). The difference in len21h was first taken from live yellow perch larvae and 
0 
again after a post death condition. Computer image analysis tools were used to determine 
the lengths of the larvae. Results show that shrinkage does occur due to both death and 
preservation. Also it was determined that the main effects of post death condition and , 
age were significant, but there was no detectable interaction between death and age. An 
estimate of shrinkage in larval yellow perch is 14% on preservation in liquid nitrogen and 
5% on death. B d. the performance of larval fish as a function of 
ecause many stu 1es assess 
length · . . f the fish Therefore, it is necessary 
.. ' It is imperative to have an accurate measure O • 
to take th . . .d t· when studying field-collected and lab-
ese shrmkage amounts mto cons1 era 100 
reared 1 . te the perfonnance or condition of arvae so as to not overestimate or undereStima 
the larvae. 
159 
Introduction 
The survival of fish larvae is dependent upon size. The bigger the larvae, the less 
likely it is to be eaten by a predator, swept away in a current that will take it out of its 
optimal habitat, or starve to death (Miller et. al. 1988). Many researchers have assessed 
performance and condition of larval fish as a function of length. Performance measures 
such as burst swimming speed (Webb and Corolla 1981) and prey selectivity (Mills et. al. 
1989) have been measured in laboratory experiments relative to length. The condition of 
wild larvae has been assessed in field studies by using RNA:DNA ratios also as a 
function of length (Suthers et. al. 1996; Rooker et al., 1997). 
Because the size of larvae is linked so closely to its performance, it is imperative 
to have an accurate measure of the larval length. Inferences about the performance of 
larvae in the field are often made from lab experiments on live fish. Often, the lengths of 
fish in the lab experiments are measured from live fish while the lengths of fish collected 
in the field are measured from preserved fish. Ichthyoplankton sampling requires towing 
a net through the water to collect larvae, typically these tows can last anywhere from 1-1 O 
minutes or longer depending on the type of research. During a tow, larvae will often die 
due to the stress on their fragile bodies, and subsequently, their body functions will shut 
doWn. Upon death, larvae lose the ability to osmoregulate and they experience a loss of 
Water from their body and tissues, leading to shrinkage. After the duration of the net tow 
the larvae ar b h b d the research vessel and preserved. With preservation the 
e roug t on oar 
larvae shrink . h' . due to the chemical fixation process of the preservative that 
agam, t 1s time 
draws the Water out of their bodies. In the lab and the field, the shrinkage due to 
160 
preservation is accounted for, however, for the fish collected in the field, the shrinkage 
due to death before preservation is neglected. 
It has been demonstrated that fish larvae shrink upon preservation (Theilacker and 
Porter 1995). This amount of shrinkage is often accounted for by adding the amount the 
larvae shrink from death to preservation, typically ten percent (Theilacker and Porter 
1995), however, this may not be adequate. Evidence from larvae of marine fish species 
suggests that fish may shrink up to twenty percent on death prior to any additional 
shrinkage due to preservation (McGurk 1985; Jennings 1991; Theilacker and Porter 
1995). Not correcting for the full amount of shrinkage of fish larvae due to death and 
preservation could result in a substantial underestimation or overestimation of the 
relationship between fish length and measures of performance of larvae in the field. 
The objectives of this study were to quantify shrinkage in different ages of yellow 
perch (per fl ) larvae under three post death conditions in an attempt to 
ca avescens 
demonstrate that accounting merely for shrinkage due to preservation is not adequate, and 
that the shrinkage due to death must also be considered. 
Methods 
Yellow perch larvae from multiple brood stocks, maintained using standard 
hatchery methods, were used to evaluate the effects of post death condition, age, and 
preservation on shrinkage. The potential interaction between post death condition and 
age Was examined. Eight ages of larvae (1 day post hatch (dph), 6 dph, 11 dph, 16 dph, 
21 dph, 27 dph, 3 1 dph, and 36 dph) were treated with three post death conditions. The 
post death d. . . d of placing the larvae in one of three conditions: sitting in 
con 1t1ons cons1ste 
161 
0Pen · air in a T . . ncame methansulfonate (MS-222) solution, sitting on ice in a MS-222 
solutio n, or placed · 1· · the m iqmd nitrogen. The liquid nitrogen treatment combined death and 
post death co . . . nd
il!on. Larvae were collected from the larval stock tank, anesthetized 
With 1 mL of a 24 2 . . . . T · 5 
mg/mL MS-222 solut10n, and videotaped for later ,magmg analysis. 
o ens ure death 1 . d ' arvae m the air and ice death treatments were treated with a lethal 
osage ofMS-2 . . 2
2, and then left in the solution of MS-222 to prevent total drymg and to 
niunic th 
h 
e coo
d
itions of field collection. The post death condition was applied for one 
alfh our and th e larvae were videotaped a second time. 
The vid · b d'ffi eotape was used to determine the amount of shrmkage Y 1 erence. A 
SNAPPYTM fr . rune grabber was used to capture the unages from the videotape. 
OT!IscsA . Image Tool (developed at the University ofTexas Health Science Center at 
San Antonio T fr ' exas and available from the Internet by anonymous FTP orn 
lllax rad6 uth . scsa.edu) was used to measure the total length of each 1arva. The shrinkage 
¾iountsd . etermined from the larvae in open air and on ice demonstrated the shrmkage 
llponct 
eath h 1· 'd ·t h d th ' w ile the shrinkage amount of the 1arvae placed in 1qlll n1 rogen s owe 
e Shr' ink.age u 1
. . . . Th nt of shrinkage due to 
pon iqmd mtrogen preservation. e amou 
Prese rvar · 
ion with 1· 'd . . d because it related to the corresponding 
fi iqm mtroaen was examme 
lelct o 
study th at was being conducted. 
Results 
Larval yellow perch did shrink in the air and ice post death conditions and 
preservation in liquid nitrogen. The larvae in the open-air post death treatment shrunk an 
average of 5 ± 0.65 % (range: 0.14 - 12.63 %), those on ice shrunk an average of 3 ± 0.61 
% (range: 0.15 - 10.95 %), and the larvae preserved in liquid nitrogen shrunk an average 
of 14 ± 0.64 % (range: 5.84 - 26.84 %). 
A significant effect of age on shrinkage was observed (Table 1 ). For air and ice, 
the amount of shrinkage increased with age (Fig. 1 and 2), while it decreased with age for 
liquid nitrogen (Fig. 3). There was no significant interaction determined between post 
death condition and age. 
There was no significant difference determined between the air post death 
condition and the ice post death condition. However, a significant difference was 
determined between the air post death condition and the liquid nitrogen post death 
condition and also a significant difference was determined between the ice post death 
condition and the liquid nitrogen post death condition. (Fig. 4, Table 2). 
Discussion 
This study shows that larval yellow perch do shrink upon death and preservation. 
Larvae left on ice shrink less than larvae left in the open air indicating that a cooler post 
death condition can diminish the amount of shrinkage upon death. An estimate of the 
shrinkage oflarval yellow perch is 14% on preservation in liquid nitrogen and 5% on 
death. However, the experience of the fish must be considered. Larval fish in laboratory 
studies are directly preserved, and therefore only shrinkage due to preservation must be 
163 
corrected for. Larval fish collected in field studies often die due to the stress on their 
bodies due to net towing and shrink upon death even before they are brought aboard the 
research vessel and preserved. Therefore, shrinkage due to death and preservation must 
be taken into consideration. 
While no significant interaction effect was found between post death condition 
and age at the alpha= 0.05 level, there was a significant interaction at the alpha= 0.10 
level. Based on the variability of data and the significant main effect of age coupled with 
the significant main effect of death, we feel there should have been a significant 
interaction between death and age Looking at the air and ice post death conditions in 
Figure 4, it is evident that there is no noticeable interaction between post death condition 
and age. However, Figure 4 does indicate that a potential interaction between post death 
condition and age does exist if you include the liquid nitrogen post death condition. This 
interaction is most likely confounded by the large variability observed in the data. 
Although this data does show that larval yellow perch do shrink upon death with 
the air and ice post death conditions, it is surprising when considering basic fish 
physiology. Many shrinkage studies have been performed on marine fish species 
(McGurk 1985; Theilacker and Porter 1994). At first, one would expect that if these 
larval fish shrank upon death, than larval yellow perch would shrink as well. However, 
marine fish have an internal concentration that is one-third that of their surrounding salt 
water environment (Moyle and Cech 1982). When they die, and their body functions 
(including osmoregulation) shut down, the laws of chemistry take over. Water in the 
hypotonic internal environment of the fish diffuses out to the external hypertonic 
environment. Therefore, the fish shrinks upon death. By contrast, while fresh water fish 
164 
also have an internal concentration that is one-third that of seawater (Moyle and Cech 
1982), the external environment is not saline. Therefore, when fresh water fish die and 
their body functions shut down, one would expect that water from the external hypotonic 
environment would diffuse into the internal hypertonic environment of the fish, and the 
fish would bloat. This study, however, did show that these larval yellow perch did shrink 
upon death. There are a few possible explanations for this. One is that the MS-222 
solution that the larvae were left to sit in for one half hour was a higher ion concentration 
than the internal environment of the fish and therefore, they shrunk. Another possible 
explanation is that water did enter the larvae ' s bodies but they were bloated in a vertical 
manner or in their width, therefore, since the larvae have a fixed amount of epidermis, 
their lengths shrunk. 
The next step in this investigation is to review the videotape used in this study and 
remeasure the larvae. Perhaps the results of shrinkage were due to reader error. Also this 
experiment should be run again with a slightly different method. The larvae should be 
placed in a lethal dosage of MS-222 and then placed in water for the remainder of the 
half hour of post death condition. This way, it can be determined if MS-222 skewed the 
results of this study. Either way, shrinking or bloating, it essential to have a correct 
measure of larval fish when studying them to accurately estimate the performance or 
condition. After this experiment is refined, a critical review of previously published data 
that assess performance as a function of length will be conducted. It will be determined 
if their results are skewed because of neglecting to account for the effects of death upon 
larval fish length. 
165 
Acknowledgements 
We would like to thank Fred Binkowski, and Tom Miller for their help and 
guidance with this project. We would also like to thank the REU program for allowing 
Emily the opportunity to conduct this research. 
166 
References 
Jennnings, S. 1991. The effects of capture, net retention and preservation upon lengths 
oflarval and juvenile bass, Dicentrarchus labrax (L.). J. Fish. Biol. 3s: 349-357. 
McGurk, M. 1985. Effects of net capture on the post preservation morphometry, dry 
weight, and condition factor of Pacific herring larvae. Trans. Am. Fish. Soc. 114: 
348-355. 
Miller, T.J. , L.B. Crowder, J.A. Rice, and E.A. Marschall. 1988. Larval size and 
recruitment mechanisms in fishes : toward a conceptual framework. Can. J. Fish. 
Aquat. Sci. 45: 1657-1670. 
Mills, E.L. , M.V. Pol, R.E. Sherman, and T.B. Culver. 1989. Interrelationships Between 
Prey Body Size and Growth of Age-0 Yellow Perch. Trans. Am. Fish. Soc. 118: 
1-10. 
Moyle, P.B. and J.J. Cech, Jr. 1982. Fishes: An Introduction to Ichthyology. Prentice 
Hall. New York, New York. p 72-73. 
Rooker, J.R. , G.J. Holt, and S.A. Holt. 1997. Condition of larval and juvenile red drum 
(Sciaenops ocellatus) from estuarine nursery habitats. Mr. Biol. 127: 387-394. 
Suthers, I.M., J.J. Cleary, S.C. Battaglene, and R. Evans. 1996. Relative RNA Content 
as a Measure of Condition in Larval and Juvenile Fish. Mar. Freshwater Res. 4T 
301-7. 
Theilacker, G.H. and S.M. Porter. 1994. Condition of larval walleye pollock, Theraora 
b 
chalcogramma, in the western Gulf of Alaska assessed with histological and 
shrinkage indices. Fish. Bull. 93: 333-344. 
167 
Webb, P.W. and R.T. Corolla. 1981. Burst swimming performance of Northern 
Anchovy, Engraulis mordax, larvae. Fish. Bull. 79: 143-150. 
168 
...... 
0\ 
'° 
Table 1. ANOV A table. Significant effect of post death condition and age ( oc = 0.05). No significant interaction between post death conditionand age ( oc = 
0.05) 
Source df F p 
Death 2 85.20 <0.0001 
Age 7 2.85 <0.05 
Death* Age 14 1.62 0.09 
Error 83 
...... 
-....) 
0 
Table 2. ANOV A table. Main effects of death . No significant difference between air and ice. Significant difference between air and liquid nitrogen and 
between ice and liquid nitrogen . 
Source 
Air vs. Ice 
Air vs. Liquid Nitrogen 
Ice vs. Liquid Nitrogen 
p 
.15 
<0.0001 
<0.0001 
Figure Legend 
Figure 1. Percent shrinkage of larval yellow perch (perca fiavescens) in air as a function 
of age. 
Figure 2. Percent shrinkage of larval yellow perch (perca fiavescens) on ice as a function 
of age. 
Figure 3. Percent shrinkage of larval yellow perch (perca fiavescens) in liquid nitrogen 
as a function of age. 
Figure 4. Percent shrinkage oflarval yellow perch (percaflavescens) by death treatment 
as a function of age. 
171 
C1) 
C> 
cu 
~ 
C: 
·-
._ 
.c 
UJ 
....... .... 
'-l C: N 
C1) 
CJ ._ 
C1) 
C. 
14 -
12 _ I. 
10 
8 --
6 • 
4 --
• 
y = 0.0402x + 3.6474 
R2 = 0.0168 
• 
• • • 
2 -I• • • 
• 
• 
• 
I 
• 
• 
• 
• • 
-e 
0 
0 
----- · -- -- --- • -- r 1--
------ -- --·i--· ------ -• . --,-- --
10 20 30 
Days Post Hatch 
• 
• 
• 
• 
• 
---- 1 
40 
12 
Cl) 
D 
C> 10 ·· D 
y = 0.053x + 2.1627 
cu 
R2 = 0.0427 D 
~ 
r: a . 
B 
·-.... 
.c 5 . U) 
...... .... 
B D D 
-...J C 4 -vJ 
'1 
Cl) 
D D 
B 
(.) 
o d D D 
.... 2 - g O ~ Cl) 
D g 
a. 
0 
------ -- -- - - O_ I --- - - - _t] _ -- --TD-
- ------ - ----r- ·- --- -- ·-- ---. - - - - - - 1 0 10 20 
Days Post Hatch 
30 40 
,...... 
-J 
.,i:::. 
• 
30 -
C1) 25 • C) 
CIS + + + y:::: -0.2145X + 18.018 
~ 20 - + + + R2 :::: 0.1889 
en 15 ·· • • • • • ·-
l: • • ~ • • • • • ! • • ~ 10 - • $ • 
l 5 - i -----, .---···· -- --1-- ------- · 
- -· - ·-·· 
40 
0 I .. --- · . ---- ---
0 
- 1- · 
10 
20 30 
Days post Hatch 
30 
Cl) 
C) 
C'G 
~ 
C 
·-... 
.c 
en 
...., 
C 
25 ~· • 
20 -• • 
- ... 
15 - • 
. . ... 
t 
• 
... 
• 
• 
... 
------ ·· --
• Air 
C Ice 
A N2 
Linear (N2) 
-
-Linear (Ice) 
........ Cl) 
--:i 
Vt CJ 
... 
- -
- Linear (Air) 
- - - - - ------
Cl) 
a. 
·1 
0 10 20. 30 40 
Days Post Hatch