ABSTRACT Title of Document: EXCLUDING MAMMALIAN PREDATORS FROM DIAMONDBACK TERRAPIN NESTING BEACHES WITH AN ELECTRIC FENCE Curtis Bennett, Sona Chaudhry, Marjorie Clemens, Lacy Gilmer, Samantha Lee, Thomas Parker, Emily Peterson, Jessica Rajkowski, Karen Shih, Sasika Subramaniam, Rachel Wells, Jessica White Directed By: Dr. Lowell Adams, Department of Environmental Science & Technology Over the past century, diamondback terrapin (Malaclemys terrapin) populations in the Chesapeake Bay region of the United States have declined from their historic abundance. One factor contributing to the decline is increased predation on terrapin nests by raccoons (Procyon lotor) and foxes (Vulpes vulpes and Urocyon cinereoargenteus). We studied the use of electric fences to deter these predators from nesting beaches along the lower Patuxent River, Calvert and St. Mary?s Counties, Maryland. Over the two-year study, the predation rate within treatment (fenced) plots was 40% (4 of 10 nests) compared to 69% (20 of 29 nests) in control plots. We believe that electric fences have potential as a conservation technique for reducing mammalian predation on diamondback terrapin nests. ii EXCLUDING MAMMALIAN PREDATORS FROM DIAMONDBACK TERRAPIN NESTING BEACHES WITH AN ELECTRIC FENCE By Team Saving Testudo Curtis Bennett, Sona Chaudhry, Marjorie Clemens, Lacy Gilmer, Samantha Lee, Thomas Parker, Emily Peterson, Jessica Rajkowski, Karen Shih, Sasika Subramaniam, Rachel Wells, Jessica White Thesis submitted in partial fulfillment of the requirements of the Gemstone Program, University of Maryland, 2009 Advisory Committee: Dr. Lowell Adams, Chair Dr. Paula Henry Dr. Willem Roosenburg Mr. Peter Sharpe Mr. Glenn Therres Dr. Katerina Thompson iii ? Copyright by Team Saving Testudo Curtis Bennett, Sona Chaudhry, Marjorie Clemens, Lacy Gilmer, Samantha Lee, Thomas Parker, Emily Peterson, Jessica Rajkowski, Karen Shih, Sasika Subramaniam, Rachel Wells, Jessica White 2009 iv Acknowledgements Faculty Mentor and Co-faculty Advisor: Dr. Lowell Adams, University of Maryland Collaborating Scientist and Co-faculty Advisor: Dr. Willem Roosenburg, Ohio University Librarian: Ms. Cindy Todd, University of Maryland Summer Field Research Leader: Marjorie Clemens, Team Saving Testudo Field Assistant: Margaret Lilly, University of Maryland Financial Support Gemstone Program, University of Maryland Fear the Turtle Fund, University of Maryland College of Agriculture and Natural Resources Undergraduate Research Program, University of Maryland Washington Biologists' Field Club, Washington D.C. John M. White of Long Island, New York Logistic and Technical Support Jefferson Patterson Park and Museum Dr. Mike Smolek Patuxent Research Refuge Mr. Holliday Obrecht, III Patuxent Wildlife Research Center Dr. John French Mr. Michael Haramis Dr. Paula Henry Trent Hall Farm Dr. Bud Virts, Owner University of Maryland Gemstone Staff Dr. Kaci Thompson Howard County, Maryland, Department of Recreation and Parks Phil Norman and Mark Raab v Table of Contents List of Tables .................................................................................................................................vii List of Figures...............................................................................................................................viii Chapter 1: Introduction....................................................................................................................1 Chapter 2: Literature Review, Field Trips, and Current Study........................................................5 Scientific Literature Review........................................................................................................5 Terrapin Biology .....................................................................................................................5 Habitat.....................................................................................................................................8 Nesting Behavior.....................................................................................................................9 Overall Factors of Decline ....................................................................................................12 Overharvesting..................................................................................................................13 Bycatch..............................................................................................................................14 Habitat Loss......................................................................................................................21 Mammalian Predators ......................................................................................................26 Electric Fences ......................................................................................................................31 Predator Exclusion Through Fencing ..............................................................................33 Conservation Efforts .............................................................................................................35 Field Trips .................................................................................................................................39 Diamondback Terrapin Working Group Workshop..............................................................39 Patuxent Wildlife Research Center .......................................................................................39 Cremona Farms .....................................................................................................................40 Jefferson Patterson Park........................................................................................................41 Current Study.............................................................................................................................42 Chapter 3: Study Areas ..................................................................................................................43 The Patuxent River ....................................................................................................................43 Cremona Farms and Trent Hall Farm........................................................................................46 Jefferson Patterson Park and Museum.......................................................................................48 Patuxent Wildlife Research Center............................................................................................49 Chapter 4: Methodology ................................................................................................................51 Summer 2007.............................................................................................................................51 Experimental Design.............................................................................................................51 Electric Fence Design ...........................................................................................................53 Experimental Procedures.......................................................................................................56 Spring 2008 ...............................................................................................................................58 Summer 2008.............................................................................................................................60 Experimental Design.............................................................................................................60 Electric Fence Design ...........................................................................................................62 Experimental Procedures.......................................................................................................64 Statistical Analysis ....................................................................................................................64 Assumptions..........................................................................................................................64 vi Extraneous/Confounding Variables ......................................................................................65 Institutional Animal Care and Use Committees ........................................................................66 Chapter 5: Results..........................................................................................................................67 Effect of Electric Fences on Predators ......................................................................................67 Effect of Electric Fences on Terrapins ......................................................................................69 Fence Design Modification and Scent Stations.........................................................................70 Chapter 6: Discussion ....................................................................................................................71 Effect of Electric Fences on Predators ......................................................................................71 Trent Hall Farm.....................................................................................................................71 Jefferson Patterson Park and Museum ..................................................................................74 Summary of Study Sites........................................................................................................75 Fence Design Modifications..................................................................................................75 Effect of Electric Fences on Terrapins ......................................................................................77 Cost Analysis.............................................................................................................................79 Conservation and Management Implications ............................................................................81 Recommendations for Future Research.....................................................................................82 Chapter 7: Summary and Conclusions...........................................................................................89 Literature Cited ..............................................................................................................................91 Appendices: ...................................................................................................................................97 Appendix A ...............................................................................................................................97 Appendix B..............................................................................................................................109 Appendix C..............................................................................................................................117 Appendix D .............................................................................................................................120 Appendix E..............................................................................................................................126 Appendix F ..............................................................................................................................141 Appendix G .............................................................................................................................144 Appendix H .............................................................................................................................146 Appendix I...............................................................................................................................147 vii List of Tables Table 1. Effectiveness of electric fences in reducing diamondback terrapin nest predation by raccoons and foxes, Trent Hall Farm, lower Patuxent River, Maryland, 2007- 2008?????????????...??????????????????? 67 Table 2. Effectiveness of electric fences in reducing diamondback terrapin nest predation by raccoons and foxes, Jefferson Patterson Park and Museum, lower Patuxent River, Maryland, 2008????????????????????????????. 68 Table 3. Diamondback terrapin nest location at all study sites, lower Patuxent River, Maryland, 2007-2008??????????????????????????????.. 69 Table 4. Cost breakdown of fence setup at Trent Hall Beach?????????????.... 80 Table 5. Cost breakdown of fence setup at Jefferson Patterson Park??????????..... 81 Table 6. Effectiveness of electric fences in reducing diamondback terrapin nest predation by raccoons and foxes, Trent Hall Farm, lower Patuxent River, Maryland, 2007????.146 Table 7. Predicted effectiveness of electric fences in reducing diamondback terrapin nest predation by raccoons and foxes, Trent Hall Farm, lower Patuxent River, Maryland, with 2 years of data equivalent to 2007.????????????...146 viii List of Figures Figure 1. Female diamondback terrapin native to the Patuxent River, Maryland. ..........................6 Figure 2. Average salinity of the Chesapeake Bay from 1985-2006. .............................................. 9 Figure 3. Uncovered terrapin nest showing a freshly laid cluster of eggs. ....................................10 Figure 4. Cross sectional diagram of a terrapin nest......................................................................11 Figure 5. Typical crab pot..............................................................................................................15 Figure 6. Terrapins caught and drowned in an underwater trap. ...................................................17 Figure 7. Typical crab pot and Roosenburg?s modified crab pot...................................................19 Figure 8. Wood?s design for a BRD. .............................................................................................21 Figure 9. Steel bulkhead along a new shoreline development. ......................................................23 Figure 10. Percentage of impervious surfaces in the Chesapeake Bay region in 2000..................25 Figure 11. Diagram of our electric fence design showing double wire system and ground wire. .32 Figure 12. Living shorelines. .........................................................................................................38 Figure 13. Location of study areas on the lower Patuxent River, Maryland..................................43 Figure 14. Location of study areas, Cremona and Trent Hall Farms............................................. 46 Figure 15. Restored shoreline at Trent Hall Farm beach with breakwater addition providing more openings for terrapins to reach the beach and nest. ..........................................................47 Figure 16. Trent Hall Farm beach schematic map of control and treatment plots, 2007. .............52 Figure 17. Fence baited with peanut butter....................................................................................54 Figure 18. Fence design at Trent Hall Farm beach, 2007. .............................................................55 Figure 19. A depredated terrapin nest............................................................................................57 Figure 20. Fence design with steel corner T-posts, 2008...............................................................59 Figure 21. Schematic design used at Jefferson Patterson Park. The treatment plot is indicated in red, and the control plot is indicated in white...................................................................61 Figure 22. Addition of chicken wire mesh to fence design at Jefferson Patterson Park and Museum, 2008. .................................................................................................................63 Figure 23. Diamondback terrapin nest depredation (in percent) by mammalian predators in treatment and control plots, lower Patuxent River, Maryland, 2007-2008.......................68 Figure 24. Diamondback terrapin nest location (in percent) between treatment and control plots, lower Patuxent River, Maryland, 2007-2008....................................................................70 Figure 25. Diagram of a typical pound net. ...................................................................................73 Figure 26. Photo of treatment plot at Jefferson Patterson Park and Museum, 2008......................78 ix Figure 27. Diagram of a recommended fence design for testing, with inner mesh along with outer fencing. .............................................................................................................................85 Figure 28. Diagram of a recommended fence design for testing, with interior wires....................86 Figure 29. Typical size and shape of terrapin egg .......................................................................144 Figure 30. Terrapin tracks in ?J? shapes......................................................................................145 1 Chapter 1: Introduction The University of Maryland?s motto, ?Fear the Turtle,? refers to the diamondback terrapin (Malaclemys terrapin), a unique reptile integrated into the history, economy, and culture of the state of Maryland and the East Coast of the United States. Unfortunately, the diamondback terrapin population is in decline, and the species is currently listed as threatened in Massachusetts and as a species of concern in at least four states along the East Coast (Natural Heritage?2008). It is clear that, in the Chesapeake Bay ecosystem, humans have had a substantial and negative impact on the terrapin and its environment. There is a great need to reverse this trend?after all, who can fear a turtle that no longer exists? The diamondback terrapin has long been part of the East Coast watershed ecosystem. The species ?terrapin? refers to a broad group of brackish water turtles that are further divided into several subspecies based on region and morphology. Terrapins inhabit the estuaries, coastal rivers, and mangrove swamps all along the East Coast and the Gulf of Mexico and have played a role in local society for over two centuries. Terrapins, once abundant, were an integral part of Native American lives and legends. Some tribes, including the Cherokee, maintain legends that include turtles in creation tales. The terrapin name comes from Native American languages; they were called ?torope? by Virginia Algonquians, ?turepe? by Abenakis, and ?turpen? by Delawares, which means ?edible? or ?good tasting.? In addition, when European colonists first arrived, terrapins, so abundant that they annoyed fishermen who ended up catching them instead of fish, were quickly noted as a good source of meat, a quality that would eventually lead to their decline (Brennessel 2006). 2 Terrapin popularity was at its highest in the 1800s when terrapins became widely used in soup and stews. As their popularity increased, so did their price, and during the 1850s, some fisheries priced a dozen large terrapins at more than $100. Considered gourmet and upscale, they were a favorite at the White House, especially under President William Howard Taft?s term. Many terrapin recipes were available, but a favorite was terrapin soup. The demand spread across the world, and terrapins were exported to Europe and South America in substantial numbers (Brennessel 2006). In the late 1800s and early 1900s, people began to create commercial terrapin farms. Terrapins were caught by ?tarpinners? who used long poles to tap in the mud for the turtles. ?Tarpinners? also used dogs and boats to drag rakes through the mud in hopes of trapping the terrapins. The commercial terrapin farms were unsuccessful, mainly due to the inability of caretakers to raise large numbers of the reptile (Brennessel 2006). After being used in fine cuisine for many decades, terrapin populations received some relief as interest in them began to decline in the 1920s. They were no longer as abundant, and the exorbitant prices made them unpopular during the harsher times of Prohibition, the Great Depression, and two major wars. The demand for the diamondback terrapin, a reptile that had been a major part of the East Coast culture for some 200 years, was gone?unfortunately, it had already harmed the diamondback terrapin population. In 1932, the University of Maryland, College Park, paid a tribute to the reptile by making it the university?s official mascot. Then, in 1994, the state of Maryland made the diamondback terrapin the state?s official reptile. Because of these new honors, the diamondback terrapin was thrust into the national spotlight and awareness of its troubled existence began to spread. 3 While they are currently well known for their historical and traditional value to the state of Maryland, terrapins also play an essential role in the ecosystem of the Chesapeake Bay watershed. The well-being of the watershed could be damaged without the diamondback terrapin population in the food chain to ensure the balance of the ecosystem. Because terrapins make up such a high percentage of the biomass within the ecosystem, they contribute to maintaining the food, nutrient, and energy balance (Appendix A). In addition, secondary consumers in the food chain, such as the terrapins, play a vital role in maintaining the balance of the salt marsh ecosystem in which they live. Terrapins prey upon primary consumers such as mollusks and snails; these primary consumers feed on salt grasses and other vegetation. If the populations of primary consumers are not regulated by secondary consumers, they grow too large and the primary consumers will destroy the marsh vegetation. Without vegetation, the salt marsh habitat can become quickly eroded, turning into uninhabitable mud flats (Silliman and Bertness 2002). There are many factors that contribute to the diamondback terrapin?s population decline in the Chesapeake Bay. First, because terrapins are considered a delicacy in some food markets and restaurants, they are being harvested for cuisine. Additionally, some fishermen use crab and eel pots for their livelihood; the unintended side effect is that these devices catch terrapins, trapping them underwater and eventually causing them to drown. Also, loss of sandy habitat due to shoreline stabilization efforts to protect human commercial and residential development makes it difficult for terrapins to nest. Lastly, terrapin populations are declining as a result of increased predation on their nests, and humans indirectly cause or contribute to this factor. Urbanization attracts predators such 4 as raccoons (Procyon lotor) and foxes (Vulpes vulpes and Urocyon cinereoargenteus), which can lead to higher predator densities in urban areas compared to rural areas (Adams 1994). The increased presence of these predators may greatly affect female terrapins during the nesting process when the turtles are highly vulnerable. Terrapins nest during the day and are in full view of any potential predators. Raccoons and foxes ravage terrapin nests and consume the eggs, depleting the possibility of a regeneration of the terrapin population. This combination of factors and several other causes have nearly taken the diamondback terrapin from the Chesapeake Bay ecosystem. Because of human actions, the future of the diamondback terrapin is uncertain. Although a moratorium passed by the Maryland legislature in early 2007 prohibits the commercial harvest of terrapins (Appendix B), the species is still in trouble. We cannot ignore the call to make a difference in the survival of this one species, especially in light of the fact that hundreds of others are vanishing each year. Our research was conducted with the underlying purpose of promoting terrapin population growth and making a small, but meaningful difference in the preservation of the culture and tradition of the state. Our interest in the terrapin led to an extensive review of the scientific literature on the species and four field trips that are discussed in Chapter 2. 5 Chapter 2: Literature Review, Field Trips, and Current Study To gain knowledge about terrapins and to formulate a research question, we conducted a scientific literature review and made field trips to a terrapin workshop, a wildlife research center, and potential research sites. Relevant studies provided useful information on the terrapin including, but not limited to, background on the species, nesting habits, and population decline factors. Some studies were reviewed but not cited in the text (Appendix C). In addition, the field trips provided us with the opportunity to discuss terrapin-related issues with experts in the field. Scientific Literature Review Terrapin Biology The diamondback terrapin?s anatomical features and biology make it a unique asset to the Chesapeake Bay ecosystem. Terrapins exhibit sexual dimorphism, physical differences between two sexes of the same species of an organism. One of the main distinguishing differences in diamondback terrapins is the great disparity in the size between mature males and females. Males mature at a much quicker rate than their female counterparts. Between the age of 4 to 7 years, males weigh an average of 275 g, with their plastron length (the bottom of the shell structure) reaching about 10 cm. Female terrapins, on the other hand, do not mature until around 8 to 13 years, when they reach an average weight of 1,000 g and an average length of 16 cm. Diamondback terrapins in the Chesapeake Bay area are known to be the largest subspecies, with females reaching 17.5 cm in length and 1,100 g (Figure 1). Males and females can also be 6 differentiated by the size of their head and tail. Females have a much larger head than their male counterparts, but their tails are shorter and narrower (Brennessel 2006). Figure 1. Female diamondback terrapin native to the Patuxent River, Maryland. (Team Saving Testudo Research Collection) The markings on their shell, unique to each individual terrapin, help researchers determine the age of the terrapins with which they work. The scutes, which can be compared to scales on other organisms, make up the pattern on the carapace and the plastron. These scutes expand and grow with the terrapins to continue to accommodate them as they mature. Due to their period of hibernation each year, which typically occurs between the months of December and April, terrapins and their scutes grow 7 inconsistently throughout the year. This leads to the deposition of keratin and pigment during each growth period, forming growth rings similar to those seen in trees. Each scute ring refers to one year of growth in a terrapin. However, as they age and the growth rate begins to decrease, the rings become more difficult to distinguish and age is harder to determine (Brennessel 2006). Terrapins prefer brackish (somewhat salty) water. The absorption of excess salt from the surrounding environment is one of the major problems marine animals face. Because diamondback terrapins live in water of varying salinity, they must rely on osmoregulation, the regulation of the water concentration in their bodies? fluids, to maintain their internal levels of fresh water and salt concentrations. Terrapins have several adaptations that allow them to live in a constantly changing environment. One adaptation is a post-orbital salt gland, known as the lachrymal gland, which allows terrapins to produce salty tears. The lachrymal gland works in a manner similar to the kidneys, allowing terrapins to produce tears with high salt concentrations, thus excreting excess salt from their body. Also, a terrapin?s skin and tissues are fairly impermeable to both sodium and water, simultaneously preventing salt from entering the body when the terrapin is in water with high salt concentrations and retaining water in the body?s tissues (Brennessel 2006). Although diamondback terrapins can excrete excess salt in various ways, they still rely on an external source of fresh water. When it is available in the environment, either through low-salinity water or rain, terrapins drink large amounts of fresh water to rehydrate (Brennessel 2006). 8 Habitat Diamondback terrapins have a relatively large geographical range. They can be found along the coast of the Atlantic Ocean and the coast of the Gulf of Mexico, from Cape Cod in Massachusetts to Corpus Christi Bay in Texas (Ernst and Bury 1982). There are seven subspecies of the diamondback terrapin that live in certain areas along the coast (Ernst et al. 1994). The northern-most subspecies, M. t. terrapin, is the terrapin that resides in the Chesapeake Bay watershed (Burger and Montevecchi 1975). These terrapins are unique among North American turtle species because they are indigenous to the brackish salt marshes along the coast (Ernst et al. 1994). Brackish water is a varying mix of salt water and freshwater that is often found along the ocean coast where freshwater rivers and streams run into the ocean. Brackish waters range in salinity from 0.5 to 30 g of salt per liter or 0.5 to 30 parts per thousand (ppt) (Por 1972). The Chesapeake Bay and other similar estuaries are prime examples of these brackish water conditions (Figure 2). 9 Figure 2. Average salinity of the Chesapeake Bay from 1985-2006. Image taken from Maryland Department of Natural Resources website: http://www.dnr.state.md.us/bay/index_2007drought.html Nesting Behavior The yearly activity cycle of the diamondback terrapin correlates with the temperature of the water. In the mid-Atlantic region, terrapins can be found hibernating in small creeks beginning in mid to late November when the temperature lowers to about 6 to 10? C. They emerge between April and May to prepare for the nesting season once more (Yearicks et al. 1981). Females prefer to lay eggs in sandy areas with no vegetation. Large sandy areas have higher soil temperatures, and the lack of vegetation reduces the chance that eggs will be destroyed by roots (Lazell and Auger 1981). However, this also increases the risk of desiccation and wind erosion so females are very cautious when choosing a nesting 10 site to protect their eggs. Terrapins occasionally nest in clusters because of natural topography constraints or the lack of suitable sites due to human construction of waterfront buildings. This limitation and the resulting nest clusters make terrapin nesting sites easier to find by both humans and predators. Terrapins lay eggs more than once per season in the Chesapeake Bay region. They have the capacity to lay eggs up to five times in a season, although two to three times is more common (Hildebrand 1932). It takes 14 to 17 days for a clutch of eggs to develop. In the Chesapeake Bay region, the diamondback terrapin prefers to nest on narrow sandy beaches where a female will, two to three times a year, deposit a clutch of about 13 eggs (Roosenburg and Dunham) (Figure 3). Figure 3. Uncovered terrapin nest showing a freshly laid cluster of eggs. (Team Saving Testudo Research Collection) 11 Females also show great fidelity to a suitable nesting site, a phenomenon known as philopatry. Once a fitting nesting beach has been found, females are known to return to it every nesting season (Brennessel 2006). Once the female has found her preferred nesting site, she smoothes out an area and begins digging a small hole about 4 cm in diameter (Brennessel 2006). She then makes a teardrop-shaped nest with a small hole at the top and a larger chamber underneath. Average depth from the surface to the bottom of the shaft is 12 cm and from the surface to the bottom of the nest, 16 cm (Figure 4). Figure 4. Cross sectional diagram of a terrapin nest. Hatchling success rates are highest in nests that are moderately deep, about 18 cm. A deeper or shallower nest will not see the same survival rates (Palmer and Cordes 1988). If the nest is too deep, the eggs have a smaller chance of surviving due to low 12 temperatures and lack of oxygen. However, if the nests are too shallow they can be easily depredated and are subject to higher temperatures and erosion. In areas with high predator populations, nesting female terrapins occasionally must chose between their mortality and the survival of the eggs they are laying. In areas with greater mammalian predator presence, it was found that terrapins nest closer to the shoreline, a less desirable nesting site in terms of nest survival rates. When female terrapins are at less risk of direct predation, they tend to nest farther from the shore, which is more ideal (Spencer 2002). The distance a female will travel onto land to nest varies by region. Terrapins along the Chesapeake Bay prefer to keep as close to the shoreline as possible, usually traveling less than 10 m to nest, whereas females in the northeast travel much farther, some making round trips of about 1,600 m across expanses of marsh and sand dunes in the Cape Cod region (Auger and Giovannone 1979). However, no matter how far the female travels, the trip is still very dangerous. Human habitation along coastal watersheds where terrapins nest has increased along with dangers associated with humans such as motor vehicles and pets that can disturb or even kill nesting females. Overall Factors of Decline Unfortunately, opportunities to learn about the diamondback terrapin may be diminishing because of population decline. The decline in their population is mainly due to human factors. In the late 19th and early 20th centuries, terrapins were heavily harvested, drastically reducing the population. Terrapins, considered a delicacy in some areas, were commercially harvested as food (Donnelly et al. 1988). They were also harvested for the pet trade. Today, they face even more and greater dangers from humans 13 according to the Maryland Diamondback Terrapin Task Force (Appendix A). Using Maryland records of terrapins from 1878 to 2001, testimony from experts, overall trends, and specific long-term research work from the Patuxent River, the Task Force concluded that the most prevalent threats to the diamondback terrapin are increased commercial harvesting, habitat loss due to human development and erosion, increasing accessibility of nests for predators, and the threat of drowning in crab pots in the Chesapeake Bay. It found that the current population is not large enough to sustain a commercial market and that the species is in decline. Overharvesting Overharvesting is a contributing factor of population decline because of a high demand for terrapins as food and pets. In the late 1800s, the terrapin was considered a delicacy in many food markets, including that of the United States. Terrapin soup and other dishes made with terrapin meat were very popular. This spurred fishermen to over- harvest the terrapin for high profits. By the turn of the century, fishermen began to notice a decline in the population of the terrapin, but the continued high value of the turtle on the food market stimulated ongoing over-harvesting. By the early 1900s, there were not enough terrapins left to fill the food supply or support fishermen. Some populations were completely wiped out, such as in Long Island, New York (Brennessel 2006). Terrapin populations were given a chance for recovery when high prices, low availability, and a failing economy drove the terrapin out of favor in the United States in the early to mid-1900s. However, interest in the terrapin as a food has re-emerged in recent years. Their renewed status as a highly valued commodity has once again begun to diminish their populations and endanger their survival (Brennessel 2006). 14 With the decline of terrapins becoming increasingly severe, legislators have begun to take notice. According to a recent study by Willem Roosenburg of Ohio University, there has been a population decline of 75% for reproductive age female terrapins in the Patuxent River in only the last decade (Staff and Wire?2007). With statistics like these and given the fact that both Virginia and Delaware had already banned the commercial harvesting of the species; a moratorium was placed on the commercial harvest of diamondback terrapins in Maryland (Appendix B). The bill passed the House and Senate with an overwhelming majority in 2007. Regulators also made it illegal to keep terrapins as pets (Staff and Wire...2007). With this ban, the responsibility for the diamondback terrapin was transferred from the Fisheries Service to the Wildlife and Heritage Service of the Maryland Department of Natural Resources. This policy change came in the winter of 2008. Because the terrapin was no longer legally commercially harvested, protection of the species was moved to the new service where there will be tighter restrictions and stricter monitoring of their take for research and educational purposes, instead of harvesting (Scott Smith, Department of Natural Resources, personal communication). It is unknown how much harvesting continues today even with the new protective legislation. It is unlikely that the ban has put an end to the harvesting. However, even if the commercial harvesting does not continue, many other factors still threaten the diamondback terrapin. Bycatch Also contributing to the population decline is bycatch as terrapins get caught in crab pots and drown. Crab pots, which are traps designed to rest on the estuary bottom, 15 are one of the major causes of the decrease in population (Figure 5). These traps are usually baited to attract crabs, but unfortunately, the bait also attracts terrapins. Once a creature enters the funnel-shaped entrance, it is unable to escape. Terrapins and other animals that are unintentionally caught in the traps are referred to as bycatch. Unlike crabs, terrapins do not have gills and cannot breathe underwater. As a result, many terrapins drown in traps before they can be released by crabbers. Additionally, ghost traps, traps that have been lost or abandoned, are deadly for the numerous terrapins that are unlucky enough to get caught in them (Bishop 1983; Roosenburg et al. 1997). These various traps tend to kill male terrapins and young females because mature females are usually too large to fit into the traps. Without these males and young females, mating occurs less often and the feasibility of replenishing the terrapin population diminishes. This population skew has unfortunate implications for the long-term survival of the population (Dorcas et al. 2007). Figure 5. Typical crab pot. 16 Image taken from North Carolina State website: http://www4.ncsu.edu/~dbeggles/education/synergy/bluecrab/bscrab.html As concerns about bycatch mortality have increased, many researchers have tried to determine the impact of crab pots on the terrapin population. Bishop (1983) measured the mortality rate for diamondback terrapins due to crab pots in Ashley Bay, South Carolina. Bishop found that turtles rarely drowned in commercial pots because crabbers usually checked their pots often enough to release terrapins before they ran out of air. He concluded that despite deaths from bycatch, the terrapin population in South Carolina was not at significant risk unless the terrapins became over-harvested by the fishing industries that hunt terrapins (Bishop 1983). Most terrapin biologists today, however, agree that the terrapin population is at risk and that crab pots are dramatically harming the terrapin population. Working in the Chesapeake Bay, Roosenburg et al. (1997) found capture rates similar to the ones Bishop found, but predicted much higher mortality rates ranging from 15% to as high as 78% of the population. Through discussions with other turtle biologists along the East Coast, Roosenburg and his associates learned that terrapin populations have been decreasing in South Carolina, New Jersey, Florida, and Louisiana ? all coinciding with an increase in the crabbing industry in those areas (Roosenburg et al. 1997). Although there is substantial evidence from these locations supporting the theory that the terrapin populations are declining, it is still difficult to prove that this decline is due to crab pot mortality. Other underwater traps also may kill terrapins, particularly if not checked frequently (Figure 6). 17 Figure 6. Terrapins caught and drowned in an underwater trap. (Team Saving Testudo Research Collection) Recently, a team of scientists published the first study that definitively connected crab pot bycatch to the terrapin population decline (Dorcas et al. 2007). Because crab pots in general only catch and kill male and young female terrapins, crab pot mortality should have a distinctive effect on population dynamics: the average body size of individuals should increase and there should be more females than males. Dorcas?s team collected data from a population in Kiawah Island, South Carolina, over a period of twenty years. They concluded that the average body size had increased for both males and females and that there was a much higher female to male sex ratio. His findings were 18 consistent with the hypothesis that crab pots were significantly affecting the terrapin population. Dorcas and associates also found that the effect of crab pots on terrapin population was most dramatic in areas with recreational fishing as opposed to commercial crab fishers. Commercial crabbers check their traps more regularly and are therefore able to release more terrapins before they drown (Dorcas et al. 2007). They are also much less likely than recreational fishers to leave ghost traps. This same pattern of terrapin mortality was observed by both Bishop (1983) and Roosenburg et al. (1997). Roosenburg et al. (1997) further noted that in the Chesapeake Bay, commercial fishers are required to set traps farther from shore in water that is deeper than the preferred habitats of the terrapin. There are no similar restrictions on recreational crabbers in the Chesapeake Bay who tend to set traps in shallow water where more terrapins are caught. Just as crab pots are harmful to the terrapin population, ghost traps are also considered detrimental, although it is difficult to determine their actual effect. Bishop (1983) found ghost pots with 15 and 28 dead turtles and Roosenburg et al. (1997) reported a ghost pot with 49 dead terrapins. Turtles tend to aggregate and may use each other to find food sources during normal foraging. Unfortunately, this means that when one or two get caught in a trap they attract others; this attraction is especially strong during the breeding season (Bishop 1983). Although it is possible to count the number of captured terrapins in ghost traps, it is difficult to quantify the full effects of ghost pots because the exact number of these abandoned traps is unknown and they are hard to locate. Crabbing is a multimillion-dollar industry and a large part of the culture in many coastal regions, especially the Chesapeake Bay. The established nature of crabbing can 19 make it difficult to balance environmental considerations with the economic and recreational desires of residents. Roosenburg et al. (1997) suggested a solution by designing a crab pot that allowed terrapins to breathe so they would not drown before they could be released by crabbers (Figure 7). The base of the pot was the same as a normal crab pot, but there was an upper story to the trap that rested above the water level so terrapins could surface. The trap was designed to be tied to a large wooden stake so that it remained upright. When tested, the design was very effective at preventing terrapin death (Roosenburg et al. 1997). Unfortunately, because these traps are substantially larger than normal traps, they create a greater inconvenience for crabbers. Therefore, it is not likely that crabbers will be willing to use this design (Dorcas et al. 2007). Figure 7. Typical crab pot and Roosenburg?s modified crab pot. (Roosenburg et al. 1997). 20 Another more feasible solution is a bycatch reduction device (BRD) (Wood 1997). Wood worked to reduce terrapin bycatch deaths in New Jersey?s Great Bay. His design was cheap and simple; the original device was constructed out of coat-hangers (Wood 1997). The BRD consisted of an inflexible wired rectangle that was inserted into the small end of the funnel-shaped openings in the crab-pots (Figure 8). This device effectively blocked most terrapins from entering while still allowing even the largest crabs to be trapped. There have been many studies that demonstrate the efficacy of BRDs. Roosenburg and Green (2000) found that a BRD with the dimensions of 4.5 cm x 12 cm was the most effective size for preventing terrapin bycatch in the Chesapeake Bay. This finding was confirmed by Butler and Heinrich (2007) who found that BRDs reduced bycatch by 73.2%. Both studies also cautioned that the optimal size for BRDs depends on the size of the individual terrapins in a particular area and may need to be adjusted for regional differences. If used properly, BRDs do not have any effect on the size, number, or sex of the crabs that are caught in the crab pots, but effectively reduce the number of terrapins becoming caught in the deathtraps. 21 Figure 8. Wood?s design for a BRD. (Wood 1997). Habitat Loss Habitat loss from construction of waterfront houses and other development makes it more difficult for terrapins to reproduce as successfully as possible because optimal beaches are eliminated. Most terrapin habitats have shrunk due to human activity, particularly urban development. The human population of Maryland has more than doubled since the 1950s and is forecasted to increase by another million individuals within the next 30 years, creating a major strain on the ecology of the Chesapeake Bay (Chesapeake Bay Watershed? [updated 2008]). Since colonial times, there has been record of the degradation of salt marshes. Salt marshes were often drained to grow salt meadow cordgrass (Spartina patens), also called salt-marsh hay, which was a valuable commodity at the time. Marshes have also been drained for other reasons including mosquito control and conversion of land to a more usable form for human recreation. Alterations to the land of this nature result in a loss of 22 habitat for terrapins because it offers protection and food (Bossero and Draud 2004). However, since the 1970s tidal wetlands have received state and federal protection. Also contributing to the loss of salt marsh habitats for terrapins is the restriction of tidal flow along shorelines. In many cases, the erosion of beaches, dunes, and marsh uplands occurs naturally, although humans may accelerate shoreline changes. Human dike construction restricts the flow of sediments and drowns marsh plants, harming and sometimes destroying the terrapins? habitat. Additionally, bulkheads, a form of light-duty seawall to protect waterfront property from erosion, create some of the largest problems for terrapins. Bulkheads are built to retain soil and prevent land from eroding and sliding towards the water (Figure 9). However, they also prevent the natural deposition of sand along the shoreline. This results in a change in the shoreline?s profile and composition, damaging marshes through active erosion, preventing access to nest sites, or destroying beaches that may have originally been considered ideal nesting sites for diamondback terrapins and many other species. 23 Figure 9. Steel bulkhead along a new shoreline development. Image taken from Waterside Construction website: http://www.watersideconstruction.com/Waterside%20Construction%20Bulkheads.htm Overall, marshes are especially valuable for terrapins because they provide adequate sources of food and cover. Terrapins are also known to make nests against bulkheads, and, as a result, their eggs are usually drowned by high tide levels (Living Shorelines?[date unknown]). The destruction of this habitat further threatens the status of the terrapin populations (Bossero and Draud 2004). The species? habitat has been put at risk by the increasing development of coastal areas. As the human population continues to grow at a high rate, there is an increased demand for adequate living space. Shorelines are a perfect example of a habitat that has been disturbed due to increased human development. As the cities and suburbs become more populated, people seek other areas to develop; shorelines have always been a 24 popular option. In Maryland, homes and other buildings have become a common sight along the shores of the Chesapeake Bay and the Patuxent and Potomac rivers. Bulkheads are used against the shores to minimize private property loss through erosion and runoff. Furthermore, as more houses are built, more construction occurs in order to provide the new residents with grocery stores, schools, and other community buildings. This leads to the paving of roads and streets, which are impervious surfaces where water can no longer infiltrate and percolate into the ground (Chesapeake Bay Watershed? [updated 2008]). Because of this, greater levels of sediment and chemicals flow into streams and create a lethal contamination of the water in some areas. Impervious surfaces have increased by more than five times since the 1990s, which could be dangerous to the overall health of the watershed because of the chemicals and foreign material that is introduced to the water (Chesapeake Bay Watershed? [updated 2008]). As human populations increase and land is developed in order to accommodate the influx of people, the Chesapeake Bay watershed is slowly being destroyed, harming terrapins and the other animals that inhabit those waters (Figure 10). 25 Figure 10. Percentage of impervious surfaces in the Chesapeake Bay region in 2000. Image taken from Woods Hole Research Center: http://www.whrc.org/midatlantic/mapping_land_cover/products/impervious_surfaces.htm Homes, seawalls, resorts, and highways have replaced traditional terrapin nesting areas. However, some efforts have been made to preserve or recreate nesting beaches for 26 terrapins. One example of successful creation of nesting beach habitat can be found at Horsehead Wetlands Center (now Chesapeake Bay Environmental Center), in Maryland (Brennessel 2006). Marshes must also be restored in order to give terrapins some of their territory back. Mammalian Predators Three of the main mammalian predators of terrapin eggs in Maryland are red and gray foxes, and raccoons (Roosenburg 1990; Roosenburg 1991). These species are found throughout the East Coast and pose a significant challenge for terrapin conservation efforts. Because these predators have such a large impact on terrapin nest survival, it is important for terrapin researchers to have, at the least, a basic understanding of these animals? behaviors and biology. Any comprehensive conservation plan must take these nest predators into consideration. The red fox is a reddish-colored canine with a pointed nose, large ears, and a long, bushy, white-tipped tail. It is about the size of a small dog, with a total length between 89 and 111 cm and a weight between 3.4 and 6.4 kg. Red foxes have benefited from the clearing of forests and have increased their numbers and the range of their habitat. They are found over most of eastern North America and are most prevalent in farmland areas that contain wooded areas, marshes, and streams. Red foxes also occupy metropolitan areas and thrive in broken land areas (Whitaker, Jr. and Hamilton, Jr. 1998). A family of red foxes may occupy an area ranging from 60 to 600 hectares. Since small prey is usually abundant in small areas, a red fox family can thrive in these areas, so much so that they may form year-round extended- family groups in their territory. The foxes? mating season usually occurs between January 27 and February, and four to 10 young are born in March or April (Whitaker, Jr. and Hamilton, Jr. 1998). The red fox usually requires about 2.3 kg of food each week and is a nocturnal predator with a very diverse diet. Because they have grown accustomed to urbanization, red foxes consume human garbage in addition to natural prey such as squirrels, mice, and rabbits (Adams 1994). They are major predators on duck nests in some areas. They will also eat fish, fruits, and seeds. In addition, foxes also consume terrapin eggs when they are available (Lariviere and Pasitschniak-Arts 1996). The gray fox is a canine with a pointed nose and ears, a bushy tail, and an overall grizzled appearance. Its fur is a mixture of white, gray, and black and consists of black tipped hairs that form a stripe down its back to its tail. Gray foxes have a total length between 80 and 112 cm and a weight between 3 and 7 kg. Typically, the male foxes are slightly larger than the females. Gray foxes live in wooded and rocky environments from Canada to South America. Like the red fox, the gray fox?s mating season usually occurs between January and February, and the young are born in March. The gray fox is a nocturnal predator with a very diverse diet that varies among regions and seasons. In the eastern United States, mammals make up the majority of the winter diet while in the summer, invertebrates and plants are most prevalent (Fritzell and Haroldson 1982). Raccoons are medium-sized mammals that have a black facial mask which covers their eyes and cheeks. They are usually a shade of gray and have a long bushy tail with black rings. Adults usually weigh between 5 and 7 kg and have an average length of 81.2 cm. Mating occurs between January and March, and female raccoons have a gestation 28 period of about 63 days. Usually, three to seven live young are born around April or May (Whitaker, Jr. and Hamilton, Jr. 1998). The population of raccoons in North America was very low during the 1930s. However, their numbers steadily grew during the 1940s, and the raccoon populations are now stable. This increase in population may be attributed to several factors. Firstly, the growth of cities is beneficial to raccoons and the mammals thrive in urban and suburban areas. Secondly, raccoons also benefit from an increase in agricultural crops such as corn. Finally, the declining populations of wolves, a natural predator of raccoons, has allowed for the replenishment of the raccoon population (Zeveloff 2002). Raccoons live in a wide variety of habitats, but are mostly found in moist or wet areas such as freshwater and saltwater marshes. The population densities of raccoons vary from site to site. For example, in North Dakota, their density ranges from 0.5 to 1 per km 2 , whereas in eastern Virginia?s tidewater region, their density may be 17.2 per km 2 (Zeveloff 2002). Raccoons are nocturnal predators and have a diverse, omnivorous diet, consuming a wide range of both plants and animals. They eat invertebrates, particularly arthropods such as insects, crustaceans, and spiders. Although crayfish is a favorite food for these animals, raccoons that live along marshes and coastlines will also eat crustaceans such as shrimp, crabs, clams, oysters, and mussels. Amphibians are usually present in raccoon habitats, although they are not consumed very often. Reptiles, such as snakes, lizards, and turtles, also do not make up a significant part of their diet but may still be consumed (Zeveloff 2002). Furthermore, a significant part of their diet consists of berries, nuts, and seeds, and they have also been known to prey on sea turtle hatchlings and their eggs 29 when available. However, because they have a varied diet, raccoons are only selective when food is overly abundant. They will eat whatever is available when food is limited (Lotze and Anderson 1979). Raccoons are also able to locate and consume new foods, and this behavior can be copied by other raccoons and passed on to later generations as a type of cultural inheritance. An example of this phenomenon would be the association of broken eggshells with terrapin nests and eggs, a source of food (Zeveloff 2002). Therefore it appears that raccoon predation of eggs is not an instinctual occurrence but rather a learned behavior. Although it is unclear how predators find terrapin nests, one study suggested that raccoons use a combination of soil disturbance patterns and scent of ocean water as indicators of nests. In another study, human scent or flags used to mark nests in scientific research did not appear to help raccoons find nests (Burke et al. 2005). Overall, it can be inferred that terrapins and their eggs comprise only a small and unimportant portion of their predators? diet. Terrapin eggs can be labeled as more of an occasional treat than a survival necessity. Such conclusions suggest that limiting the availability of terrapin eggs would not harm the populations of either raccoons or foxes but would greatly protect the terrapin population. Although some places have recorded a predation rate of up to 90%, the exact impact of predators on terrapin populations is unknown (Feinburg and Burke 2003). Recently, more people have begun to move to beaches and build homes there, destroying the terrapins? natural habitat. When there is less habitat available, the nesting density increases as terrapins nest closer together in the remaining accessible habitat. However, a higher nesting density also leads to higher predation rates because there are more nests in 30 a smaller area for predators to destroy. Therefore, limited space for terrapin nests may lead to an increase in predation rate (Roosenburg 1991). One study concerning the Australian fresh water turtle (Emydura macquarii) suggested that foxes use both chemical detection of eggs and slight soil disturbances to locate nests (Spencer 2002). Another comprehensive experiment supported this conclusion (Burke et al. 2005). These investigators also examined the effects of humans on nest detection by predators. After recording which nests had been marked or visited by humans, the researchers observed the nests for signs of predation. They concluded that flags did not increase the likelihood for predation and the presence of human scent actually lowered the predation rate. Other studies suggest a link between human development and an increase in the number of predators in an area. For instance, Hoffmann and Gottschang (1977) concluded that as areas become more developed, the density of predators in anthropological areas increases. Another study found that construction of roads and bridges makes accessing otherwise-secluded beaches easier. Therefore, humans facilitate the increased predation of terrapins by essentially drawing animals such as raccoons to those areas which were once unknown to the predators (Roosenberg 1991). Finally, one study concluded that eggs are susceptible to predation because with the increase in human population comes an abundance of associated nest predators such as raccoons, skunks, dogs, and cats (Chambers 2000). The connection between developed areas and predator populations may be attributed to an increase in food and habitat availability for these predators. Raccoons and foxes in more developed areas are recognized as subsidized predators, meaning people either intentionally or accidentally provide food and other habitat requirements. By 31 receiving such aid from people, populations of these predators can be maintained at higher than natural levels. Densities of both raccoons and foxes are typically higher in more developed areas than in more rural areas (Adams 1994; Whitaker, Jr. and Hamilton, Jr. 1998). In some areas, raccoons have become so overabundant that they could be detrimental to the continued survival of their prey (Garrott et al. 1993). Because raccoons and foxes pose a possible threat to the status of some of their prey, many researchers are studying and developing possible predator control methods. For instance, Engeman et al. (2003) conducted long-term field research that centered on the main predators (including raccoons) of three threatened or endangered species of sea turtles. These investigators tested the impact of monitoring predators over a long period of time on the effectiveness of predation control. From their study, the researchers found that monitoring predators can successfully indicate the best times and locations for predator-removal methods. They also showed that there were low levels of predators before the nesting season and then an increase in the number of raccoons during the nesting season. Electric Fences Electric fences are commonly used for the exclusion of animals from vegetation, livestock, and nesting areas (LaGrange et al. 1995; Reidy et al. 2008). An electric fence consists of three major components: a hot wire fence, an energizer to supply the power to the fence, and a ground system. The ground system usually consists of a ground rod that is simply a metal rod stuck into the earth and attached to the ground side of the energizer via a wire. For current to flow through a circuit, there needs to be a complete, unbroken connection between the positive and negative ends of the battery. The positive end of the 32 battery is attached to the energized wire while the negative end of the battery is attached to the grounding system. The fence creates an open circuit because the energized wire is not connected to the ground or the ground wire. When an animal touches the fence and the ground at the same time, it completes the circuit through the ground, and current is pulled through the animal, shocking it (Figure 11). Electric fences used on animals are also typically high-voltage, low-amperage systems. Figure 11. Diagram of our electric fence design showing double wire system and ground wire. Image taken from Gallagher Animal Management Systems website: http://www.gallagher.co.nz/electric-fence-brochures.aspx 33 Typically, an energizer sends out the power in pulses. This is because when a body receives an electrical shock, the muscles tend to tense up and contract, sometimes causing an animal to grab on to the wire and not be able to let go. A pulsing signal allows time for the animal to let go of the fence. The shock does not harm the animal, but scares it and presumably makes it less likely to try to cross the fence in the future. Predator Exclusion Through Fencing Many studies have tested the efficacy of various types of fencing in predator exclusion, but none have been conducted specifically with diamondback terrapins. When we started the process of designing our fences, we considered two different methods of fencing: those that protected individual nests after they were laid and those that covered the beach but still allowed terrapins to move in and out. However, we were concerned with creating a fence design that was simple, easy to assemble, affordable, and maximized effectiveness. To determine the best way to do this, we consulted several studies, some that focused on fencing as a means to protect birds? nests and some that focused on fencing to keep pests out of crops. A long-term study was conducted in Iowa between 1978 and 1990 and focused on the use of electric fences for the exclusion of striped skunks and raccoons from duck nests (LaGrange et al. 1995). After the duck eggs were laid, an electric fence exclosure was placed over the nests. The fence was constructed of wires placed at ground level and of five strands of alternating electrified and grounded wires placed at 69, 76, 86, 97, and 109 cm above the ground. A charged trip wire and 5 cm of non-electrified poultry netting were also added as additional barriers. Most importantly, the design of the fencing allowed hens and hatchlings to easily get out of the enclosures. The fences led to a 19% 34 increase in nest success rate and a 21% reduction in nest predation rate (LaGrange et al. 1995). Another study, conducted in Australia, was designed to test fences for general protection of threatened species rather than for a specific species. Both electric wire fences and netting fences were tested, and feral cats, foxes, and rabbits were placed inside the fences (Moseby and Read 2006). To prevent predators from digging under the fences, the fences were extended underground, and to prevent predators from jumping over the fences, the fences were extended to an upside-down ?U? at the top (Moseby and Read 2006). At the beginning, wooden posts were used, but the cats easily climbed up the posts and escaped. After the posts were replaced with metal ones, the fences were more effective (Moseby and Read 2006). Overall, the electric wire fences were found to be ineffective unless combined with a physical barrier, such as a netting fence, to ensure that the predators paused long enough to receive a shock (Moseby and Read 2006). In the United Kingdom, Poole and Mckillop (2002) studied the effectiveness of electric and non-electric fences in excluding red foxes. Wire fences with alternating electric and ground wires at various heights above the ground were used with a 6kV maximum output energizer, as well as electrified netting fences and non-electrified wire fences. Foxes that had been raised in captivity were placed inside the fences. As seen through video recordings, the foxes only ever crossed the electric fences during maintenance but frequently crossed the plain wire fences. The netted fences were crossed less frequently than the wire fences (Poole and Mckillop 2002). In a similar study conducted in the United Kingdom, researchers tried to exclude badgers using both non- electric fences and electric fences with various voltages. The non-electric fences were 35 found to be almost wholly ineffective. The efficacy of the fences increased as the voltage increased, but the lowest voltage was significantly effective (Poole et al. 2004). Around alkali lakes in North Dakota and Montana, Murphy et al. (2003) studied the effectiveness of predator exclusion fences to protect endangered shorebirds. A wire mesh barrier was placed over the nests, and an electric fence was placed around the whole area. Potential predators included coyotes, red foxes, raccoons, badgers, skunks, squirrels, and various birds. The electric fence, with the wire mesh barrier, was not found to be effective enough to justify its cost. In addition, these investigators determined that, although the individual nest enclosures were effective in some of the areas in which studies were set up, they were not consistently effective (Murphy et al. 2003). Therefore, while the use of enclosures around individual nests is effective in some circumstances, it is also labor-intensive, expensive, and slightly unreliable. If electric fences are effective in reducing predation on nesting beaches, then large nesting areas could be protected more easily. Conservation Efforts The diamondback terrapin's threatened existence received national awareness after the turtle gained its official status as the state reptile of Maryland and the University of Maryland, College Park's official mascot. After this rise in awareness, state laws began to emerge in efforts to protect the species, and more political action was taken by researchers, environmentalists, and Maryland residents to prevent a vital part of the Chesapeake Bay ecosystem from becoming a threatened species. Currently there are a limited number of studies concerning conservation efforts for the diamondback terrapin due to a lack of general knowledge on the species and its 36 behaviors. For decades, researchers have conducted multiple studies to obtain basic information on terrapin ecology. Ultimately, these data can be used to develop conservation strategies for the terrapin. For instance, Roosenburg et al. (1997) studied the mortality rates and declining population numbers of terrapins in the Chesapeake Bay in an effort to learn more about the species and its plight. Other researchers concentrated on collecting information regarding hatchlings and the nesting behavior of the terrapin (Burger and Montevecchi 1975; Burger 1976; Burger 1977). Overall, past research conducted on the species has pinpointed a decline in population and has identified several factors, many of them related to human behavior, affecting this population trend. After researchers noted a decline in the diamondback?s population, individuals formed institutes and programs to educate the public about the terrapin?s predicament and to, on a smaller scale, physically protect the animal. For example, the Terrapin Conservation Project at the New Jersey Wetlands Institute incubates and hatches terrapin eggs recovered from terrapins killed on roads, eventually releasing the hatchlings into the wild (Appendix D). These endeavors are major firsts in addressing the population decline. For the past several years, the diamondback terrapins? decreasing numbers have captured the interest of the public and governmental administrations. In 2006, the Maryland General Assembly passed legislation regarding commercial harvesting of terrapins in the state (Appendix E), and the Maryland Department of Natural Resources developed regulations to implement the law (Maryland Department of Natural Resources 2006). These new regulations affected the commercial harvesting season, the selection of terrapins, and the legal process associated with the harvest of terrapins. Firstly, fishermen 37 and trappers now could only commercially harvest terrapins from the beginning of August through the end of October, six months shorter than the previous nine-month terrapin harvesting season. This protected terrapins in the winter when they are known to hibernate together in areas known as hibernacula. Secondly, only terrapins between 10.2 and 17.8 cm could be caught, protecting reproducing female terrapins and hatchlings and small juveniles. Lastly, those who want to harvest the diamondback terrapin were now required to obtain permits in advance and provide information to the Department of Natural Resources about their catch. Overall, these new regulations were a first step in addressing the population decline in the terrapin. Most recently in Maryland, the diamondback terrapin made the news when Maryland Governor Martin O?Malley in April 2007 signed into effect a ban on commercial harvesting of terrapins (Wagner 2007). Under the Natural Resources ? Diamondback Terrapin ? Take and Possession Act (Appendix B), a person cannot catch a terrapin for commercial purposes (Dyson 2007). This act had great implications for the conservation of the diamondback terrapin; terrapins could no longer easily be caught to meet demand in China and American Chinese restaurants, where they are considered a delicacy. Other legislation has also recently been brought to the attention of the Maryland government; these acts will protect numerous species including the diamondback terrapin. More specifically, in spring of 2008, the Maryland Senate passed the Living Shoreline Protection Act of 2008 (Appendix F) to address habitat destruction by shoreline development. Under this act, shorefront lot owners would be required, whenever possible, to use nonstructural shoreline stabilization methods to prevent erosion and marsh destruction. Among the suggested methods include living shorelines in which 38 natural elements such as plants, stone, and sand are deliberately placed along the shore to protect vegetation and habitats (Chesapeake Bay Foundation 2007) (Figure 12). Figure 12. Living shorelines. Image taken from Jefferson Patterson Park and Museum website: http://www.jefpat.org/Living%20Shorelines/lsmainpage.htm This living shorelines approach has the potential to protect the terrapin?s natural habitat and offset some of the damage done by shoreline development, and several states including North Carolina and Virginia have already begun to implement the method (Welcome to Maryland?[date unknown]). While the diamondback terrapin may hold a special significance to the state of Maryland, other states in which the reptile can be found are also, like Maryland, using political means to protect the species. 39 Field Trips Diamondback Terrapin Working Group Workshop On March 2, 2007, several team members and our mentor attended the Mid- Atlantic Region?s Diamondback Terrapin Working Group Annual Meeting, where we met biologists, herpetologists, and researchers from Maryland, Virginia, Delaware, Pennsylvania, New Jersey, and Ohio and several members of the state governments of Maryland, Virginia, and Delaware. One of the main areas of discussion was regulatory legislation, including the proposed moratorium on terrapin harvesting that Maryland was considering at the time. We learned a great deal about current, or not-yet-published, research in the field including habitat creation projects, tagging, and the creation of a DNA database to track terrapin populations. We also looked at several studies aimed at eliminating or lessening crab pot bycatch, specifically in Texas and New Jersey. We discussed research needs concerning terrapins with experts in the field, including Dr. Paula Henry of Patuxent Wildlife Research Center, Dr. Willem Roosenburg of Ohio University, and Dr. Roger Wood of Richard Stockton College. We learned that predation by foxes and raccoons on terrapin nests was a significant mortality factor of concern to researchers and conservationists. Patuxent Wildlife Research Center On March 9, 2007, we visited Patuxent Wildlife Research Center in Laurel, Maryland, and met with researchers to discuss the use of electric fences to deter mammalian predators. The Center maintains many experimental animals in pens and cages and uses electric fences to deter both avian and mammalian predators. 40 After we presented our initial fence design concept to these experts, we were able to thoroughly discuss with them the feasibility and quality of our experiment. The first issue that arose concerned the number of strands of wire included in the design; initially, the team had focused on a one-strand fence, but we were made aware of the heights and jumping capabilities of the various predators, so we broadened our design concept to include at least three strands of wire at different heights. The researchers also suggested baiting the wires or using an attractant (lure) to ensure that the predators would touch their noses to the fence and receive that initial shock; this had the potential to quickly and effectively deter predators away from the nests and contribute to the success of the fence as a conservation effort. While we eventually decided to bait the fences, the Center?s staff raised the question of whether or not we wanted to aggressively attract predators to the nests, a question we would debate months later. After discussing our interest in using fences to deter mammalian predators from diamondback terrapin nesting beaches, the Patuxent researchers invited us to visit a research site to see how the Center was using large electric cages to protect American kestrels, Falco sparverius, a small falcon. Our trip to the Center helped to further develop our experimental design. Cremona Farms On May 23, 2007, we visited potential study sites along the Patuxent River near Mechanicsville, Maryland. The lower Patuxent River offered several potential advantages for our study. Dr. Willem Roosenburg, of Ohio University, had studied terrapins in the area for close to 30 years and maintained a field research station at Cremona Farms with housing accommodations. In addition, Dr. Roosenburg had agreed to collaborate with us 41 as a co-faculty advisor. Lastly, it was known that terrapins used beaches along the river for nesting. Several of us and our mentor returned to Mechanicsville to examine two beaches: one at Trent Hall Farm and another at Burton?s Beach. Both were viable nesting beaches, and the beach at Trent Hall was specifically restored to create more terrapin nesting habitat as restitution for a nearby oil spill that occurred in April 2000 (Holliday et al. 2008). Trent Hall Beach, an area estimated to be about 230 m 2 , had varying levels of habitat, from loosely packed sand and sparse vegetation to dense vegetation and rocky soil. Burton?s Beach was a smaller beach located off of Washington Creek. It was long and very narrow, with loose sand and dense vegetation, including small trees and poison ivy patches. After visiting both sites with our two faculty advisors, we decided to work only at the Trent Hall Farm beach during the summer of 2007. Because our project design required multiple treatment and control plots of at least 5 m x 5 m each, Burton?s Beach would not have provided sufficient area due to the very narrow nature of the beach. Jefferson Patterson Park In 2007, we also visited several other possible nest sites, one of which was Jefferson Patterson Park and Museum (JPP), along the Patuxent River. At JPP we observed a very high terrapin nest density and also saw high predator activity. An informal survey of the beach on one day revealed at least ten depredated nests. We felt this would be an ideal location for further testing of our fence design and ultimately received permission to use the beach at JPP for the spring of 2008. The park was very 42 accommodating and supportive of our research and requested that we construct our fences as far away as possible from areas with heavy public traffic. Current Study The current study was conceived and designed following review of the scientific literature concerning diamondback terrapin populations and the field trips outlined above. We focused our efforts on excluding mammalian predators from diamondback terrapin nesting beaches with an electric fence. We developed three questions we hoped to answer with our research: (1) What effect, if any, does electric fencing have on predation of terrapin nests? (2) Do terrapins discriminate based on presence of fences when choosing nesting sites? (3) Are electric fences viable (in terms of cost, durability, environmental protection, etc.) as a widespread conservation technique? In Chapter 3, we present the study sites we selected to conduct our research. 43 Chapter 3: Study Areas This study was conducted on diamondback terrapin nesting beaches at Trent Hall Farm and Jefferson Patterson Park and Museum along the lower Patuxent River, north of Solomons, Maryland, and at Patuxent Research Refuge in Laurel, Maryland. Trent Hall Farm and Jefferson Patterson Park and Museum had similar terrain and were known to provide quality nesting areas for terrapins (Figure 13). Figure 13. Location of study areas on the lower Patuxent River, Maryland. Satellite image from Google Earth. The Patuxent River The Patuxent River watershed drains from seven counties in Maryland. It is one of the major river basins in the state that empty into the Chesapeake Bay, the largest 44 estuary in the nation. The Patuxent runs along from the Maryland Piedmont in between Frederick, Carroll, Montgomery, and Howard counties, emptying out into the Chesapeake Bay at Solomons, Maryland. It is 177 km long and covers a geographical area of 2290 km 2 , making it the largest river entirely bounded by the state of Maryland (Breitburg et al. 2003). It lies between the major population centers of Washington, D.C. and Baltimore, Maryland, and traverses rural-agricultural and urban-suburban land uses. This makes the Patuxent River an ideal modeling system for the effects of urban, rural, and agricultural activity on various aspects of the watershed. The river?s topography is also extremely variable, ranging in depth from 3.1 to 39.6 m at its deepest and never becoming wider than 3.7 km. The estuarine portion of the river consists of brackish wetlands and marshes (USACE 1996). Most likely, the majority of terrapins can be found in these areas of the Patuxent River. This non-homogenous landscape presents a unique opportunity to study multifaceted approaches to management of the leading causes of decline to the overall health of the river and the bay (Breitburg et al. 2003). Since colonial days, the Patuxent has served as an anthropological base for many European settlers. Quotes from the early 1800s depict the Patuxent as a clear and thriving ecosystem with a healthy seafood market (Breitburg et al. 2003). However, the arrival of agriculture and development of human settlements along the banks and surrounding fields of the Patuxent have severely altered its ecosystem and food web ecology (Bockstael 1996). In 1994, roughly 50% of the shoreline was natural vegetation while 30% was invested in agriculture and 15% in residential use. About 5% of the river was in industrial use. The intensive use of agriculture, in companion with the clearing of over 85% of the forests surrounding the Patuxent River Basin, increased the amount of 45 nitrogen and sediment runoff to nearly 5 times pre-European settlement levels, while phosphorus increased 20 times pre-European settlement conditions (Breitburg et al. 2003). Dissolved oxygen has also reached catastrophically low levels, making it very difficult for fish and invertebrates to survive during the summer months (Breitburg et al. 2003). However, due to the concern of local citizens, the Patuxent is the focus of a major water quality and habitat restoration effort aimed at reducing the impact of fertilizers and sediment runoff (Bockstael 1996). The industrial impact on the Chesapeake Bay came into sharp focus on April 7, 2000, when almost 530,000 liters of crude and fuel oil leaked into the River at Swanson?s Creek from the Chalk Point power plant and contaminated 27.4 km of shoreline. About 64.4 km of shoreline and creeks were affected; the oil caused substantial damage to wetlands, beach shorelines, and wildlife. A comprehensive study attempted to assess the total number of terrapin-years lost in the oil spill. The estimate was based on the 122 adult and juvenile terrapins killed directly by the oil spill, the successive loss in the next generation due to these deaths, and an estimate of the loss of hatchlings made by experts working in the field. The researchers estimated that 5,244.6 terrapin-years were lost during the 2000 oil spill (Byrd et al. 2002). Another study was conducted to determine if the spill caused the uptake of polycyclic aromatic hydrocarbons (PAH) in the eggs of diamondback terrapins. After careful chemical analysis of eggs collected at different sites, it was determined that nest site itself did not present a direct correlation between degree of oil contamination and the levels of PAH in the eggs. Researchers concluded that it was more likely maternal transfer that accounted for varying levels in the eggs. It is convenient to study the localized effects on certain populations because of the abundance 46 of terrapins in the Patuxent River area and their tendency to return to the same nesting sites (Holliday et al. 2008). Cremona Farms and Trent Hall Farm Cremona was a 394.2-ha historic farm in St. Mary?s County, Maryland, along the Patuxent River. A conservation easement was established in 2001 for Cremona Farms to protect forests, wetlands, farmland, historic buildings, and wildlife habitat, including that of the bald eagle and the diamondback terrapin (Maryland Environmental Trust 2001). Our field research station was located on Cremona Farms, providing us easy access to study sites along the lower Patuxent River. Trent Hall was a farm similar to Cremona that was located a few kilometers north and also bordering the Patuxent River (Figure 14). Figure 14. Location of study areas, Cremona and Trent Hall Farms. Satellite image from Google Earth. 47 Both farms were located a few kilometers downstream from PEPCO?s (now Mirant Corporation) Chalk Point Power Plant. In response to the oil spill of April 2000, restoration plans were developed for Washington Creek, a tributary of the Patuxent located just south of Chalk Point. Restoration and enhancement of beach shoreline was the primary method implemented to help increase terrapin nesting habitat. The shoreline was restored with increased beach sand and a gradual slope was developed from the shoreline to ?high beach? areas to make it easier for terrapins to find nest sites. Beach grasses were planted to help stabilize the area from erosion. Our study site was located on one such enhanced shoreline located adjacent to Washington Creek (Figure 15). Figure 15. Restored shoreline at Trent Hall Farm beach with breakwater addition providing more openings for terrapins to reach the beach and nest. (Team Saving Testudo Research Collection) 48 Jefferson Patterson Park and Museum Jefferson Patterson Park and Museum was a large, 226.6-ha tract of land along the Patuxent River and St. Leonard Creek in Calvert County, Maryland. Mary Patterson, the wife of Jefferson Patterson, donated the property to the state of Maryland in 1983, and it was then quickly turned into a park. Today, Jefferson Patterson Park and Museum offers various special programs throughout the season including heritage celebrations, children's activities, tours, concerts, dances, lectures, and educational programs (Jefferson Patterson -History?[date unknown]). Jefferson Patterson Park houses a museum that studies the changing cultures and environment of the Chesapeake Bay region over the past 12,000 years. It also has an archeology research laboratory that has curated over 4.5 million artifacts from all over the state of Maryland, dealing with such topics as Native American life, archaeology, history, agriculture, historic agriculture, historic architecture, and the identification and conservation of artifacts. Along with the museums, Jefferson Patterson Park also has a mission to preserve and study the environment. This has led to the park?s introduction of various educational opportunities for children to learn about the Chesapeake Bay ecosystem (Jefferson Patterson -Visitor?[date unknown]). Currently, the park emphasizes the issue of erosion control. Some techniques, like groins and bulkheads, often have unintended negative consequences for shorelines. More recent developments such as ?living shorelines? are better for the environment, intended to create or restore coastal wetlands and beach strand habitats (Living Shorelines?[date unknown]). These efforts have made Jefferson Patterson Park a safer area for 49 diamondback terrapins to nest. The abundance of diamondback terrapins, along with the park?s pursuit of knowledge, made it an ideal place to test the effects of electric fencing on excluding mammalian predators. Patuxent Wildlife Research Center The Patuxent Wildlife Research Center was originally known as the Patuxent Research Refuge and its inception was due mainly to the efforts of President Franklin D. Roosevelt. On December 16, 1936, Roosevelt signed an executive order which gave the Department of Agriculture 1080.5 ha of land located in both Anne Arundel County and Prince George?s County in Maryland. This land was designated as a wildlife experiment and research refuge, and it eventually became known as the Patuxent Research Refuge. The area was ultimately renamed the Patuxent Wildlife Research Center in 1956 (Perry 2004). Since then the size of the Patuxent Wildlife Research Center has grown substantially, providing more land for wilderness area and for the construction of state- of-the-art research facilities. Through tenures of various directors, the research center has remained focused on its goal of using wildlife research to gain a better understanding of wildlife. Over the years, scientists have studied many species, including bald eagles, box turtles, black rat snakes, red-shouldered hawks, several species of ducks, and most notably, the whooping crane (Perry 2004). Some of the research dates back to the 1930s and continues today. The center, located near the University of Maryland, provided us an excellent opportunity to conduct our research on electric fence design. Patuxent Wildlife Research Center?s long history of wildlife research was one of the many factors that interested us and persuaded us to pursue the possibility of 50 conducting our own research at this facility. We spoke with many officials from the Patuxent Wildlife Research Center, and they provided us with valuable information regarding the best way to conduct our research. Additionally, in the spring of 2008, we conducted a small experiment at this location to test possible fence modifications. Because there were no terrapins there, we used cod liver oil to attract the same predators that were present at our beach study sites. 51 Chapter 4: Methodology This study focused on the exclusion of mammalian predators from diamondback terrapin nesting beaches through the use of electric fences. We collected data on the effectiveness of electric fences in deterring mammalian predators from terrapin nests and determined whether electric fences have the potential to be used as a viable conservation technique to increase diamondback terrapin populations. The basic concept of our study focused on surrounding small sample plots of beach with electric fencing. We observed whether terrapins nested inside the fenced plots and then compared nest depredation rates inside the fences to rates for random control plots of unprotected beach. Summer 2007 For the first summer of experimentation, we limited our study to one beach on Trent Hall Farm. Experimental Design At Trent Hall Farm, we randomly established six control and six treatment plots. Each plot measured 25 m 2 (5 m x 5 m) (Figure 16). Each treatment plot was surrounded by an electric fence, whereas control plots were marked with plain wooden stakes at each corner. We determined the locations of treatment and control plots by using stratified random sampling. After gridding off the entire beach into 25 m 2 plots, we divided the beach into five sections based on distance from the water. The sections corresponded to different rows in our grid. To ensure that differences in the conditions and locations of the various beach sections did not affect our results, we established an equal number of 52 treatment and control plots in each section. Due to the shape of the available nesting habitat, some sections were able to accommodate only one control and treatment pairing while other sections could fit two pairings. To assign treatment and control plots within the grid, we numbered each 5-m x 5- m square and then used a computer program to randomly select which plots would be used as control and treatment areas. We then flipped a coin to determine whether the selected plot would be a treatment or control plot. As we assigned plots, we disqualified those that were touching the selected plot and reworked the random number generator so that it only selected from the remaining squares. As a result, we had the same number of control and treatment plots at various distances from the water, the control and treatment conditions were randomly assigned within each of these stratifications, and no treatment or control plot was ever within 5 m of other treatment or control plots (Figure 16). Figure 16. Trent Hall Farm beach schematic map of control and treatment plots, 2007. (Team Saving Testudo Research Collection) 53 Electric Fence Design Our electric fence design was based on recommendations in the literature (Boggess 1994; Phillips and Schmidt 1994; Hadidian et al. 1997). Boggess (1994) recommended a 1-wire (15.2 cm from the ground) or 2-wire (15.2 cm and 30.5 cm from the ground) fence for deterring raccoons. Phillips and Schmidt (1994) stated that a 3-wire fence (15.2 cm, 30.5 cm, and 45.7 cm from the ground) ?can repel foxes,? and Hadidian et al. (1997) recommended an electric fence for exclusion of foxes and raccoons from unwanted areas. Our basic design consisted of six single wires placed in pairs at three heights. The sets of wires were 15.2, 30.5, and 45.7 cm above the ground. For each pair of wires, there was one wire which was electrified and another wire that served as a ground wire. These wires were only 2.54 cm apart. A ground wire was necessary with each ?hot? wire because dry sand does not conduct electricity well (see Electric Fences section in the Literature Review). In the pairs of wires at 15.2 and 30.5 cm off the ground, we electrified the lower wire in each pair, but in the highest pair, we electrified the top wire. This arrangement may have maximized the chances of a fox or raccoon touching the electrified wire. Fence posts were necessary at each corner and in the center of each side of the plots to hold the wires taut and separate. The following fence equipment was purchased from Premier Sheep Supplies Ltd. in Washington, Iowa. 1 The fence posts (fiberglass rods) were 0.9 m long and had a diameter of 1 cm. Insulators were plastic snap-on units, and the electric wire was MaxiShock. The insulated wire we used was MaxiShock Double Insulated Cable. Additionally, we used a lightning diverter to protect the energizer, which was a Kube 1 Reference to company names and products does not imply endorsement of those companies or products. 54 Argus 250. The fence was powered by an EzePower 160-165 amp 9 volt battery. Our ground rod was galvanized, 0.9 m long, and 1.3 cm in diameter. We baited the wire by applying peanut butter directly onto the 15.2 cm hot wire (Figure 17). We wound a bit of extra electric wire to the fence to give the wire more width and then put the peanut butter on the fence and wrapped tinfoil loosely around it. When all of the fences were connected, the energizer delivered about 4,000 volts of electricity to each fence. Figure 17. Fence baited with peanut butter. (Team Saving Testudo Research Collection) We found the fiberglass rods were too flexible to allow the wires enough tension to stay separated without significant reinforcement. To reinforce the rods, we secured them with a combination of plastic tent stakes and rebar stakes that were 1.2 m long and 1 55 cm in diameter. The rebar stakes were necessary in areas where the ground was too soft for the tent stakes to hold. The fence posts were tied to the stakes or rebar with about 0.9 m of generic twine. The control plots were marked by 0.9-m wooden stakes that were hammered into the ground about 0.3 m deep. For set-up, we needed the following tools. First, a large sledge hammer was used to pound in the fence posts, stakes, and rebar. We used regular hammers and an electric screwdriver to attach the lightning diverter and perform various smaller jobs. Small wire cutters were used to cut and strip the wires. We cut back the vegetation with hand-held grass clippers. Lastly, the entire fence tightening was done by hand (Figure 18). Figure 18. Fence design at Trent Hall Farm beach, 2007. (Team Saving Testudo Research Collection) 56 Experimental Procedures Starting on May 28, 2007 we checked the beach daily and recorded nest and fence conditions. We generally collected data in the evening after the daily peak terrapin nesting time was over, searching the beach for new nests and noting conditions of old nests. New nests and newly depredated nests were recorded in our data notebook. We kept a record of all nests laid on the beach, but only included nests within treatment and control plots in our data analysis. The locations of nests were recorded using a hand-held GPS unit. To locate terrapin nests, we used a technique recommended by Roosenburg (Ohio University, personal communication). Roosenburg trained us to locate nests by following the females? tracks from the water (Appendix G). We walked along the edge of the water and when we found terrapin tracks, we followed them back to the nest. Slight disturbances in the sand, loose soil, or a typical terrapin ?sand angel? pattern made when the female used her plastron to tamp down the reburied sand indicated the existence of an intact nest. Depredated nests were easier to find, marked by predator tracks, a dug-out hole, and broken eggshells (Figure 19). We recorded both nest location and conditions in a small journal and kept a daily log of other pertinent observations such as weather conditions and maintenance issues with the fences. 57 Figure 19. A depredated terrapin nest. (Team Saving Testudo Research Collection) We monitored fence conditions carefully, checking the fence voltage with a small voltage meter (Horizont Six Light Tester) to ensure that voltage remained above 4,000 volts. We immediately repaired any shortages or breaks in the fence and replaced the bait on fences every few weeks. We also used hand-held grass clippers to keep the beach grass and other vegetation away from the fences. To make it easier to find nests and follow tracks, we smoothed out the sand inside and around our treatment and control plots with a garden rake. 58 Spring 2008 During the spring of 2008, we conducted research at the Patuxent Research Refuge and focused on improving the design of our electric fences and studying effectiveness of the improved design in deterring raccoons and foxes. Two sites were selected for the Patuxent study based on their close proximity to water and to edge habitat, which we believed would have a greater concentration of predators within the research site. We randomly selected one site for the treatment plot and another for the control plot. At the control site, we measured a 5-m x 5-m plot and marked the corners with wooden stakes. In the center of the plot we constructed a scent station in accordance with Travaini et al. (1996) and Lowell Adams (University of Maryland, personal communication). The station consisted of a circle of sifted soil 1 m in diameter; in the center, we placed a scent attractant (a cotton ball saturated with cod liver oil) on a wooden stake. To prepare the station, grass and other vegetation were removed from the site and the soil was sifted through a 3.2 mm mesh screen to prepare it for track impressions. Some sand was mixed with the soil and the station was covered with about 6.4 mm of the sifted earth. On the day before a survey, stations were groomed and the attractant was added. Stations were checked early the following morning and data recorded included site number, the presence of tracks (and the species leaving tracks), and any notes. This process was repeated the following night. A light boot imprint was placed at the circle boundary to indicate whether or not the station was operative or inoperative the previous night. If the boot imprint remained visible, it was assumed that the station was operative; a washed-out boot imprint resulted in an inoperative night. If 59 the station was inoperative, the soil in the circle was raked, the attractant was replaced (if necessary), and another boot imprint was placed at the station boundary. Tracks of a single species at a station on any day were recorded as one visit, regardless of the number and size of tracks of that species, and only tracks located inside the circle were recorded. After data were recorded for a particular day, any tracks present were removed by lightly raking or brushing the dirt within the circle with a whisk broom. At the treatment site, we set up one 5-m x 5-m electric fence plot. The electric fence design was similar to the design from the summer of 2007. However, instead of using the original fiberglass rods as our corner posts, we used steel corner posts (Figure 20) to increase the strength and durability of the fence. Figure 20. Fence design with steel corner T-posts, 2008. (Team Saving Testudo Research Collection) 60 The set-up of the scent station was similar to the set-up detailed in the previous paragraph. We collected data on weekends from April 18 to May 18. On Fridays, we turned on the fence in preparation for gathering data over the weekend. On Saturdays and Sundays, we traveled to the research site to check for footprints in either of the plots. It is also important to note that we checked to see if the scent stations were in good operating conditions, consistently raked the sand to ensure that any footprints would still be distinguishable, and replaced the cod liver oil to ensure that our attractant remained strong. At the end of the weekend, we removed the attractant and turned off the electric fence in order to conserve the battery. Summer 2008 During the summer of 2008, we conducted research on two beaches: Trent Hall Farm and Jefferson Patterson Park and Museum. Experimental Design At Trent Hall Farm, we set up our experiment in the same manner as we did in 2007: there were six treatment and six control plots distributed randomly within stratified sections of the beach. At Jefferson Patterson Park, we divided the experimental beach in half and had one large treatment plot and one large control plot. The treatment plot at Jefferson Patterson Park was our largest plot (Figure 21). 61 Figure 21. Schematic design used at Jefferson Patterson Park. The treatment plot is indicated in red, and the control plot is indicated in white. Image created with Google Maps. To create a coordinate grid to record the locations of nests, we used the fence line farthest from the water (the back line) as the x-axis and then measured the y-axis straight down from the back line. We designated the origin of the system as the upper left-hand corner of the fence (when facing away from the water). After we constructed the electric fence to fit to the shape of the beach, we measured the perimeter and estimated the area. From that information we created a control plot on the other side of the beach with the same area as the treatment plot. We used the same coordinate grid for the control plot. Because this coordinate grid measured over a much longer length than at Trent Hall, we recorded the distances in meters instead of centimeters. 62 Electric Fence Design The basic fence design remained the same for our second year of data collection except for the replacement of fiberglass rods with steel posts at fence corners. This design change resulted from our research during Spring 2008 at Patuxent Wildlife Research Center. At Trent Hall Farm, we used one fiberglass rod between the steel corner posts to keep the wires from sagging and touching each other, and at Jefferson Patterson Park, we placed fiberglass rods 3 m to 3.5 m between steel corner posts. At the landowner?s request, we used larger electric fence warning signs in 2008 than in 2007. Other electric fence materials were the same as used in 2007. In mid-June, we added about 0.6 m of non-electrified chicken wire (a light galvanized wire netting of hexagonal mesh) to the top of the fence at Jefferson Patterson Park in an effort to reduce fox penetration of the fence (Figure 22). 63 Figure 22. Addition of chicken wire mesh to fence design at Jefferson Patterson Park and Museum, 2008. (Team Saving Testudo Research Collection) There was enough height left on the steel posts above the original electric wires to allow us to attach the chicken wire directly to these posts. Because the steel posts were spaced too far apart to hold up the chicken wire, we also had to make wooden stakes, which were 1.4 m tall, to add support between the steel posts. The chicken wire was attached to the steel posts and the wooden stakes with zip ties. We used 15.2 cm of chicken wire at the top of the fence to build an overhang by curling the chicken wire back over the outside of the fence. The chicken wire addition was positioned 2.5 to 3.8 cm above the top wire of the electric fence. 64 Experimental Procedures The procedures to check the beaches for nests and to check the condition of fences were the same for both study sites and were the same as used in 2007, with three exceptions. One, we did not use a GPS unit to mark the locations of nests; instead we relied exclusively on the coordinate systems of our grids. Two, we checked the plots at Jefferson Patterson Park every other day rather than daily because it was located too far from our field headquarters on Cremona Farm to check daily. Three, because Jefferson Patterson Park was a public park, there were a few days when the park required us to turn off the fence when special events, with large crowds of people, were held. Statistical Analysis We used two different one-tailed statistical tests to analyze our data. Firstly, because of small sample sizes, we used Fisher?s Exact Test to analyze the differences in terrapin nest depredation between control and treatment plots. Additionally, the Chi Square Goodness of Fit Test procedure (Johnson and Kuby 2005) was used to analyze the differences in terrapin nesting between control and treatment plots. Assumptions To complete the experiment, we made several necessary assumptions. First, we assumed that terrapins were nesting at the selected locations and that nesting occurred at the same time at the two locations. Second, we assumed that predators existed on the beaches and attempted to destroy terrapin nests (Burger 1977; Spencer 2002; Feinberg and Burke 2003; Butler et al. 2004; Draud et al. 2004; Burke et al. 2005). Third, we assumed that mammalian predators such as the fox and raccoon contributed to the 65 predation rate. Fourth, we assumed that terrapin nesting behavior was not affected by the electrical fence enclosures because the fences were placed high enough off the ground so that the turtles could fit underneath and continue with their normal nesting behavior. Fifth, we assumed that the fence shocked the predators enough to deter them from further efforts at accessing the nests in treatment plots. Lastly, we assumed that the predators did not dig under the fence. We considered these assumptions while designing the fence, recording data, and interpreting results. Extraneous/Confounding Variables In order to ensure accuracy of our results, we addressed several extraneous and confounding variables that could affect the successful completion of our project. A substantial amount of information concerning the actions and behaviors of terrapins and predators was unknown. Additionally, the fence was not designed to protect terrapins from all predators. For example, if the terrapin nests were preyed upon by birds or other air-borne predators, our electric fences would not be able to protect the nests. The behavior of the terrapins could have greatly affected the research. Terrapins will most likely have different reactions to different predators, affecting how and where they nest. For example, it may be easier for terrapins to notice a fox than a bird, making them more likely to nest in different locations when each animal is present. In addition, terrapins may not fear some predators as much as they fear other predators. Also, the presence of other animals that are not necessarily predators could affect where terrapins nest on beaches. It is also possible that terrapins might display different types of behavior at the two sites. Terrapins from different beaches may have different nesting tendencies and their reaction to human presence on the beaches and nesting sites may differ. For 66 example, if they are habituated to human presence, they may be more or less likely to nest than if they were not habituated to humans. There were also variables that could affect where females lay their eggs including soil type, sun and/or shade, moisture, and the depth of sand on the beach where they were nesting. The types of soil and soil temperatures, as well as simple geographical differences, could also be responsible for the terrapins making their nests in different locations. These environmental differences may account for variance in nesting locations. Lastly, there was the possibility of equipment failure and human error. With regards to the electric fences, sources of error included periodic fence shortages due to tangled wires or overgrown vegetation and battery depletion. Additionally, sources of human error include misreading the volt meter or failure to record or observe a recently laid nest. More specifically, there were instances in which we suspected a nest had been laid but were unable to locate it; if there had been nests, these oversights could have affected our overall data. Institutional Animal Care and Use Committees This project was conducted under the auspices and approval of the University of Maryland Institutional Animal Care and Use Committee permit number R-07-24 and the United States Geological Survey?s Patuxent Wildlife Research Center Animal Care and Use Committee. 67 Chapter 5: Results Effect of Electric Fences on Predators No significant differences were noted in mammalian predation rates of terrapin nests between treatment and control plots for Trent Hall Farm (Fisher?s Exact Test, P=0.55) or Jefferson Patterson Park (Fisher?s Exact Test, P=0.25) (Tables 1-2). Sample sizes were small, especially in the treatment plots at both study sites. Although no significant differences were noted in predation rates between treatment and control plots, a pattern of lower predation rates in the treatment plots for both study sites seems to suggest that electric fences may have influenced predation rates (Figure 23). Table 1. Effectiveness of electric fences in reducing diamondback terrapin nest predation by raccoons and foxes, Trent Hall Farm, lower Patuxent River, Maryland, 2007-2008. Study Plots Terrapin Nests Treatment (fenced) Control (unfenced) Not Depredated 4 nests 6 nests Depredated 1 nest 3 nests Total Nests 5 nests 9 nests Percent Depredated 20% 33.3% 68 Table 2. Effectiveness of electric fences in reducing diamondback terrapin nest predation by raccoons and foxes, Jefferson Patterson Park and Museum, lower Patuxent River, Maryland, 2008. Study Plots Terrapin Nests Treatment (fenced) Control (unfenced) Not Depredated 2 nests 3 nests Depredated 3 nests 17 nests Total Nests 5 nests 20 nests Percent Depredated 60% 85% Figure 23. Diamondback terrapin nest depredation (in percent) by mammalian predators in treatment and control plots, lower Patuxent River, Maryland, 2007-2008. 0 10 20 30 40 50 60 70 80 90 Pe r c e n t Trent Hall JPP Study Site Treatment Control 69 Effect of Electric Fences on Terrapins A significant difference was noted in the number of nests within treatment and control plots at Jefferson Patterson Park (? 2 = 9.0, df = 1, P <0.01 ) (Table 3, Figure 24). No difference was found between the number of nests within treatment and control plots at Trent Hall Farm ( ? 2 = 1.14, df = 1, P 0.25> 16-201. (a) A person who is the owner of land bounding on navigable water is entitled to any natural accretion to the person's land, to reclaim fast land lost by erosion or avulsion during the person's ownership of the land to the extent of provable existing boundaries. The person may make improvements into the water in front of the land to preserve that person's access to the navigable water or, subject to subsection (c), protect the shore of that person against erosion. After an improvement has been constructed, the improvement is the property of the owner of the land to which the improvement is attached. A right covered in this subtitle does not preclude the owner from developing any other use approved by the Board. The right to reclaim lost fast land relates only to fast land lost after January 1, 1972, and the burden of proof that the loss occurred after this date is on the owner of the land. (b) The rights of any person, as defined in this subtitle, which existed prior to July 1, 1973 in relation to natural accretion of land are deemed to have continued to be in existence subsequent to July 1, 1973 to July 1, 1978. (c)(1) Improvements to protect a person's property against erosion shall consist of nonstructural shoreline stabilization measures that preserve the natural environment, such as marsh creation, except: (i) In areas designated by Department mapping as appropriate for structural shoreline stabilization measures; and (ii) In areas where the person can demonstrate to the Department's satisfaction that such measures are not feasible, including areas of excessive erosion, areas subject to heavy tides, and areas too narrow for effective use of nonstructural shoreline stabilization measures. 143 (2)(i) Subject to subparagraph (ii) of this paragraph, in consultation with the Department of Natural Resources, the Department shall adopt regulations to implement the provisions of this subsection. (ii) Regulations adopted by the Department under subparagraph (i) of this paragraph shall include a waiver process that exempts a person from the requirements of paragraph (1) of this subsection on a demonstration to the Department's satisfaction that nonstructural shoreline stabilization measures are not feasible for the person's property. SECTION 2. AND BE IT FURTHER ENACTED, That this Act shall take effect October 1, 2008. Approved April 24, 2008. 144 Appendix G Additional Images Figure 29. Typical size and shape of terrapin egg (Team Saving Testudo Research Collection) 145 Figure 30. Terrapin tracks in ?J? shapes Image taken from Capital Gazette Newspapers website: http://www.hometownannapolis.com/nat_terrapins.html 146 Appendix H Doubling of Trent Hall Data from Summer 2007 Table 6. Effectiveness of electric fences in reducing diamondback terrapin nest predation by raccoons and foxes, Trent Hall Farm, lower Patuxent River, Maryland, 2007. Study Plots Terrapin Nests Treatment (fenced) Control (unfenced) Not Depredated 4 nests 2 nests Depredated 0 nests 3 nests P=0.119048 Table 7. Predicted effectiveness of electric fences in reducing diamondback terrapin nest predation by raccoons and foxes, Trent Hall Farm, lower Patuxent River, Maryland, with 2 years of data equivalent to 2007. Study Plots Terrapin Nests Treatment (fenced) Control (unfenced) Not Depredated 8 nests 4 nests Depredated 0 nests 6 nests p=0.0113122 147 Appendix I Additional Nest Information from Study Sites, 2007-2008 Key: Control = nest found inside control plot Treatment = nest found inside treatment plot Out = nest found outside of treatment or control plots Whole = nest was found intact Dep = nest was found depredated Nest information from Trent Hall Farm Beach, lower Patuxent River, Maryland, 2007 Nest Date Found Date Depredated Location Condition 1 5/28/07 Control Whole 2 6/6/07 6/19/07 Out Dep 3 6/6/07 6/6/07 Out Dep 4 6/6/07 6/6/07 Out Dep 5 6/6/07 Out Whole 6 6/6/07 6/21/07 Out Dep 7 6/6/07 6/6/07 Out Dep 8 6/6/07 6/6/07 Out Dep 9 6/6/07 Out Whole 10 6/6/07 Out Whole 11 6/6/07 6/6/07 Out Dep 12 6/6/07 6/6/07 Out Dep 14 6/6/07 Out Whole 15 6/7/07 Out Whole 16 6/8/07 Out Whole 17 6/8/07 6/25/07 Out Dep 18 6/9/07 Out Whole 19 6/9/07 Control Whole 20 6/14/07 6/18/07 Out Dep 21 6/14/07 Out Whole 22 6/14/07 Out Whole 23 6/17/07 Control Whole 24 6/17/07 6/19/07 Out Dep 25 6/18/07 Out Whole 26 6/19/07 6/19/07 Out Dep 148 27 6/19/07 Out Whole 28 6/19/07 Out Whole 29 6/19/07 Out Whole 30 6/19/07 6/19/07 Out Dep 31 6/24/07 Out Whole 32 6/24/07 Treatment Whole 33 6/25/07 6/25/07 Out Dep 34 6/25/07 6/25/07 Out Dep 35 6/30/07 6/30/07 Control Dep 36 7/1/07 7/1/07 Out Dep 37 7/1/07 Out Whole 38 7/1/07 Treatment Whole 39 7/2/07 Control Whole 40 7/4/07 Out Whole 41 7/5/07 Treatment Whole 42 7/7/07 Out Whole 43 7/7/07 Treatment Whole 44 7/8/07 7/8/07 Out Dep 45 7/8/07 7/8/07 Out Dep 46 7/11/07 Out Whole 47 7/15/07 Treatment Whole 48 7/18/07 Out Whole Nest information from Trent Hall Farm Beach, lower Patuxent River, Maryland, 2008 Nest Date Found Date Depredated Location Condition 1 6/1/08 Out Whole 2 6/2/08 Control Whole 3 6/8/08 6/8/08 Out Dep 4 6/10/08 Out Whole 5 6/10/08 Control Whole 6 6/11/08 Control Whole 7 6/11/08 Out Whole 8 6/23/08 Out Whole 9 7/3/08 Out Whole 10 7/3/08 Out Whole 11 7/4/08 Control Whole 12 7/8/08 7/8/08 Out Dep 13 7/8/08 Out Whole 14 7/8/08 7/8/08 Out Dep 15 7/13/08 7/13/08 Treatment Dep 149 Nest information from Jefferson Patterson Park and Museum, lower Patuxent River, Maryland, 2008 Nest Date Found Date Depredated Location Condition 1 6/1/08 Control Whole 2 6/4/08 Out Whole 3 6/8/08 Treatment Whole 4 6/12/08 Control Whole 5 6/12/08 6/12/08 Control Dep 6 6/20/08 6/21/08 Control Dep 7 6/20/08 6/20/08 Control Dep 8 6/20/08 6/20/08 Control Dep 9 6/20/08 6/20/08 Control Dep 10 6/20/08 Treatment Whole 11 6/23/08 6/23/08 Control Dep 12 6/25/08 6/25/08 Treatment Dep 13 6/27/08 Control Whole 14 6/27/08 6/27/08 Control Dep 15 6/27/08 6/27/08 Treatment Dep 16 6/27/08 6/27/08 Treatment Dep 17 6/30/08 7/3/08 Control Dep 18 7/1/08 Out Whole 19 7/1/08 7/1/08 Out Dep 20 7/3/08 7/3/08 Control Dep 21 7/7/08 7/7/08 Control Dep 22 7/12/08 7/12/08 Out Dep 23 7/12/08 7/12/08 Control Dep 24 7/12/08 7/12/08 Control Dep 25 7/12/08 7/12/08 Out Dep 26 7/12/08 7/12/08 Treatment Dep 27 7/15/08 7/15/08 Out Dep 28 7/15/08 7/15/08 Control Dep 29 7/15/08 7/15/08 Control Dep 30 7/18/08 7/18/08 Control Dep 31 7/22/08 7/22/08 Control Dep 32 7/22/08 7/22/08 Control Dep 33 7/23/08 7/23/08 Out Dep 34 7/28/08 7/28/08 Out Dep