<%BANNER%>

The Breeding Ecology of Endangered Snail Kites (Rostrhamus Sociabilis Plumbeus) on a Primary Nesting Site in Central Flo...

Permanent Link: http://ufdc.ufl.edu/UFE0045445/00001

Material Information

Title: The Breeding Ecology of Endangered Snail Kites (Rostrhamus Sociabilis Plumbeus) on a Primary Nesting Site in Central Florida, USA
Physical Description: 1 online resource (135 p.)
Language: english
Creator: Olbert, Jean M
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2013

Subjects

Subjects / Keywords: abandonment -- behavior -- breeding -- cameras -- collapse -- conservation -- dsr -- ecology -- elaphe -- endangered -- florida -- incubation -- lake -- lotor -- nestlings -- nests -- nocturnal -- obsoleta -- oryzomys -- palustris -- pomacea -- predation -- procyon -- quadrivittata -- raptor -- rostrhamus -- shaffer -- sociabilis -- toho -- tohopekaliga
Wildlife Ecology and Conservation -- Dissertations, Academic -- UF
Genre: Wildlife Ecology and Conservation thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The snail kite (Rostrhamus sociabilis plumbeus) is a federally endangered raptor whose population in Florida has recently undergone precipitous declines. The remaining population remains heavily dependent upon the Kissimmee Chain of Lakes in central-Florida for nesting, particularly Lake Tohopekaliga.  These lakes are subject to many anthropogenic influences, including water and vegetation management. Understanding what affects snail kite nest success on Lake Tohopekaliga will help sustain breeding activity on this highly managed system. I studied causes of nest failure, habitat influences on nest success, and snail kite breeding behavior in response to temperature stress on Lake Tohopekaliga during the 2010 and 2011 breeding season.  Predation was found to be the primary cause of nesting failure, with the yellow rat snake (Elaphe obsoleta quadrivittata) as the most common predator. Additionally, results indicate that habitat characteristics of snail kite nesting areas were found to have an effect on predation events with different predators influenced by different habitat variables. For some terrestrial predators nest access was affected by distance of the nest patch to the shore, water depth, and nest height. Aquatic predators were influenced by distance of the nest patch to shore, water depth, and minimum daily temperature. Finally, it was determined that kites responded to the historically cold temperatures in 2010 by delaying their breeding season until conditions allowed for them to provision their young at a consistent rate.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Jean M Olbert.
Thesis: Thesis (M.S.)--University of Florida, 2013.
Local: Adviser: Kitchens, Wiley Mirf.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2013
System ID: UFE0045445:00001

Permanent Link: http://ufdc.ufl.edu/UFE0045445/00001

Material Information

Title: The Breeding Ecology of Endangered Snail Kites (Rostrhamus Sociabilis Plumbeus) on a Primary Nesting Site in Central Florida, USA
Physical Description: 1 online resource (135 p.)
Language: english
Creator: Olbert, Jean M
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2013

Subjects

Subjects / Keywords: abandonment -- behavior -- breeding -- cameras -- collapse -- conservation -- dsr -- ecology -- elaphe -- endangered -- florida -- incubation -- lake -- lotor -- nestlings -- nests -- nocturnal -- obsoleta -- oryzomys -- palustris -- pomacea -- predation -- procyon -- quadrivittata -- raptor -- rostrhamus -- shaffer -- sociabilis -- toho -- tohopekaliga
Wildlife Ecology and Conservation -- Dissertations, Academic -- UF
Genre: Wildlife Ecology and Conservation thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The snail kite (Rostrhamus sociabilis plumbeus) is a federally endangered raptor whose population in Florida has recently undergone precipitous declines. The remaining population remains heavily dependent upon the Kissimmee Chain of Lakes in central-Florida for nesting, particularly Lake Tohopekaliga.  These lakes are subject to many anthropogenic influences, including water and vegetation management. Understanding what affects snail kite nest success on Lake Tohopekaliga will help sustain breeding activity on this highly managed system. I studied causes of nest failure, habitat influences on nest success, and snail kite breeding behavior in response to temperature stress on Lake Tohopekaliga during the 2010 and 2011 breeding season.  Predation was found to be the primary cause of nesting failure, with the yellow rat snake (Elaphe obsoleta quadrivittata) as the most common predator. Additionally, results indicate that habitat characteristics of snail kite nesting areas were found to have an effect on predation events with different predators influenced by different habitat variables. For some terrestrial predators nest access was affected by distance of the nest patch to the shore, water depth, and nest height. Aquatic predators were influenced by distance of the nest patch to shore, water depth, and minimum daily temperature. Finally, it was determined that kites responded to the historically cold temperatures in 2010 by delaying their breeding season until conditions allowed for them to provision their young at a consistent rate.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Jean M Olbert.
Thesis: Thesis (M.S.)--University of Florida, 2013.
Local: Adviser: Kitchens, Wiley Mirf.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2013
System ID: UFE0045445:00001


This item has the following downloads:


Full Text

PAGE 1

1 THE BREEDING ECOLOGY OF ENDANGERED SNAIL KITES ( Rostrhamus sociabilis plumbeus ) ON A PRIMARY NESTING SITE IN CENTRAL FLORIDA, USA By JEAN M. OLBERT A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIR EMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2013

PAGE 2

2 2013 Jean M. Olbert

PAGE 3

3 To wildlife c onservation

PAGE 4

4 ACKNOWLEDGMENTS I am incredibly grateful to my advisor and mentor, Wiley Kitchens for his unwavering support throughout this process. His remarkable wealth of knowledge about the system and his faith in my ability to succeed helped me to grow tremendously as an ecologist. I truly appreci ate his ability to be present and encouraging while trusting me with the freedom to make crucial decisions regarding the direction of my research. I would like to thank my committee members Peter Frederick and Scott Robinson for their thoughtful guidance throughout this process. Although not officially on my committee, Rob Fletcher provided a great deal of help and direction through the analyses process for which I am truly grateful. I would also like to thank James Colee in the stats department for his st atistical guidance. I am truly indebted to all of the field assistants whose hard work and dedication played a vital role in this research. A big thank you to Nick Belfry, Emily Butler, Dan Cavanaugh, Emily Evans, Megan Ford, Siria Gamez, Ashley Holmes, C arley Jennings, Amanda Lee, Kristen Linner, and Jeremy Wood. Thank you for watching endless images of snail kites and for never hesitating to follow me into murky water while dragging heavy equipment. I owe a great deal of gratitude to the United States Fish and Wildlife Service and the Aquatic Habitat Restoration and Enhancement section of the Florida Fish and Wildlife Conservation Commission for their support and funding of this research. Paul Souza, Sandra Sneckenberger, and Zach Welch all played an in tegral role in facilitating this partnership and for that I am very grateful. I would like to thank all of my friends and colleagues at the Florida Coop for their insightful discussions about snail kites and ecology. Brian Reichert, Zach Welch, Christa

PAGE 5

5 Zw eig, Chris Cattau, Natalie Williams, Lara Drizd, Melissa Desa, Ellen Robertson, and Rebecca Wilcox all helped to make this research better and to keep me sane. Kyle Pias has been the single most important person helping and encouraging me through this pro cess. Given that that his research coincided with mine we were able to spend countless hours in the field discussing our project goals and how we could better our research. He is an incredibly intelligent, hard working, caring, and capable individual whose unequivocal passion for conservation is inspiring. Finally, I thank my parents who have always encouraged me to follow my heart.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF FIGURES ................................ ................................ ................................ ........ 10 ABSTRACT ................................ ................................ ................................ ................... 13 CHAPTER 1 BACKGROUND ................................ ................................ ................................ ...... 15 Population Decline ................................ ................................ ................................ .. 15 Breeding Behav ior ................................ ................................ ................................ .. 16 Historical Information on Nesting Failure ................................ ................................ 18 Research Objectives ................................ ................................ ............................... 19 2 SOURCES OF SNAIL KITE NEST MORTALITY ON LAKE TOHOPEKALIGA, FL ................................ ................................ ................................ ........................... 21 Introduction ................................ ................................ ................................ ............. 21 Methods ................................ ................................ ................................ .................. 23 Study Area ................................ ................................ ................................ ........ 23 Field Methods ................................ ................................ ................................ ... 24 Data Recording ................................ ................................ ................................ 26 Documenting Non lethal Nest Visitors ................................ .............................. 27 Data Analysis ................................ ................................ ................................ ... 28 Results ................................ ................................ ................................ .................... 28 Camera Effects on Nestin g ................................ ................................ ............... 28 Nest Outcomes ................................ ................................ ................................ 29 Timing of Predation ................................ ................................ .......................... 31 Non lethal Nest Visitors ................................ ................................ .................... 32 Discussion ................................ ................................ ................................ .............. 32 Nest Predators ................................ ................................ ................................ 32 Predatory Partial Loss of Nest Contents ................................ .......................... 35 Timing of Predation ................................ ................................ .......................... 36 Non Predation Nest Failure ................................ ................................ .............. 37 Non Predatory Partial Loss of Nest Contents ................................ ................... 39 Non lethal Nest Visitors ................................ ................................ .................... 40 Camera Effects on Nesting ................................ ................................ ............... 42 3 RELA TIONSHIPS BETWEEN HABITAT CHARACTERISTICS AND SNAIL KITE NEST PREDATORS ON LAKE TOHOPEKALIGA, FL ................................ .. 55 Introduction ................................ ................................ ................................ ............. 55 Methods ................................ ................................ ................................ .................. 59 Study Area ................................ ................................ ................................ ........ 59

PAGE 7

7 Field Methods ................................ ................................ ................................ ... 60 Data Recording ................................ ................................ ................................ 62 Habitat Variables ................................ ................................ .............................. 64 Data Analysis ................................ ................................ ................................ ... 64 Results ................................ ................................ ................................ .................... 67 Nest Outcomes ................................ ................................ ................................ 67 Nest Predation Models ................................ ................................ ..................... 67 Predator Specific Models ................................ ................................ .................. 68 Nest Substrates ................................ ................................ ................................ 70 Discussion ................................ ................................ ................................ .............. 71 4 SNAIL KITE (ROSTRHAMUS SOCIABILIS) BREEDING BEHAVIORS IN RELATION TO TEMPERATURE STRESS ON LAKE TOHOPEKALIGA, FL ......... 94 Introduction ................................ ................................ ................................ ............. 94 Methods ................................ ................................ ................................ .................. 99 Study Area ................................ ................................ ................................ ........ 99 Field Methods ................................ ................................ ................................ ... 99 Data Recording ................................ ................................ .............................. 101 Data Analysis ................................ ................................ ................................ 103 Results ................................ ................................ ................................ .................. 105 Nesting Outcome ................................ ................................ ............................ 105 Clutch Initiation Dates ................................ ................................ .................... 106 Clutch S ize ................................ ................................ ................................ ..... 106 Number of Young Fledged ................................ ................................ ............. 106 Number of Abandoned Nests ................................ ................................ ......... 107 Number of Deserted Nests ................................ ................................ ............. 107 Provisioning Rates ................................ ................................ ......................... 108 Discussion ................................ ................................ ................................ ............ 108 5 CONCLUSIONS ................................ ................................ ................................ ... 125 LIST OF REFERENCES ................................ ................................ ............................. 127 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 135

PAGE 8

8 LIST OF TABLES Table page 2 1 Snail kite nest predators of individual eggs and nestlings recorded on Lake Toho, FL from 2010 2011. ................................ ................................ .................. 46 2 2 Snail kite nesting outcomes recorded on Lake Toho, Fl from 2010 2011. .......... 46 2 3 Source of partial egg/nestling loss of a total of 34 individual snail kite eggs and nestlings from a total of 24 nests on Lake Toho, FL from 2010 2011. ......... 47 2 4 Observed non lethal snail kite nest visitors documented by cameras during the active nest stage, and five days post failure or fledge in the 2010 and 2011 b reeding seasons on Lake Toho, FL. ................................ ........................ 48 3 1 Description of variables included in the predation models. ................................ 77 3 2 Model codes in relation to variables. ................................ ................................ .. 77 3 3 Support for models predicting snail kite nest survival for depredated and non depredated nests on Lake Toho, Florida from 2010 2011 (n=1,878 days observed). ................................ ................................ ................................ ........... 78 3 4 Model averaged parameter estimates for factors hypothesized to affect snail kite nest survival for depredated and non depredated nests on Lake Toho, FL in 2010 and 2011. ................................ ................................ .......................... 79 3 5 Support for models predicting snail kite nest survival for depredated and non depredated nests on Lake Toho, Florida in 2010 (n=592 days observ ed) ........ 80 3 6 Model averaged parameter estimates for factors hypothesized to affect snail kite nest survival for depredated and non depredated nests on Lake Toho, FL in 2010. ................................ ................................ ................................ .......... 80 3 7 Support for models predicting snail kite nest survival for depredated and non depredated nests on Lake Toho, Florida in 2011 (n=1,169 days observed). ...... 81 3 8 Model averaged parameter estimates for factors hypothesized to effect snail kite nest survival for depredated and non depre dated nests on Lake Toho, FL in 2011. ................................ ................................ ................................ .......... 81 4 1 Differences in nest survival rates by month of initiated on Lake Toho, FL 2010 and 201 1.. ................................ ................................ ................................ 112 4 2 Desertion month, sex, nestling age, and nest fate of the nests that experienced mate desertion in 2010 and 2011 on Lake Toho, FL. ................... 113

PAGE 9

9 4 3 Differences of nestling feeding rates analyzed by number of young in the nest for Lake Toho, FL 2010 and 2011. ................................ ............................ 113

PAGE 10

10 LIST OF FIGURES Figure page 1 1 Lake Tohopekaliga in Osceola County, Florida. ................................ ................. 20 2 1 Kissimmee Chain of Lakes with the relevant nesting lakes East Lake Tohopekaliga, Lake Tohopekaliga, Lake Hatchineha, Lake Ki ssimmee, and Lake Jackson outlined in red. ................................ ................................ ............. 49 2 2 Lake Tohopekaliga in Osceola County, Florida. ................................ ................. 50 2 3 Age of snail kite at time of depredation on the 58 day cycle (1 28 eggs, 29 58 nestlings) in 2010 and 2011 on Lake Toho, FL. ................................ .................. 51 2 4 Predicted survival rates and 95% confidence intervals of snail kite nests in relation to nest stage on Lake Toho, Florida in 2010 2011. ................................ 52 2 5 Time of day of predation on snail kite nest contents in 2010 and 2011 on Lake Toho, FL. ................................ ................................ ................................ ... 53 2 6 Time of year of snail kite nest predation represented by Julian date in 2010 and 2011 on Lake Toho, FL. ................................ ................................ .............. 54 3 1 Kissimmee Chain of Lakes with the relevant nesting lakes East Lake Tohopekaliga, Lake Tohopekaliga, Lake Hatchineha, Lake Kissimmee, and Lake Jackson outlined in red. ................................ ................................ ............. 82 3 2 Representation of the yearly water stage schedule for Lake Toho, FL in 2010 and 2011. ................................ ................................ ................................ ........... 83 3 3 Location of Lake Toho within the state of Florida ................................ ............. 84 3 4 Predicted survival rates with 95% confidence intervals for snail kites nests in relation to year on Lake Toho, FL in 2010 and 2011. ................................ ......... 85 3 5 Daily survival rates of depredated snail kite nests in 2010 and 2011 with 95% confidence intervals on Lake Toho, FL. ................................ .............................. 85 3 6 Daily survival rates of depredated snail kite nests partitioned by the egg and nestling stages in 2010 and 2011 with 95% confidence intervals on Lake Toho, FL. ................................ ................................ ................................ ............ 86 3 7 Average distances of nest patches/trees to shore (m) for nests with and without predator observations in 2010 and 2011. ................................ ............... 87 3 8 Average nest patch areas (ha) for nests with and without predator observations in 2010 and 2011. ................................ ................................ .......... 88

PAGE 11

11 3 9 Average distance of nest structure to the edge of the nesting patch or tree cover (m) for nests with and without predator observations in 2010 and 2011. .. 89 3 10 Average nest height above the waterline (cm) for nests with and without predator observations in 2010 and 2011. ................................ ........................... 90 3 11 Average water depths (cm) beneath nests with and without predator observations in 2010 and 2011. ................................ ................................ .......... 91 3 12 Average minimum daily temperatures recorded for nests with and without predator observations in 2010 and 2011. ................................ ........................... 92 3 13 Predicted survival rates with 95% confidence intervals for snail kites nests in relation to nesting substrate on Lake Toho, FL in 2010 and 2011. ..................... 93 4 1 Kissimmee Chain of Lakes with the relevant nesting lakes East Lake Tohopekaliga, Lake Tohopekaliga, Lake Hatchineha, Lake Kissimmee, and Lake Jackson outlin ed in red ................................ ................................ ............ 114 4 2 Daily water stage data (NGDV Feet) for Lake Toho from January to September in 2010 and 2011 on Lake Toho, FL. ................................ ............. 115 4 3 Daily minimum temperatures ( C) from January to September in 2010 and 2011 on Lake Toho, FL. ................................ ................................ ................... 116 4 4 Location of Lake Toho within the state of Florida. ................................ ............ 117 4 5 A breeding female snail kite feeding her 9 10 day old young snail meat on Lake Toho in 2011. ................................ ................................ ........................... 118 4 6 A picture of a recently banded 30 day old young with an entire extracted snail in its mouth on Lake Toho in 2011. ................................ ................................ ... 118 4 7 Breeding female returning to her nest of two 21 day old young with a non extracted snail minus the operculum. Photograph taken on Lake Toho, Fl in 2011. ................................ ................................ ................................ ................ 119 4 8 Proportions of all nests initiated by month (January September) for the 2010 (n=63) and 2011 (n=79) breeding season on Lake Toho, FL. .......................... 119 4 9 Survival probability of nests initiated each month from January to September of 2010 and 2011 on Lake Toho, Fl. ................................ ................................ 120 4 10 Average monthly clutch size from January to September of 2010 and 2011 on Lake Toho, Fl. ................................ ................................ .............................. 120 4 11 Number of individual young fledged from one, two, and three young camera monitored nests in 2010 (n=10) and 2011 (n=25) on Lake Toho, FL. ............... 121

PAGE 12

12 4 12 Estimates of the number of snails fed to individual young per daylight hour from nest ages 29 58 (or nestling age of 1 30 days old) on Lake Toho, FL in 201 0 and 2011. ................................ ................................ ................................ 122 4 13 Estimates of the number of snails fed to individual young per daylight hour from nests containing 1 3 young wit h nest ages 29 58 (or nestling age of 1 30 days old) on Lake Toho, FL in 2010 and 2011. ................................ ........... 123 4 14 Average number of snails fed to ind ividual young each day separated by number of young in the nest for 2010 and 2011 on Lake Toho, FL. ................. 124

PAGE 13

13 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science THE BREEDING ECOLOGY OF ENDANGERED SNAIL KITES ( Rostrhamus sociabilis plumbeus ) ON A PRIMARY NESTING SITE IN CENTRAL FLORID A, USA By Jean M. Olbert May 2013 Chair: Wiley Kitchens Major: Wildlife Ecology and Conservation The snail kite ( Rostrhamus sociabilis plumbeus ) is a federally endangered raptor whose population in Florida has recently undergone precipitous declines. T he remaining population remains heavily dependent upon the Kissimmee Chain of Lakes in central Florida for nesting, particularly Lake Tohopekaliga. These lakes are subject to many anthropogenic influences, including water and vegetation management. Unders tanding what affects snail kite nest success on Lake Tohopekaliga will help sustain breeding activity on this highly managed system. I studied causes of nest failure, habitat influences on nest success, and snail kite breeding behavior in response to tempe rature stress on Lake Tohopekaliga during the 2010 and 2011 breeding season. Predation was found to be the primary cause of nesting failure, with the yellow rat snake ( Elaphe obsoleta quadrivittata ) as the most common predator. Additionally, results indic ate that habitat characteristics of snail kite nesting areas were found to have an effect on predation events with different predators influenced by different habitat variables. For some terrestrial predators nest access was affected by distance of the nes t patch to the shore water depth, and nest height. Aquatic predators were influenced by distance of the nest patch to shore, water depth, and minimum daily

PAGE 14

14 temperature. Finally, it was determined that kites responded to the historically cold temperatures in 2010 by delaying their breeding season until conditions allowed for them to provision their young at a consistent rate

PAGE 15

15 CHAPTER 1 BACKGROUND Population Decline The Florida snail kite ( Rostrhamus sociabilis plumbeus ) is a critically endangered mid sized raptor that is restricted to freshwater wetlands and lakes throughout the central and south Florida (Martin et al. 2006) As an extreme dietary specialist, snail kites forage almost exclusively on freshwater apple snails ( Pomacea spp. ) (Snyder and Snyder 1969; Sykes 1987b; Rawlings et al. 2007) and are restricted to areas where apple snails are not only present but available on emergent vegetation K ite abundance, apple snail abundance, nesting substrate, and demography are all currently influenced by water, plant management activities, and habitat degradation (Darby 2006; Martin et al. 2008) The snail kite population has declined considerab ly since 1999, with preliminary population viability analyses predicting a 95% probability of extinction within 40 years (Reichert et al. 2011) Decades of landscape fragmentation and hydroscape alterations ,000 km), by more than half (Sykes et al. 1995) Currently the population is restricted to the Everglades watershed, Lake Okeechobee, Loxahatchee Slough, the Kissimmee Chain of Lakes, and the Upper St. Johns River of the central and southern peninsula. In addition to the loss of habitat and the population decline there has been a decrease in the number of nesting attempts and the number of young fledged annually (Reichert et al. 2011) Historically Water Conservation Area 3A (WCA3A) was a critical breedin g habitat for nesting snail kites. Due to extended periods of droughts and long term habitat degradation, recent reproduction within WCA3A has all but ceased, with no successful

PAGE 16

16 reproduction in 2005, 2007, 2008 or 2010. This has resulted in much of the pop ulation now heavily concentrated in and dependent upon the Kissimmee Chain of Lakes (KCOL) (Fig. 1), particularly Lake Tohopekaliga (Lake Toho). Toho accounted for 41% of all documented successful nests and 57% of all fledged throughout the state from 2005 2010 (Reichert et al. 2011). Lake Toho, is a shallow lake covering approximately 8,176 ha (2009 Remetrix bathymetry map), in northwest Osceola County, Florida (Fig. 1 1 ). Much of the lake shore has been altered to accommodate houses, docks, shoreline v egetation removal, and cattle grazing (HDR Engineering. 1989) Similar to other lakes in Florida, Lake Toho is considered a highly eutrophic lake with mixed emergent littoral vegetation covering (Welch 2004) Snail kite nesting has been documented lake wide although more commonly on the northern half of the lake. Breeding Behavior Snail kite breeding has been document during every month of the year, although, not necessarily in the same year. Most initiated nesting attem pts take place from December through July (Sykes 1987c; Sykes et al. 1995) The start of the breeding season varies temporally and regionally from year to year in relation to water levels and temperature (Sykes et al. 1995; Bennetts and Kitchens 1997) Sna il kites nest in either loose colonies or independently (Sykes et al. 1995) Kites have been documented using an array of woody and herbaceous nesting substrates when nesting statewide, including but not limited to willow ( Salix caroliniana ), cattail ( Typ ha sp.), pond apple ( Annona glabra ), bulrush ( Scirpus californicus ), maidencane ( Panicum hemitomon ), cypress trees ( Taxodium sp.) and sawgrass ( Cladium jamaicensis ) (Snyder et al. 1989). When nesting on lakes kites regularly build

PAGE 17

17 in herbaceous vegetation, principally cattail and bulrush (Rodgers 1998) with 24% of nests in bulrush and 54% of nests in cattail on Lake Tohopekaliga in 2010 and 2011. Nests are almost always built over water and are constructed out of sticks and available green vegetation (Sykes 1987c; Sykes et al. 1995) Nests can vary in structure and size and are approximately 25 58 cm in outer diameter with a height of 8 44 cm (Nicholson 1926; Sykes et al. 1995) Males are the predominant nest builders prior to initiation (Sykes et al. 1995) and may build several courtship structures before settling on one location (Dreitz et al. 2001) Snail kites have been known to lay anywhere from 1 6 eggs with an average clutch size of 2.66 or 2.92 depending on the study (Sykes 1987c; Snyder et al. 1989) Both sexes take turns incubating the eggs, adding vegetation to the nest, and provisioning the young, with the female performing the majority of the nocturnal incubation (Sykes 1987c) Snail kites are unique in that it is not uncommon for either sex to de sert its mate near the time of fledging, in some cases in order to pursue another mate for further nesting. This is especially true in years of high snail abundance (Beissinger 1987a, b; Beissinger and Snyder 1987) In addition to ambisexual mate desertio n, snail kites increase their annual productivity by renesting after failure and attempting multiple broods. Snail kites have been recorded making multiple successful breeding attempts in a year (Snyder et al. 1989) Furthermore, snail kites are a long liv ed species and have successfully produced young up to 18 years of age (Reichert et al. 2010b) This allows for extended beneficial years of breeding by experienced breeders (Forslund and Prt 1995)

PAGE 18

18 Historical Information on Nesting Failure Earlier researc h on central Florida lakes estimates collapse, abandonment, and predation to be the leading cause of snail kite nest failure (Sykes and Chandler 1974; Sykes 1987c; Snyder et al. 1989; Rodgers 1998) An 18 year study conducted on Lake Okeechobee (primarily) Lake Kissimmee, and Lake Toho by Snyder et al. (1989) reported structural collapse to be responsible for 47% (n=120) of all nest failures Additionally, they estimated apparent (n=38) and probable (n=35) desertion to be the cause of 29% of all nest fail ures. Predation, both apparent (n=13) and probable (n=27) were reported to be cause of 16% of all nest failures. A six year study conducted by Rodgers (1998) on East Lake Toho (n=59), Lake Toho (n=170), Lake Kissimmee (n=250), and Lake Okeechobee (n=417) f ound abandonment to be the leading cause of nest failure (35.2%) followed by nest collapse (n=18.3%). Twenty nests were removed from the study after they were deemed to be at risk of collapse, and were installed into nest baskets. Although the data was pr incipally recorded from wetland regions of south Florida, Sykes (1987c) determined over a nine year period that out of 82 nests, predation (44%) was the primary cause of nesting failure, along with adverse weather (22%), weak nesting substrate (16%) and ot her miscellaneous causes. Collectively Everglade rat snakes ( Elaphe obsolete rossalleni ), cottonmouths ( Agkistrodon piscivorus ), raccoons ( Procyon lotor ), boat tailed grackles ( Quiscalus major ), fish crows ( Corvus ossifragus ) great horned owls ( Bubo virgi nianus ), larval Dermestes nidum, and ants ( Crematogaster sp. ) were all recorded as predators or potential predators of kite nests throughout the state of Florida (Chandler 1974; Sykes 1987c; Snyder et al. 1989)

PAGE 19

19 More recently researchers have questioned t he accuracy of determining nest failure from nest remains (Larivire 1999) Due to the cryptic nature of predators and the aquatic nature of snail kite nests, it is difficult to determine reasons for kite nest failure. In the situations where predation had clearly occurred (i.e. broken shells present or blood and feathers collected) it may be possible to identify the predator to broad categories such as reptilian versus mammalian predators, however implicating a particular species is not possible (Larivire 1999) Additionally, when nests are found collapsed or upside down in the water there is no way to determine if the collapse was due to predation, insufficient nesting substrate, or nest abandonment. This suggests that there may be a gap in our understand ing of snail kite breeding biology with regards to aspects of nesting failure. Research Objectives Given that Toho has accounted for 41% of all successful nests range wide since 2005 (Reichert et al. 2011) it is imperative that the lakeshore is managed in a way that allows for the highest nest success possible. Currently there is a disconnect with our understanding of the true causes of snail kite nest failure in this system. Therefore, the first objective of this study is to accurately establish the reasons of nesting failure. The second objective is to determine how temporal, environmental, and habitat variables influence the risk of nest failure. The third and final objective of this study is to assess th e snail kites breeding behavior in response to the two very different years of nesting success on Lake Toho in 2010 and 2011.

PAGE 20

20 Figure 1 1 Lake Tohopekaliga in Osceola County, Florida.

PAGE 21

21 CHAPTER 2 SOURCES OF SNAIL KIT E NEST MORTALITY ON LAKE TOHOPEKA LIGA, FL Introduction Typically multiple factors can affect a breeding population of birds over the course of a season depending upon the species and its nesting behavior (i.e. ground nester, cavity nester, etc.). For conservation purposes it is essential to have an accurate understanding of what is causing nest failure in the population in order for managers to properly manage for nest success. Although potential reasons for nest failure may range anywhere from flooding to abandonment it is most often pred ation that is found to be the primary cause of nesting failure among species (Ricklefs 1969; Nilsson 1984; Ohlendorf et al. 1989; Hartley and Hunter 1998; Richardson et al. 2009) It is estimated that predation can account for 80% of nesting failure for ma ny species in different habitats and geographic locations (Ricklefs 1969; Martin 1993; Cain III et al. 2003; Hoover 2006) populations, especially when reproduction is a limiti ng factor in their recovery (Cade and Temple 1995; Ct and Sutherland 1997; Whitehead et al. 2008) One such species is the Florida snail kite ( Rostrhamus sociabilis plumbeus ), which is both critically endangered and experiencing a recent decline of >80% population growth rate attributed to reduced adult fertility (Martin et al. 2008) The snail kite is a federally endangered mid sized raptor that is restricted to freshwater wetlands and lakes throughout the central and south Florida (Martin et al. 2006) As a dietary specialist, they forage almost exclusively on freshwater apple snails ( Pomacea spp. ) (Snyder and Snyder 1969; Sykes 1987b; Rawlings et al. 2007)

PAGE 22

22 The snail kite population has declined considerably since 1999, with preliminary population viabi lity analyses predicting a 95% probability of extinction within 40 years (Reichert et al. 2011) Decades of landscape fragmentation and hydroscape alterations (Sykes et al. 1995) Currently the population is restricted to the Everglades watershed, Lake Okeechobee, Loxahatchee Slough, the Kissimmee Chain of Lakes, and the Upper St. Johns River of the central and southern peninsula. In addition to the loss of habitat and the populat ion decline there has been a decrease in the number of nesting attempts and the number of young fledged annually (Reichert et al. 2011) Historically Water Conservation Area 3A (WCA3A) was a critical breeding habitat for nesting snail kites. Due to extended periods of droughts and long term habitat degradation, recent reproduction within WCA3A has all but ceased, with no successful reproduction in 2005, 2007, 2008 or 2010. This has resulted in much of the population now heavily concentrated in and d ependent upon the Kissimmee Chain of Lakes (KCOL) (Fig. 2 1), particularly Lake Tohopekaliga (Lake Toho). Toho accounted for 41% of all documented successful nests and 57% of all fledged throughout the state from 2005 2010 (Reichert et al. 2011). This re cent shift to lake habitats also means a shift in nesting substrate, as kites nesting in lakes tend to use cattail or bulrush, as opposed to woody species used in the Everglades habitats. Earlier research estimated collapse as the number one cause nest fai lure on central Florida lakes, accounting for 47% (n=120) of all snail kite nest failures in non woody substrates (Sykes and Chandler 1974; Sykes 1987c; Snyder et al. 1989; Rodgers 1998) While predation was thought to be far less important than nest colla pse,

PAGE 23

23 determining the rate and cause of failure of snail kite nests on Lake Toho is of activity occurs in habitats previously thought to have low success rates. However, d ue to the cryptic nature of predators and the aquatic nature of snail kite nests, it is difficult to determine exact reasons for kite nest failure. Current nest monitoring procedures require technicians to search in and around all failed nests to locate a nd collect any signs of dead young and depredated egg shells (Reichert et al. 2011) ; a method since found to be ineffective and unreliable (Larivire 1999) Additionally, when nests are found collapsed or upside down in the water there is no way to determi ne if the collapse was due to the action of a heavy bodied predator, weak nesting substrate, or simply nest abandonment. Technological advances have now led to readily accessible and relatively inexpensive methods of remotely monitoring nests, allowing researchers to accurately and unambiguously determine the reasons of nesting failure with cameras. Nest cameras not only ide ntify the primary predators (Weatherhead and Blouin Demers 2004) but can record reasons for nesting failure related to nest sites; including habitat and temporal/environmental variation over the course of the breeding season (Thompson 2007; Richardson et al. 2009; Cox et al. 2012) The goal of this chapter is to better identify and understand the causes of snail kite nest failure in one of the ir primary breeding habitats in Florida, Lake Tohopekaliga. Methods Study Area In 2010 and 2011 I monitored nests o n Lake Toho, a shallow lake covering approximately 8,176 ha (2009 Remetrix bathymetry map), in northwest Osceola County,

PAGE 24

24 Florida (Fig. 2 2). Much of the lake shore has been altered to accommodate houses, docks, shoreline vegetation removal, and cattle graz ing (HDR Engineering. 1989). Similar to other lakes in Florida, Lake Toho is considered a highly eutrophic lake with (Welch 2004). Snail kite nesting has been documented lake wide although nesting occurs more commonly on the northern half of the lake. Field Methods Nest searching was performed by airboat and occurred on the Lake Toho from November to August in 2010 and 2011. Nest locations were determined through the nest sear ching protocol currently in place for the state wide snail kite population monitoring project (Reichert et al. 2011) Nests were detected when an adult was seen flushing from a potential nest site and/or when an adult was observed exhibiting a breeding beh avior (i.e. nest defense, stick carrying, copulating, flying with snail meat). At every new nest a mirror pole with measurements on the pole was used to determine nest contents, water depths, and nest heights. Nest coordinates were recorded at the nest usi ng a GPS (Garmin GPSmap 76Cx). After a full lake nest search was completed nests were randomly selected for camera placement and cameras were deployed within four days or greater from the original nest check. All cameras were set on nests with at least on e egg to minimize the chance of abandonment during the nest building phase (Richardson et al. 2009) When possible, cameras were set up on nests in the egg stage, although cameras were set on nests with young when no other nests were available. Cameras wer e set up on nests as fast as possible to minimize disturbance. Typically the entire set up took approximately 10 minutes. Cameras were placed in a variety of substrates including cattail ( Typha

PAGE 25

25 spp. ), giant bulrush ( Schoenoplectus californicus ), willow ( Sa lix caroliniana ), etc. based upon their availability. Once cameras were deployed nest checks were preformed every seven days. Airboats were used to reach the nest site location at which point the person checking the nest would wade or swim the remaining 12 to 46 meter distance through the dense vegetation to access the nest. Prior to the camera set up and for nests without cameras, nest checks were performed every 21 days. Water depths and nest heights were recorded at every nest check. Additionally, camera batteries and memory cards were changed during nest checks, and cameras were realigned with the nest if necessary. A total of 15 Reconyx HyperFire Low Glow Semi Covert Infrared cameras were used on Lake Toho throughout the study, including both the PC8 5 and PC800 models (Reconyx Inc., Holmen WI). These cameras allowed for 3.1 megapixel images to be taken in color during the day and monochrome infrared by night. Cameras were set to trigger automatically when a combination of heat and motion passed into t he field of view. The camera would then take 5 rapid fire images with only one to three seconds between images in order to ensure that the camera would fully capture any predation events. Along with the automatic trigger, cameras were set to take one pictu re every two minutes in order to ensure the capture of events such as snake predation, which may not consistently activate the automatic trigger. Anywhere from a few hundred to seventy or eighty thousand images were captured in one week per camera dependin g upon the nest stage and camera set. Cameras were mounted on handmade steel tripods that could vary in height depending on the nest height. Tripods were constructed so that the bottom half could be pounded into the sediment with the three stabilizing

PAGE 26

26 legs remaining below the water. The top half which consisted of a single steel post could slide onto the tripod base and lock into place. Cameras were set within three meters of the nest and were attached to the tripod using a basic mount that allowed for the camera to be angled downward into the nest. Any vegetative structure that thoroughly impeded the view into the nest was tied back using twine. All equipment was camouflaged using paint and the surrounding vegetation. Data Recording All images were viewed using FastPictureViewer Professional software. Data acquisition was considered to have started on the first day the camera was set and ended when the young reached fledge age or 24 hours after a predation event or abandonment occurred. The nest success per iod was determined on a 58 day nest cycle with 28 days for laying and incubation and 30 days for the nestling stage (Sykes 1987c) Given that snail kites can build several courtship structures prior to laying eggs, failures at this stage represent failures of courtship and not the nest, therefore, only initiated nests are considered for this paper (Steenhof 1987; Dreitz et al. 2001) Nests were considered successful when at least one young reached the fledge age of 30 days old (Sykes et al. 1995) This time frame was selected as it coincided with the typical age that young started leaving the nest for extended periods of time during the study. Nests were considered to be failed when all nest contents were determined to be abandoned, unviable, removed or destr oyed. A predation event was defined as a partial loss of either eggs and/or young or a complete loss of all nest contents due to a predator. Nests were determined to be abandoned once both adults stopped tending the nest during the viable incubation or you ng stage. Nest collapse was determined to be any nest which shifted position so far that the eggs/young were dislodged. Eggs were

PAGE 27

27 determined to be unviable when incubation was observed for 33 days or more and were eventually abandoned in the egg stage. Th e date of initiation (laying of the first egg) was either back calculated 28 days from the hatch date of the first young if the nest reached hatch age or back calculated depending on the number of eggs observed at the first nest check. For example, if a nest was found with 3 eggs the initiation date was estimated to be 14 days prior to the nest check. If a nest was found with 2 eggs the initiation date calculated to 4 days prior to the nest check. If a nest was found with only 1 egg the initiation date wa s estimated to be 2 days prior This method was based on previous literature regarding the laying intervals of snail kites (Sykes 1987c) When looking at the time of day that a predation event took place I determined nocturnal or "night" predation to occur when a predator removed nest contents after sunset or before sunrise. A diurnal or "day" predation event occurred after sunrise but before sunset. Occasional technical issues occurred when the camera turned away from the nest or the nest shifted out o f view before a cause of failure could be determined. Similarly this occurred in instances where the batteries died early or the memory card filled prior Documenting Non lethal Nes t Visitors All non lethal nest visits of species that were seen either directly on the nest or noted within several meters of the nest were recorded. In addition I took note of all species that were observed to scavenge remaining eggs from nests that had b een abandoned. There was no attempt to identify individuals T herefore the number of times a species was documented is not an accurate representatio n of the surrounding

PAGE 28

28 population but instead represents presence/absence. Non lethal nest visitors were documented throughout all active nesting periods and for 5 days after the nests were considered to be failed or fledged. I selected five days as it was still likely to observe fledging young around the nest during this time a nd allowed for scavengers to remove remaining eggs. When eggs remained in the nest for greater than a week post abandonment the camera was removed for reuse prior to the egg removal. Data Analysis We modeled daily survival rates (DSR) with Shaffer's logist ic exposure method (Shaffer and Burger 2004) using Proc GENMOD in SAS 9.3 (SAS Institute 1989; Rotella 2004) This method of analysis models the probability of the nest surviving between nest checks and allows for intervals that vary in length. For nests w ith cameras our nest check intervals were set at 24 hours (daily) provided the recording had not failed for the week in which case the interval was set for the following nest check. Essentially if a nest was monitored by a working camera it was considered to be nest checked daily. Additionally, daily survival rates were used to determine if there were any effects from the camera monitoring system. This was done by comparing the daily survival rates of nests monitored by cameras and nests not monitored by ca meras. Models were ranked using Akaike's information criterion (AICc) to correct for a small sample size (Burnham and Anderson 2002) Results Camera Effects on Nesting There was an average return time of 34.65 5.6 minutes to the nests following camera p lacement, including the time it took for the camera deployment. The overall daily survival rate of the 75 camera monitored nests (95% CI = 0.9717 0.9846) was not

PAGE 29

29 significantly higher than the daily survival rate of the 65 nests without cameras (0.9680 0.98 21, x = 0.41; P = 0.52). The daily survival rate of the 75 camera nests in the incubation stage (95% CI = 0.9539 0.9782) was not significantly higher than the daily survival rate of the 65 not monitored by camera (0.9492 0.9779, x = 0.03; P = 0.87). Simila rly I found the daily survival rate of the 49 camera nests that reached the nestling stage (95% CI = 0.9782 0.9220) was not higher than the daily survival rate of the 46 nests with young not monitored by camera (0.9704 0.9869, x = 1.53; P = 0.22). Nest Ou tcomes Cameras were set up on 75 of the 142 Snail Kite nests located on Lake Toho during the 2010 (n=32/63) and 2011 (n=43/79) breeding season (January October). A total of 1,751 days of data were collected and reviewed over the two years. Of the 75 moni tored nests, three of the cameras did not record the cause of nest failure due to either dead batteries or a full memory card. As a result, these nests have been removed from the following nesting outcome results. We observed a clutch size of 1 4 eggs in 2 010 with an average of 2.75 eggs per clutch (n=32). The clutch size in 2011 was 1 3 eggs with an average of 2.74 eggs per nest (n=43). Nests fledged anywhere from 1 3 young during both years of the study, with an average of 2.2 young fledged per nest in 20 10 (n=10) and an average of 1.8 young per nest in 2011 (n=25). A total of 22 young reached the fledge age of 30 days old in 2010 and a total of 47 young reached fledge age in 2011. Over the course of the study I recorded a total of 32 predation events (5 7 individual eggs or young) where there was either a partial or complete loss of nest contents (Table 2 1). The observed predator community included yellow rat snakes ( Elaphe obsoleta quadrivittata ), marsh rice rats ( Oryzomys palustris ), raccoons

PAGE 30

30 ( Procyon lotor ), American alligators ( Alligator mississippiensis ), a great horned owl ( Bubo virginianus ), fish crow ( Corvus ossifragus ), and a purple gallinule ( Porphyrio martinica ). Yellow rat snakes were the most common predator to consume both eggs and young (Fi g. 2 3). No instances of adult snail kite predation were observed. I found predation (n=21) to be the leading cause of nesting failure on Lake Toho. Other sources of nesting failure resulted from abandonment of eggs (n=10), unhatched eggs (n=3), accidental egg or young loss (n=2), and nest collapse (n=1). Of the 72 nests, 49% (n=35) of the nests were successful having fledged at least one young (Table 2 2 ). Of the 177 eggs laid over the course of the two years 60% of the eggs hatched (n=106) and of the hatc hed individuals, 63% fledged (n= 67). In all but one depredated nest, yellow rat snakes caused complete or partial loss of nest contents with one visit to the nest. Alligators, marsh rice rats, raccoons, and the great horned owl(s) were observed ret urning repeatedly to the same nests either to remove remaining nest contents or to recheck the empty nest for missed food. Partial loss of nest contents and unhatched eggs that did not directly lead to the failure of the nest were observed at 24 of the 72 nests. The recorded reasons for partial loss of nest contents were from weak or starved young (n=8), yellow rat snake predation (n=4 young), fish crow predation (n=2 eggs), and an alligator predation (n=1 young) (Table 2 3). I found a total of 10 eggs that remained unhatched and 9 eggs/young went missing for unknown reasons. Many of the unknown egg/young losses were due to the difficulty of accounting for missing eggs (i.e. possibly rolled out) although on one occasion a small young went missing suddenly fr om between images without evidence

PAGE 31

31 of what caused the loss. Additionally, some egg/young losses were missed during gaps in recording due to battery failure. Timing of Predation From the nests that were monitored by camera (n=75) I found that the DSR was s ignificantly lower in the incubation stage (95% CI = 0.9529 0.9782) than nests with young (0.9782 0.9220, x= 7.89; P = 0.005) (Fig. 2 3 2 4). This result suggests that nest success was most limited during the egg stage of the nesting cycle. Out of the 30 predation events where the predator was successfully recorded (52 individual eggs/young), seven events (23%) occurred diurnally and 23 events (77%) were nocturnal (Fig. 2 5). All 12 of the snake predation events (28 individual eggs/young) and 4 of the marsh rice rat predation events (6 individual eggs/young) occurred at night. The single predation event by the fish crow (n=2 eggs) and the single predation event by the purple gallinule (n=1 egg) both occurred during the day. Raccoons and the gr eat horned owl were observed to take young both at night and in the early morning hours of daylight. Alligators were observed taking young indiscriminately throughout the day and night. No crepuscular or "twilight" predation was observed. On Lake Toho yel low rat snake predation events occurred from late February (2/24) through early June (6/10). Raccoon predation was observed at the very end of May and throughout the month of June (5/31 6/15). Marsh rice rat predation was observed later in the breeding sea son between the months of August and October (8/2 10/10). The purple gallinule, fish crow, and great horned owl predation events occurred in March, May, and August respectively (Fig. 2 6).

PAGE 32

32 Non lethal Nest Visitors The most common species found to visit i n and around active snail kite nests were boat tailed grackles ( Quiscalus major ), red winged blackbirds ( Agelaius phoeniceus ), marsh rice rats ( Oryzomys palustrisI ), and intruding snail kites not associated with the active nest. Fish crows (n=9), marsh ric e rats (n=2), purple gallinule (n=1), and a yellow rat snake (n=1) were all observed scavenging abandoned or recently depredated eggs (Table 2 4). Discussion Nest Predators Yellow rat snakes were the primary predator observed on Lake Toho and were seen ta king both eggs and young. Typically they ate young between the ages of 1 5 days old, and usually ate all of the young or eggs present. On one occasion a 23 day On anot her occasion a yellow rat snake ate one of the young at a nest but did remove a second or third (both 1 5 days old); either from satiation or adult nest defense. In this example, however, the second young died the following day from what appeared to be eit her injury or shock. Over the course of the study I recorded marsh rice rats depredating two nests with eggs, both when the nests were left unattended for a night by the adults. Given that both adults returned the following days after the predation the f ailure was attributed to predation and not abandonment. Similarly a final nest was depredated with 3 young ranging from 2 9 days old when the female neglected to sit on the nest overnight. All 3 young were pulled over the edge of the nest by one or more ra ts (only one was seen by the camera at any given time). What exactly happened to the young once dragged

PAGE 33

33 below is unknown but the rats were certainly the cause of failure. All of the marsh rice rat predation events occurred late in the breeding season (Augu st October) and in approximately 58 126 cm of water. On a separate occasion in March of 2011 a marsh rice rat was again seen biting and pulling on three young in a nest approximately 20 days old. The rat attempted to pull one young over the edge of the nest and succeeded in getting it as far as the outer rim of the nest. The rats attempt at removing the young eventually failed possibly due to the large size of the young. Adults were not present for this event which was common for a nest with young this age as adults typically stopped tending nests during the night when the young were an average of 20 days old (20.00 0.6). Given that marsh rice rats were occasionally observed around nests throughout the breeding season either after a nest had failed or fledged suggests that they are present in many nest patc hes but do not always cause harm to the nest contents. It is possible that the rats may be acting opportunistically given that all affected nests were without adults actively present. This suggests that nests may be more vulnerable to such types of predati on with less attentive parents or where disturbance has flushed the adult during the night. American alligators were another predator of nestlings, but were only recorded taking older young from 21 to 29 days old (a total of 5 you ng from 3 nests). All alli gator depredated nests were located in water depths between 80 and 140 cm at the time of failure. Nest heights above the water surface were 5, 18, and 31 cm at the time of failure. Alligators accessed these nests typically by lunging upward to grab the nes t

PAGE 34

34 contents. Structurally nests appeared to be disheveled and tilted towards the water after such a predation. The great horned owl depredation was the only predation event to be determined without direct images of the predator. The predator was determined due to the speed by which each young was removed from the nest (within 20 seconds), the large size of the almost fledged young, and the fact that the young were removed at night and very early morning hours (Houston et al. 1998) Unfortunately each young was removed so quickly that the camera did not trigger fast enough to record the predator. Unlike most snail kite nests in the study sample this nest was located at the very top of a small Chinese tallow tree and was exposed to the dense tree li ne that surrounded it. After the predation event, great horned owl feathers where collected from below cypress trees in close proximity to the nest. A purple gallinule depredated a nest very early in the season in 2010 when temperatures were still low. T he female flushed when the bird approached and the gallinule simply walked off with the egg. Although purple gallinule predation is not unheard of for avian community they are typically opportunistic and remove eggs from unoccupied nests (Frederick and Col lopy 1989) This lack of defense may be attributed potentially from an infestation of mites or other unseen bugs or skin condition (Philips 2007) and spent little ti me actually sitting on the nest incubating the egg. This level of obvious skin irritation was never again observed in the 2010 or 2011 season. Given that the female was not banded it is unknown whether or not she survived through the season.

PAGE 35

35 Predatory Par tial Loss of Nest Contents After starvation, yellow rat snakes were recorded as the second most common reason for partial loss of eggs and nestlings at nests. Typically the snakes would eat and/or critically injure one or two of the young in a nest before leaving. Although aggressive displays by the female may have encouraged the snakes to leave before eating all of the contents in the nests it is difficult to know how much of an effect this had on the snake. We observed a fish crow removing two eggs from an active nest in 2010 when the adults left the nest unattended for a few minutes. At the point that the eggs were depredated by the fish crow the nest was technically determined to have unviable eggs as it was well beyond the hatch date. I decided to list the eggs as depredated given that both adults appeared to still be invested in the nest and spent much of their time tending and sitting on the eggs. From personal observations it is not unusual to see fish crows attempt to access snail kite nests during the season or scavenge eggs from abandoned nests, it was however unique to see them succeed at depredating an active nest on this lake in 2010 and 2011. The unhatched eggs and nest age may have influenced the opportunity for this fish crow predation. The adults may have been less attentive at the nest given that the eggs were now 47 days into the nesting cycle. Although still tending the nest, the adults had gone from leaving the nest unoccupied 0 minutes each day to approximately 1 hour over a 24 hour per iod. This decreased investment in the nest may have allowed for increased predation opportunities. As was discussed earlier, an alligator was responsible for the loss of a 23 day old young. The nestling was eaten as it perched low near the water away from the nest on

PAGE 36

36 some strands of cattail. This type of predation and the low perching away from the nest may be a greater unforeseen issue for young as they start fledging from the nest and are not yet fully competent at flight Timing of Predation There are several potential reasons why the daily survival rates of nests were significantly lower in the incubation stage than the nestling stage. The top predator, yellow rat snake s consumed 49% more eggs than young. It is possible that they were limite d by the size of young they could eat, but not by eggs. It is also possible that the nests depredated earlier in the 58 day period of the nest (i.e. incubation) may have been located in poor nest sites that were either easier to locate or in a predator hea vy areas, lending to faster predation (Martin et al. 2000). Additionally, after predation (57%), I recorded abandonment (27%) and unhatched eggs (8%) to be most common reasons of nesting failure. Given that all abandonment events and unhatched eggs occurre d in the egg stage this likely added to the limited nesting success during incubation. The timing of predation is biologically relevant given that each recorded predator is likely going to be affected by either the time of day or time of year in some way. It is also possible that the cues from the snail kites on the nest may be different depending on the time of day with regards to their activity level and movement. Great horned owls, raccoons, and marsh rice rats are all predominantly nocturnal predators (Houston et al. 1998; Whitaker and Hamilton 1998) and will most likely be a threat to snail kite nests during the night regardless, especially when the nest has been left unoccupied for the night. Yellow rat snakes were only observed depredating nests duri ng the night and may have done so in order to avoid mobbing by the snail kites and locally nesting boat

PAGE 37

37 tailed grackles and red winged blackbirds during the day (Hensley and Smith 1986; Stake and Cimprich 2003) Although I recorded alligators eating kites at random times throughout the day they too can be limited by the temperature and time of year as they stop eating when the ambient temperature drops below 16C. Non Predation Nest Failure Abandonment often began with one adult decreasing their time inves tment on the nest until they no longer returned, forcing the other adult to eventually abandon as well. Nests were abandoned at different intervals throughout the season for both years. Ultimately I do not know the exact reasons why snail kites abandon the ir nests, although it is a commonly recorded phenomena (Chandler 1974; Bennetts et al. 1988; Snyder et al. 1989; Rodgers 1998) Literature suggests that factors such as food stress, lack of water beneath the nest, cold weather events, and unhatched eggs ma y be to blame (Snyder et al. 1989) Additionally two banded females were observed to have abandoned nests in both the 2010 and 2011 breeding season (one event without a camera but was observed by another researcher) without indicating any visible concern r egarding the camera set up, suggesting perhaps that some individuals may be prone to this behavior. Abandonment was never recorded after a partial loss of nest contents or t or actual predation had occurred. Adults were considered to be invested in a nest when they were incubating, brooding, adding nesting material or feeding young. All failures from nest abandonment occurred with nests in the egg stage. Unhatched eggs that are believed to be infertile resulted in a total of 3 nest failures over the course of 2010 and 2011. I observed adults tending the failed nests up to 45, 53, and 56 days into the 58 day nesting cycle before finally deserting the nests.

PAGE 38

38 Similarly to the a bandoned active nests, either the male or female would decrease their investment at the nest initially before eventually abandoning the unviable eggs a day or two prior to the remaining mate. Accidental failures, as described earlier, were instances wh ere either an egg or a young were seen to fall out of bounds of the nest and then left to die or get eaten. During the breeding season a 5 day old young fell out of the nest, landing on flattened cattail next to the nest. Although the parents continued to shade the unviable egg in the nest and bring in snail meat to the empty nest they did not seem to notice the young on the edge who died shortly after (approximately 1 hr), most likely from heat exposure. Neither adult seemed to be aware of the location of the young. In another instance a male was seen tipping the remaining egg in the nest out and over the edge. The male was recorded sitting next to the one egg in the nest after it had either been knocked or rolled out to the edge of the nest, shortly after he was seen to knock the egg entirely over the edge while straightening sticks in the nest. Although this male seemed unaware of the precarious situation of his nest contents a different nest observed during the season showed two adults acutely becoming a ware of an unhatched egg that had rolled out of the nest onto a flat rack of cattail. By the next image the adults had managed to roll the egg back into the nest suggesting that it may be lack of individual experience or awareness that allows for such acci dental nest failures to occur, but not necessarily a trait of the overall population. The one recorded collapse failure occurred in 2011 with a nest placed 485 cm from the ground on the branches of a willow. At the time of failure the nest contained two young approximately nine and eleven days old. As the nest started to slide from its

PAGE 39

39 placement on the braches the adults continued to bring in food and sticks until the nest fell completely to the ground. I found the nest empty in the mud at the base of the tree on the following nest check. Although the nest started shifting out of place two days prior to it falling out of the tree there is no real indication by the images or the available wind data from the DBHYDRO website to suggest that extreme weather wa s to blame. Both the average wind speed and wind gusts were recorded as 4.61 mph for that day. Non Predatory Partial Loss of Nest Contents Starvation of one or two of the young from an otherwise successful nest was the leading cause of partial nest cont ent losses. Starvations were documented on Lake Toho from the beginning of March to the beginning of September. There did not seem to be a spatial or temporal pattern to these events. Typically starvation occurred when the young were anywhere from 1 5 days old in a nest with limited provisioning. It was difficult in such instances to know if the low provisioning was accidental or an intentional effort to reduce the brood size to match the capabilities of the parents and resources available. Typically the st arved young were the last hatched and thus smaller than the remaining young. The adults removed all small dead nestlings from the nest almost immediately. Young that were approximately 15 days old and older were left in the nest and were eventually flatten ed by the remaining young. Two of the older young that starved on Lake Toho were 15 and 22 days old and appeared to have died as a resulted of extended asynchronous hatching. The 15 day old hatched four days after the first young in the nest and was not la rge enough to compete for food. The 22 day old young hatched three days later and grew noticeably slower than the other two young. In this case the young did not appear to lack for provisioned snails but fell behind once the

PAGE 40

40 adults started dropping off ful l sized extracted snails instead of ripping up the snail and feeding it. We observed one starvation event where a 15 day old young died after the female deserted the nest when the two young were only 8 and 10 days old. Although it is not atypical to obser ve desertion by an adult during the nestling stage, it is highly unusual for them to do so with such small young in the nest (Beissinger and Snyder 1987) After the female deserted, the male was unable to feed the young as consistently as the pair had done together throughout the day. Additionally the young were still small enough to require shading in the afternoon on hot days and brooding during the night. Eventually the smaller young was no longer able to compete with the larger one for meat and died. T ypically the remaining unhatched eggs remained in the nest for several days/weeks after the other eggs had hatched. Eventually these eggs either rolled out or became buried in the nesting material by the general movement of the young and adults in the nes t. There did not seem to be a spatial or temporal pattern relating the nests with unviable eggs present. Non lethal Nest Visitors Previous literature has been mixed regarding the threat of predation by boat tailed grackles (Chandler 1974; Sykes 1987c; Snyd er et al. 1989) During the course of this two year study I did not observed a single predation event or apparent attempt at by a boat tailed grackle (BTGR) even though they were the most common visitor in and around the nests. On at least one occasion a l arge male BTGR was left alone for a moment with recently hatched young and although he paused to look in the nest he did not touch the young. On several occasions BTGR were observed foraging around abandoned eggs even taking the time to shift them aside wh ile look beneath for other

PAGE 41

41 prey items. It is possible that they may exhibit predatory behavior towards snail kite nests in years where food is limited or perhaps previous observations were the result of a localized behavior as was suggested by Sykes (Sykes 1987c) That being said I observed both boat tailed grackles and red winged blackbirds mobbing, chasing, and pecking adult snail kites as they entered in and exited from their shared nest patch (Beissinger 1987b) It is possible that the presence of these Icterids may be detrimental or beneficial in ways not recorded by this study. On 14 different occasions I observed intruding adult snail kites and recently fledged young entering onto foreign active nests. Several times intruding adults were observed tentatively sitting on eggs in the active nest; although typically they would just land on or near the nest and get chased off by the tending adults. The intruding young of the year observed entering nests were possibly looking to take snail meat still bei ng provisioned to the young in the nest. All intruding young were aggressively chased out by the male or the female tending the nest, often with the adult grabbing the young by the head. Snail kites were observed to remove sticks from neighboring nests aft er failure in order to add to their own structures. Purple gallinules were recorded stealing snail meat from young on occasion when the adults were no longer around. Similarly marsh rice rats would remove any remaining snail meat from the nest while fledg ing young were away. Of the recorded non lethal nest visitors at active nests I recorded four nest visitors that were considered to be potential nest predators based on their diets, a great blue heron ( Ardea herodias ), a barn owl(s) ( Tyto alba ), a barred owl ( Strix varia ), and a river otter ( Lontra canadensis ). A barn owl was observed on two separate occasions in the

PAGE 42

42 2010 breeding season. On one occasion the owl landed directly on one of the three young in the nest (22 24 days old) before flying away for u nknown reasons leaving the young intact. On another occasion a barn owl was observed landing on the flattened rack of cattail next to a nest to dismember and eat an American coot ( Fulica americana ), again leaving the soon to be fledged young (who was hidde n) unharmed. The barred owl perched on a tree limb of a cypress tree where a female was incubating her eggs for the night. Shortly before the owl landed the female flushed from the nest leaving the eggs exposed. If the nest had contained young at the time it is possible that the owl may have been a potential predator considering their diet include birds (Mazur and James 2000) I recorded the great blue heron standing on the edge of a nest in a cypress tree that contained two young (20 days old) with no dire ct interaction with the young. Finally, a river otter was recorded as it came across a low lying nest in a bulrush patch with three young ranging from 13 14 days old. With the three young trying to move further back in the nest the otter leaned into the ne st looking at the young for approximately one minute before moving on. It is unknown how these interactions may have changed, if at all, if the nest had been in a different stage. Camera Effects on Nesting The exact impacts of cameras and people visiting nests cannot be determined, but these methods were consistent with those used in the annual population study (Reichert et al. 2011). Only on one occasion did a predator (a raccoon) directly look at a camera, but even then it was only after eating all of the nest contents. There was never an indication that an avian predator used the camera pole as a perch although the poles were typically lower than the surrounding vegetation purposefully limiting the likelih ood of this. The return rate of the adults after deployment and visual response of the bird

PAGE 43

43 depended on the individual. Some of the adults would stare at the camera for a few moments, while others would ignore the camera once on the nest. I occasionally ob served kites perching on top of the camera pole before entering or exiting a nest. The effects of this behavioral change are unknown. In addition to the camera deployment I visited the camera nests every 7 days compared to the non camera nest checks whic h occurred every 21 days. It is important to note that even with this increased level of disturbance at camera nests I found no significant difference between the daily survival rates of the nests. This may suggest that increased monitoring of the nests di d not detrimentally affect the nesting outcome. That being said the scope of this study did not account for monitored nests versus completely unmonitored nest success. Conservation Implications: Throughout the study it became clear that cause of nest fai lure is difficult to determine based solely on evidence left behind at the nest. Perhaps one of the most interesting observations was the almost complete lack of failure from nest collapse even though the majority of our monitored nests (91%) occurred in h erbaceous substrates. Previous publications have stated that non woody vegetation nests collapse twice as frequently as nests in woody vegetation on the central Florida lakes (Rodgers 1998) Earlier estimations of collapse rates on the lakes were documente d as being one of the leading causes of failure, if the not the leading cause of nest failure on the lakes (Sykes and Chandler 1974; Sykes 1987c; Snyder et al. 1989; Rodgers 1998) Nest collapse was determined to be any nest which shifted in a manner that dislodged the eggs/young. Nests that were found tipped over or broken into the water

PAGE 44

44 when visited were in fact often the result of a predator once cameras were checked. Cameras also revealed that damaged or severely shifted nests still containing eggs or young are structurally rebuilt by the adults to avoid total collapse. It is likely that previous estimates of nest collapse may have been inflated due to either predator damage or lack of nest maintenance after nest failure. Surprisingly, the singular fail ure from collapse during the two year study occurred in a willow tree, not the seemingly less stable herbaceous vegetation. Although it is typical for nests in herbaceous substrates to drop as the young age there was no relationship between the nest height s of failed nests when compared to successful nests as long as the nest remains upright ( see next chapter ). Previously it was believed that either lower water levels or high wind events (>15 mph) could seriously undermine the structural integrity of herba ceous nest substrates, especially cattail (Sykes 1987c; Rodgers 1998) Although it is unclear what constituted a low water level in earlier publications, 21% (n=4) of bulrush nests and 17% (n=8) of cattail nests in our sample were located in 50 cm of water or less by their final fledge or fail date. It is possible that the low water levels of previous studies allowed for easier wide scale predator access to nests. During the course of this study I found raccoon predation to most mimic nest collapse and I fo und that raccoons significantly preferred accessing nests in shallower water ( see next chapter Personal Obs.). Over the course of the study 88% of the 1751 recorded days experienced wind gusts greater than 15 mph with the wind gust speed ranging from 15 t o 50 mph (averaging 22.47mph). On days with wind over 15 mph anywhere from 1 to 92 gusts of wind were recorded to occur each day (averaging 18 gusts per day) according the available wind data on the

PAGE 45

45 DBHYDRO website. While nest failure from collapse should certainly not be dismissed altogether, it appears that earlier estimates of nest collapse may have been overestimated, particularly for non woody substrates. With the snail kite population at risk of extinction, it is important to focus on mitigating the causes of nest failure caused by predation. In the next chapter I will address the habitat and environmental variables that play a role in snail kite nests success on Lake Toho in 2010 and 2011.

PAGE 46

46 Table 2 1. Snail kite nest predators of individua l eggs and nestlings recorded on Lake Toho, FL from 2010 2011. Predator Eggs Consumed % Eggs Consumed Nestlings Consumed % Nestlings Consumed Total % of Total Consumed Yellow Rat Snake ( Elaphe obsoleta quadrivittata ) 20 63 7 28 27 46 Marsh Rice Rat ( Oryzomys palustris ) 4 13 3 12 7 13 Common Raccoon ( Procyon lotor ) 4 13 2 8 6 11 Unknown Predator 1 3 5 20 6 11 American Alligator ( Alligator mississippiensis ) 0 0 5 20 5 9 Great Horned Owl ( Bubo virginianus ) 0 0 3 12 3 5 Fish Crow ( Corvus ossifragus ) 2 6 0 0 2 4 Purple Gallinule ( Porphyrio martinica ) 1 3 0 0 1 2 Table 2 2. Snail kite nesting outcomes recorded on Lake Toho, Fl from 2010 2011. Nest Outcome 2010 2011 Total % Total Fledged 10 25 35 49 Depredated 11 10 21 29 Abandoned 6 4 10 14 Unviable Eggs 2 1 3 4 Accidental/Negligence 2 0 2 3 Nest Collapse 0 1 1 1

PAGE 47

47 Table 2 3. Source of partial egg/nestling loss of a total of 34 individual snail kite eggs and nestlings from a total of 24 nests on Lake Toho, FL from 2010 2011. Partial Egg/Nestling Loss Eggs Nestlings Total Unhatched Eggs 10 0 10 Unknown 6 3 9 Starvation 0 8 8 Yellow Rat Snake ( Elaphe obsoleta quadrivittata ) 1 3 4 Fish Crow (Corvus ossifragus) 2 0 2 American Alligator (Alligator mississippiensis) 0 1 1

PAGE 48

48 Table 2 4. Observed non lethal snail kite nest visitors documented by cameras during the active nest stage, and five days post failure or fledge in the 2010 and 2011 breeding seasons on Lake Toho, FL. Non lethal Nest Visitors Active Post Failed Post Fledged Scavenged Nest Remains Boat tailed Grackle ( Quiscalus major ) 52 11 8 Red winged Blackbird ( Agelaius phoeniceus ) 24 Marsh Rice Rat ( Oryzomys palustris ) 21 3 7 2 Snail Kite ( Rostrhamus sociabilis ) 14 4 Purple Gallinule ( Porphyrio martinica ) 9 0 2 1 Marsh Wren ( Cistothorus palustris ) 6 13 Carolina Wren ( Thryothorus ludovicianus ) 6 Unknown Avian Sp. 5 1 Yellow Rat Snake ( Elaphe obsoleta quadrivittata ) 4 1 1 1 Tree Swallow ( Tachycineta bicolor ) 4 Limpkin (Aramus guarauna ) 3 9 Common Yellowthroat ( Geothlypis trichas ) 3 2 Treefrog ( Hyla spp. ) 3 Unknown Snake Sp. 2 1 Raccoon ( Procyon lotor ) 2 3 Barn Owl ( Tyto alba ) 2 Common Moorhen ( Gallinula chloropus ) 2 Common Moorhen or American Coot 2 Great Blue Heron ( Ardea herodias ) 1 1 Mouse spp. 1 1 Barred Owl ( Strix varia ) 1 Turtle spp. 1 Tricolored Heron ( Egretta tricolor ) 1 River Otter ( Lontra canadensis ) 1 Ribbon snake ( Thamnophis spp. ) 1 Common Grackle ( Quiscalus quiscula ) 1 Anhinga ( Anhinga anhinga ) 9 Fish Crow ( Corvus ossifragus ) 4 9 Great Egret ( Ardea alba ) 2 White Ibis ( Eudocimus albus ) 1 American Coot ( Fulica americana ) 1

PAGE 49

49 Figure 2 1. Kissimmee Chain of Lakes with the relevant nesting lakes East Lake Tohopekaliga, Lake Tohopekaliga, Lake Hatchineha, Lake Kissimmee, and Lake Jackson outlined in red.

PAGE 50

50 Figure 2 2 Lake Tohopekaliga in Osceola County, Florida.

PAGE 51

51 Figure 2 3. Age of snail kite at time of depredation on the 58 day cycle (1 28 eggs, 29 58 nestlings) in 2010 and 2011 on Lake Toho, FL.

PAGE 52

52 Figure 2 4. Predicted survival rates and 95% confidence intervals of snail kite nests in relation to nest stage on Lake Toho, Florida in 2010 2011. 0.90 0.91 0.92 0.93 0.94 0.95 0.96 0.97 0.98 0.99 1.00 Incubation Nestling Daily Survival Rate

PAGE 53

53 Figure 2 5. Time of day of predation on snail kite nest contents in 2010 and 2011 on Lake Toho, FL.

PAGE 54

54 Figure 2 6. Time of year of snail kite nest predation represented by Julian date in 2010 and 2011 on Lake Toho, FL.

PAGE 55

55 CHAPTER 3 RELATIONSHIPS BETWEE N HABITAT CHARACTERI STICS AND SNAIL KITE NEST PREDATORS ON LAKE TO HOPEKALIGA, FL Introduction considering that much of avian nest success is driven by a high frequency of nest predation (Ricklefs 1969; Nilsson 1984; Hartley and Hunter 1998) Given that nest success plays an essential role in recruitment, it is important to accurately assess nest survival estimates and to identify reasons for nesting failure (e.g. primary predators, disturbances, nest collapse). This is especially true for imperiled species where conservation of key habitat features effecting breeding productivity and survival can be crucial (Martin 1993) Commonly, apparent nest survival is estimated from the proportion of observed successful nests. This may overestimate nest survival because not all nests are found on the first day of initiation and early nest failures are easily missed (Mayfield 1975; Shaffer and Burger 2004) With more robust and flexible methods of estimating daily nest survival rate (Dinsmore et al. 2002; Rotella 2004; Shaffer and Burger 2004) we are now able to create biologically realistic models in order to determine the imp ortance of environmental and temporal variables of interest in relation to nest survival (Butler 2009) Prior to the availability of nest cameras, researchers often used the appearance of the nest and any egg or nestling remains to estimate reasons of fail ure; a method since found to be ineffective and unreliable (Larivire 1999) Technological advances have led to readily accessible and relatively inexpensive methods of remotely monitoring nests with cameras, allowing researchers to accurately and unambigu ously determine the

PAGE 56

56 proximate causes of nesting failure. Nest cameras can identify not only the primary predators, but other causes of nesting failure related to nest sites, including habitat use habitat characteristics, and temporal/environmental variatio n over the course of the breeding season (Thompson 2007; Richardson et al. 2009; Cox et al. 2012) This method is also useful for studying threatened or endangered species, where remote monitoring allows for intense data collection with little nest disturb ance. One such species is the Florida snail kite ( Rostrhamus sociabilis plumbeus ), which is both critically endangered and has recently shifted the majority of its breeding attempts into habitats previously considered as high risk for failure (Sykes 19 87a) The snail kite is a federally endangered hawk that resides on the freshwater wetlands and shallow lakes of central and south Florida (Martin et al. 2006) As a dietary specialist, snail kites forage almost exclusively on freshwater apple snails ( Poma cea spp. ) (Snyder and Snyder 1969; Sykes 1987b; Rawlings et al. 2007) and are restricted to areas where apple snails are not only present but available on emergent vegetation. K ite abundance, apple snail abundance, nesting substrate, and demography are all influenced by water, plant management activities, and habitat degradation (Darby 2006; Martin et al. 2008) The snail kite population has declined considerably since 1999, wi th preliminary population viability analyses predicting a 95% probability of extinction within 40 years (Reichert et al. 2011) Decades of landscape fragmentation and hydroscape alterations e than half (Sykes et al. 1995) Currently the population is restricted to the Everglades watershed, Lake Okeechobee, Loxahatchee Slough, the Kissimmee Chain of Lakes, and the Upper St. Johns River of the central and southern peninsula. In addition to the loss of habitat and

PAGE 57

57 the population decline there has been a decrease in the number of nesting attempts and the number of young fledged annually (Reichert et al. 2011) Historically Water Conservation Area 3A (WCA3A) was a critical breeding habitat for nes ting snail kites. Due to extended periods of droughts and long term habitat degradation, recent reproduction within WCA3A has all but ceased, with no successful reproduction in 2005, 2007, 2008 or 2010. This has resulted in much of the population now heavi ly concentrated in and dependent upon the Kissimmee Chain of Lakes (KCOL) (Fig. 3 1), particularly Lake Tohopekaliga (Lake Toho). Toho accounted for 41% of all documented successful nests and 57% of all fledged throughout the state from 2005 2010 (Reichert et al. 2011). Males typically build the nests and using an array of woody and herbaceous nesting substrates, including willow ( Salix caroliniana ), cattail ( Typha sp.), pond apple ( Annona glabra ), bulrush ( Scirpus californicus ), maidencane ( Panicum hemi tomon ), cypress trees ( Taxodium sp.) and sawgrass ( Cladium jamaicensis ) (Snyder et al. 1989). Snail kite nests in the Everglades typically occur in woody substrates, but in lake habitats kites regularly build in herbaceous vegetation, principally cattail a nd bulrush (Rodgers 1998) The difference in substrate use between the two habitat types is thought to be the result of the shallow water depths(high predation risk) typically associated with woody substrates on the shoreline of lakes (Sykes 1987c) Cattail and bulrush patches can be found anywhere within the expansive littoral zones of the shallow lakes in central and south Florida, ranging from several centimeters to more than a meter in depth, depending on the proximity of the patch to the shore an d the time of year. Lake water levels typically decrease from winter to

PAGE 58

58 summer months, getting shallower throughout the snail kite breeding season (Fig. 3 2). This is due to water management and scheduled draw downs for flood control, agricultural and muni cipal uses, recreation, and to supply water to the greater Everglades (Sykes et al. 1995) Nests built in herbaceous substrates often shift and drop in height over time with wind, rain, and the weight of the young (Sykes et al. 1995, see previous chapter ) Similarly, water depths beneath the nest may change rapidly depending on precipitation or water management schedules. When nesting in cattail or bulrush snail kites fold over multiple strands of vegetation prior to adding sticks and available green veget ation to form the structure of their nest. Nests are approximately 25 58 cm in outer diameter with a height of 8 44 cm and are almost always built over water (Sykes 1987c; Sykes et al. 1995) Males may build several courtship structures before settling on one location (Dreitz et al. 2001) Earlier research suggested nest collapse, abandonment, and predation were the primary reasons for snail kite nest failures range wide, but nest collapse was thought to be especially high in lake habitats where nests were in herbaceous substrates (Sykes and Chandler 1974; Sykes 1987c; Rodgers 1998) However, recent research on Lake Toho ( see previous chapter ) found predation to be the primary cause of nesting failure, with almost no instances of nest collapse. Predators we re identified as yellow rat snakes ( Elaphe obsoleta quadrivittata ), marsh rice rats ( Oryzomys palustris ), raccoons ( Procyon lotor ), American alligators ( Alligator mississippiensis ), great horned owls ( Bubo virginianus ), fish crows ( Corvus ossifragus ), and purple gallinules ( Porphyrio

PAGE 59

59 martinica ). Yellow rat snakes were found to be the most common predator in the system and were observed eating both eggs and young. Since predation was identified as the leading cause of nest failure on Lake Toho, efforts to re duce predator access in this critical breeding habitat are important to snail kite population recovery. I hypothesize that habitat characteristics of nest patches will affect nest predation. For example, terrestrial based predators should be affected by va riables like water depth, nest substrate, patch area, distance of the nest to the edge of the patch, or the distance of the nest patch to the shore (Picman et al. 1993; Jobin and Picman 1997; Weatherhead and Blouin Demers 2004; Albrecht et al. 2006) Avian predators will be primarily affected by nest patch area (Bowman and Harris 1980; Martin 1993; Liebezeit and George 2002) Aquatic predators, on the other hand, should be affected by variables like nest height, patch area, and minimum daily temperat ure (Lance 2003) By identifying the relevant landscape and habitat patch factors affecting snail kite nest success in lake habitats, we can provide managers with effective conservation plans to improve reproductive efforts for this critically endangered species. Methods Study Area In 2010 and 2011 I monitored nests on Lake Toho, a shallow lake covering approximately 8,176 ha (2009 Remetrix bathymetry map), in northwest Osceola County, Florida (Fig. 3 3). Much of the lake shore has been altered to accom modate houses, docks, shoreline vegetation removal, and cattle grazing (HDR Engineering. 1989) Similar to other lakes in Florida, Lake Toho is considered a highly eutrophic lake with

PAGE 60

60 mixed emergent littoral vegetation covering approximately 25% of the lak (Welch 2004) Snail kite nesting has been documented lake wide although more commonly on the northern half of the lake. Field Methods Nest searching was conducted every 21 days by airboat from November to August of 2010 and 2011 on Lake Toho. I u sed the nest searching protocol currently in place for the state wide snail kite population monitoring project (Reichert et al. 2011) to determine nest locations. Nests are typically detected when a snail kite flushes from dense vegetation and exhibits a b reeding behavior (i.e. nest defense, stick carrying, copulating, and flying with snail meat) or a general unwillingness to leave the area. At every new nest a mirror pole with measurements on the pole was used to determine nest contents, water depths, and nest heights. Nest height measurements were recorded from the top of the water to the bottom of the nest, if the nest was tilted or at an angle I measured to the lowest corner of the nest. Three water depth measurements were taken under and next to the nes t from the ground to the waterline. The three measurements were then averaged together to get one primary water depth in an attempt to account for uneven lake sediment. Nest coordinates were recorded at the nest using a GPS (Garmin GPSmap 76Cx) and the pri mary nesting substrate under and around the nest was noted. After a full lake nest search was completed nests were randomly selected for camera placement and cameras were deployed within four days or greater from the original nest check. Cameras were set on nests with at least one egg to minimize the chance of abandonment during the nest building phase (Richardson et al. 2009) When possible, cameras were set up on nests in the egg stage, although cameras were set on

PAGE 61

61 nests with young when no other nests we re available. In an attempt to minimize disturbance I aimed to set up cameras as fast as possible and were typically back on the boat and idling away from the nest site after 10 minutes. Cameras were placed in a variety of substrates including cattail ( Typ ha spp. ), giant bulrush ( Schoenoplectus californicus ), willow ( Salix caroliniana ), etc. based upon their availability. We attached cameras to tree limbs near the nest when available (n=2) or mounted them on handmade steel tripods that could vary in height depending on the height of the nest (n=73). The tripods were constructed so that the bottom half (always 1.82 meters) could be pounded into the sediment with the three stabilizing legs remaining below the water. The top half which consisted of a single st eel post of varying height could slide onto the tripod base and was locked into place using a hitch pin. Cameras were set within three meters of the nest and were attached to the tripod using a basic mount that allowed for the camera to be angled downward into the nest. Any vegetative structure that thoroughly impeded the view into the nest was tied back using twine. Before deployment the tripods and mounts were spray painted to match the surrounding vegetation. Once the cameras were set up the surrounding vegetation was used, when available, to camouflage the cameras in the nest patch. Once there was a camera present, nest checks were preformed weekly. Airboats were used to reach the nest site location at which point the person nest checking would wade or swim the remaining 12 to 46 meter distance through the dense vegetation to access the nest. Prior to the camera set up and for nests without cameras, nest checks were performed every 21 days. Water depths and nest heights measurements were taken at every nest check. Additionally, camera batteries and memory cards (28 GB)

PAGE 62

62 were changed during nest checks, and cameras were realigned with the nest if necessary. Over the course of the study 15 Reconyx HyperFire Low Glow Semi Covert Infrared cameras were depl oyed on Lake Toho, including both the PC85 and PC800 models (Reconyx Inc., Holmen WI). The cameras recorded 3.1 megapixel images in color during the day and monochrome infrared by night. The cameras were set to trigger in response to heat and motion passi ng in front of the field of view. Once triggered the camera would take five rapid fire images with a delay of one to three seconds between images in order to ensure that the camera would fully capture any predation events. In addition to the automatic trig ger, cameras were set to take one picture every two minutes all day long to ensure the capture of events such as snake predation, which may not consistently activate the automatic trigger. Anywhere from a few hundred to eighty thousand images were recorded in one week per camera depending upon the nest stage and camera set. Data Recording We used FastPictureViewer Professional software to view all of the images. Data entry was considered to have started on the first day the camera was set and ended when th e young reached fledge age or 24 hours after a predation event or abandonment occurred. The nest success period was based on a 58 day nest cycle with 28 days for laying and incubation and 30 days for the nestling stage (Sykes 1987c) Given that snail kite s can build several courtship structures prior to laying eggs, failures at this stage represent failures of courtship and not the nest, therefore, only initiated nests are considered for this paper (Steenhof 1987; Dreitz et al. 2001) Nests were considered successful when at least one young reached the fledge age of 30 days

PAGE 63

63 old. This timeframe was selected as it coincided with the typical age that young start leaving the nest for extended periods of time. Nests were considered failed when all nest contents were determined to be abandoned, unviable, removed or destroyed. The nest age was determined by back calculating date of initiation by 28 days prior to the hatch date of the first young if the nest reached hatch age. If eggs did not hatch then the initiat ion date was back calculated depending on the number of eggs observed at the first nest check. If a nest was found with 3 eggs the initiation date was estimated to be a median of the 28 day incubation period, 14 days prior to the nest check. If a nest was found with 2 eggs the initiation date calculated to 4 days prior to the nest check. If a nest was found with only 1 egg the initiation date was estimated to be 2 days prior. This method was based on previous literature regarding the laying intervals of sna il kites (Sykes 1987c) Failure from abandonment, unviable eggs, collapse, and accidental deaths were onment to have taken place once both adults stopped tending the nest during the viable incubation or young stage. Any nest which shifted in a manner that dislodged the eggs/young was considered to be collapsed. Eggs were determined to be unviable when incu bation was observed for 33 days or more and were eventually abandoned in the egg stage. Technical issues would occasionally occur when the nest would shift out of view or the camera would turn away from the nest prior to failure. Similarly there were time s where the batteries died or the SD card filled up prior to the nest check. The outcome of

PAGE 64

64 Habitat Variables Feature Analyst software was used to delineate the herbaceous nest patches on Lake Toho using 4 ban d aerial imagery (0.2 meter pixel resolution) collected on June 7 th 2010 and June 6 th 2011. Once the nest patches were outlined, ArcGIS 9.3 was used to calculate the area of each patch, the distance of nest to the nearest edge of the patch, as well as th e distance of the nest patch to the nearest vegetated shoreline (including islands). Daily water depth measurements under the nest were calculated by taking weekly nest check water depths and adding or subtracting from the depth using lake stage data avail able for that date on DBHYDRO. Additionally, average wind speed and maximum wind gust data was obtained from the South Florida Water Management District database (DBHYDRO) ; station = TOHOW_H. In order to estimate a daily nest height I compensated for any r ecorded drop in height by applying the difference of the weekly measurements linearly across the week. Data Analysis I modeled daily survival rates with Shaffer's logistic exposure method (Shaffer and Burger 2004) fitted using Proc GENMOD in SAS 9.3 (SAS I nstitute 1989; Rotella 2004) This method of analysis models the probability of the nest surviving between nest checks (everyday for camera monitored nests) and allows for intervals that vary in length. A binomial response distribution was used to fit the models, 1 if the interval nest fate was successful or 0 if the nest had failed during the interval. For nests with cameras our intervals were set at 24 hours provided the recording had not failed for the week in which case the interval was set for the foll owing nest check. Daily survival rates were used to evaluate support of models for all nests in relation to our a priori hypotheses and predictions regarding environmental variables, temporal variables, and

PAGE 65

65 vegetative variables over the course of the study Similarly this approach was used to determine if there were any effects from the camera monitoring system. Models that comprised the top 90% of the total weight were considered to be well supported (Burnham and Anderson 2002) When evaluating models for nest predation I compared all nests that experienced predation reasons as well as successful nests. I distinguished between the response variables using 1 for depredated and 0 for other (Cox et al. 2012) When comparing predation rates to the surrounding habitat variables I included nests that had experienced even partial predation nest loss over the tw o years (n=3) and excluded any nests where an unknown egg/young loss occurred (n=5) with a total of 1,611 observed days. When including both years of data I evaluated support for 23 models describing daily nest survival rates that included main effects mod els of year, nest age, minimum daily temperatures (C), nest height (cm), nest patch area (ha), distance of nest patch to shore (m), distance of nest to edge of patch (m), and water depth (cm) (Tables 3 1 3 2). I predicted that predation would increase wi th increased nest age, and cooler temperatures. Additionally I expected predation risk and nest access to diminish with decreased nest height, nest patch area, water depth, and distance of nest to edge of patch and patch to shore. Additive effects of nest height and distance to shore as well as nest height and water depth were considered. For the predation models breeding years 2010 and 2011 were run independently to evaluate support for 13 models in each year. When both years of data were combined I contro lled for the significant differences in daily survival rates

PAGE 66

66 between the years by including year as an effect in all of the models. Models were ranked using Akaike's information criterion (AIC c ) to correct for a small sample size. Model support was assesse c and Akaike weights ( w i ) (Burnham and Anderson 2002) Given that nesting locations on the lake differed by the habitat variables accounted for (i.e. distance from shore, nest substrate, water depth, etc.) I did not test c > 2 were considered to be equally pla usible, best fit models. I model averaged estimates for parameters including in the c > 2) and compared their effects to zero using 95% CI. Relationships between the most frequent predators (yellow rat snakes, marsh rice rats, a nd raccoons) and habitat variables including nest patch area (ha), distance of nest patch to shore (m), distance of nest to edge of patch (m), nest height (cm), water depth (cm), and minimum temperature (C) were tested using generalized linear models (GLM Presence (1) or absence (0) of the predators was used as the response variable, a normal distribution, and a identity link. Nest patch area, distance of nest to edge of the patch, and distance of nest patch to the shore were all log transformed for no rmality. Due to the log transformation the results from these models have asymmetric depredating partial or complete nest contents or were recorded entering active or recently fled ged nests (up to 5 days post fledge) without causing predation. Non predation nest visits were included to increase the predator sample size and to address predator access into nests in addition to predation. Given the small sample sizes for predator obser vations we pooled data across 2010 and 2011 and did not evaluate for

PAGE 67

67 year effect. We only included predator observations of n=5 and greater so that single predator events were not included given the small sample size. We analyzed the herbaceous and woody nesting substrates using generalized characteristics. For herbaceous nest substrates I compared the size of nest patches and the distance of the nest to the edge of patch. Dista nce of nest site to shore was evaluated for nests in both woody and herbaceous vegetation. When comparing percentages of predation events in each nesting substrates only predation events were included. Results Nest Outcomes Cameras were set up on 75 of the 142 Snail Kite nests located on Lake Toho during the 2010 (n=32/63) and 2011 (n=43/79) breeding season (January October). A total of 1,751 days of data were collected and reviewed over the two years. Predation was found to be the leading cause of nestin g failure (n=21) ( see previous chapter ). Of the 75 nests with cameras, eight were built in trees, 19 were built in bulrush, and 48 were built in cattail. Tree species observed during the study include cypress ( Taxodium sp. ), willow ( Salix caroliniana ), and a Chinese tallow ( Triadica sebifera ). Nest Predation Models Of the three nest predation model sets (2010, 2011, and both years combined) only the 2010 model set had covariates with significant results. The top model for the combined 2010 and 2011 nest p redation models included the covariates of minimum temperature, year, and a categorical variable indicating if predation had occurred (1) or had not occurred (0). The weight for this model was low at 0.359 and the confidence

PAGE 68

68 intervals spanned zero (Tables 3 3 3 4). In the 2010 predation model set the effects of minimum temperature and predation/non predation best explained daily survival rates for the depredated nests with an AIC c weight of 0.94 (Tables 3 5 3 6). Given that the AIC c weight was greater than 0.90 for the top model, averaging was not performed (Burnham and Anderson 2002) The top model for the 2011 predation model set included nest height, distance of nest patch to shore, and predation/non predation. This model was not s ignificant as it was weighted 0.498 with a confidence interval that spanned zero (Tables 3 7 3 8). Although daily survival rates were significantly lower in 2010 than 2011 ( 2 0.05 =7.65, df=1893, p =0.01) (Fig. 3 4), I did not find a significant difference in the daily survival rates for depredated nests in 2010 when compared to depredated nests in 2011 ( 2 0.05 =0.03, df=1607, p =0.86). Depredated nests in 2010 had a daily survival rate of 0.83 0.1 while dep redated nests in 2011 were 0.91 0.0 (Fig. 3 5). Of the 29 nests depredated over the combined 2010 and 2011 breeding seasons there was not a significantly different daily survival rate of nests in the egg stage (95% CI = 0.9229 0.9757) compared with those in the nestling stage (0.8891 0.9611, X = 1.11; P = 0.2926) (Fig. 3 6). Predator Specific Models There was a significant difference in the distance of the nest patch/tree to the 2 0.05 =7.00, df=66, p 2 0.05 =12.31, df=66, p =0.001). Marsh rice rats were found to be present in nests with an average distance of m (71.38 285.48 m) from shore but were absent in nest sites approximately m (25.11 12.61 m)

PAGE 69

69 from shore. Alternatively, yellow rat snakes were observed to be an average m (2.65 15.65 m) from shore while nest sites m (13.68 89.18 m) from shore did not have yellow rat snakes. The distance of the nest patch or tree to the nearest 2 0.05 =0.62, df=66, p =0.43) (Fig. 3 7) We did not find a difference of marsh rice rats ( 2 0.05 =1.84, df=66, p =0.18), raccoons ( 2 0.05 =0.73, df=66, p =0.39), or yellow rat snake presence ( 2 0.05 =0.50, df=66, p =0.48) in relation to nest patch area (Fig. 3 8). Similarly there was no diff erence of presence or absence of marsh rice rats ( 2 0.05 =0.55, df=66, p =0.46), raccoons ( 2 0.05 =0.19, df=66, p =0.66), or yellow rat snakes ( 2 0.05 =1.81, df=66, p =0.18) in relation to the distance of the nest to the edge of the cattail patches, bulrush patches, or tree (Fig. 3 9). There was a significant difference in nest height for the nests that were accessed by raccoons when compared to the nests without racco ons ( 2 0.05 =6.5, df=66, p =0.01). Raccoons were found to enter nests at an average of 337.22 49.3 cm while the nests that did not experience raccoon intrusion were an average of 206.58 13.9 cm tall. There was no difference betwe en the presence or absence of marsh rice rats ( 2 0.05 =0.89, df=66, p =0.35) or yellow rat snakes ( 2 0.05 =0.05, df=66, p =0.83) with regard to nest height access (Fig. 3 10). Marsh rice rats were observed in nests with significantly deeper water underneath with approximately 115.74 11.3 cm compared to nests where rats were absent with 78.52 4.4 cm of water ( 2 0.05 =9.39, df=66, p =0.002). Raccoons were observed accessing significantly shallower nests with an average of 18.79 14.7 cm of water

PAGE 70

70 rather than nests placed above 86.11 4.1 cm of water ( 2 0.05 =19.36, df=66, p =0.<0.001). Water depth beneath the n est did not affect the likelihood of yellow rat snakes ( 2 0.05 =1.71, df=66, p =0.19) accessing a nest (Fig. 3 11). When taking the minimum daily temperature into account marsh rice rats were observed entering nests during significan tly cooler temperatures (16.61 C 8.4) than 2 0.05 =4.88, df=66, p =0.03). Minimum daily temperature did not appear to affect the likelihood for raccoons ( 2 0.05 =2.43, df=66, p =0.12) or yellow rat snakes ( 2 0.05 =0.34, df=66, p =0.56) to appear at a nest location (Fig. 3 12). Nest Substrates Although bulrush had the highest daily survival rate (0.9862 0.005), there was no significant difference in the daily survival rates of nests built in bulrush (n=19, 95% CI =0.9727 0.9931), cattail (n=18, 95% CI = 0.9672 0.9846), or trees (n=8, 95% CI = 0.9274 0.9849, X = 2.96, p = 0.2274) (Fig. 3 13). Nest patch areas of bulrush patches (0.15 0.7 ha) were signif icantly smaller than nest patches of cattail (2.44 0.4 ha) ( 2 0.05 = 8.41, df=65, p =0.003). Additionally, nests were placed significantly further from the edge of the patch when nesting in cattail (8.65 1.1 m) than when nesting in 2 0.05 =5.31, df=65, p =0.02). Nesting trees were significantly closer to shore (12.38 45.6 m) than cattail nest patches (114.85 18.6 m, 2 0.05 =4.33, df=72, p =0.04), but not sign ificantly closer than bulrush patches (101.68 29.6 m, p =0.10) There was no significant difference between cattail and bulrush distances to shore (101. 68 29.6 m, 2 0.05 =0.14, df=72, p =0.71). Bulrush patches used for nesting were not placed at a significantly different distance from trees or cattail patches

PAGE 71

71 Approximately 82% of yellow rat snake predation events took place in cattail patches (n=9) with the remaining 18% of predations taking place in bulrush (n=1) and a cypress tree (n=1). Yellow rat snakes were observed in 6% of nests built in bulrush (1 out of 18 nests), 14% of nests built in trees (one out of seven nests), and 27% of nests built in cattail (13 out of 48 nests). All marsh rice rat predations occurred in bulrus h patches (n=3). Alligators attacked young 67% of the time in cattail (n=2) and 33% of the time in bulrush (n=1). Raccoons depredated nests 67% of the time in cypress trees (n=2) and 33% of the time in cattail (n=1). Both the purple gallinule and fish crow predation events occurred in cattail while the owl predation occurred in a Chinese tallow tree. Of the 75 monitored nests, three of the cameras did not record the cause of nest failure due to either dead batteries or a full memory card ( see previous chap ter ). Of the 72 known fate nests 57% of nests in trees experienced a predation event (4 nests out of 7 nests), 30% in cattail (14 out of 46 nests), and 26% in bulrush (5 out of 19 nests). Discussion Habitat characteristics of snail kite nesting areas were found to affect predation events in this study. Different predators were found to be affected by different habitat variables. For some terrestrial predators (e.g. yellow rat snakes and raccoons) nest access was influenced by the distance of the nest patch to the shore, nest height, or water depth, while aquatic predators (marsh rice rats) were affected by distance of nest patch to shore, water depth, and minimum daily temperature. Avian predators were not addressed due to the small sample size of predation events. Yellow rat snakes were the most common predator of snail kite nests ( see previous chapter ), and the majority (82%) of these predation events took place in cattail and in areas close to shore (

PAGE 72

72 m) (2.65 15.65 m). Although yellow rat sna kes are capable swimmers they are primarily terrestrial and face an array of potential predators while moving through lake littoral zones (Allen and Neill 1950) Snakes found in aquatic habitats in the southeast are readily eaten by otter, mink, raptors, w ading birds, alligators, other snakes, and species of fish such as gar, catfish, and bass (Gibbons and Dorcas 2005) Given that there was no significant difference between the distance of bulrush patches and cattail patches from shore there are likely othe r reasons driving yellow rat snakes apparent preference for cattail. One possible explanation for this might be the structural differences between the two substrates. Structurally cattail stems tend to be wider and flatter than bulrush stems which are smal ler in diameter and cylindrical. This difference could make traveling and climbing through the bulrush more difficult and thus less desirable for these predatory snakes. occurred in bulrush patches that tended to be in significantly deeper water (115.74 11.3 cm) and further from the shore ( cm) (71.38 285.48 cm). Althou gh the marsh rice rats are susceptible to predation threats similar to the yellow rat snakes while swimming, the difference may be they are aquatic year round breeders who primarily colonize over water and would not have to return to shore for breeding and foraging as a terrestrial predators such as the yellow rat snake would. Inhabiting nest patches further from shore in deeper water allows for the rats to avoid terrestrial predators, including the yellow rat snake and raccoons that might commonly forage t he shoreline of the lake and shallower vegetation. Although their diet consists of immature flies, crabs, and snails (Post 1981) they are also voracious predators of bird nests consuming

PAGE 73

73 both eggs and young of boat tailed grackles, red winged blackbirds, a nd marsh wrens (Bancroft 1986) Considering their habitat and foraging preferences it is clear why they would depredate unattended snail kite nests when given the opportunity. In addition to their habitat preferences marsh rice rats were documented enteri ng nests in significantly cooler temperatures (16.61 C 8.4 ) than when they were not observed ( 36.63 C 3.3 ). Although the exact reason for this is unknown it may have to do with their energy requirements and increased foraging during colder temperatur es. Alternatively they may be reducing their risk of alligator predation by foraging and traveling into bulrush patches in cooler temperatures as alligators reportedly stop eating when ambient temperatures drop below 16 C (Lance 2003) Given that marsh ri ce rats are nocturnal and are exceedingly capable of swimming under water for periods of time to avoid above water predation it would be beneficial for them to be most active in and near the water during temperatures that would eliminate such a notable thr eat. This is especially true since as alligator have been documented to eat aquatic rats (Whitaker and Hamilton 1998; Rice et al. 2007) Raccoons were recorded in significantly taller nests averaging 337.22 49.3 cm in height and in shallower water app roximately 18.79 14.7 cm in depth. It is most likely water made snail kite n ests located in trees especially vulnerable to their foraging. Although the raccoons seemed to prefer shallower water they did not necessarily limit the distances from shore they were willing to travel in order to access a nest or foraging site. The raccoo ns may have been less cautious about accessing nests further from

PAGE 74

74 shore in shallower water given that their size decreases the risk of wading bird or raptor predation, unlike the yellow rat snakes. That being said they may have selected to avoid deeper wat ers in an attempt to limit their interactions with alligators (Frederick and Collopy 1989; Rice et al. 2007) Furthermore the herbaceous substrates of the deeper nests may not have support the weight of a raccoon therefore making the predation risk of swim ming not worth the trip. There were too few predation events by alligators to draw m any conclusions (n=3), though there are likely influenced by tempor al or temperature effects. When ambient temperatures reach 16 C or lower, for example, their metabolism is inhibited and the assimilation of food is no longer possible (Lance (2003) Therefore, it is reasonable to assume that alligator depredation would likely increase in the latter part of the breeding season. Beyo nd nest patch characteristics, this study found an effect of minimum temperature on nest survival. The 2010 predation model, minimum daily temperature and whether or not the nest was depredated explain most of the variation in the daily survival rates. It appears as though in 2010 nest survival increased with warmer temperatures for nests that experienced not only predation but non predation failures as well. The effect of temperature on survival was most likely exacerbated in January of 2010 when the state of Florida experienced historically cold weather. Temperatures were recorded to be well below the normal averages for the state (Leftwich et al. 2010) According to DBHYDRO minimum temperatures on Lake Toho hovered around 0 C from January 4 th to January 13 th with water temperatures remaining below 10 C from January to April of 2010. Additionally temperatures remained 15 26 degrees cooler in

PAGE 75

75 late February and early March than the temperatures recorded in 2011. Considering that the low temperatures negativ ely affected the survival of both depredated and non depredated nests there may have been multiple factors at play during this time. With the first six nests of the season failing due to abandonment and predation it is likely that breeding kites may have b een plagued by decreased foraging success and in turn decreased nest investment. It is difficult to know how, if at all, such severe temperatures impacted the breeding adults for the rest of the breeding season. It has been documented that apple snails ( P omacea spp. ) commonly respond to temperatures of less than 15 C by becoming inactive and burying themselves into the sediment when temperatures reach 10C. This behavior renders many of them immobile and inaccessible to the foraging adults (Stevens et al. 2002) Snyder et al (1989) recorded several instances of entire snail kite colonies abandoning their nests after several days of cold weather. This was documented following their observations of greatly depressed capture rates of snails. Similarly rates o f abandonment were highest in the start of the 2010 breeding season when the temperatures were unusually low ( see next chapter ). In addition to inducing abandonment these findings suggest that prolonged cold temperatures may create underfed, unhealthy, and possibly less invested adults. In the previous chapter I reviewed one of the early 2010 predation events by a purple gallinule in which the female spent much of her time scratching her feathers on the edge of the nest rather than incubating the nest conte nts. This same female was flushed by a purple gallinule who then ate and removed the only egg in the nest. During the course of the 2010 and 2011 breeding seasons females were often observed posturing defensively against purple gallinules and other non rap torial birds

PAGE 76

76 been a normal response to such a predator. Conservation Implications: Given the impact of nest predation on breeding snail kites it is important that the nest predators and their relationship to nesting sites are properly identified. This study identified the primary predation threats on Lake Toho and how they relate to nest habitat characteristics, providing managers with important information about how to increase nest success in the region. Considering that yellow rat snakes were the most common predator of both eggs and nestlings managing for nest site locations further from the shore should reduce predation. Additionally, since marsh rice rats were only observed in nests without adults present it may be important to limit nocturnal disturbance (airboats, etc.) near the nests in order to prevent flushing the adults at night. Predator guards might also be used when kites nest in woody substrates close to s hore or when water levels are expected to dry out the area. This study also highlighted the danger of having such a large portion of the snail Florida. Although the ext reme cold weather in 2010 affected much of the state, snail kites remain especially vulnerable to such events while nesting at the northern extent of their range on Lake Toho. These lakes are also some of the most heavily managed and heavily recreated in t he state, increasing the potential for disturbances and nest failures where kite activity has become concentrated. In addition to managing for higher nest success on the KCOL, restoring habitat in the southern end of the state will likely provide the best long term viability for the snail kite population.

PAGE 77

77 Table 3 1. Description of variables included in the predation models. Variables Descriptions Area of Patch Area of the patch of vegetation that the nest was located in (ha) Distance to Edge Distance of the nest to the closest edge of the patch (m) Distance to Shore The shortest distance of the edge of the nest patch to the vegetation on the shoreline (m) Julian Date Continuous count of days representing the real date starting on January 1st Minimum Temp Lowest recorded temperature of the day ( C) Nest Age 1 58 days, from nest initiation to fledge Nest Height Height measured from the water to the bottom of the nest (cm) Nest Substrate Vegetative structure that the nest was built in, categorized as trees, bulrush, or cattail Precipitation Total daily rainfall (cm) Predation vs Other Depredated (1) or not (0) Water Depth Measured from the sediment to the waterline beneath the nest (cm) Year 2010 or 2011 Table 3 2. Model codes in relation to variables. Variables Model Code Area of Patch Area Distance to Edge DistEdge Distance to Shore DistShore Julian Date JDate Minimum Temperature MinTemp Nest Age NestAge Nest Height NestHeight Nest Substrate NestSub Precipitation Precip Predation vs Other PvsO Water Depth WaterDepth Year Year

PAGE 78

78 Table 3 3. Support for models predicting snail kite nest survival for depredated and non depredated nests on Lake Toho, Florida from 2010 2011 (n=1,878 days observed) Predation Models K AICc w i PvsO + MinTemp + Year 4 356.476 0.000 0.359 PvsO + DistEdge + DistShore + Year 5 359.393 2.918 0.084 PvsO + DistEdge + Year 4 359.893 3.418 0.065 PvsO + Year 3 360.119 3.643 0.058 PvsO + DistEdge + DistShore 4 360.144 3.668 0.057 PvsO + Area + Year 4 360.361 3.886 0.051 PvsO + DistEdge 3 360.689 4.213 0.044 PvsO + DistShore + Year 4 360.699 4.224 0.043 PvsO + NestAge + Year 4 361.582 5.106 0.028 PvsO + NestSub + Year 5 361.815 5.340 0.025 PvsO 2 361.851 5.376 0.024 PvsO + WaterDepth + Year 4 361.898 5.423 0.024 PvsO + NestHeight + Year 4 362.127 5.652 0.021 PvsO + Area 3 362.303 5.827 0.020 PvsO + DistShore 3 362.490 6.014 0.018 PvsO + NestHeight + DistShore + Year 5 362.666 6.191 0.016 PvsO + MinTemp 3 362.780 6.304 0.015 PvsO + NestAge 3 363.352 6.876 0.012 PvsO + WaterDepth 3 363.684 7.209 0.010 PvsO + NestHeight 3 363.712 7.236 0.010 PvsO + NestHeight + WaterDepth + Year 5 363.860 7.385 0.009 PvsO + NestSub 4 364.404 7.929 0.007 Null 1 381.716 25.240 0.000

PAGE 79

79 Table 3 4. Model averaged parameter estimates for factors hypothesized to affect snail kite nest survival for depredated and non depredated nests on Lake Toho, FL in 2010 and 2011. Parameter Level Estimate SE Low CI Upper CI WaterDepth 0.0000 0.0001 0.0002 0.0002 Intercept 2.5667 0.8564 0.8538 4.2795 MinTemp 0.0366 0.0507 0.0647 0.1379 Nestsub Bulrush 0.0181 0.0395 0.0609 0.0972 Nestsub Tree 0.0083 0.0254 0.0425 0.0591 Nestsub Typha 0.0000 0.0000 0.0000 0.0000 PvsO 0 1.5058 0.3591 0.7877 2.2239 PvsO 1 0.0000 0.0000 0.0000 0.0000 NestAge 0.0004 0.0009 0.0015 0.0022 WaterDepth 0.0001 0.0003 0.0006 0.0005 Year 0 0.6742 0.5137 1.7015 0.3532 Year 1 0.0000 0.0000 0.0000 0.0000 Area 0.0059 0.0127 0.0313 0.0195 DistEdge 0.0593 0.1011 0.2614 0.1428 DistShore 0.0254 0.0464 0.0675 0.1183

PAGE 80

80 Table 3 5. Support for models predicting snail kite nest survival for depredated and non depredated nests on Lake Toho, Florida in 2010 (n=592 days observed) Predation Models K AICc w i PvsO + MinTemp 3 177.490 0.000 0.940 PvsO + Area 3 186.651 9.162 0.010 PvsO + NestSub 4 186.781 9.292 0.009 PvsO 2 186.824 9.334 0.009 PvsO + NestAge 3 187.897 10.407 0.005 Null 1 187.920 10.430 0.005 PvsO + DistEdge 3 188.255 10.765 0.004 PvsO + NestHeight 3 188.259 10.769 0.004 PvsO + DistShore 3 188.545 11.056 0.004 PvsO + WaterDepth 3 188.628 11.138 0.004 PvsO + NestHeight + WaterDepth 4 189.568 12.078 0.002 PvsO + DistEdge + DistShore 4 189.711 12.222 0.002 PvsO + NestHeight + DistShore 4 189.991 12.501 0.002 Table 3 6. Model averaged parameter estimates for factors hypothesized to affect snail kite nest survival for depredated and non depredated nests on Lake Toho, FL in 2010. Parameter Level Estimate SE Low CI Upper CI NestHeight 0.0000 0.0000 0.0001 0.0000 Intercept 1.5801 1.5248 4.6297 1.4695 MinTemp 0.2103 0.0720 0.0663 0.3544 Nestsub Bulrush 0.0097 0.0206 0.0315 0.0510 Nestsub Tree 0.0024 0.0084 0.0144 0.0192 Nestsub Cattail 0.0000 0.0000 0.0000 0.0000 PvsO 0 1.2370 0.4956 0.2457 2.2282 PvsO 1 0.0000 0.0000 0.0000 0.0000 NestAge 0.0001 0.0002 0.0003 0.0005 WaterDepth 0.0000 0.0001 0.0002 0.0001 Area 0.0012 0.0027 0.0066 0.0041 DistEdge 0.0011 0.0028 0.0066 0.0045 DistShore 0.0005 0.0014 0.0023 0.0033

PAGE 81

81 Table 3 7. Support for models predicting snail kite nest survival for depredated and non depredated nests on Lake Toho, Florida in 2011 (n=1,169 days observed). Predation Models K AICc w i PvsO + NestHeight + DistShore 4 167.299 0.000 0.498 PvsO 2 171.031 3.733 0.077 PvsO + NestHeight + WaterDepth 4 171.203 3.904 0.071 PvsO + DistShore 3 171.532 4.233 0.060 PvsO + DistEdge 3 171.548 4.250 0.059 PvsO + DistShore + DistEdge 4 171.582 4.284 0.058 PvsO + NestHeight 3 171.953 4.655 0.049 PvsO + WaterDepth 3 172.813 5.515 0.032 PvsO + MinTemp 3 172.962 5.663 0.029 PvsO + NestAge 3 173.035 5.736 0.028 PvsO + Area 3 173.038 5.739 0.028 PvsO + NestSub 4 174.969 7.671 0.011 Null 1 187.971 20.672 0.000 Table 3 8. Model averaged parameter estimates for factors hypothesized to effect snail kite nest survival for depredated and non depredated nests on Lake Toho, FL in 2011. Parameter Level Estimate SE Low CI Upper CI NestHeight 0.0086 0.0086 0.0086 0.0257 Intercept 1.4007 1.4450 1.4892 4.2907 MinTemp 0.0004 0.0020 0.0036 0.0044 Nestsub Bulrush 0.0017 0.0083 0.0148 0.0183 Nestsub Tree 0.0004 0.0089 0.0173 0.0182 Nestsub Cattail 0.0000 0.0000 0.0000 0.0000 PvsO 0 2.0670 0.5298 1.0075 3.1265 PvsO 1 0.0000 0.0000 0.0000 0.0000 NestAge 0.0000 0.0006 0.0011 0.0012 WaterDepth 0.0012 0.0025 0.0038 0.0062 Area 0.0000 0.0027 0.0054 0.0054 DistEdge 0.0258 0.0534 0.1326 0.0810 DistShore 0.2116 0.2126 0.2137 0.6368

PAGE 82

82 Figure 3 1. Kissimmee Chain of Lakes with the relevant nesting lakes East Lake Tohopekaliga, Lake Tohopekaliga, Lake Hatchineha, Lake Kissimmee, and Lake Jackson outlined in red.

PAGE 83

83 Figure 3 2. Representation of the yearly water stage schedule for Lake Toho, FL in 2010 and 2011.

PAGE 84

84 Figure 3 3. Location of Lake Toho within the state of Florida

PAGE 85

85 Figure 3 4. Predicted survival rates with 95% confidence intervals for snail kites nests in relation to year on Lak e Toho, FL in 2010 and 2011. Figure 3 5. Daily survival rates of depredated snail kite nests in 2010 and 2011 with 95% confidence intervals on Lake Toho, FL. 0.88 0.90 0.92 0.94 0.96 0.98 1.00 2010 2011 Daily Survival Rate 0.88 0.90 0.92 0.94 0.96 0.98 1.00 2010 2011 Daily Survival Rate

PAGE 86

86 Figure 3 6. Daily survival rates of depredated snail kite nests partitioned by the egg and nestling stages in 2010 and 2011 with 95% confidence intervals on Lake Toho, FL. 0.90 0.91 0.92 0.93 0.94 0.95 0.96 0.97 0.98 0.99 1.00 Eggs Nestlings Daily Survival Rates

PAGE 87

87 Figure 3 7. Average distances of nest patches/trees to shore (m) for nests with and without predator observations in 2010 and 2011.

PAGE 88

88 Figure 3 8. Average nest patc h areas (ha) for nests with and without predator observations in 2010 and 2011.

PAGE 89

89 Figure 3 9. Average distance of nest structure to the edge of the nesting patch or tree cover (m) for nests with and without predator observations in 2010 and 2011.

PAGE 90

90 Fi gure 3 10. Average nest height above the waterline (cm) for nests with and without predator observations in 2010 and 2011.

PAGE 91

91 Figure 3 11 Average water depths (cm) beneath nests with and without predator observations in 2010 and 2011.

PAGE 92

92 Figure 3 12 Average minimum daily temperatures recorded for nests with and without predator observations in 2010 and 2011.

PAGE 93

93 Figure 3 1 3 Predicted survival rates with 95% confidence intervals for snail kites nests in relation to nesting substrate on Lak e Toho, FL in 2010 and 2011. 0.90 0.91 0.92 0.93 0.94 0.95 0.96 0.97 0.98 0.99 1.00 Bulrush (n=19) Cattail (n=48) Trees (n=8) Daily Survival Rate

PAGE 94

94 CHAPTER 4 SNAIL KITE (ROSTRHAMUS SOCIABILIS) BREEDING BEHAVIORS IN RELATION TO TEMPERATURE STRESS ON LAKE TOHOPEKALIGA, FL Introduction Breeding avian populations may experience fluctuating rates of success from year to year depending on prey availability. Resource availability plays a vital role in a breeding season as birds must be healthy enough and have enough resources to partake in b reeding activities (Martin 1987; Safina et al. 1988; Schoech 1996) When resources are altered due to sudden environmental perturbations such as extreme temperatures, hurricanes, or droughts the output of the breeding season is dependent upon the breeding behavior responses of the individuals in the population. There is value to understanding how individual breeding birds respond to such perturbation as it may have implications for the longevity of the population (Reichert et al. 2010b) When breeding prod uctivity is reduced in a given year, efforts should be made to determine the cause. This may help managers and researchers identify key factors in the system that affect productivity, and help create better long term conservation goals. Breeding behaviors such as clutch initiation date, clutch size, nest abandonment/ desertion, number of young fledged, and provisioning rates may all provide insight to breeding productivity from year to year. Clutch initiation date may be a primary indicator of food availability prior to egg laying, due to the nutritional reserve requirements of the female prior to laying (Perrins 1970b; Martin 1987; Safina et al. 1988; Nilsson and Svensson 1993) Given that so me species experience increased chick survivorship earlier in the season food availability earlier in the season can be of critical importance for a breeding population (Perrins 1970a; Pietz 1987)

PAGE 95

95 Clutch size has also been found to decrease during times of food scarcity and can other environmental influences (Steenhof and Newton 1987) In addition to food scarcity, exposure to extreme temperatures may limit avian repro duction as drastic temperature fluctuations often reduce semen production, egg laying, and copulation attempts (Blanco et al. 1987) For these reasons clutch size is a possible indicator of the A lthough nest abandonment occurs for a variety of reasons (parent mortality, nest parasites, predator presence, human disturbance) it is thought to be primarily induced by harsh weather conditions and low food availability (Anderson et al. 1982; Marzluff 19 88; Snyder et al. 1989; Huhta et al. 2004) For example, higher rates of nest abandonment would be expected during years of increased environmental stresses, like extreme temperatures and/or low food resources. Abandonment by either parent in early stages almost always ensures nest failure. In contrast, late stage nestling desertion by either parent may be an indication of increased prey abundance and favorable system conditions (Beissinger 1986; Beissinger 1987a, b; Fujioka 1989; Pilastro et al. 2001) Th ese conditions allows for the deserting parent to increase their fitness by re mating or avoiding parental duties, while the remaining parent is left to care for the young, rarely resulting in offspring mortality. Provisioning rates of young vary with r esource abundance, availability, habitat quality, and parental foraging ability (Riddington and Gosler 1995; Goodbred and Holmes 1996; Van Oort et al. 2007) Although predation may have the greatest effect on the number of young fledged from a nest, provis ioning rate may have the greatest

PAGE 96

96 impact on fledgling weight and thus long term juvenile health (Martin 1987; Ringsby et al. 1998) The endangered Florida snail kite ( Rostrhamus sociabilis plumbeus ), is a neotropical hawk that resides on the freshwater we tlands and shallow lakes of central and south Florida (Martin et al. 2006) As a dietary specialist, snail kites forage almost exclusively on freshwater apple snails ( Pomacea spp. ) (Snyder and Snyder 1969; Sykes 1987b; Rawlings et al. 2007) and are restric ted to areas where apple snails are not only present but available on emergent vegetation. Snail kites have been documented to breed in every month of the year, with most initiated nesting attempts taking place from December through July (Sykes et al. 1995 ; Bennetts and Kitchens 1997) The start of the breeding season varies temporally and regionally from year to year in relation to water levels and temperature. Peak egg laying occurs from February through April (Sykes 1987c) Both sexes take turns incubati ng the eggs, adding vegetation to the nest, and provisioning the young, with the female performing the majority of the nocturnal incubation (Sykes 1987c) Snail kites are unique in that it is not uncommon for either sex to desert its mate near the time of fledging, in some cases in order to pursue another mate for further nesting during the same season. This is especially true in years of high snail abundance (Beissinger 1987a, b; Beissinger and Snyder 1987) It is important to note that throughout the bree ding season snail kites may make multiple successful breeding attempts and will renest after a failure (Snyder et al. 1989) Clutch sizes have been documented to range anywhere from one to six eggs, with clutches of four to five eggs common before 1940 (Be issinger 1986; Sykes 1987a) and only one

PAGE 97

97 reported clutch with six eggs ever described (Sykes et al. 1995) Currently the average clutch size is 2.66 or 2.92 eggs depending on the study with a range of 1.4 2.5 young fledged per nest (Sykes 1987c; Snyder et al. 1989) Historically, Water Conservation Area 3A (WCA3A) was a critical breeding habitat for nesting snail kites (Bennetts and Kitchens 1997) Due to extended periods of droughts and long term habitat degradation, recent reproduction within WCA3A ha s been almost non existent, with no successful reproduction in 2005, 2007, 2008 or 2010. With the habitat conditions in the region remaining unsuitable for kite nesting, much of the population is now heavily concentrated in and dependent upon the Kissimmee Chain of Lakes (KCOL) (Fig. 4 1), particularly Lake Tohopekaliga (Lake Toho), which has accounted for 41% of all documented successful nests and 57% of all fledged young on a range wide basis from 2005 2010 (Reichert et al. 2011). Although lacustrine syst ems tend to provide more consistent water levels than the palustrine wetlands of the everglades, there is still a great deal of fluctuation throughout the season with water levels decreasing from winter to summer months. This is primarily due to water mana gement and scheduled draw downs for flood control, agricultural and municipal uses, recreation, and to supply water to the greater Everglades (Sykes et al. 1995) (Fig. 4 2). There were substantially fewer nesting attempts and lower apparent nest success range wide in 2010 compared to 2011 (Reichert et al. 2010a, 2011) The most notable environmental difference between the 2010 and 2011 breeding seasons was the historically cold temperature that gripped Florida in early 2010 (Leftwich et al. 2010) On Lake Toho, minimum temperatures hovered around 0 C from January 4th to January

PAGE 98

98 13th with water temperatures remaining below 10 C from January to April of 2010 (DBHYDRO 2012) Temperatures remained 15 26 C cooler in late February and early March than the tem peratures recorded in 2011 (Fig. 4 3). Such low temperatures can have a serious impact on the snail kites prey availability as apple snails ( Pomacea spp .) become inactive at temperatures of < 13 C (Stevens et al. 2002) making them inaccessible to foragi ng adults. In addition to the cold weather event 2010 and 2011 differed in the amount of hydrilla ( Hydrilla verticillata ) that was present on Lake Toho (Florida Fish and Wildlife Conservation Commission, unpub. data). This invasive submerged vegetation was (Langeland 1996) and is often snails are accessible to kites (Pias 2012) This is important since kites are limite d to accessing snails within their reach, approximately 16 cm below the water surface (Sykes et al. 1995) The majority of the snails eaten by kites on Lake Toho are the exotic island apple snail (Cattau 2008; Baker et al. 2010; Pias et al. 2012) ( Pomacea insularum ), and this species readily consumes hydrilla (Baker et al. 2010) Therefore, snail kites frequently forage in hydrilla mats within and just beyond the emergent littoral zone on Lake Toho (Pias 2012) Due to the combination of unusually low temper atures, known to inhibit hydrilla growth (Van et al. 1978) and the regularly scheduled aquatic plant management treatments (FFWCC 2011) there was very little hydrilla foraging substrate available for snail kites on Lake Toho in 2010 (Pias 2012) The objective of this chapter is to assess snail kite breeding behavior differences between 2010 and 2011 on Lake Toho. While the previous two chapters identified direct

PAGE 99

99 causes of nest failure, this chapter will address large scale factors and indirect eff ects by evaluating how clutch initiation date, clutch size, fledge rates, nest abandonment, mate desertion, and provisioning rates changed from 2010 to 2011. Methods Study Area In 2010 and 2011 we monitored nests on Lake Toho, a shallow lake covering appr oximately 8,176 ha (2009 Remetrix bathymetry map), in northwest Osceola County, Florida (Fig. 4 4). Much of the lake shoreline has been modified, either through routine lake management (vegetation removal and herbiciding), development, and/or cattle grazi ng (HDR Engineering. 1989) Similar to other lakes in Florida, Lake Toho is considered a highly eutrophic lake with mixed emergent littoral vegetation covering (Welch 2004) Snail kite nesting has been documented lake w ide although more commonly on the northern half of the lake. Field Methods Nest searching was conducted every 21 days by airboat from November to August of 2010 and 2011 on Lake Toho. Nest locations were determined using the nest searching protocol curren tly in place for the state wide snail kite population monitoring project (Reichert et al. 2011). Nests are typically detected when a snail kite flushes from dense vegetation and exhibits a breeding behavior (i.e. nest defense, stick carrying, copulating, a nd flying with snail meat) or a general unwillingness to leave the area. At every new nest a mirror pole was used to determine the number of eggs or young present in the nest. Coordinates were recorded at the nest using a GPS (Garmin GPSmap 76Cx).

PAGE 100

100 After a full lake nest search was completed nests were randomly selected for camera placement and cameras were deployed within four days or greater from the original nest check. Cameras were set on nests with at least one egg to minimize the chance of abandonment during the nest building phase (Richardson et al. 2009) When possible, cameras were set up on nests in the egg stage, although cameras were set on nests with young when no other nests were available. Reducing disturbance during camera installation was a p riority, and our total time at the nest patch was typically 10 minutes or less. Cameras were set on a tripod within three meters of the nest at a height that allowed for good visibility into the nest. Vegetation that impeded the view was tied back out of view while the camera was in use. The tripods and mounts were camouflaged to match the surrounding vegetation before deployment. Cameras were camouflaged with surrounding vegetation, if available, to make them less visible in the nest patch. Once there wa s a camera present, nest checks were preformed weekly. Airboats were used to reach the nest site location at which point we would wade or swim the remaining 12 to 46 meter distance to access the nest. Prior to the camera set up and for nests without camera s, nest checks were performed every 21 days. Camera batteries and memory cards (28 GB) were changed weekly during nest checks, and cameras were realigned with the nest if necessary. Over the course of the study we used a total of 15 Reconyx HyperFire Lo w Glow Semi Covert Infrared cameras on Lake Toho, including both the PC85 and PC800 models (Reconyx Inc., Holmen WI). The cameras recorded 3.1 megapixel images in color during the day and monochrome infrared by night. Cameras triggered in response

PAGE 101

101 to heat and motion in the field of view. Once triggered, five rapid fire images separated by one to three seconds each were taken to fully capture any provisioning events. Between several hundred to eighty thousand images were recorded in one week per camera depen ding upon the nest stage and camera set. Data Recording FastPictureViewer Professional software was used to view all of the images. Data entry was considered to have started on the first day the camera was set and ended when the young reached fledge age or 24 hours after a predation event or abandonment occurred. I based the ne st success period on a 58 day nest cycle with 28 days for laying and incubation and 30 days for the nestling stage (Sykes 1987c) Given that snail kites can build several courtship structures prior to laying eggs, failures at this stage represent failures of courtship and not the nest, therefore, only initiated nests are considered for this paper (Steenhof 1987; Dreitz et al. 2001) Nests were considered successful when at least one young reached the fledge age of 30 days old. We selected this timeframe as it coincided with the typical age that we observed young to start leaving the nest for extended periods of time. Nests were considered failed when all nest contents were determined to be abandoned, unviable, removed or destroyed. The nest age was determi ned by back calculating date of initiation by 28 days prior to the hatch date of the first young if the nest reached hatch age. If eggs did not hatch then the initiation date was back calculated depending on the number of eggs observed at the first nest ch eck. If a nest was found with 3 eggs the initiation date was estimated to be a median of the 28 day incubation period, 14 days prior to the nest check. If a nest was found with 2 eggs the initiation date calculated to 4 days prior to the nest check. If a

PAGE 102

102 n est was found with only 1 egg the initiation date was estimated to be 2 days prior. This method was based on previous literature regarding the laying intervals of snail kites (Sykes 1987a) Clutch size was recorded during each nest check and updated from t he images if the camera was set prior to clutch completion. I determined abandonment to have taken place once both adults stopped tending the nest during the viable incubation or young stage. A mate was considered to have deserted once it stopped appearing at the nest and provisioning the young in nestling stage. Although indications of mate desertion and abandonment (i.e. less nest visits and provisioning) may have started prior to the adult(s) leaving it was only officially documented after their last vis it to the nest. It is important to note that abandonment always results in nest failure and occurs when both parents leave the nest. In contrast, mate desertion primarily occurs in the late nestling stage by one parent and typically results in successful n ests. Nest abandonment is considered to be the result of poor system conditions or other factors, while desertion is considered to be an indication of increased prey abundance (Beissinger 1986; Beissinger 1987a, b; Fujioka 1989; Pilastro et al. 2001) In order to obtain a provisioning rate for a nest I recorded every time a snail was brought to the nest starting from the first day a nestling hatched to the fledged date of the oldest young (30 days). If the meat was entirely eaten by the adult or accidental ly dropped over the edge it was not added to the nestling feeding rates. Although adults consistently appeared to feed all young in the nest equally, no attempt was made to document exact rates and quantities of snail meat provided to each individual young (Fig. 4 5).

PAGE 103

103 Feeding rates per young were determined by dividing the total number of snails brought to the nest in one day by the number of live young present in the nest on that day. Given that the majority of snails brought to the nest had already been extracted from their shell it was difficult to compare the exact size of each snail or how much the adults had eaten prior to bringing it into the nest (Fig. 4 6). Only once the young were close to fledge age did the adults occasionally bring in complete u n extracted snails with the operculum removed for the young to eat (Fig. 4 7). For comparison of feeding rates I recorded the daylight hours for each day to account for nests that may have had more daylight hours available to feed their young. From this I created a rate of snails provisioned per young per hour. If vegetation blocked the view for part of the day or the camera failed during daylight hours the entire day was removed from the sample. Data Analysis The proportion of clutch initiation dates was determined using all nests on Lake Toho in 2010 (n=63) and 2011 (n=79) This allowed for a larger sample size given that the data was comparable for both non camera and camera nests. The survival probability of nests initiated each month was assessed usin g only the camera data set for 2010 (n=32) and 2011 (n=43) to increase accuracy in the results. I modeled the (Shaffer and Burger 2004) fitted using Proc GENMOD in SAS 9.3 (SAS Institute 1989) This method of analysis models the probability of the nest surviving between nest checks (everyday for camera monitored nests) and allows for intervals that vary in length. A binomial response distribution was used to fit the models, 1 if the interval nest fate was successful or 0 if the nest had failed during the interval.

PAGE 104

104 Clutch sizes were compared between 2010 (n=31) and 2011 (n=43) using a generalized linear model (GLM) in Proc GENMOD in SAS 9.3 with a normal distribution and an identity link (SAS Institute 1989) One nest was excluded from the 2010 data sample since there was ambiguity in the final number of eggs prior to nest depredation. Number of young fledged was evaluated for camera nests for 2010 (n=32) and 2011 (n=43) using the same type as GLM stated prev iously. All camera nests were included in the analysis to compare number of young fledged by the total nesting attempts. Total nests abandoned and deserted were compared in 2010 and 2011 using logit link. A random effect for site was added in to account for any habitat effects. Given that all nests were capable of experiencing abandonment the sample size remained at 32 nests in 2010 and 43 nests in 2011. The probability of a snail kite abandonin g its nest based on the month of initiation was evaluated with (Shaffer and Burger 2004) fitted using Proc GENMOD in SAS 9.3 (SAS Institute 1989) When comparing nests for mate desertion, only nests with young as old as th e youngest observed deserted nests (10 days old) were included for comparison. This was done to remove nests that had failed for other reasons but that might have experienced mate desertion if the young had reached a certain age. This left a sample size of 13 total nests in 2010 and 28 nests in 2011. Provisioning rates were compared in 2010 (190 observation days, n=14 nests) to rates in 2011 (616 observation days, n=29) using generalized linear mixed models was included for both nest identification and nest site location to account for association of feeding rates within

PAGE 105

105 nest and for any spatial correlation of foraging grounds. Provisioning rates were obtained by dividing the total number of snails delivered to the nest by the number of by the number of daylight hours observed for that day. This accounted for increased or decreased feeding rates due to day length. Result s Nesting Outcome Of the 258 nest structures documented throughout Florida in 2010, only 190 nests reached egg laying stage, with 50% of nests on the KCOL (34% on Lake Toho). In 2011, 253 nests reached egg laying stage out of 293 structures statewide, wit h 69% of nests on the KCOL (33% on Lake Toho). Less than 100 young were fledged from nests in 2010 range wide compared to the approximately 200 young fledged in 2011. Over this two year period of time Lake Toho consistently produced 34% and 33% of the succ essful nests range wide. On Lake Toho nest cameras were set up on approximately half the nests in 2010 (32 out of 63 total nests) and 2011 (43 out of 79 total nests). With this data it was determined that the nest survival probability on this lake was sig nificantly lower in 2010 (0.11 0.05) than in 2011 (0.45 0.08, 2 0.05 =9.87, df=1749, p =0.001). The primary cause of egg and nestling mortality was found to be predation for both years with no significant difference in the predati on rates between the two years ( see previous chapter ). The secondary reason for nest failure was due to abandonment of nests during the egg stage. Late stage mate desertion was observed in both years to varying degrees.

PAGE 106

106 Clutch Initiation Dates The first initiated nest(s) on Lake Toho occurred in early February for 2010 (2/9/2010) and early January for 2011 (1/12/2011). The final nest initiation date was in late October (10/25/2010) for 2010 and late September (9/22/2011) for 2011. Consequently the breedin g season ended up being approximately nine months long for both years on Lake Toho. The number of initiated nests on Lake Toho peaked in April of 2010 with approximately 50 percent of all nesting attempts occurring within April. Similarly, the highest numb er of nesting attempts in 2011 occurred in April as well. The difference being that the nesting attempts were distributed more evenly throughout the breeding season in 2011 (Fig. 4 8). Some trends appeared when comparing the survival rates of nests initiat ed by month and year. The survival rates of nests initiated in February of 2010 were significantly lower than all other months of initiation in both 2010 and 2011. Nests initiated in March of 2010 experienced significantly lower rates of survival than the majority of nests initiated in 2010 and 2011 (Table 4 1, Fig. 4 9). Clutch Size Observed clutch sizes ranged from 1 4 eggs in 2010 and 1 3 eggs in 2011. There was no significant difference in the average clutch size of nests in 2010 (2.81 0.1) when comp ared to nests in 2011 (2.74 0.1, 2 0.05 =0.22, df=72, p =0.64) (Fig. 4 10). Number of Young Fledged There was no significant difference in the number of young fledged per nesting attempt in 2010 (0.69 0.2) when compared to 2011 (1.09 0.2, 2 0.05 =2.56, df=73, p=0.11). Additionally there was no significant difference in number of young fledged per

PAGE 107

107 successful nest as successful nests in 2010 (n=10) fledged an average of 2.20 0 .2 young and successful nests in 2011 (n=25) fledged an average of 1.88 0.1 young ( 2 0.05 =1.40, df=33, p=0.24). Nests fledged anywhere from 1 3 young during both years of the study with 22 young fledged from 10 nests in 2010 and 4 7 young fledged from 25 nests in 2011 (Fig. 4 11). Number of Abandoned Nests The probability of a snail kite abandoning its nest based on the month of clutch initiation was highest for February (0.67 0.3) and March (0.50 0.3) of 2010. There was no significant difference in number of nests abandoned in 2010 compared to 2011 (F Value=1.37, df=55, p =0.25). In 2010, 19% of nests (n=6) were abandoned while in 2011 only 9% of nests (n=4) were abandoned. All nests were left during the incubation stage. Mal es (n=3) and females (n=3) initiated abandonment equally in 2010 while more females initiated abandonment in 2011 (n=3) with the final nest abandoned simultaneously by both the male and female. Number of Deserted Nests There was no significant difference in the number of nests deserted in 2010 when compared to 2011 (F value=0.79, df=25, p =0.38). In 2010, 15% of nests (n=2) experienced mate desertion while 29% of nests (n=8) were deserted by a parent in 2011. Only females deserted in 2010 while both males (n=2) and females (n=6) deserted in 2011 (Table 4 2). One nest loss one of two young from starvation after the female deserted, while another nest failed entirely when the remaining fledgling was eaten by an alligator at 28 days old.

PAGE 108

108 Provisioning Rates There was no significant difference between the overall feeding rates of individual young fed per daylight hour in 2010 and 2011 (F value =0.11, df=38.87, p =0.74) (Fig. 4 12). However, in both years, individual young in ne sts with only one young were fed significantly more per hour than nests with two and three young. In 2011, individuals in nests with two young were also fed significantly more than in nests with three young (Table 4 3). In 2010, nests with one young were f ed approximately 1.03 0.1 snails per hour (15 obs. days), individuals in two young nests were fed 0.56 0.1 snails per hour (71 obs. days), and individuals in three young nests were fed 0.54 0.1 snails per hour (104 obs. days). In 2011, nests with one young were fed approximately 0.84 0.1 snails per hour (182 obs. days), individuals in two young nests were fed 0.67 0.1 snails per hour (272 obs. days), and individuals in three young nests were fed 0.55 0.1 snails per hour (162 obs. days) (Fig. 4 1 3). See Figure 14 for averages of total snails fed per young per day for both years. Discussion Although the duration of the kites breeding season remained essentially the same (9 months) on Lake Toho for 2010 and 2011 there appeared to be some vital diffe rences between the two seasons. In response to the extended cold snap in 2010 the kites started initiating nests one month later than what was observed in 2011. Considering the effects of cold temperatures on availability of apple snails (Stevens et al. 20 02) this was likely influenced by the lack of available prey at that time. In addition to later clutch initiation dates in 2010, kites in 2010 experienced significantly reduced early season nesting success (Fig. 4 9) Furthermore, nests initiated in February and March of the 2010 nesting season were more likely to be abandoned than any other

PAGE 109

109 nests observed during this study. However, kite breeding efforts increased dramatically in late March and April of 2010, with almo st half of all nests initiated in April (46%). Once temperatures, and thus snail activity, reached an optimal level for breeding, kites apparently attempted to fledge as many young as possible to make up for the late start. This approach was markedly diffe rent from what was observed in 2011, where initiation rates were much more evenly distributed throughout the season (Fig. 4 8). Remarkably, the delayed breeding response by snail kites in 2010 did not affect annual measures of clutch size or number of you ng fledged per successful nest, in comparison to 2011. Beissinger (1986; 1987b) also found no evidence of snail kite clutch size adjustment in relation to annual variations in food abundance. He hypothesized that snail kites respond to variations in prey a vailability through differential rates of mate desertion; when prey are abundant desertion increases, allowing for multiple breeding attempts, and when prey are scarce, desertion is rare. While I did not find significantly higher rates of desertion in 2011 this may have been due to increased desertion rates later in 2010 as snail availability increased with increasing temperatures. The initiation of the 2010 breeding season may have been delayed until kites were able to provision young at rate similar to th at seen in 2011. For example, young were not hatched until April 22 nd in 2010, compared to February 9 th in 2011. Given that there was no difference in individual feeding rates from year to year or rate of productivity per nest it is likely that kites delay ed initiations in 2010 until snail availability was high enough for successful breeding. However, there is no way to know if the breeding adults consumed fewer snails in order to maintain provisioning rates to young if food was scarce during provisioning.

PAGE 110

110 Conservation Implications: Given that Toho has accounted for 41% of all successful nests range wide from 2005 2011 (Reichert et al. 2011) it is critical that the lake is carefully managed during the breeding season for increased productivity ( see prev ious chapter ). It is important to note that early season (first half of the breeding season) productivit y is generally higher than late season nesting attempts. Early nesting is beneficial so that recently fledged young can learn to forage on their own in warmer temperatures when apple snails are most active. Pias (2012) found hydrilla to be the third most commonly used foraging substrate by kites on Lake Toho in 2011, while hydrilla foraging was negligible in 2010 due to its limited presence on the lake. I t is likely that hydrilla provides the kites with an important foraging substrate for late season nests. This finding would suggest that hydrilla management should be minimal during this critical time for the kites (Pias 2012) While the 2010 cold snap appeared to significantly delay breeding activity on Lake Toho early in the nesting season, any additional environmental perturbati on in April could have negatively affected a large portion of the kites breeding efforts for that year. Thus, when unusually cold winter/early spring temperatures appear to stymie kite nesting, managers should be aware that amplified nesting activity will likely take place when conditions improve. Efforts should be made to delay any management activities or to reduce recreationa l disturbance near nests following exceptionally cold winters in lost time as temperatures warm. It is also breeding range an d that restoring wetlands in the southern portion of the kites range could provide a breeding refuge when temperatures drop on the KCOL.

PAGE 111

111 With the effects of global warming becoming increasingly evident it is likely that temperatur e fluctuations and extreme weather events, such as Florida's 2010 cold snap, will become progressively more common (IPPC 2012). Considering how severely snail kite nesting success was affected in 2010 the population may be especially susceptible to future extreme temperatur es The results in this chapter suggest that while the snail kites flexible breeding strategy of multiple nesting attempts and prolonged breeding season has proven to be beneficial, they rem ain extremely vulnerable to increased climatic variation.

PAGE 112

112 Table 4 1. Differences in nest survival rates by month of initiated on Lake Toho, FL 2010 and 2011. Plus signs (+) indicate a positive significant difference in survival rates. Minus signs ( ) imply a negative significant difference in survi any initiated nests have been removed. Feb (10) Mar (10) Apr (10) May (10) Jun (10) Aug (10) Sep (10) Jan (11) Feb (11) Mar (11) Apr (11) May (11) Feb (10) Mar (10) Apr (10) May (10) Jun (10) Aug (10) ls ls ls ls ls Sep (10) . ls Jan (11) ls Feb (11) ls Mar (11) ls + Apr (11) ls May (11) . ls . July (11) ls .

PAGE 113

113 Table 4 2. Desertion month, sex, nestling age, and nest fate of the nests that experienced mate desertion in 2010 and 2011 on Lake Toho, FL. Year Desertion Month Deserting Mate Age of Oldest Nestling Nest Fate 2010 May Female 10 days old Successful 2010 May Female 20 days old Successful 2011 March Female 24 days old Successful 2011 April Male 13 days old Successful 2011 April Female 25 days old Successful 2011 April Female 21 days old Successful 2011 April Female 17 days old Successful 2011 May Male 24 days old Successful 2011 May Female 21 days old Successful 2011 August Female 19 days old Failed Table 4 3. Differences of nestling feeding rates analyzed by number of young in the nest for Lake Toho, FL 2010 and 2011. Plus signs (+) indicate a positive significant difference in survival rates. Minus signs ( ) imply a negative significant difference in survival rates (Alpha > 0.05). Dots (.) imply no difference. 2010 2011 1 Young 2 Young 3 Young 1 Young 2 Young 3 Young 1 Young + + + + 2010 2 Young + 3 Young + 1 Young + + + + 2011 2 Young + + + 3 Young + + +

PAGE 114

114 Figure 4 1. Kissimmee Chain of Lakes with the relevant nesting lakes East Lake Tohopekaliga, Lake Tohopekaliga, Lake Hatchineha, Lake Kissimmee, and Lake Jackson outlined in red

PAGE 115

115 Figure 4 2. Daily water stage data (NGDV Feet) for Lake Toho from January to September in 2010 and 2011 on Lake Toho, FL. 51.50 52.00 52.50 53.00 53.50 54.00 54.50 55.00 55.50 56.00 1 Jan 1 Feb 1 Mar 1 Apr 1 May 1 Jun 1 Jul 1 Aug 1 Sep NGDV Feet 2010 2011

PAGE 116

116 Figure 4 3. Daily minimum temperatures ( C) from January to September in 2010 and 2011 on Lake Toho, FL. 10 5 0 5 10 15 20 25 30 1 Jan 1 Feb 1 Mar 1 Apr 1 May 1 Jun 1 Jul 1 Aug 1 Sep Daily Minimum Temperature (C) 2010 2011

PAGE 117

117 Figure 4 4. Location of La ke Toho within the state of Florida.

PAGE 118

118 Figure 4 5. A breeding female snail kite feeding her 9 10 day old young snail meat on Lake Toho in 2011. Figure 4 6. A picture of a recently banded 30 day old young with an entire extracted snail in its mouth on L ake Toho in 2011.

PAGE 119

119 Figure 4 7. Breeding female returning to her nest of two 21 day old young with a non extracted snail minus the operculum. Photograph taken on Lake Toho, Fl in 2011. Figure 4 8. Proportions of all nests initiated by month (January September) for the 2010 (n=63) and 2011 (n=79) breeding season on Lake Toho, FL. 0 0.1 0.2 0.3 0.4 0.5 0.6 1 2 3 4 5 6 7 8 9 Proportion of Initiated Nests Month 2010 nests 2011 nests

PAGE 120

120 Figure 4 9. Survival probability of nests initiated each month from January to September of 2010 and 2011 on Lake Toho, Fl. Figure 4 10. Average monthly clutch size from January to September of 2010 and 2011 on Lake Toho, Fl. 0.00 0.20 0.40 0.60 0.80 1.00 1.20 0 2 4 6 8 10 Survivial Probability Month 2010 nests 2011 nests

PAGE 121

121 Figure 4 11. Number of individual young fledged from one, two, and three young camera monitored nests in 2010 (n=10) and 2011 (n=25) on Lake Toho, FL. n=2 n=4 n=4 n=8 n=12 n=5 0 5 10 15 20 25 30 1 Young Nests 2 Young Nests 3 Young Nests Number of Young Fledged 2010 2011

PAGE 122

122 Figure 4 12. Estimates of the number of snails fed to individual young per daylight hour from nest ages 29 58 (or nestling age of 1 30 days old) on Lake Toho, FL in 2010 and 2011.

PAGE 123

123 Figure 4 13. Estimates of the number of snails fed to individual young per daylight h our from nests containing 1 3 young with nest ages 29 58 (or nestling age of 1 30 days old) on Lake Toho, FL in 2010 and 2011.

PAGE 124

124 Figure 4 14. Average number of snails fed to individual young each day separated by number of young in the nest for 201 0 and 2011 on Lake Toho, FL. 0 2 4 6 8 10 12 14 16 18 1 Young 2 Young 3 Young Snails /Young / Day 2010 2011

PAGE 125

125 CHAPTER 5 CONCLUSIONS The whole purpose of this research was to increase our knowledge of the snail Lake Toho, an essential breeding grou nd. This study has highlighted a significant disconnect in what was originally perceived to be the primary cause of nesting failure on the central Florida lakes and what is currently happening to this breeding population. Regardless of the accuracy of hist orical causes of nesting failure it is clear that the majority of nests on Lake Toho failed due to nest predation, not to high rates of nest collapse. The breeding behavior of the kites revealed a consistent tendency towards continuous nest structure maint enance even after catastrophic egg or young loss by predators or as a nest was shifting on its base vegetation. With the knowledge that snail kites are at high risk of nest predation managers on the central Florida lakes can focus on reducing nest access o f key predators. Predators such as the yellow rat snake could be dissuaded by managing for nest patches further from shore or placing predator guards on trees to limit tree access to raccoons and snakes. Additionally, this study highlighted the danger of having such a large portion of the central Florida. Although the extreme cold weather in 2010 affected much of the state, snail kites remain especially vulnerable to such events while nesting at the northern extent of their range on the KCOL. These lakes are also some of the most heavily managed and heavily recreated in the state, increasing the potential for disturbances and nest failures where kite activity has become co ncentrated (Sykes 1987c, Snyder et al 1989, Personal obs.).

PAGE 126

126 While the 2010 cold snap appeared to significantly delay breeding activity on Lake Toho early in the nesting season, any additional environmental perturbation in April could have decreased a larg e portion of the kites breeding efforts for that year. Thus, when unusually cold winter/early spring temperatures appear to stymie kite nesting, managers should be wary that amplified nesting activity will likely take place when conditions improve. Efforts should be made to delay any management activities or to reduce recreational disturbance following exceptionally cold winters in order to allow higher nest success on the KCOL, restoring habitat in the southern end of the state will likely provide the best long term viability for the snail kite population

PAGE 127

127 LIST OF REFERENCES ALBRECHT, T., D. HO K, J. KREISINGER, K. WEIDINGER, P. KLVA A, AND T. C. MICHOT. 2006. Factors determining P ochard nest predation along a wetland gradient Journal of Wildlife Management 70:784 791. ALLEN, E. R., AND W. T. NEILL 1950. The life history of the Everglades Rat Snake, Elaphe obsole ta rossalleni. Herpetologica 6:109 112. ANDERSON, D. W., F. GRESS, AND K. F. MAIS 1982. Brown Pelicans: i nfluence of food supply on reproduction Oikos 39:23 31. BAKER, P., F. ZIMMANCK, AND S. M. BAKER 2010. Feeding rates of an introduced freshwater ga stropod (Pomacea insularum) on native and non indigenous aquatic plants in Florida. Journal of Molluscan Studies 76:138 143. BANCROFT, G. T. 1986. Nesting success and mortality of the Boat Tailed Grackle. The Auk 103:86 99. BEISSINGER, S. 1986. Demograph y, environmental uncertainty, and the evolution of mate desertion in the Snail Kite. Ecology 67:1445 1459. BEISSINGER, S. R 1987a. Anisogamy o vercome : female strategies in Snail Kites. The American Naturalist 129:486 500. BEISSINGER, S. R 1987b. Mate d esertion and reproductive effort in the Snail K ite. Animal Behaviour 35:1504 1519. BEISSINGER, S. R ., and N. F. R. Snyder. 1987. Mate desertion in the Snail Kite Animal Behaviour 35:477 487. BENNETTS, R. E., M. W. COLLOPY, AND S. R. BEISSINGER 1988. Nesting ecology of Snail Kites in Water Conservation Area 3A. Page 62. University of Florida, Florida Cooperative Fish and Wildlife Research Unit, Gainesville, FL. BENNETTS, R. E., AND W. M. KITCHENS 1997. The Demography an d movements of Snail Kit es in Florida. Annual Report. USGS Florida Fish and Wildlife Research Unit, University of Florida, Gainesville, FL. BLANCO, J., D. M. BIRD, AND J. H. SAMOUR 1987. Physiology In Raptor Research And Management Techniques. Institute for Wildlife Research, Washington, D.C. BOWMAN, G. B., AND L. D. HARRIS 1980. Effect of spatial heterogeneity on ground nest depredation. The Journal of Wildlife Management 44:806 813.

PAGE 128

128 BURNHAM, K. P., AND D. R Anderson. 2002. Model Selection and Multimodel Inference: A Prac tical Information Theoretic Approach, 2nd ed. Springer Science and Business Media, New York. BUTLER, R. 2009 Sources of nest failure in Mississippi Sandhill Cranes, Grus canadensis pulla: nest survival modeling and predator occupancy M.S. thesis Univer sity of New Orleans New Orleans CADE, T. J., AND S. A. TEMPLE 1995. Management of threatened bird species: evaluation of the hands on approach. Ibis 137:S161 S172. CAIN III, J. W., M. L. MORRISON, AND H. L. BOMBAY. 2003. Predator activity and nest succ ess of Willow Flycatchers and Yellow Warblers. The Journal of Wildlife Management 67:600 610. CATTAU, C.E. 2008. Effects of the invasive exotic apple s nail (Pomacea insularum) on the Snail Kite (Rostrhamus sociabilis plumbeus) in Florida, USA. M.S. thesis University of Florida, Gainesville CHANDLER, R. 1974. Notes on Everglade K ite reproduction. American birds 28:856. CT, I. M., AND W. J. SUTHERLAND. 1997. The effectiveness of removing predators to protect bird populations Conservation Biology 11:39 5 405. COX, W. A., F. R. THOMPSON, AND J. FAABORG 2012. Spec ies and temporal factors affect predator specific rates of nest predation for forest songbirds in the midwest The Auk 129:147 155. DARBY, P. C 2006. Apple s nail densities in habitats used by foraging Snail Kites Florida Field Naturalist 34:37. DBHYDRO. 2012. South Florida Water Management District. Available at www.sfwmd.gov DINSMORE, S. J., G. C. WHITE, AND F. L. KNOPF. 2002. Advanced techniques for modeling avian nest survival. Ecology 83:3476 3488. DREITZ, V. J., R. E. BENNETTS, B. TOLAND, W. M. KITCHENS, AND M. W. COLLOPY 2001. Spatial and temporal variability in nest success of Snail Kites in Florida: a meta analysis The Condor 103:502 509. FLORIDA FISH AND WILDLIFE CONSERVATION COMMIS S IO N 2012. FWCC hydrilla management position statement. [Online] Available at http://myfwc.com/fis hing/freshwater/black bass/first year updates/hydrilla/ FORSLUND, P., AND T. PRT. 1995. Age and reproduction in birds hypotheses and tests. Trends in Ecology & Evolution 10:374 378.

PAGE 129

129 FREDERICK, P. C., AND M. W. COLLOPY 1989. The role of predation in d etermining reproductive success of colonially nesting wading birds in the Florida Everglades. The Condor 91:860 867. FUJIOKA, M. 1989. Mate and nestling desertion in colonial Little Egrets. The Auk 106:292 302. GIBBONS, J. W., AND M. E. DORCAS 2005. Sna kes of the S outheast. University of Geor gia Press, Athens, Georgia GOODBRED, C. O. N., AND R. T. HOLMES. 1996. Factors affecting food provisioning of nestling Black Throated Blue Warblers. The Wilson Bulletin 108:467 479. HARTLEY, M. J., AND M. L. HUNTE R. 1998. A meta analysis of forest cover, edge effects, and artificial nest predation rates Conservation Biology 12:465 469. HDR Engineering. 1989. Technical Report of the development of a Surface Water Improvement and Management (SWIM) plan for Lake Tohopekaliga/East Lake Tohopekaliga. in Final Report. South Florida Water Management District, West Palm Beach, FL. HENSLEY, R. C., AND K. G. SMIT H. 1986. Eastern Bluebird responses to nocturnal Black Rat Snake nest predation The Wilson Bulletin 98:602 603. HOOVER, J. P 2006. Water depth influences nest predation for a wetland dependent bird in fragmented bottomland forests. Biological Conservati on 127:37 45. HOUSTON, C. S., G. S. DWIGHT, AND C. R. ROHNER 1998. Great Horned Owl (Bubo virginianus). In The Birds of North America Online, no. 372 (A. Poole, Ed.) Ithica: Cornell Lab of Ornithology; Retrieved from the Birds of North America Online. [O nline] Available at http://bna.birds.cornell.edu/ bna/species/372 doi:10.2173/bna.372 HUHTA, E. S. A., T. AHO, A. R. I. J NTTI, P. SUORSA, M. KUITUNEN, A. R. I. NIKULA, AND H. HAKKARAINEN 2004. Forest fragmentation increases nest predation in the Eurasian Treecreeper Conservation Biology 18:148 155. IPPC. 2012. Managing the risks of extreme events and disasters to advance climate change adapt ation A special report of working groups I and II of the Intergovernmental panel on climate change [FIELD, C.B., V. BARROS, T.F. STOCKER, D. QIN, D.J. DOKKEN, K.L. EBI, M.D. MASTRANDREA, K.J. MACH, G.K. PLATTNER, S.K. ALLEN, M. TIGNOR, AND P.M. MINDGLEY ( eds.)]. Cambridge University Press, Cambridge, UK, and New York, NY, USA, 582 pp. JOBIN, B., AND J. PICMAN 1997. Factors affecting predation on artificial nests in marshes. The Journal of Wildlife Management 61:792 800.

PAGE 130

130 LANCE, V A. 2003. Alligator physiology and life history: the importance of temperature. Experimental Gerontology 38:801 805. LANGELAND, K. A. 1996. Hydrilla verticillata (L.F.) Royle (Hydrocharitaceae), the perfect aquatic weed ". Castanea 61:293 304. LARIVIRE, S. 1999. Reasons wh y predators cannot be inferred from nest remains. The Condor 101:718 721. LEFTWICH, P., D. ZIERDEN, AND M. GRIFFIN 2010. Climate s ummary for Florida January 2010. Florida Climate Center Tallahassee, FL. Available at www.climatecenter.fsu.edu LIEBEZEIT, J. R., AND T. L. GEORGE 2002. Nest predators, nest site selection, and nesting success of the Dusky Flycatcher in a managed Ponderosa Pine forest The Condor 104:507 517. MARTIN, J., W. M. KITCHENS, M. OLI, AND C. E. CATTAU. 2008 Relative importance of natural disturbances and habitat degradation on Snail Kite population dynamics. Endangered Species Research 6:25 39. MARTIN, J., J. D. NICHOLS, W. M. KITCHENS, AND J. E. HINES 2006. Multiscale patterns of m ovement in fragmented landscapes and consequences on demography of the Snail Kite in Florida. Journal of Animal Ecology 75:527 539. MARTIN, T. E 1987. Food as a limit on breeding birds: a life history perspective. Annual Review of Ecology and Systematics 18:453 487. MARTIN, T. E. 1993. Nest predation and nest sites BioScience 43:523 532. MARTIN, T. E., S. JASON, AND C. MENGE. 2000. Nest predation increases with parental activity: separating nest site and parental activity effects. Proceedings: Biologic al Sciences 267:2287 2293. MARZLUFF, J. M. 1988. Do Pinyon J ays alter nest placement based on prior experience? Animal Behaviour 36:1 10. MAYFIELD, H. F. 1975 Suggestions for calculating nest success The Wilson Bulletin 87:456 466. MAZUR, K. M., AND P. C. JAMES 2000. Barred Owl (Strix varia). In The Birds of North America Online no. 508 (A. Poole Ed.) Ithica: Cornell Lab of Ornithology; Retrieved from the Birds of North America Online. [Online] Available at http://bna.birds.cornell.edu/bna/species/508 doi:10.2173 /bna.508

PAGE 131

131 NICHOLSON, D J. 1926. Nesting habits of the Everglade Kite in Florida. The Auk 43:62 67. NILSSON, J.., AND E. SVENSSON. 19 93. Energy constraints and ultimate decisions during egg laying in the Blue Tit. Ecology 74:244 251. NILSSON, S. G. 1984. The evolution of nest site selection among hole nesting birds: the importance of nest predation and competition Ornis Scandinavica 1 5:167 175. OHLENDORF, H. M., R. L. HOTHEM, AND D. WELSH 1989. Nest success, cause specific nest failure, and hatchability of aquatic birds at selenium contaminated Kesterson Reservoir and a reference site The Condor 91:787 796. PERRINS, C. M. 1970a. Th Ibis 112:242 255. PHILIPS, J. R. 2007. Raptor Research and Management Techniques. Hancock House Publishers, Washington, D.C. PIAS, K., Z. WELCH, AND W. KITCHENS 2012. An artificial perch to help Snail Kites handle an exotic apple s nail. Waterbirds 35:347 351. PIAS, K. E. 2012. Foraging ecology of breeding Snail Kites (Rostrhamus sociabilis plumbeus) on Lake Tohopekaliga, Florida, USA. M.S. thesis University of Florida Gainesville PICMAN, J., M. L. MILKS, AND M. LEPTICH 1993. Patterns of predation on passerine nests in marshes: effects of water depth and distance from edge The Auk 110:89 94. PIETZ P. J. 1987. Feeding and nesting ecology of sympatric South Polar and Brown Skuas. The Auk 104:617 627. PI LASTRO, A., L. BIDDAU, G. MARIN, AND T. MINGOZZI. 2001. Female brood desertion increases with number of available mates in the Rock Sparrow. Journal of Avian Biology 32:68 72. POST, W. 1981. The influence of Rice R ats Oryzomys palustris on the habitat use of the Seaside Sparrow Ammospiza maritima. Behavioral Ecology and Sociobiology 9:35 40. RAWLINGS, T., K. HAYES, R. COWIE, AND T. COLLINS 2007. The identity, distribution, and impacts of non native apple s nails in the continental United States. BMC Evolutionary Biology 7:97.

PAGE 132

132 REICHERT, B., C. CATTAU, W. KITCHENS, R. FLETCHER, J. OLBERT, K. PIAS, C. ZWEIG, AND J. WOOD. 2010a. Snail Kite D emography Annual Report 2010. Annual Report. U.S.G.S. Florida Fish and Wildlife Cooperative Research Unit, Universi ty of Florida, Gainesville, FL. REICHERT, B., C. CATTAU, W. KITCHENS, R. FLETCHER, J. OLBERT, K. PIAS, C. ZWEIG, AND J. WOOD. 2011. Snail Kite De mography Annual Report 2011. Annual Report. U.S.G.S. Florida Fish and Wildlife Cooperative Research Unit, Univ ersity of Florida, Gainesville, FL. REICHERT, B. E., J. MARTIN, W. L. KENDALL, C. E. CATTAU, AND W. M. KITCHENS. 2010b. Interactive effects of senescence and natural disturbance on the annual s urvival probabilities of Snail K ites. Oikos 119:972 979. RICE, A. N., J. P. ROSS, A. R. WOODWARD, D. A. CARBONNEAU, AND H. F. PERCIVAL. 2007. Alligato r diet in relation to a lligator m ortality on Lake Griffin, FL. Southeastern Naturalist 6:97 110. RICHARDSON, T. W., T. GARDALI, AND S. H. JENKINS 2009. Revie w and met a analysis of camera effects on avian nest success The Journal of Wildlife Management 73:287 293. RICKLEFS, R. E. 1969. An analysis of nesting m ortality in birds. Smithsonian Contributions to Z oology 9:1. RIDDINGTON, R., AND A. G. GOSLER 1995. Differences in reproductive success and parental qualities between habitats in the Great Tit Parus major. Ibis 137:371 378. RINGSBY, T. H., B. E. STHER, AND E. J. SOLBERG 1998. Factors affecting juvenile Survival in House Sparrow Passer domesticu s. Journal of Avian Biology 29:241 247. RODGERS, J. A. 1998. Fate of artificially supported Snail Kite Rostrhamus sociabilis nests in central Florida, U.S.A. Bird Conservation International 8:53 57. ROTELLA, J. J. 2004 Modeling nest survival data: a com parison of recently developed methods that can be implemented in MARK an d SAS. Animal Biodiversity and C onservation 27:187. SAFINA, C., J. BURGER, M. GOCHFELD, AND R. H. WAGNER 1988. Evidence for prey limitation of Common and Roseate Tern r eproduction. T he Condor 90:852 859. SAS 1989. SAS/STAT User's Guide, 4th ed ition SAS Institue, Cary, NC

PAGE 133

133 SCHOECH, S J. 1996. The effect of supplemental food on body condition and the timing of reproduction in a cooperative breeder the Florida Scrub Jay. The Condor 98:234 244. SHAFFER, T. L., AND A. E. BURGER. 2004. A unified approach to analyzing nest success The Auk 121:526 540. SNYDER, N., AND H. SNYDER. 1969. A comparative study of mollusc predation by limpkins, Everglade Kites and Boat tailed G rackles. Living Bird 8:177 223. SNYDER, N. F. R., S. R. BEISSINGER, AND R. E. CHANDLER 1989. Reproduction and d emography of the Florida Everglade (Snail) Kite. The Condor 91:300 316. STAKE, M. M., AND D. A. CIMPRICH. 2003. Using video to monitor predation at B lack capped Vireo n ests. The Condor 105:348 357. STEENHOF, K 1987. Assessing raptor reproductive success and productivity. Pages 157 170 in Raptor Management Techniques (B.A.G. Penleton, B.A. Millsap, K.W.Kline, and a. D. M. Birds, Eds.). National Wildli fe Federation Scientific and Technical Series, Washington, D.C. STEENHOF, K., AND I. NEWTON 1987. Assessing nesting success and productivity. Pages 181 191 in Raptor Research and Management Techniques (D. M. Bird, and K. L. Bildstein, Eds.). Institute for Wildlife Research, Washington, D.C. STEVENS, A. J., Z. C. WELCH, P. C. DARBY, AND H. F. PERCIVAL. 2002. Temperature effects on Florida a pple snail activity: implications for Snail Kite f oraging success and distribution Wildlife Society Bul letin 30:75 81. SYKES, P. W. 1987a. Snail Kite nesti ng ecology in Florida. Florida Field N aturalist 15:57 70. SYKES, P. W., JR. 1987b. The feeding h abits of the Snail Kite in Florida, USA. Colonial Waterbirds 10:84 92. SYKES, P. W., JR 1987c. Some aspe cts of the breeding biology of the Snail Kite in Florida. Journal of Field Ornithology 58:171 189. SYKES, P. W., JR., AND R. CHANDLER 1974. Use of artificial nest structures by Everglade Kites. The Wilson Bulletin 86:282 284. SYKES, P. W., JR., J. A. RO DGERS, JR., AND R. E. BENNETTS 1995. Snail Kite (Rostrhamus sociabilis) In The Birds of North America Online no. 171 (A. Poole Ed.) Ithica: Cornell Lab of Ornithology; Retrieved from the Birds of North America Online. [Online] Available at h ttp://bna.birds. cornell.edu/ bna/species/ 171 doi:10.2173/bna.171

PAGE 134

134 THOMPSON, F. R. 2007. Factors affecting nest predation on forest songbirds in North America. Ibis 149:98 109. VAN OORT, H., K. A. OTTER, K.T. FORT, AND Z. O. E. MCDONELL 2007. Habitat, dominance, and the phenotypic quality of male Black Capped Chickadees. The Condor 109:88 96. VAN, T. K., W. T. HALLER, AND L. A. GARRARD 1978. The effect of day length and temperature on hydrilla growth and tuber production. Journal of Aquatic Plant Management 16:57. WEATHERHEAD, P. J., AND G. BLOUIN DEMERS 2004. Understanding avian nest predation: why ornithologists should study snakes. Journal of Avian Biology 35:185 190. WELCH, Z. 2004. Littoral vegetation of Lake Tohopekaliga: community descriptions prior to a large scale fisheries habitat enhancement project M.S. thesis University of Florida, Gainesville. WHITAKER, J. O., AND W. J. HAMILTON 1998. Rats and mice, Muridae. Pages 278 282 in Mammals of the Eastern United States 3rd edition. Comstock Publishing Associates, Ithaca, N.Y. WHITEHEAD, A. L., K. A. EDGE, A. F. SMART, G. S. HILL, AND M. J. WILLANS 2008. Large scale predator control improves the productivity of a rare New Zealand riverine duck. Biological Conservation 141:2784 2794.

PAGE 135

135 BIOGRAPHICAL SKETCH Jean received her BS in wildlife b iology from Colorado State University in Fort Collins, Colorado. She is interested in conservation biology and e cology and spent several years working with endangered avian species before pursuing an MS with the University of Florida in 2009.