Geospatial Modeling of Common Raven Activity in Snowy
Total Page:16
File Type:pdf, Size:1020Kb
GEOSPATIAL MODELING OF COMMON RAVEN ACTIVITY IN SNOWY PLOVER HABITATS IN COASTAL NORTHERN CALIFORNIA By Matthew Joseph Lau A Thesis Presented to The Faculty of Humboldt State University In Partial Fulfillment of the Requirements for the Degree Master of Science in Natural Resources: Wildlife Committee Membership Dr. Mark A. Colwell, Committee Chair Dr. Daniel Barton, Committee Member Dr. William T. Bean, Committee Member Dr. Yvonne Everett, Graduate Coordinator December 2015 ABSTRACT GEOSPATIAL MODELING OF COMMON RAVEN ACTIVITY IN SNOWY PLOVER HABITATS IN COASTAL NORTHERN CALIFORNIA Matthew Joseph Lau The Common Raven (Corvus corax) poses a conservation dilemma because as a native predator it can negatively affect populations of other native species, including the Western Snowy Plover (Charadrius nivosus nivosus). In Humboldt County, the Common Raven is one of the primary causes of low reproductive success in Snowy Plovers. To better understand Common Ravens, I investigated their activity and distribution in Snowy Plover habitats using 11 years of point count data (2004-2014). Furthermore, I analyzed several landscape factors known to influence raven activity at three spatial scales and related them to Common Raven activity using Generalized Additive Models (GAMs). Common Raven distribution varied appreciably across Snowy Plover habitats and this spatial patterning was consistent across the 11 years. Moreover, Common Raven activity was highest in Snowy Plover habitats that were near more agricultural lands and low- intensity urban areas at all scales (small and large scale). Common Ravens were found to be in high abundance coinciding with areas of high Snowy Plover breeding activity, which warrants prioritizing predator management in these beach habitats. ii ACKNOWLEDGMENTS I would first like to thank my graduate advisor, Dr. Mark Colwell, for the incredible amount of knowledge and inspiration that he has passed on to me, both as his Master’s student and his undergraduate advisee. He has provided me the stepping stones to achieve all the success I have acquired and has stimulated my love for birds. I would also like to tremendously thank my committee members, Dr. Daniel Barton and Dr. William T. Bean, for offering the time and energy to assist me in spatial analyses and statistical modeling. I want to acknowledge my fellow graduate students in the Shorebird Ecology Lab for their companionship, input, and survey effort: Allie Patrick, Dana Herman, Matt Brinkman, Stephanie Leja, David Orluck, Alexa DeJoannis, and Teresa King. I am grateful for the numerous field surveyors that have helped collect data: Kayla Bonnette, Aaron Gottesman, Chloe Joesten, Garrett Moulton, Elizabeth Feucht, Derek Harvey, Jasmin Ruvalcaba, Grayson Sandy, and Maryjean Greitl. I would also like to extend my thanks to the following from California State Parks: Amber Transou, Jay Harris, Carol Wilson, Mark Morrisette, Casey Ryan, and Tony Kurz. Additionally, I am thankful to Jim Watkins of the U.S. Fish and Wildlife Service and Sean McAllister. I would also like to thank Jim Graham and Jeff Dunk for helping me with understanding the complexities of Generalized Additive Models. Anthony Desch and the rest of the Wildlife faculty were also a crucial part of my success as a graduate student. My research was funded by: Humboldt State University, U.S. Fish and Wildlife Service, Bureau of iii Land Management, California State Parks, the Asian Pacific Islander Organization, the Sonoma County Fish and Wildlife Commission, Stockton Sportsmen’s Club, and Marin Rod and Gun Club. Lastly, but most importantly, I would like to thank my family, for their everlasting support and endurance. I am eternally indebted to all those above and my success is not without them. iv TABLE OF CONTENTS ABSTRACT ........................................................................................................................ ii ACKNOWLEDGMENTS ................................................................................................. iii TABLE OF CONTENTS .................................................................................................... v LIST OF TABLES ............................................................................................................ vii LIST OF FIGURES ......................................................................................................... viii LIST OF APPENDICES .................................................................................................... ix INTRODUCTION .............................................................................................................. 1 METHODS ......................................................................................................................... 6 Study Area ...................................................................................................................... 6 Field Methods ................................................................................................................. 6 Analytical Methods ......................................................................................................... 8 Analyses of Common Raven Distribution and Activity ............................................. 8 Geospatial Modeling of Common Raven Activity ................................................... 10 RESULTS ......................................................................................................................... 16 Common Raven Distribution and Activity ................................................................... 16 Geospatial Modeling ..................................................................................................... 18 DISCUSSION ................................................................................................................... 25 Common Raven Distribution and Activity ................................................................... 25 Landscape Correlates of Common Raven Activity ...................................................... 27 MANAGEMENT IMPLICATIONS ................................................................................ 31 LITERATURE CITED ..................................................................................................... 33 v Appendix A ....................................................................................................................... 38 Appendix B ....................................................................................................................... 39 Appendix C ....................................................................................................................... 40 Appendix D ....................................................................................................................... 41 vi LIST OF TABLES Table 1. Predictor variables, their definitions, and data sources used in geospatial modeling of Common Raven activity using point count data from 2004-2014. .............. 11 Table 2. Model selection results evaluating relationships between Common Raven activity and anthropogenic variables at three spatial scales: 500 m, 1,450 m, and 3,590 m. ........................................................................................................................................... 19 Table 3. Summary of effects of each covariate on Common Raven activity at three focal spatial scales (500 m, 1,450 m, and 3,590 m), based on top models. A (+) indicates a positive effect on Common Raven activity and a (-) indicates a negative relationship. A (+/-) either indicates no significant effect or no clear effect. ............................................ 27 vii LIST OF FIGURES Figure 1. Map of ocean-fronting beach and gravel bars of the lower Eel River in Humboldt County, California, where observers collected point count data to quantify Common Raven numbers during surveys for Snowy Plovers, 2004 – 2014. ..................... 7 Figure 2. An example of 500 m x 1,500 m (north-south) grids overlaying Clam Beach with associated point count locations from 2004-2014. ................................................... 13 Figure 3. Map of average Common Ravens detected per grid cell from Gold Bluffs Beach south to gravel bars of the Eel River, calculated from point count data, 2004-2014........ 17 Figure 4. Response curve outputs for each of the covariates from the top Generalized Additive Model (GAM) of the 500 m scale (RAVENS ~ HUM + ROAD + AGR + URBL + WATER). The curves show the modeled effects as solid lines, with 95% Bayesian credible intervals in shaded grey. The y-axis is on that of the linear predictor, but due to identifiability constraints, they are presented in a mean-centered fashion. The labels indicate the predictor variable with its associated effective degrees of freedom (Wood 2006). ................................................................................................................................ 20 Figure 5. Response curve outputs for each of the covariates from the top Generalized Additive Model (GAM) of the 1,450 m scale ( RAVENS ~ AGR + URBL + URBH + WATER + FOR). The curves show the modeled effects as solid lines, with 95% Bayesian credible intervals in shaded grey. The y-axis is on that of the linear predictor, but due to identifiability constraints, they are presented in a mean-centered fashion. The labels indicate the predictor variable with its associated effective degrees of freedom (Wood 2006). ...................................................................................................................