Black-Shouldered Kite - Small Mammal - Vegetation Relationships in Northwestern California
Total Page:16
File Type:pdf, Size:1020Kb
Black-shouldered kite - Small mammal - Vegetation Relationships in Northwestern California by Jeffrey R. Dunk A Thesis Presented to The Faculty of Humboldt State University In Partial Fulfillment of the Requirements for the Degree Master of Science February 1992 BLACK-SHOULDERED KITE - SMALL MAMMAL - VEGETATION RELATIONSHIPS IN NORTHWESTERN CALIFORNIA by Jeffrey R. Dunk Approved by the Master's Thesis Committee Robert J. Cooper, Chairman David W. Kitchen Timothy E. Lawlor Director, Natural Resources Graduate Program Date 92/W-235/02/14 Natural Resources Graduate Program Number Approved by the Dean of Graduate Studies Susan H. Bicknell ABSTRACT This thesis is written in 2 sections to divide the major themes of the study. The second section is partially based on the work presented in the first. I studied the relationships between populations of both California voles (Microtus californicus) and Western harvest mice (Reithrodontomys megalotis) and vegetation structure and composition in northwestern California. Both species exhibited annual fluctuations in abundance during the 19 months of the study. In general, M. californicus populations were largest in areas with lower cover height and larger percent brown grass cover. This finding is contrary to the majority of Microtus spp. studies that have examined population - vegetation relationships. The discrepancy is likely due to the fact that all areas that I trapped had fairly abundant cover (no areas were grazed). The increase of M. californicus populations (from the annual low) coincided with grass seeds being more available (i.e., dropping off of grasses). To my knowledge I recorded the largest annually fluctuating M. californicus population ever found (range = 0 - 1414 ind/ha, Jolly-Seber estimate). I propose that resource limitation is a primary factor in M. californicus population fluctuations. R. megalotis populations fluctuated synchronously iii iv with those of M. californicus, but at much reduced numbers. Increasing populations of R. megalotis were associated with an increased availability of grass seeds. The largest populations of R. megalotis were associated with relatively low cover height of vegetation. Each month I developed discriminant function analysis models to predict both species of small mammal abundances using vegetation (structure and composition) variables. Mean percent correct classification was 70.03% and 88.2% for M. californicus and R. megalotis, respectively (n = 19). I also studied the relationship between black- shouldered kite (Elanus caeruleus) territory size and both prey abundance and competitor abundance in northwestern California. Kite territory size ranged from 1.6 - 21.5 ha (n = 26). Estimated mean number of M. californicus per territory was 1483 (SE = 163, n = 25). Competitor abundance (i.e. total raptor abundance) ranged from 4.8 - 31.0 individuals/km2 and was strongly correlated with abundance of M. californicus. Both estimated prey abundance and competitor abundance were negatively correlated with kite territory size. After developing a multiple regression model using both variables, partial correlation analysis revealed that once the effects of prey abundance were statistically V controlled, competitor abundance continued to be significantly correlated with kite territory size. When the effects of competitor abundance were statistically controlled, prey abundance was no longer significantly correlated with kite territory size. I conclude that kite territory size is proximately regulated by competitor abundance and ultimately regulated by prey abundance because M. californicus abundance regulates numbers of raptors. ACKNOWLEDGMENTS I am indebted to so many people for their help, support, and encouragement prior to and throughout this entire experience that I am not quite sure where to begin. I will start by thanking my parents, Bill and Barbara Dunk, for their support (v and $) from the first mention of this endeavor. They continued this support even after they found out that scientists have to pay to have their work published. I thank Grandma Lou, Uncle Mike and Aunt Pat, and Bill and Marilyn Gillaspy for introducing me to natural areas at a young age and showing me how to enjoy them. It is very unlikely that I would be in this field had it not been for them. I thank G. Monroe of California Department of Fish and Game for allowing me access to state property to conduct my study. I also thank H. and P. Hunt for their thoughtful consideration of my project during their activities on the study area. Dr. James R. Koplin (deceased) first introduced me to the system and animals that I studied. For those that knew him, his influence on me will be obvious. My friends TallChief A. Comet, Douglas G. Leslie, and I had many discussions on black-shouldered kites before this project began. Their support, enthusiasm, and ideas are largely responsible for my enthusiasm and my thesis topic. vi vii I thank my friends S. Beatty, J. Browning, K. Bradley, T. Burgess, D. Call, T. Comet, R. Cooper, S. Dunk, L. Ellis, R. Grosz, B. and D. Kristan, D. Leahy, D. Leslie, R. Lightfoot, R. Long, G. Roemer, S. Steinberg, L. Tisue, C. Verhey , and N. Weisman for providing help and companionship in the field. Thanks also to A. Franklin for letting me sub-permit under his banding permit. I thank my committee members Drs. T. E. Lawlor and D. W. Kitchen for their many helpful comments on drafts of my thesis. I would also like to thank K. Moon and R. Brown for their assistance whenever it was asked. Sabra Steinberg helped during all phases of my graduate life. I thank her for her support, encouragement, companionship, and tolerance (of my pre-dawn ventures into the field each month) throughout. I am indebted to my major professor, Dr. Robert J. Cooper, for his encouragement, friendship, guidance, and support in every phase of this project. He went well beyond his "duties" as a major professor. Partial funding for this project came from 2 Humboldt State University Research and Creative Activity grants to Dr. Cooper. Lastly, I would like to thank the Wildlife Department at HSU for attracting such high caliber students. It was while interacting with these people that my greatest educational experiences occurred. TABLE OF CONTENTS Page ABSTRACT . • . iii ACKNOWLEDGMENTS . • • vi LIST OF TABLES . ix LIST OF FIGURES . • . xi GENERAL INTRODUCTION • • • . • • • • 1 STUDY AREA . 2 SMALL MAMMAL - VEGETATION RELATIONSHIPS . 4 INTRODUCTION . 4 METHODS . • 9 Small mammal abundance . 9 Vegetation characteristics . 11 RESULTS . • • 13 Microtus californicus . 13 Reithrodontomys megalotis . 33 DISCUSSION . 41 Microtus californicus . 41 Reithrodontomvs megalotis . 48 TERRITORY SIZE REGULATION IN BLACK-SHOULDERED KITES. 50 INTRODUCTION . 50 METHODS . 53 RESULTS . • • 58 DISCUSSION • • • • • • • • . • 63 LITERATURE CITED . 69 viii LIST OF TABLES Table Page 1 Comparison of 4 methods of estimating California vole (Microtus californicus) populations. All estimates are rounded to the nearest whole number. Numbers in parentheses are standard errors. (na) means that the calculation could not be made. Fay Slough Wildlife Area, Eureka, California, June 1989 - December 1990. 15 2 Sex ratios of California voles (Microtus californicus) (percent female) caught on the Fay Slough Wildlife area, Eureka, California. June 1989 - December 1990. Total number of individuals (male and female) sexed in parentheses. Dashes (-) represent no individuals caught or sex not determined. 19 3 Jolly-Seber estimates of monthly survival probabilities (and SE) for Microtus californicus caught on the Fay Slough Wildlife Area, Eureka, California, from June 1989 - December 1990. Dashes (-) represent parameters not estimated by the model. Blanks represent a grid not being trapped that month. 21 4 Four-group stepwise discriminant analysis of Microtus californicus abundance in ungrazed grassland habitat. Only variables entered into the model are represented. Numbers presented are structure coefficients for the first two discriminant functions (DF1 and DF2). Fay Slough Wildlife Area, Eureka, California. June 1989 - December 1990. 29 5 Evaluation of Stepwise discriminant analysis models ability to predict Microtus californicus abundance based on monthly vegetation characteristics in small mammal grids. Observations were considered correctly classified if they were classified into the grid they came from or a grid with similar M. californicus abundance (similar = within 5 individuals). Fay Slough Wildlife Area, Eureka, California, June 1989 - December 1990. 33 ix • x 6 Stepwise-regression analysis results using Microtus californicus abundance (naive) as the dependent variable and vegetation parameters as independent variablesa. Analyses included increase phases in abundance of M. californicus, decrease phases, and all data regardless of population phase. Only variables entered into models are presented. Results where no variables were entered into models are not included. 34 7 Stepwise-regression analysis results using Reithrodontomys megalotis abundance (naive) as the dependent variable and vegetation parameters as independent variables. Analyses included increase phases in abundance of R. megalotis, decrease phases, and all data regardless of population phase. Only variables entered into models are presented. Results where no variables were entered into models are not included. • • • • 38 8 Evaluation of Stepwise discriminant analysis models ability to predict Reithrodontomys megalotis abundance based on monthly vegetation