COS 89-4 - Comparing models of effective dispersal to assess causal mechanisms of sex-biased dispersal: Wild turkeys as a model

Thursday, August 6, 2009: 9:00 AM
Santa Ana, Albuquerque Convention Center
Richard S. Phillips1, Warren B. Ballard1, Mark C. Wallace1, Ernest B. Fish1, Nancy E. McIntyre2 and Stephen J. DeMaso3, (1)Natural Resources Management, Texas Tech University, Lubbock, TX, (2)Department of Biological Sciences, Texas Tech University, Lubbock, TX, (3)Texas Parks and Wildlife, Austin, TX
Background/Question/Methods

Several studies use effective dispersal distributions (EDDs), distance distributions scaled by home range size, to investigate the relative role of competition as a causal factor in dispersal.  Although numerous studies report evidence of competition via conformity to the geometric distribution, others suggest sampling issues, not process, are responsible for such patterns.  Using known-fate data, we compare dispersal distances of 394 yearling wild turkeys to six commonly cited geometric models of dispersal.  

Results/Conclusions

Sex-specific effective dispersal distances differed among study sites for yearling females, 3.7 to 6.5, but did not differ among yearling males, averaging 2.8 home ranges traversed across all sites.  As expected, data support unequal underlying distributions between sexes as well as general trends of larger raw dispersal distances by females in birds.  Geometric models described 3 of 4 male distance distributions and 1 of 3 female distance distributions.  Data suggesting a dichotomy also exists within sex classes based upon short- and long-distance dispersal.  Although short-distance dispersal follows predictions regarding competition-driven dispersal, long-distance dispersal deviates substantially from the geometric distribution possibly indicating inbreeding avoidance as a causal factor.  No geometric model approximated long-distance dispersal by the female cohort.  Given our results, inference based on sex-specific EDDs may be an artifact of unknown fate sampling and confound causal mechanisms associated with dispersal.  Superficially, such errors would hinder our ability to model dispersal among populations.  Fundamentally, current knowledge of causal mechanisms associated with dispersal, particularly long distance dispersal, may be flawed.  Despite logistical difficulties, we suggest a greater need for known-fate datasets addressing dispersal questions.

Copyright © . All rights reserved.
Banner photo by Flickr user greg westfall.