Behavioral Ecology Vol. 14 No. 4: 583-591
© 2003 International Society for Behavioral Ecology
The effect of intra- and interspecific aggression on patch residence time in Negev Desert gerbils: a competing risk analysis
a Department of Life Sciences, Ben-Gurion University of the Negev, Beer-Sheva, Israel b School of Forestry and Environmental Studies, Yale University, 370 Prospect Street, New Haven, CT 06511, USA
Address correspondence to O. Ovadia. E-mail: ofer.ovadia{at}yale.edu.
We observed patch-use behavior by two gerbil species in a field setting and investigated how aggression and intrinsic decision-making interact to influence patch residence times. Results were interpreted by using a competing risk analysis model, which uniquely enabled us to estimate the intrinsic patch-leaving decisions independently of external interruptions of foraging bouts by aggression. The experiment was conducted in two 1-ha field enclosures completely surrounded by rodent-proof fences and included allopatric (only Gerbillus andersoni allenbyi) and sympatric (G. a. allenbyi and G. pyramidum) treatments. We predicted that increased food patch quality (i.e., habitat quality) should decrease intrinsic patch-leaving rates and increase rates of aggressive interactions involving the forager feeding in the patch (i.e., the occupant individual). We also anticipated that increasing population density should result in an increase in the rate of aggressive interactions involving the occupant individual. Our results supported the first two predictions, indicating a trade-off between foraging and aggression. However, the third prediction was realized only for G. a. allenbyi in allopatry. Furthermore, in allopatry, occupant G. a. allenbyi individuals with high competitive ranks were involved in aggressive interactions at lower rates than those with low competitive ranks. However, in sympatry, patch-use behavior of occupant G. a. allenbyi individuals was mainly influenced by aggressive behavior of G. pyramidum, which did not respond to their competitive rank. Thus, it should pay less for G. a. allenbyi to be aggressive in sympatric populations. The observed reduction in intraspecific aggression among individual G. a. allenbyi in the presence of G. pyramidum supports this assertion. We suggest that this reduction likely weakens the negative effect of intra- and interspecific density on the per capita growth rate of G. a. allenbyi. Because this would change the slope of the isocline of G. a. allenbyi, it could be an important mechanism promoting coexistence when habitat selection is constrained.
Key words: aggression, competing risk method, intrinsic decision-making, patch residence time, patch use, survival analysis.