Asymmetric responses of two larval salamanders to varying competitor density in a response surface design

Background/Question/Methods: Competition is an important community interaction that is affected by many factors including competitor density.  While using equal proportions of competitors (substitution design) is common, often overlooked is the influence that the relative proportions of competitors have on competitive outcomes.  A response surface design manipulates both overall and relative competitor density to more accurately assess the varying competitor densities that occur in natural populations.  The objective of this research was to determine how different relative proportions of larval salamander competitors would affect the outcome of competition.  Spotted (Ambystoma maculatum) and mole (A. talpoideum) salamanders were manipulated in twelve combinations of competitor density at three different overall density levels within cattle tank mesocosms.  Analysis of growth rates, size at metamorphosis, and metamorph production was conducted for each species. 

Results/Conclusions: Both species exhibited significant intraspecific density-dependent growth, confirming previous experiments for these species.  Ambystoma maculatum showed no significant differences in any response comparing treatments that held only conspecifics versus treatments where both species were present.  Similarly, A. talpoideum showed few responses that were different in conspecific only treatments versus tanks that held congeners.  These findings are difficult to interpret, as intra- and interspecific competitor density were confounded.  Using a response surface design allowed further analysis which showed an interaction between intra- and interspecific competition for A. talpoideum that was not apparent for A. maculatum.  At higher intraspecific density, A. talpoideum experienced greater negative effects from increasing A. maculatum density for all growth responses. Size and production of metamorphs were not different between density treatments.  This experiment shows the importance of incorporating multiple density combinations, as different findings can be obtained compared to using substitution designs. Yet, response surface designs are still infrequently used in competition experiments.  Furthermore, a large portion of information on competition relies on substitution design data which may not adequately reflect natural conditions.  Because an important goal of ecology is to understand how competition operates in natural populations, a critical look at experimental design which better mimics natural conditions is warranted.