Cannibalism in Two Subtropical Lady Beetles (Coleoptera: Coccinellidae) as a Function of Density, Life Stage, and Food Supply

Abstract

Cannibalism is an important factor influencing both immature survival and adult reproductive success in aphidophagous lady beetles. This study employed three series of laboratory experiments to characterize life stage-specific cannibalism responses of Coccinella undecimpunctata L. and Cydonia vicina nilotica Mulsant to 1) different conspecific densities, with and without prey, 2) other life stages, and 3) various densities of prey, Myzus persicae Sulzer. All larval instars of both species cannibalized more in the absence of prey than in its presence at all conspecific densities, but in general, cannibalism increased with conspecific density only in the absence of prey, and more strongly for third and fourth instar C. undecimpunctata than for their C. vicina nilotica counterparts. Adults contributed the most cannibalism of any life stage, and eggs were the most vulnerable. In addition to cannibalizing their own and earlier instars, second and third instar C. undecimpunctata sometimes cannibalized third and fourth instars, respectively, and fourth instars occasionally ate pupae. Larvae of C. vicina nilotica were only preyed upon by the same or later stages and pupae, by adults, not fourth instars. A relative vulnerability index was calculated for each life stage based on its net vulnerability to cannibalism by all life stages and plotting these indices revealed species-specific patterns of diminishing vulnerability to cannibalism as a function of life stage. Relative species vulnerability to cannibalism, considering all life stages, was higher for C. undecimpunctata (0.55) than for C. vicina nilotica (0.45). Finally, linear regression was used to characterize the change in propensity for cannibalism between same-instar larval pairs as a function of prey density, which enabled determination of a theoretical upper prey threshold for each larval instar, i.e., the prey density beyond which no cannibalism would be expected. In both species, regressions for third and fourth instars did not intercept the X-axis, suggesting that some cannibalism by these stages was inevitable within the range of prey densities tested.