Animals of a wide range of taxonomic groups mix various food sources to achieve a nutritionally balanced diet. The strategies they adopt to balance multiple nutrients depend on their availability in the environment. Behavioural and physiological adaptations to forage for nutrient-differing food sources have rarely been investigated in respect to nutrient availability in the environment. We developed a simulation model to explore the strategy consumers should adopt in response to the abundance of two nutritionally complementary food types. Results show that (1) consumers should invest more effort in detecting the scarce resource; (2) there is an optimized negative relationship between effort foragers should allocate to find the two types of food; (3) consumers should exhibit higher selectivity when the proportion of food types in the habitat deviates from their optimal ratio in the diet. These findings have important implications for pest control using predators that benefit from plant-based food supplements.

Coupling of multiple barriers to gene-flow, such as divergent local adaptation and reproductive isolation, facilitates speciation. However, alleles at loci that contribute to barrier effects can be dissociated by recombination. Models of linkage between diverging alleles often consider elements that reduce recombination, such as chromosomal inversions and alleles that modify recombination rate between existing loci. In contrast, here, we consider the evolution of linkage due to the close proximity of loci on the same chromosome. Examples of such physical linkage exist in several species, but in other cases, strong associations are maintained without physical linkage. We use an individual-based model to study the conditions under which the physical linkage between loci controlling ecological traits and mating preferences might be expected to evolve. We modelled a single locus controlling an ecological trait that acts also as a mating cue. Mating preferences are controlled by multiple loci, formed by mutations that are randomly placed in the "genome", within varying distances from the ecological trait locus, allowing us to examine which genomic architectures spread across the population. Our model reveals that stronger physical linkage is favoured when mating preferences and selection are weaker. Under such conditions mating among divergent phenotypes is more frequent, and matching ecological trait and mating preference alleles are more likely to become dissociated by recombination, favouring the evolution of genetic linkage. While most theoretical studies on clustering of divergent loci focus on how physical linkage influences speciation, we show how physical linkage itself can arise, establishing conditions that can favour speciation.

The desirable characteristics of effective natural enemies and the causes for failure of biological control efforts have been discussed extensively in the literature, yet predicting which collection site may yield efficient natural enemies remains a challenge. Insect characteristics, such as morphology, physiology, life history and behavior, often vary across geographic cline and location. These variations may reflect phenotypic plasticity across environments, or genetically based local (demic) adaptation. Parameters such as body size, photoperiod response, thermal tolerance and genetic diversity may greatly influence the outcome of biological control efforts. Therefore, geographic variation in such characteristics may be used to optimize the collection site of efficient enemies to be employed in biological control programs. The first step towards this goal is compilation of data on the trait diversity of promising natural enemies across their geographic distribution range. For example, we used published information to compile a database on the geographic distribution of various traits of 92 Orius species (Heteroptera: Anthocoridae), a genus known for its potential contribution to biological control in IPM systems. We discuss how the widespread distribution of this genus in different ecozones should enable the collection of species and populations that differ in various geographically dependent traits relevant to biological control. Finally, we suggest a quantitative method to optimize collection efforts of natural enemies. This approach balances the effects of several natural enemy traits that vary geographically. Lastly, we demonstrate the use of this method by evaluating the potential employment of two geographically distinct populations of O. albidipennis.