Trait-Specific Indirect Effects Underlie Variation in the Response of Spiders to Cannibalistic Social Partners.

AbstractIn cannibalistic species, selection to avoid conspecifics may stem from the need to avoid being eaten or to avoid competition. Individuals may thus use conspecific cues to modulate their behavior to such threats. Yet the nature of variation for such cues remains elusive. Here, we use a half-sib/full-sib design to evaluate the contribution of (indirect) genetic or environmental effects to the behavioral response of the cannibalistic wolf spider Lycosa fasciiventris (Dufour, 1835) toward conspecific cues. Spiders showed variation in relative occupancy time, activity, and velocity on patches with or without conspecific cues, but direct genetic variance was found only for occupancy time. These three traits were correlated and could be lumped in a principal component: spiders spending more time in patches with conspecific cues moved less and more slowly in those areas. Genetic and/or environmental components of carapace width and weight loss in the social partner, which may reflect the quality and/or quantity of cues produced, were significantly correlated with this principal component, with larger partners causing focal individuals to move more slowly. Therefore, environmental and genetic trait variation in social partners may maintain trait diversity in focal individuals, even in the absence of direct genetic variation.

Food web engineering: ecology and evolution to improve biological pest control.

If we are to sustainably provide food to a rapidly growing human population, biological pest control (BPC) should integrate food web theory and evolution. This will account for the impacts of climate warming on the complex community settings of agroecosystems. We review recent studies looking for top-down augmentative pest control being hampered/promoted by biotic (community contexts) and/or abiotic (climate) drivers. Most studies found either positive or neutral effects on BPC. However, most ignored potential evolutionary responses occurring in the environments under study. We propose engineering food webs by engaging in a continuous feedback between ecological and evolutionary data, and individual-based modelling of agroecosystems. This should speed up the procurement of strains of efficient natural enemies better adapted to warming.

The sources of variation for individual prey-to-predator size ratios.

The relative body size at which predators are willing to attack prey, a key trait for predator-prey interactions, is usually considered invariant. However, this ratio can vary widely among individuals or populations. Identifying the range and origin of such variation is key to understanding the strength and constraints on selection in both predators and prey. Still, these sources of variation remain largely unknown. We filled this gap by measuring the genetic, maternal and environmental variation of the maximum prey-to-predator size ratio (PPSRmax) in juveniles of the wolf spider Lycosa fasciiventris using a paternal half-sib split-brood design, in which each male was paired with two females and the offspring reared in two food environments: poor and rich. Each juvenile spider was then sequentially offered crickets of decreasing size and the maximum prey size killed was determined. We also measured body size and body condition of spiders upon emergence and just before the trial. We found low, but significant heritability (h2 = 0.069) and dominance and common environmental variance (d2 + 4c2 = 0.056). PPSRmax was also partially explained by body condition (during trial) but there was no effect of the rearing food environment. Finally, a maternal correlation between body size early in life and PPSRmax indicated that offspring born larger were less predisposed to feed on larger prey later in life. Therefore, PPSRmax, a central trait in ecosystems, can vary widely and this variation is due to different sources, with important consequences for changes in this trait in the short and long terms.

New Litter Trap Devices Outperform Pitfall Traps for Studying Arthropod Activity.

Soil fauna play a key role in nutrient cycling and decomposition, and in recent years, researchers have become more and more interested in this compartment of terrestrial ecosystems. In addition, soil fauna can act as ecosystem engineers by creating, modifying, and maintaining the habitat for other organisms. Ecologists usually utilize live catches in pitfalls traps as a standard method to study the activity of epigeic fauna in addition to relative abundance. Counts in pitfall traps can be used as estimates of relative activity to compare among experimental treatments. This requires taking independent estimates of abundance (e.g., by sifting soil litter, mark-recapture), which can then be used as covariates in linear models to compare the levels of fauna activity (trap catches) among treatments. However, many studies show that the use of pitfall traps is not the most adequate method to estimate soil fauna relative abundances, and these concerns may be extensible to estimating activity. Here, we present two new types of traps devised to study activity in litter fauna, and which we call "cul-de-sac" and "basket traps", respectively. We experimentally show that, at least for litter dwellers, these new traps are more appropriate to estimate fauna activity than pitfall traps because: (1) pitfall traps contain 3.5× more moisture than the surrounding environment, potentially attracting animals towards them when environmental conditions are relatively dry; (2) cul-de-sac and basket traps catch ca. 4× more of both meso- and macrofauna than pitfall traps, suggesting that pitfall traps are underestimating activity; and (3) pitfall traps show a bias towards collecting 1.5× higher amounts of predators, which suggests that predation rates are higher within pitfall traps. We end with a protocol and recommendations for how to use these new traps in ecological experiments and surveys aiming at estimating soil arthropod activity.