Abstract1
In temporally correlated environments, current and past conditions provide information that is predictive about future conditions. Selection may then favour the ability to acquire and respond to such information, leading to adaptive plasticity. In some cases, the time needed to obtain sufficient information for reliable prediction will extend beyond that of a generation, or beyond the time window individuals have before the plastic response must take place. Selection could then favour the transmission of information regarding current conditions to later generations, so that developmental or behavioural decisions of individuals can depend also on inherited information. Epigenetic factors, intimately connected to trait expression, have been shown to transmit across several generations and could play the role as information bearer. We have formulated a simple but general model in order to study the circumstances under which the transmission of acquired information across generations is adaptive. The model allows the degree of transmission to evolve: individuals acquire information of current environmental state that with some probability is passed on to offspring, and to their offspring in turn. The model takes into account two major sources of uncertainty: that inherent to the process of acquiring information and that resulting from incomplete temporal correlation of environmental conditions. In the simplest cases, the model allows for some analytic insight that connects with theories of bet-hedging. We extend upon these analytic results by means of computer simulations and find that it can be adaptive to transmit information across several generations, and discuss this in relation to various epigenetic factors.
Abstract2
Many prey use aposematic signals to warn predators that they are toxic, and predators can quickly learn to avoid aposematic prey. If the prey vary in toxicity, the predators may instead learn to capture and taste them carefully before ingesting or rejecting them (go-slow behavior). It has long been held that increases in prey toxicity will decrease predation on prey populations. However, many toxins are also distasteful, and while prey with higher toxin loads are more harmful to ingest, they can also be easier to recognize and reject due to higher distastefulness, which might facilitate a taste-sampling foraging strategy in the predators. Here, a foraging model is presented that accounts for the separate effects of taste and toxicity on predator preferences. The model is applicable to automimicry and Batesian mimicry. A key prediction is that a moderately toxic model population may provide a mimicry complex with better protection against predation than a highly toxic one, everything else being equal. Taste mimicry can make a mimicry complex unprofitable to the predator and thus increase protection from predation. In contrast, an increase in the concentration of a distasteful non-toxin in the model prey can reduce the protection obtained at the population level, even if the increase is favored by individual selection. I discuss how the taste-sampling process may ultimately affect the evolution of the warning signals.