A predictive adaptive response (PAR) is a type of developmental plasticity where the response to an environmental cue is not immediately advantageous but instead is later in life. The PAR is a way for organisms to maximize fitness in varying environments. Insects living in seasonal environments are valuable model systems for testing the existence and form of PAR. Previous manipulations of the larval and the adult environments of the butterfly Bicyclus anynana have shown that individuals that were food restricted during the larval stage coped better with forced flight during the adult stage compared to those with optimal conditions in the larval stage. Here, we describe a state-dependent energy allocation model, which we use to test whether such a response to food restriction could be adaptive in nature where this butterfly exhibits seasonal cycles. The results from the model confirm the responses obtained in our previous experimental work and show how such an outcome was facilitated by resource allocation patterns to the thorax during the pupal stage. We conclude that for B. anynana, early-stage cues can direct development toward a better adapted phenotype later in life and, therefore, that a PAR has evolved in this species.
number of larvae, pupae and adults (figure 2 from paper)
In this file are numbers of larvae, pupae and adults shown from one forward simulation. There are 10 columns. The first indicates time in days. Then three columns indicate total number of larvae, pupae and adults followed by number of larvae, pupae and adults in good patches (g_) and bad patches (b_).
individuals.txt
behavior
This file lists the number of individuals that feeding, flying or reproducing in good patch (g_) and bad patches (b_) dependent on time in the year. This data is used to produce figure 3A, 3B and 3C in the paper
traits
This file lists the average fat (fat), abdomen (abd) and thorax (tho) weight and biological age (age) in good (g_) and bad pacthes (b_) of model simulations dependent on time. This data is used to create figure 3D, 3E and 3F.
individual survival data
This file contains simulation data from individuals listing that state, optimal behavior and optimal reallocation to maintenance and repair decision. Data is listed dependent on year (only 1), time in the year, patch quality (pq, 0=good, 1=bad), individual number (ind), fat (fat), abdomen (abd) and thorax (tho) weight, fecundity (fec), age in number of days (age), biological age (bdamage), optimal behavior decision (beh), optimal reallocation strategy (real), proportion of fat used for repair (rfat), absolute fat used for repair (ufat) and generation of which the adult belongs to. This all is used for figure 3G.
thorax ratio and abdomen weight
File is the summary of thorax weight and fecundity data from both model and earlier performed experiment. Source (lab experiment or model simulation), food treatment (ad libitum or restricted), trait (thorax ratio or fecundity in experiment and abdomen weight in model) and mean and standard deviation (sd) of these variables are listed. This data is used for figure 4 of the paper.
data experiment PAR
Summarizing data from an experiment in which larvae and adults were stressed and optmally treated from both a laboratory experiment and a model simulation. Source indicates where data is from (lab experiment, 1st, 2nd or 3rd generation butterfly simulation), which trait is measured (trait), larval and flight stress treatments and means and standard deviations of the measurements. This data is used for figure 5 of the paper
pupal weight data
This file contains the individual pupal weight data at the initiation of pupation from simulations. Listed are the individual number (ind), time, patch quality (pq, 0=good, 1=bad) and actual pupal weight in mgs.
allocation strategy pupae
This file lists data from allocation strategies of pupae from the simulations. Columns list indvidual number (ind), time, generation the individual belongs to, allocation to fat (fat), abdomen (abd) and thorax (tho) and patch quality (pq, 0=good, 1=bad). This data was used to produce figure A4.