Orthopteran insects are characterized by high variability in body coloration, in particular featuring a widespread green-brown color polymorphism. The mechanisms that contribute to the maintenance of this apparently balanced polymorphism are not yet understood. To investigate whether morph-dependent microhabitat choice might contribute to the continued coexistence of multiple morphs, we studied substrate choice in the meadow grasshopper Pseudochorthippus parallelus. The meadow grasshopper occurs in multiple discrete, genetically determined color morphs that range from uniform brown to uniform green. We tested whether three common morphs preferentially choose differently colored backgrounds in an experimental arena. We found that a preference for green backgrounds was most pronounced in uniform green morphs. If differential choices improve morph-specific performance in natural habitats via crypsis and/or thermoregulatory benefits, they could help to equalize fitness differences among color morphs and potentially produce frequency-dependent microhabitat competition, though difference appear too small to serve as the only explanation. We also measured the reflectance of the grasshoppers and backgrounds and used visual modelling to quantify the detectability of the different morphs to a range of potential predators. Multiple potential predators, including birds and spiders, are predicted to distinguish between morphs chromatically, while other species, possibly including grasshoppers themselves, will perceive only differences in brightness. Our study provides the first evidence that morph-specific microhabitat choice might be relevant to the maintenance of the green-brown polymorphisms in grasshoppers and shows that visual distinctness of color morphs varies between perceivers.

There are three datasets:

1) Data on substrate choice has been collected by visual checks at irregular intervals. Date and time of recording are given in the file. Each row refers to a record of an individual at a specific point in time. Patch number and patch color are readouts. There are repeated measurements per individual and attributes of individuals (sex, color morph, developmental stage) are given. There are many missing values (indicated by NA) when an individual was still alive, but not sitting on the floor.

2) Data on patch temperatures. Records refer to temperature recordings of green and brown experimental substrates and are always paired (one green and one brown patch recorded).

3) Data on spectrometric reflectance measurements of grasshoppers and substrates (recorded by a hand-held spectrophotometer, AvaSpec-2048, Avantes). Data from the recording device has been compiled in a single table and rounded to 3 digits. The data are somewhat oddly formatted, but are ready to be used for analysis by the pavo package in R. Each row refers to a wavelength (given in the first column) and all other columns refer to individual records (reflectance of a particular body part or substrate at a particular wavelength). Attributes of the record are encoded in the column header (see readme file for details).

There are readme files for all datasets. For the reflectance data there are five repeated measurements per patch (indicated by the last two digits of the column header) and these have later been averaged in the analysis. There are very few cases with fewer measurements.