Data from: Repeated evolution of reduced visual investment at the onset of ecological speciation in high-altitude Heliconius butterflies
Data files
Sep 08, 2025 version files 65.69 KB
Abstract
Colonisation of new habitats is typically followed by divergent selection acting on traits that are immediately important for fitness. For example, differences between sensory environments are often associated with variation in sensory traits critical for navigation and foraging. However, the extent to which the initial response to novel sensory conditions is mediated by phenotypic plasticity, and the contribution of sensory or neural adaptation to early species divergence remains unclear. We took advantage of repeated cases of speciation in Heliconius butterflies with independent allopatric distributions in the west of the Colombian and Ecuadorian Andes. Using volumetric brain measurements, we analysed patterns of investment in primary sensory processing areas of the brain across different localities and habitats. We find that a higher-altitude species, Heliconius chestertonii, differs in levels of investment in visual and olfactory brain components compared to its lower altitude relative H. erato venus, mainly attributable to broad-sense heritable variation as inferred from comparisons between wild and common-garden reared individuals. We provide evidence that this variation is consistent with divergent selection, and compare these shifts with those reported for another high-altitude species, H. himera, and its parapatric lowland counterpart, H. erato cyrbia, to demonstrate parallel reductions in the size of specific optic lobe neuropils. Conversely, for the antennal lobe, we detected different trait shifts in H. himera and H. chestertonii relative to their lowland H. erato neighbours. Overall, our findings add weight to the adaptive potential of neuroanatomical divergence related to sensory processing during early species formation.
https://doi.org/10.5061/dryad.02v6wwqdg
Description of the data and file structure
The data in the file Supplementary_Tables-Repeated_evolution_of_reduced_visual_investment_at_the_onset_of_ecological_speciation_in_hig-altitude_Heliconius_.xlsx contain the following:
- Raw and log-transformed volumetric measurements of neuropils and allometric controls for wild-caught and insectary-reared Heliconius erato venus and Heliconius chestertonii individuals (S1). Column 1 indicates the individual's butterfly species, 2 its origin (wild caught or insectary reared), 3 its sex (male or female), 4 the year of collection and 5 its unique identification number. The neuropils (structures of the brain, as described in the article) shown are the medulla (ME), the lobula plate (LOP), the lobula (LO), the ventral lobula (vLO), the accessory medulla (AME), the anterior optic tubercle (AOTu), the antennal lobe (AL), the laminated (LA), the central complex (CC), the protocerebral bridge (PB), the posterior optic tubercle (POTu), the mushroom body (MB) and the central brain (CBR). The volumes of these structures are presented raw (cubic micrometers, columns 6-18) and log-transformed (columns 19-32),
- Log-transformed volumetric measurements of individuals of Heliconius chestertonii, Heliconous erato venus, Heliconius himera and Heliconius erato cyrbia (S2). Columns 1-5 convey species, origin, sex, year of collection and unique identification number. Column 7 indicates habitat type (high or low altitude forest) and column 8 indicates locality (Colombia or Ecuador). Columns 8-16 indicate neuropils selected for parallel evolution analyses (note this is a subset of, rather than all of the neuropils, as described in the article).
- Results of allometric scaling analyses of brain variation between Heliconius erato venus and Heliconius chestertonii across groups of different origin (wild S3 and insectary S4) as well as within each species (wild vs insectary Heliconius chestertonii, S5 and wild vs insectary Heliconius erato venus, S6). In each case, it is indicated, for the neuropil under comparison, its location in the brain (optic lobe or central brain) and the parameters of tests for differences in slope, elevation and major-axis in scaling (each of which is described fully in the main article).
- Results of Pst (neutral phenotypic variation)-Fst (neutral genomic variation) comparisons (S7) for different neuropils. For each neuropil, a Pst value is presented together with Pst estimate confidence intervals and the proportion of genomic Fst that lies under the Pst value. Columns 3-7 show Pst under strict sense heritability scenarios (squared h) ranging from 1 (column 3) to 0.5 (column 7).
- Tests of correlation in log-transformed neuropil size in the medulla, the lobula and the lobula plate (S8).
- Neuropil volumes predicted from the allometric scaling function fitted for each species (S9). The parameters shown are the slope (column 5), the intercept (column 6) the mean volume of the allometric control (column 7), the lower confidence interval for mean volume of the allometric control (column 8), the upper confidence interval for the mean volume of the allometric control (column 9), the predicted neuropil volume (column 10), the lower confidence interval for neuropil volume (column 11) and the upper confidence interval for neuropil volume (column 12)
The file "Supplemental Information.docx" contains details of an ancestral character reconstruction of altitude across butterfly species.
Code/software
The annotated scripts include SMATR regressions and Pst estimations as well as code for the variance partitioning approach used to infer levels of paralellism.
Tables 1-9 contain raw data and analyses results. Scripts are attached separatedly
Supplementary information 1 contains supplementary methods, results and discussion related to an ancestral state reconstruction of altitude in Heliconius species*.*