Locally adapted populations are often used as model systems for the early stages of ecological speciation, but most of these young divergent populations will never become complete species. The maintenance of local adaptation relies on the strength of natural selection overwhelming the homogenizing effects of gene flow; however, this balance may be readily upset in changing environments. Here I show that soapberry bugs (Jadera haematoloma) have lost adaptations to their native host plant (Cardiospermum corindum) and are regionally specializing on an invasive host (Koelreuteria elegans), collapsing a classic and well-documented example of local adaptation. All populations that were adapted to the native host-including those still found on that host today-are now better adapted to the invasive host in multiple phenotypes. Weak differentiation remains in two traits, suggesting that homogenization across the region is incomplete. This study highlights the potential for adaptation to invasive species to disrupt native communities by swamping adaptation to native conditions through maladaptive gene flow.
J.haematoloma morphology in 1988 & 2013-2015
Morphological data from 1988 (collected by SP Carroll, first published in Carroll & Boyd 1992) and 2013-2015 (collected by ML Cenzer). Column heading definitions: pophost=host plant species from which individuals were collected in the field; population=location name (city or island name); sex=M for male, F for female; beak=length of beak in mm; thorax=width of pronotum at widest point in mm; wing=length of forewing at longest point in mm; body=length from anterior tip of tylus to distal tip of closed forewing in mm (S indicates short-winged individuals - body length measures were not taken for these); month=collection month; year=collection year. Not all measures were taken for all individuals; NAs and blanks both indicate the measure was not taken. Data from two host plants (Cardiospermum microcarpum & Koelreuteria paniculata) not reported on in the manuscript are also included in this data file.
morphology data for DRYAD.csv
J.haematoloma survival in 2014 (F1, cross-rearing experiment)
Survival data for J.haematoloma in cross-rearing experiment conducted on F1 laboratory generation in 2014. Column heading interpretation: Number=individual identifier (because the datasheet was generated before hatching, some numbers were not used for individuals that never hatched); Family=maternal identifier (field-collected, paternity uncertain); bug.pop=location, either city or island name, from which the parents of that individual were collected; host.pop=location, either city or island name, from which the natal host seeds for that individual were collected; pophost=host species from which parents were collected; nathost=host species on which that individual was reared; hatch date=date the individual hatched; eclosion date=date the individual reached adulthood; death date=date an individual died if they did not reach adulthood (mortality post-adulthood was generally not recorded); sex=M for male, F for female if individuals reached adulthood. Blanks and NAs indicate data types that were not collected.
survival for DRYAD.csv
J.haematoloma development times (F1 2014, F2 2014, F1 2013)
Development time from hatching to adulthood for J.haematoloma combined from 3 experiments. Column heading interpretation: Number=individual identifier (unique between experiments); Family=maternal identifier: NOT unique across experiments (explained in more detail below); bug.pop=location, city or island name, from which the individuals parents or grandparents were collected; host.pop=location, city or island name, from which natal host seeds were collected (not available for K.elegans in 2013); pophost=host species from which parents or grandparents were collected in the field; nathost=host species on which the individual was reared; hatch date=date an individual hatched (not available for 2013 data); eclosion date=date an individual reached adulthood (not available for 2013 data); sex=M is male, F is female; Dvpt.time=number of days after hatching before an individual reached adulthood; Experiment=which experiment the individual is from (F1 2014, F2 2014, or F1 2013); mate=identifier (Number) of the individual with whom each F1 2014 individual was mated (some individuals did not have suitable mates available or died before they became available; not applicable for F2 2014 or F1 2013 experiments). FAMILY elaboration: The family number for individuals in the F2 2014 experiment indicates the Number of the mother, all of whom were members of the F1 2014 experiment. Paternal identifiers can be determined from the mate of the mother. Family identifiers were reused between experiments.
development time for DRYAD.csv
Simulated development time code
This file includes: 1) Code for data extraction of development times from Figure 3 of Carroll et al 1997; note that the exact pixel values will differ depending on how you size and crop the image. 2) Code for simulating 1000 (or however many you want) datasets using the means and standard deviations of 1988 and 2013-2014 data. 3) Code for statistical comparison of all 1000 datasets for the complete dataset and within each host plant. Code for model selection is not included (but could be added if requested).
Dvpt.time code for DRYAD.R