Rapid wing size evolution in African fig flies (Zaprionus indianus) following temperate colonization

Abstract

Invasive species often encounter novel selective pressures in their invaded range, and understanding their potential for rapid evolution can offer insight toward evolutionary processes and the factors that drive invasion success. Zaprionus indianus is an invasive drosophilid native to Africa that reached Florida in 2005 and likely reestablishes temperate North American populations each year. We addressed two evolutionary questions in this system: First, do populations evolve phenotypic changes in the generations immediately following colonization of temperate environments? Second, does Z. indianus evolve directional phenotypic changes along a latitudinal cline? We established isofemale lines from wild collections and measured twelve ecologically relevant phenotypes, using a reference strain as a control. Z. indianus evolved smaller wings following colonization, and we found evidence of significant postcolonization evolution when considering all phenotypes simultaneously. We found little evidence for latitudinal clines. However, we documented substantial laboratory evolution and large effects of the laboratory environment across multiple phenotypes, emphasizing the importance of controlling for both possibilities in common-garden studies. Our results demonstrate the potential for rapid evolution in Z. indianus, which could contribute to its ongoing expansion, and offer insights toward the types of rapid evolutionary changes that might occur in invasive insects.

Keywords: adaptation; life-history evolution; local adaptation; natural selection; trait evolution.

Figure 1.

Overview of experimental design and timeline. The diagram indicates the relative timing of fly capture and experimental assays for the postcolonization (orange), latitude (gray), and lab evolution (green) experiments. The postcolonization evolution experiment compares the phenotypes of isofemale lines captured early and late in the season. The lab evolution experiment compares phenotypes of lines captured early in the season after 3–4 generations of lab rearing and 7–8 generations of lab rearing. The latitude experiment compares lines collected from locations spanning 15° latitude in North America. The inbred control line (purple) serves as a reference to test for lab environment or assay variation between experiments conducted at different times.

Figure 2.

Z. indianus wing size evolves rapidly in the wild and in the lab. Wing size is the geometric mean of the major and minor radii of an ellipse fitted to the wing. Early and late describe when the flies were assayed; for the lab evolution experiment, the same lines were assayed at two different generations. Points illustrate all individuals measured (n = 1079 total); boxplots show median and quantiles. Asterisks indicate Bonferroni-corrected p < .05 in a mixed-effect linear model (see Supplementary Table S1 for full model results).

Figure 3.

Postcolonization evolution, lab evolution, and lab environment effects for 12 phenotypes in Z. indianus. The endpoint of each line shows the least-squared model-fitted mean for each timepoint, and error bars show model standard errors. Wild-derived lines (orange) were collected in August (early season) and November (late season) and measured for each phenotype after 3–4 generations of laboratory culture. Lab evolution lines (green) were collected and measured in the early season and remeasured alongside the late season flies after several additional generations in the lab. Inbred controls (purple) are a single inbred line that was assayed alongside all experimental flies. Bold lines indicate significant differences between early and late season phenotypes in mixed-effect models after Bonferroni correction (see Table 2 and Supplementary Table S1).

Figure 4.

Limited latitudinal phenotypic variation in Z. indianus. Points represent the mean of all individuals from each population for a given phenotype, and error bars represent standard error of the mean. No phenotypes showed significant changes with latitude in the final mixed-effect models after Bonferroni correction.

Figure 5.

Multidimensional analysis of phenotypes. (A) Principal components analysis (PCA) for wild-derived flies caught early in the season, late in the season, and from four different latitudes (combined here for visual clarity). Each point represents one isofemale line. Triangles represent the inbred control line phenotyped alongside each group. Data ellipses were drawn using a multivariate t-distribution. (B) PCA of the four latitudinal populations. (C) Ordination plot for the 12 phenotypes used for the PCA; PCs were calculated using normalized line means for each phenotype from the combined seasonal and latitudinal experiments. (D) MANOVA results for both the seasonal and latitudinal analysis.