Abstract
Human disturbances play a pivotal role in the evolution of populations, exerting strong selective pressures that frequently drive rapid adaptive responses. Variation can be expressed across a gradual cline, whereby shifts are associated with local environmental gradients generated by human activities, for instance, gradients in forest cover due to deforestation.
The remarkable loss of forest upon human settlement in New Zealand has left an indelible mark on the Clutha region (southern South Island) which currently comprises a variety of ecosystems, ranging from fully forested (south) to completely deforested (north). Low-lying tributaries of the Clutha River host populations of Zelandoperla fenestrata, a widespread native stonefly that naturally displays wing polymorphism driven by the environment, whereby adults exhibit a reduced-winged phenotype in streams situated above the alpine tree line or a fully winged phenotype in streams situated below the tree line.
I aimed to test whether human-led deforestation could have rapidly driven adaptive wing reduction events in Clutha populations of Z. fenestrata at sites where forest cover was recently removed. Using morphological analysis coupled with geographic cline analysis, the wing morphologies of 769 Z. fenestrata were characterized across nine sites in the Clutha region experiencing a gradient in forest cover across a latitudinal cline. A strong association was found between shifts in Z. fenestrata wing reduction and shifts in forest cover, with a gradual increase in reduced-winged individuals as forest cover decreased.
Additionally, a genotyping-by-sequencing approach retaining 68,011 genome-wide variants assessed population genetic structure among 171 fully winged and reduced-winged Z. fenestrata sampled across the Clutha region. Overall, the differentiation among populations was minimal, likely reflecting neutral variants responding to isolation by distance patterns. Notably, a lack of differentiation was observed among reduced-winged and fully winged Z. fenestrata at sites where they co-occur, suggesting the recent emergence of the reduced-winged phenotype. Accordingly, genome-wide patterns were found to fit with a scenario of anthropogenically-driven wing reduction, whereby swift forest loss in the Clutha region driven by recent human activity could have led to rapid wing reduction events in populations of Z. fenestrata, likely underpinned by repeated sorting of standing variation.
Interestingly, the emergence of a primarily reduced-winged clade experiencing heightened structure indicates that wing reduction can remarkably impact divergence, despite ongoing gene flow.
Furthermore, I aimed to explore the genomic mechanism underpinning wing reduction events in Z. fenestrata populations, which has so far been challenging to resolve. I used two independent outlier approaches and found a number of SNPs across the genome that differentiated reduced-winged and fully winged stoneflies. Several of the 11 genomic regions that were significantly differentiated among alternative wing types (q < 0.05) have a known role in wing development in Drosophila, and thus potentially have a role in driving wing loss. I also amplified a small fragment of a coding region that was previously identified as a strong wing reduction candidate in South Island populations of Z. fenestrata. Strikingly, this region consistently differentiated reduced-winged and fully winged Clutha individuals, providing strong evidence that this gene is involved in adaptive wing reduction.