Genetic isolation among mountains but not between stream types in a tropical high-altitude mayfly

Abstract

Glaciers that directly feed high-altitude streams create unique environmental conditions that contribute substantially to regional-scale lotic habitat diversity and biodiversity, including intra-specific genetic diversity (as population structure) between glacier-fed and other types of streams (e.g. groundwater-fed). However, these population-genetic patterns are thus far only understood for macroinvertebrates in the temperate zone, where strong seasonality and narrow temporal windows for emergence and mating could help drive patterns of genetic differentiation between streams with contrasting temperature, flow, or other environmental characteristics influencing life-history patterns. Our primary objective was to assess population-genetic structure between groundwater-(GW) and glacier runoff-fed (RO) streams in high-altitude tropical (relatively aseasonal) basins of the Ecuadorian Andes. Our focal species was Andesiops peruvianus, a baetid mayfly confamilial with well-studied temperate alpine mayflies. We pursued secondary objectives of evaluating broader scale population-genetic patterns across mountain ranges for the first time in high-altitude tropical streams, and evaluating genetic evidence for recovery from population bottlenecks in this volcanically active region. For the primary objective, we collected A. peruvianus (mean N = 16.5 per reach) and a suite of environmental variables from six intra-basin pairs of GW/RO stream reaches at altitudes 4000-4300 m a.s.l. on three glaciated volcanoes representing two parallel sub-ranges of the Ecuadorian Andes. We tested for significant GW/RO pairwise differences in haplotype distribution and genetic diversity obtained by sequencing the barcoding region of the mitochondrial cytochrome oxidase I gene. For the broader scale sub-objectives, we added two unpaired populations (total N = 231) and evaluated genetic structure at nested spatial scales of streams/basins/mountains, and we tested for differences between mountains. We also measured Tajima's D and Fu's FS to evaluate evidence for demographic instability at the scale of individual mountains, each with a different volcanic history. We found no evidence for population structure between GW and RO streams within basins. Population structure among basins within mountains was significant, but only in areas where streams occupied deep, physically isolating canyons. Comparisons between all possible pairs of the three mountains revealed significant structure, but pairwise ?ST was an order of magnitude greater between pairs of mountains occupying different ranges than for the pair in the same range. Indeed, no haplotypes were shared between the two Andean sub-ranges. All three mountains, regardless of recent volcanic history, showed a significant signature of recovery from recent bottleneck. Our results suggest that strong environmental differences between glacial runoff and groundwater stream types do not isolate these tropical, high-altitude mayfly populations. Rather, populations are panmictic within basins. Broader scale patterns among mountains suggest that dispersal and gene flow in these tropical streams proceed similarly to temperate alpine systems; that is, relatively strong isolation among mountains but reasonable capacity for gene flow between headwaters in close proximity on a single mountain. A notable difference from the temperate studies is that mayfly populations in Ecuadorian high-altitude streams appear to be demographically unstable, regardless of the recent volcanic eruption history of the mountain they occupy. Frequent eruptions in this volatile region might affect streams across areas more extensive than a single mountain.