Population differentiation of the African cyprinid Barbus neumayeri across dissolved oxygen regimes

Abstract

Population level response to hypoxia has become an issue of global significance because of increased frequency and intensity of hypoxic events worldwide, and the potential for global warming to exacerbate hypoxic stress. In this study, we sequenced two nuclear intronic regions and a single mitochondrial region across seven populations of the African cyprinid, Barbus neumayeri from two river drainages in Uganda: the Rwembaita Swamp-Njuguta River System and the Dura River. We then examined two indices of population structure, GST and Jost's D, to detect links between oxygen availability and genetic variation and to determine if population genetic structure was associated with (i) dissolved oxygen regime (hypoxia or normoxia), (ii) geographical distance, or (iii) a combination of dissolved oxygen regime and geographical distance. Our results indicate that over a large scale (between drainages), geographical distance significantly affects the genetic structure of populations. However, within a single drainage, dissolved oxygen regime plays a key role in determining the genetic structure of populations. Within the Rwembaita-Njuguta system, gene flow was high between locations of similar oxygen regimes, but low between areas characterized by divergent oxygen regimes. Interestingly, GST analyses appear to yield less realistic measures of population structure than Jost's D, suggesting that caution must be taken when interpreting and comparing the results from different studies. These results support the idea that aquatic dissolved oxygen can act as a selective force limiting gene flow among populations of aquatic species and therefore should be considered when implementing conservation plans and assessing environmental impact of human activities.

The marine ‘great wall’ of China: local‐ and broad‐scale ecological impacts of coastal infrastructure on intertidal macrobenthic communities

“…Pairwise Φ ST s and Jost's D of COI sequences were calculated to investigate interpopulation genetic differentiation (Jost, ; Portnoy et al ., ; Harniman et al ., ). Populations on artificial structures were genetically different from populations on the northern natural rocky shore (NNRS), with few exceptions (Fig.…”

Low diversity of fishes in high elevation Afrotromontane streams renders them unsuitable for biomonitoring

“…Small‐bodied Enteromius and Clarias spp. are rhithronic species with preferences for cover and limnophilic features (Fugère, Mehner, et al, 2018; Harniman et al, 2013; Muñoz‐Mas et al, 2019). The longitudinal distribution of fishes in the Sondu‐Miriu River is also influenced by the Odino Waterfalls, which prevent the upstream migration of fishes.…”

Impacts of deforestation‐induced warming on the metabolism, growth and trophic interactions of an afrotropical stream fish

“…Agricultural streams were deforested over 50 years ago, which should provide ample time for adaptation given that evolution in physiological traits can occur over only a few generations in some fishes (Barrett et al, 2011;Handelsman et al, 2013). Gene flow could prevent adaptive TA B L E 2 Results of ANCOVA or ANOVA models testing effects of body size, stream land use and watershed on field growth rates divergence, yet our mark recapture study indicated no apparent dispersal among populations (see also Harniman et al, 2013). The absence of population differences could also be due to low heritability of RMR (but see Burton, Killen, Armstrong, & Metcalfe, 2011), "cryptic adaptive evolution" caused by opposing effects of adaptation and phenotypic plasticity (Conover & Schultz, 1995;Grether, 2005) or, perhaps most likely, no selection on RMR because metabolic costs of warming do not influence fitness, as our results for natural growth rates suggest.…”

“…In each watershed, the farm and forest fish populations are separated by a hypoxic papyrus swamp with a dissolved oxygen concentration generally <1 mg O 2 L −1 (watershed 1: Chapman, Chapman, Brazeau, McLaughlin, & Jordan, ; watershed 2: L. Chapman, unpublished data; Figure S1b). Genetic and mark–recapture studies have indicated that stream‐adapted populations of E. neumayeri very rarely disperse through papyrus swamps (Chapman et al., ; Harniman, Merritt, Chapman, Lesbarreres, & Martinez, ), suggesting low gene flow between farm and forest populations within each watershed. Molecular studies have also established that populations from watersheds 1 and 2 are genetically distinct (Harniman et al., ), confirming that each forest–farm population pair constitutes an independent replicate to study thermal adaptation to deforested (warmer) conditions.…”

“…Figure S1b). Genetic and mark-recapture studies have indicated that stream-adapted populations of E. neumayeri very rarely disperse through papyrus swamps (Chapman et al, 1999;Harniman, Merritt, Chapman, Lesbarreres, & Martinez, 2013), suggesting low gene flow between farm and forest populations within each watershed. Molecular studies have also established that populations from watersheds 1 and 2 are genetically distinct (Harniman et al, 2013), confirming that each forest-farm population pair constitutes an independent replicate to study thermal adaptation to deforested (warmer) conditions.…”