Date Published: October 25, 2018
Publisher: Public Library of Science
Author(s): Kjartan Østbye, Eivind Østbye, Anne May Lien, Laura R. Lee, Stein-Erik Lauritzen, David B. Carlini, Sylvie Rétaux.
Cave animals provide a unique opportunity to study contrasts in phenotype and life history in strikingly different environments when compared to surface populations, potentially related to natural selection. As such, we compared a permanent cave-living Gammarus lacustris (L.) population with two lake-resident surface populations analyzing morphology (eye- and antennal characters) and life-history (size at maturity, fecundity and egg-size). A part of the cytochrome c oxidase subunit I gene in the mitochondrion (COI) was analyzed to contrast genetic relationship of populations and was compared to sequences in GenBank to assess phylogeography and colonization scenarios. In the cave, a longer life cycle was implied, while surface populations seemed to have a shorter life cycle. Egg size, and size at maturity for both sexes, were larger in the cave than in surface populations, while fecundity was lower in the cave than in surface populations. The cave population had longer first- and second antennae with more articles, longer first- and second peduncles, and fewer ommatidia than surface populations. The cold low-productive cave environment may facilitate different phenotypic and life-history traits than in the warmer and more productive surface lake environments. The trait divergences among cave and surface populations resembles other cave-surface organism comparisons and may support a hypothesis of selection on sensory traits. The cave and Lake Ulvenvann populations grouped together with a sequence from Slovenia (comprising one genetic cluster), while Lake Lille Lauarvann grouped with a sequence from Ukraine (comprising another cluster), which are already recognized phylogenetic clusters. One evolutionary scenario is that the cave and surface populations were colonized postglacially around 9 000–10 000 years ago. We evaluate that an alternative scenario is that the cave was colonized during an interstadial during the last glaciation or earlier during the warm period before onset of the last glaciation.
Caves represent evolutionary model systems to study trait changes in the cave living fauna as compared to their relatives on the surface [1–3]. Striking apparent adaptations in morphology and life-history in cave animals have intrigued scientists for centuries [4–6]. Such cave dwelling populations, or species, often display trait reductions (often termed regressive evolution), reduced sight or even the complete loss of eyes and pigmentation, but attenuation of other appendages, as well as enhancement of extra-optic sensory structures, when compared to relatives outside the cave environment [7,8]. Since most obligate cave species (i.e. permanent residents of caves) often lack extant surface populations, or close taxonomic relatives, being separated for extended evolutionary time, they may not constitute the most ideal study objects for assessment of the very early stages of cave associated trait shifts and evolutionary changes. Facultative cave species (i.e. that also occupy surface areas) may better represent early stages in evolution of cave associated trait alterations where gene flow with surface populations still occurs or have recently been terminated [9–11]. The frequently observed infertility between cave and surface populations may also provide a possibility to analyze the genetic basis of complex traits [12,13]. Thus, cave animals are excellent model systems for studying speciation mechanisms at different stages in a long process [3,14–16].
According to our hypotheses that life-history parameters and phenotypic traits were similar in the cave and surface populations, the results suggest that most of the hypotheses should be rejected. The results suggest that cave and surface environments harbor populations with different life-history and phenotypic traits. First, the life-history cycle in the cave appeared longer with delayed sexual maturation in contrast to surface populations with a shorter life-cycle and early sexual maturation. Secondly, the gonad allocation appeared altered in the cave environment resulting in lower fecundity and larger eggs than in surface populations. Further, the antennae (first and second antennae) were larger in the cave than in the surface populations. Finally, eye traits (number of ommatidia and eye-area) in the cave were reduced compared to the surface populations who each had larger eyes and more ommatidia. These phenotypic patterns emerged when traits were scaled by body size. As the Sandågrotta cave to our knowledge is the only “known” permanent cave-living population of G. lacustris worldwide we should merit this system special protection both nationally and internationally. In the following we discuss mechanisms behind our finds and compare results with similar pairs of surface—cave dwelling populations in a geographically remote and phylogenetically independent habitat specialization in North American G. minus evaluating if trait divergences and convergences to cave—surface environments are replicated in two genetic sister clades.