Molecular and adaptive variation in the Caledonian Pine, Pinus sylvestris (L.)

Abstract

The remnants of the Caledonian Pine Forest represent the north western boundary of the Eurasian Pinus sylvestris (L.) distribution. Remnant populations occupy a diverse range of environments within Scotland, subject to a steep rainfall gradient, and previous investigations have found evidence of local adaptation. Additionally, studies of biochemical and molecular markers have indicated that Scotland’s native pinewoods originated from more than one glacial refugium. Whole-genome-shotgun (WGS) sequencing was employed for the discovery of mitochondrial (mt) variants that may provide further insight into the origins of P. sylvestris populations both in Scotland and mainland Europe. DNA extractions were performed on megagametophyte tissue from Scottish, Finnish, and Spanish populations. Three members of the closely related P. mugo species complex were also sequenced. Using similarity-based approach, 160kbp of putative mitochondrial sequence was recovered by comparison of de novo assembled contigs with the mtgenome of the gymnosperm Cycas taitungensis. In total, 16 novel variants were identified among samples, which may be used in future phylogeographic studies. A study of needle characters was performed for eight native populations of P. sylvestris in an outdoor provenance/progeny trial of 192 saplings. A negative correlation was detected between longitude and the number of stomatal rows present on needle surfaces. It was posited that this may be an adaptive response to lower water availability in eastern pinewoods, possibly in conjunction with increasing altitude. The west coast of Scotland is one of the wettest regions in Europe: western pinewoods may receive in excess of 3,000mm of rainfall in a year, compared with an average of 800mm eastern sites. To determine whether native pinewoods are differentially adapted to waterlogging, a glasshouse based provenance/progeny trial of 432 saplings from nine native populations was undertaken, in which 50% were subject to a long-term waterlogging treatment, and the remainder used as a control. Two studies were then conducted. In the first, responses to the treatment were assessed in terms of phenological and growth traits. Bud flush was delayed in response to waterlogging, and growth was impeded relative to the control. Although population differences were observed, treatment × population interactions were not detected. In the second study physiological traits known to be sensitive to plant stress and water balance were measured at intervals throughout the experiment. Prior to the commencement of the treatment needle δ13C was found to exhibit interpopulation differentiation, and was positively correlated with longitude. This seems likely to represent differential selection for water use efficiency between eastern and western pinewoods. Photochemical efficiency and stomatal conductance were found to be reduced by waterlogging, and needle δ13C was increased. After generalising populations into ‘high’ and ‘low’ rainfall groups (monthly averages of 214.9mm and 72.8mm, respectively), high rainfall populations were observed to maintain consistently higher photochemical efficiency under waterlogging the low rainfall populations. In addition, the low rainfall group exhibited greater variability in response to flooding (in terms of phenotypic and additive genetic variance) which may be indicative of a lack of past selection pressure.