The genetics of thermal plasticity in Plantago lanceolata

Phenotypic plasticity, an individual’s phenotypic response to environmental change, is a fundamental characteristic of all life on earth that plays a central role in adaptation, phenotypic differentiation, and speciation. Temperature-sensitive phenotypic plasticity, i.e. thermal plasticity, often increases with latitude, suggesting an increasingly adaptive role of thermal plasticity in predominantly cool, thermally variable environments. While the hypothesis is reasonable, it has not been thoroughly tested. Demonstrating local adaptation of thermal plasticity requires showing that: 1) thermal plasticity increases fitness in high latitude environments, 2) clinal variation arises from natural selection, and not by chance alone, 3) differences in thermal plasticity persist in the presence of gene flow, 4) thermal plasticity has a genetic basis and varies genetically along a latitudinal gradient, 5) thermal plasticity is heritable, and 6) thermal plasticity is a derived phylogenetic character. Today, little is known about the genetic properties of thermal plasticity. I took advantage of natural geographic variation in a widespread perennial herb, Plantago lanceolata to improve our understanding of adaptation along latitudinal clines by examining the genetic features of thermal plasticity. With genetic data I address the questions: 1) Is clinal variation in thermal plasticity best explained by natural selection driven by environmental differences among populations, neutral genetic evolution, or both? 2) What is the genetic architecture of thermal plasticity and single-environment trait variation, and how are they related? 3) Do genetic properties of thermal plasticity mirror phenotypic patterns along a latitudinal gradient? Among 14 European populations of Plantago lanceolata I estimated differentiation in temperature-sensitive floral reflectance plasticity (QST/PST), neutral genetic differentiation (FST & Jost’s D) of AFLP markers, and between-population differences in aspects of the reproductive environment. I used phenotypic QST (PST) vs. FST comparisons to investigate the evolutionary forces responsible for geographic patterns of thermal plasticity, and to determine if differences brought about by neutral evolutionary forces are sufficient to explain these patterns. My data supported the hypothesis that natural selection, driven by environmental properties of the reproductive season, particularly the duration and proportion of time at cool temperatures, has contributed to geographic patterns of thermal plasticity. As between-population differences in these environmental variables increased, differences in thermal plasticity increased more quickly than did neutral genetic differences. To determine the genetic architecture of thermal plasticity I produced an F2 mapping family from parents derived from distant northern and southern European populations that exhibited high (northern parents) and low (southern parents) thermal plasticities of floral reflectance. I then grew parents and offspring in two environments (cool and warm) mimicking what plants would encounter in nature. I attained genetic markers via genotype-by-sequencing (ddRADseq), produced a recombination map and performed QTL mapping of thermal plasticity and single-environment trait values for six traits: floral reflectance, flowering time, rosette diameter, leaf length, leaf fresh mass, and leaf area. My data provides critical genetic support for the hypothesis temperature-sensitive floral reflectance plasticity in P. lanceolata is adaptive in high latitude environments where growing seasons are cool and short. My data confirms thermal plasticity in P. lanceolata has a genetic basis as I found one single QTL underlying the thermal plasticities of three traits, floral reflectance, flowering time and leaf length. Floral reflectance plasticity and flowering time plasticity QTLs colocalized with, and shared phenotypic effects with corresponding single environment QTLs. The leaf length plasticity QTL did not colocalize with any single-environment QTLs, and was influenced by cytoplasm. I did not find evidence plasticity QTLs of different traits were pleiotropic. Additionally, genotypic differences at plasticity QTLs paralleled patterns of plasticity along latitudinal clines. At plasticity QTLs northern genotypes (Danish and Swedish) increased the magnitude of thermal plasticity, while southern genotypes (French and Italian) decreased plasticity