Genetic differentitation of pedunculate oak (Quercus robur L.) provenances along a latitude gradient in responses to drought and rewatering

Acorns were collected in 2013 in nine Quercus robur L. populations along a latitudinal gradient from Estonia to Italy. Seedlings were raised during 2014 in a nursery and individually transplanted in 50l plastic pots with homogenised soil from a local oak forest. A greenhouse trial was established in spring 2015 according to RCBD split plot design with two main plots. Plants in the first plot did not receive any water from April 1st till July 21st (drought treatment), while plants in the second plot (control) were constantly kept at 45-50% of soil moisture content (SMC). Plants in the drought plot were rewatered on July 21st and kept at 45-50% of SMC till the end of the growing season. Heights and photosynthetic rates (PhRs) were periodically measured on all plants. The aim of the study was determining phenotypic responses of the provenances to drought and rewatering and assessing amount and pattern of among provenance genetic variation (Bogdan et al 2015ab, Sever et al 2015).Repeated measures ANOVA were performed using proc MIXED of the SAS software. Effects of 53 time, treatment, provenance and their interactions were determined. Statistical significance of differences between various effect levels was determined by Post-hoc Tukey test. Multivariate regression tree (MRT) analyses using MVpart procedure in R were conducted to reveal a pattern of genetic differentiation among the provenances (De’Ath et al 2002). MRT were done combining phenotypic data from the trial and mean annual climate data of the provenance original sites from the ClimateEU software (Wang et al 2012, Hamann et al 2011).Height growth of all provenances decreased as response to drought. Mean heights of most provenances were more than 50% lower, compared to the control on the peak of drought. However, some provenances showed milder decrease (e.g. Hungarian, Lithuanian and Estonian, whose growth decreased 35, 40 and 4%, respectively). Mean PhR of provenances in the drought treatment was also significantly lower, compared to control. For both traits, between provenance variations were highly significant (p