TREELIM – Population processes and evolution

B1 Reproductive potential (seeds and viability)

Seeds collected from the leading cold edge of current broad-leaved tree species occurrence do not show any indication of reduced quality. Germination hardly differed between provenances from low and high elevation. Three species germinated poorly irrespective of provenance (origin) and treatment (Tilia, Quercus, Ilex). In the case of Quercus this was related to fungi. We conclude that seed properties offer no explanation for the current tree species limit (paper published by Kollas et al. 2012).

Proper seed formation constitutes one major prerequisite in successful tree establishment at the boundary of tree species distribution. To exclude insufficient seed quality from the list of possible constraints in trees reproduction, we conducted this study by comparing low elevation to high elevation seed sources.

The study took place in the two main study regions of the project, Western Swiss Alps (near Martigny) and Eastern Swiss Alps (near Chur). Both regions were surveyed for seed bearing tree individuals. From summer to autumn 2009 we visited both regions on a weekly basis. We collected the ripe seeds of 9 mostly widely distributed broad-leaved tree species beginning with low elevations seed sources and early ripening species (e.g. Prunus avium) to high elevation sources and late ripening species (e.g. Quercus petraea).

The sampling design comprised of two regions (as replicates), two elevations (close to optimal growth conditions at low elevation and marginal stands at high elevation) and 5 mother tree individuals per site. 300 to 1000 seeds per tree individual were collected and transferred to Basel. One subsample of 20 seeds per mother tree was used for morphological measurements, another 100 seeds were used for a germination trial and the large amount of remaining seeds was stratified to serve as source of tree seedlings for the reciprocal common garden experiment (see section B.3) within TREELIM. We measured seed size, seed weight, concentrations of non-structural carbohydrates and nitrogen and viability via Tetrazolium-test. Furthermore we conducted a greenhouse germination trial with 100 seeds of each of the c. 180 sampled tree individuals (9 species, 2 regions, 2 elevations, 5 mother trees). We used generalized linear models and mixed effect models to evaluate whether high elevation seed sources differ significantly from low elevation sources (Table 1).

In the broad-leaved tree taxa studied here seed quality does not constitute a serious constraint in reproduction at their high-elevation limit (under current climatic conditions). These results contradict the hypothesis of reduced quality of seeds in trees at their high-elevation range limits and support the findings of the study on population dynamics in the same regions (section B.2).