Recruitment Traits Could Influence Species' Geographical Range: A Case Study in the Genus Saxifraga L

Abstract

The reasons why some species occur widespread, while related species have restricted geographical ranges have been attributed to habitat specialization or ecological niche breadth. For species in the genus Saxifraga, habitat specialization alone cannot explain the distributional differences observed. We hypothesize that recruitment traits (i.e., germination, emergence, and survival) may account for differences in geographical ranges and that early life stages correlate to survival. We studied recruitment responses in 13 widespread and 12 narrow-ranged Saxifraga species in the laboratory and common garden experiments using seeds collected from 79 populations in the European Alps. We found that in the laboratory cold temperature led to higher germination percentages compared with warm temperature for both distribution groups. This represents an exception to the general assumption that alpine species require warm cues for germination. In warm laboratory temperatures, widespread species germinated better than narrow-ranged species, indicating a greater tolerance of warm temperatures for the former. Subsequent to germination, recruitment traits between the two distribution groups were lower or null in the common garden, suggesting that the impact of recruitment on species' geographical ranges occurs at the earliest life stage. Mean time to emergence of narrow-ranged species showed lower variability than that of widespread species. Consistently, intraspecific variation of mean annual temperatures between seed collection sites was lower for narrow-ranged species, indicating a close relationship between home sites and emergence time. Emergence percentage was a strong predictor of survival only for widespread species, underlining that seed and seedling functional traits differ between distribution groups, which require further research. Our results support the view that early life stages are critical to population dynamics and thus can influence species' geographical ranges. The wider responses to climatic conditions in widespread species may have facilitated their spread across the Alps. Our results also suggest that all Saxifraga species face a considerable threat from climate warming due to their overall cold-adapted recruitment niche.

Keywords: common garden; germination; intraspecific variation; phenology; regeneration niche; seed sowing experiment.

FIGURE 1

(A) Köppen-Geiger climate map (open source) of the European Alps and seed collection sites (points). In total, 79 populations of Saxifraga spp. were collected from Southwest (France) to the East (Austria). As some points overlap in the overview, detailed location maps are provided in Supplementary Figure 1. The climatic zones can be classified as arid (Bsk), warm temperate (Cfa, Cfb, Cfc. Csa, Csb), boreal (Dfb, Dfc), and alpine (ET, EF). (B) Histogram of elevations of seed collection sites per population, in descending order of elevation.

FIGURE 2

(A) Cumulative germination (%) in the laboratory at two temperature regimes (warm 25/15°C, cold15/5°C), grouped according to distribution ranges (wide, narrow). Points denote the cumulative mean on each day of observation, error bars denote ± standard errors. Boxes in (B) germination (%) and (C) MTG (days) illustrate the interquartile range with the median as line; whiskers illustrate the minimum and maximum; points denote outliers; x denotes the mean. Points without border in (D) germination (%) and (E) MTG denote the observed values colored according to the distribution range along the ecological indicator value W (alternate soil humidity); points with border denote the estimated means and colored ribbons denote confidence intervals from the models. Note different maxima on y-axis between panels. Stars denote significances at α = 0.01 (**) and α = 0.001 (***).

FIGURE 3

(A) Emergence percentage in the common garden, grouped in distribution (wide, narrow), at each recording date (i.e., every 2 weeks). Bars denote mean at each recording date, error bars denote ± standard errors. Boxes in (B) emergence percentage and (C) MTE (recording date) illustrate the interquartile range with the median as line; whiskers illustrate the minimum and maximum; points denote outliers; x denotes the mean. Points without border in (D) emergence percentage (E,F), MTE, denote the observed values colored according to distribution along ecological indicator values H (humus) and T (Temperature value); points with border denote the estimated means per ecological indicator value life form (c, chamaeophyte; h, hemicryptophyte; t, therophyte), and colored ribbons denote confidence intervals from the models [missing in (D) to simplify the graph]. Note different maxima on y-axis between panels. Stars denote significances at α = 0.01 (**) and α = 0.001 (***).

FIGURE 4

(A) Seedling survival of the growing season (“gs”) in the common garden, grouped according to distribution (wide, narrow), for the years of observation (2017, 2018), and (B) overwinter seedling survival (“ow”). Boxes in (A) survival “gs” and (B) survival “ow” illustrate the interquartile range with the median as line; whiskers illustrate the minimum and maximum; points denote outliers; x denotes the mean. Points without boarder in (C) survival “ow” denote the observed values colored according to distribution along the ecological indicator value F (humidity), and colored ribbons denote confidence intervals from the model. Note different maxima on y-axis between panels. Stars denote significances at α = 0.001 (***).

FIGURE 5

Recruitment responses at species level. (A) Germination percentage under warm and cold laboratory temperature (N = 584). (B) Emergence percentage in the common garden under natural seasonal climate (N = 236). (C) Survival percentage “gs” (N = 79), and (D) survival percentage “ow” (N = 52). Species are listed in alphabetical order within each distribution. Boxes illustrate the interquartile range with the median as line; whiskers illustrate the minimum and maximum; points denote outliers; x denotes the mean.

FIGURE 6

(A) Intraspecific variation per species represented by the coefficient of variation (CV) of response variables: Germination percentage in warm (G_w) and cold (G_c) temperature in the laboratory, emergence percentage (E), mean time to germination in warm (MTG_w) and cold (MTG_c) temperature, mean time to emergence (MTE), and average annual temperature at the seed collection sites (Tave). (B) Comparison of mean CV of response variables per distribution range, done by linear regressions. Note different maxima on y-axis between panels. Stars denote significances at α = 0.05 (*).