Search

Aim Rainfall reliability has been neglected as a determinant of plant trait convergence and divergence in mediterranean-climate ecosystems. This paper reports on patterns of rainfall reliability — quantified as interannual variation in monthly and seasonal rainfall, and as the frequency of individual events in terms of their size, duration and intensity — for four fire-prone mediterranean-climate ecosystems. Location The four mediterranean-climate regions of the world with fire-prone ecosystems, namely SW Cape (South Africa), SW Australia, California and the Mediterranean Basin (Andalusia, Spain). Methods Using long-term monthly rainfall data from stations dispersed across the four regions, we computed monthly means and interannual variation for each month of the year — the latter quantified as the coefficient of variation (CV) — and divided these into winter and summer seasons. We also computed the mean number of rainfall events, the mean frequency in various categories of event duration (days), the amount of rainfall per event (mm) and the rainfall intensity per event (mm/day) per year for winter and summer seasons for a subset of the rainfall stations. Results The fraction of rain falling in summer was lowest in California (5%) and similarly low ( c. 25%) in the other three regions. The hierarchy of values of coefficient of variation (CV) of monthly rainfall during the winter period was as follows: California > Andalusia >> SW Cape > SW Australia; results for summer were: California > > Andalusia >> SW Australia ∼ SW Cape. SW Australian sites experienced the greatest frequency of short, small and low-intensity rainfall events in both seasons; patterns in the SW Cape were intermediate between Australia and the two northern hemisphere sites which both received fewer, larger and more intense events. Overall, the two southern hemisphere regions (SW Australia and the SW Cape) had significantly more reliable regimes than the two northern hemisphere ones (Mediterranean Basin and California). Main conclusions These differences in rainfall reliability regimes may provide a novel perspective on the distribution of certain plant life-history traits in mediterranean-climate ecosystems. Less reliable regimes would select for germination and seedling survival traits that enable persistence of genets in the face of uncertain moisture conditions during the winter and spring establishment phase. Study systems that accommodate for phylogenetic constraints, namely invasive species derived from mediterranean-climate ecosystems, as well as shared lineages, provide good opportunities to develop and test hypotheses on the implication of different rainfall reliability regimes. One of the novel implications of this study is that the distinctive trait of assemblages in the southern hemisphere regions may be a consequence not so much of their shared nutrient-poor soils as of their similarly reliable rainfall regimes. [ABSTRACT FROM AUTHOR]

ABSTRACT Aim Classic theory suggests that species-rich communities should be more resistant to the establishment of exotic species than species-poor communities. Although this theory predicts that exotic species should be less diverse in regions that contain more native species, macroecological analyses often find that the correlation between exotic and native species richness is positive rather than negative. To reconcile results with theory, we explore to what extent climatic conditions, landscape heterogeneity and anthropogenic disturbance may explain the positive relationship between native and exotic plant richness. Location Catalonia (western Mediterranean region). Methods We integrated floristic records and GIS-based environmental measures to make spatially explicit 10-km grid cells. We asked whether the observed positive relationship between native and exotic plant richness ( R2= 0.11) resulted from the addition of several negative correlations corresponding to different environmental conditions identified with cluster analysis. Moreover, we directly quantified the importance of common causal effects with a structural equation modelling framework. Results We found no evidence that the relationship between native and exotic plant richness was negative when the comparison was made within environmentally homogeneous groups. Although there were common factors explaining both native and exotic richness, mainly associated with landscape heterogeneity and human pressure, these factors only explained 17.2% of the total correlation. Nevertheless, when the comparison was restricted to native plants associated with human-disturbed (i.e. ruderal) ecosystems, the relationship was stronger ( R2= 0.52) and the fraction explained by common factors increased substantially (58.3%). Main conclusions While our results confirm that the positive correlation between exotic and native plant richness is in part explained by common extrinsic factors, they also highlight the great importance of anthropic factors that - by reducing biotic resistance - facilitate the establishment and spread of both exotic and native plants that tolerate disturbed environments. [ABSTRACT FROM AUTHOR]