For Those Who Do Not Grasp Dabrafenib Today or You Will Despise Your Self Later

We fit linear and quadratic regression models to the moving-window coefficients for each trait pair and selected the best model using Akaike's information criterion (AIC). We constructed generalized linear models with plant functional type, latitude, elevational range, minimum elevation and type of variation (within or among species) as predictors, then used a stepwise model http://www.selleckchem.com/products/DAPT-GSI-IX.html selection procedure based on AIC to find the best reduced models. We conducted z-tests for effect size heterogeneity (Borenstein et?al. 2009) to compare the weighted mean effect sizes among groups of studies. Finally, we assessed publication bias using a number of tests. We found only limited evidence for publication bias in favour of positive results in LMA studies, and no evidence for bias in Nmass or Narea studies (see Appendix S4, Supporting information). All analyses were done using R 2.14.1 (R Development Core Team 2011), including the packages meta (Schwarzer 2012) and raster (Hijmans & van Etten 2013). We did not conduct a quantitative meta-analysis of the common garden and reciprocal transplant studies due to low availability of published data. Instead, we determined whether each study reported significant genetic effects among http://www.selleckchem.com/products/Adriamycin.html elevations, using F-statistics from analyses of variance or correlation coefficients from linear regressions. We used a vote-counting approach (DeCoster 2004) to qualitatively assess the genetic basis of variation in LMA, Nmass and Narea across elevations. Overall, we found that (i) LMA and leaf N content varied with mean annual temperature along elevational gradients in similar fashion among plant species, (ii) both intraspecific and interspecific variation in these traits are of similar magnitude across disparate and extensive elevational gradients and (iii) much intraspecific variation in leaf traits along elevational gradients may be explained by convergent evolution. Across 46 elevational gradients spanning a total of over 4800?m, the mean effect of modelled MAT on LMA was negative [mean r?=??0��51, 95% CI?=?(?0��30, ?0��68), P?=?1?��?10?6, Fig.?1a]. For Nmass, the mean effect size did not differ significantly from zero (P?=?0��84, Fig.?1b). On average for each gradient, there was a significantly negative relationship between Narea and MAT [mean r?=??0��55, 95% CI?=?(?0��40, ?0��67), P? http://www.selleck.cn/products/dabrafenib-gsk2118436.html all gradients (Fig. S2, Supporting information). The strengths of the correlations among each of the three trait pairs changed significantly with increasing mean annual temperature, as revealed by moving-window regression analyses (Fig.?2). A change in the magnitude or direction of pairwise trait relationships across different environments represents strong evidence for environmental filtering across elevations. A quadratic least-squares regression model fit the pairwise trait correlation data best for all three pairs.