Scaling-Up Methods Influence on the Spatial Variation in Plant Community Traits: Evidence Based on Leaf Nitrogen Content

Plant functional traits have received increasing attention at a community scale. There are three methods, for scaling-up from the organ to the community: species arithmetic mean (SAM), community-weighted mean (CWM), and per unit land area community-weighted mean (PLA-CWM). To date, it is unclear how these scaling-up methods influence the spatial variation of these plant community traits. Here, we consistently measured the leaf nitrogen content (N) of 1,844 plant species from 22 natural communities along a 5,100 km transect with a temperature and precipitation gradient and explored the potential effect of different scaling-up methods on leaf N in the community as N-SAM, N-CWM, and NPLA-CWM. Our results demonstrated that leaf N at the community scale showed a different spatial variation among the three methods and was dependent on climate or plant community types. N-SAM and N-CWM decreased significantly with the increasing of mean annual precipitation (MAP), and NPLA-CWM showed the opposite trend. Furthermore, N-SAM overestimated the leaf nitrogen content at the community level compared to N-CWM. SAM underestimated leaf nitrogen in a dry or cold environment and vice versa in a wet or warm environment compared to CWM. With MAP greater than 300 mm or with mean annual temperature higher than 0.24 degrees C, N-SAM and N-CWM were significantly different, which should be emphasized. Based on these findings, over- or underestimation from different scaling-up methods should be tested in other plant community traits. In summary, community-weighted methods (CWM and PLA-CWM), covering community information on species composition and traits, implicating many important ecological processes, should be helpful for evaluating the spatial variation of plant community traits on a large scale in the future.