Community leaf dry matter content predicts grassland production

Plant growth correlates with values of collinear (covarying) traits from
the leaf economics spectrum.  Environmental variation and
differences in community composition may alter contributions of these
traits to plant production and thereby limit the consistency of
trait-based growth predictions among years and plant
communities.  We tested effects of interannual variation in
precipitation and differences in grassland community composition (planted
monoculture of switchgrass, Panicum virgatum, and mixture of perennial
herbaceous species) on the utility of two traits from the leaf economics
spectrum (leaf dry matter content [LDMC] and plant [N]) to predict
aboveground net primary production (ANP) during spring of 6
years.  Spatial and temporal variation in spring production
correlated with community-scale (species-abundance weighted) values of
both traits, but community LDMC explained 66% of the variance in
production and accounted for ≥89% of the variance in ANP explained by the
two traits combined.  The ANP response to trait variation and the
variance in ANP explained by trait values differed with precipitation and
between communities.  Greater precipitation increased the
production response to trait variation by increasing slopes of ANP-trait
regression relationships and increased the variance in ANP explained by
trait values.  Communities differed in response to precipitation
variation and in the role of annual variation in the [N]-LDMC relationship
in explaining variance in ANP.  Results indicate that mean trends
in grassland production can be predicted using community-scale values of
LDMC.  Trait-based predictions of grassland production could be
improved, however, by accommodating precipitation and community effects on
production-trait relationships.