Functional trait metrics, such as the community weighted mean (CWM), are increasingly used as proxies for the effects of anthropogenic modification of local community composition on ecosystem fluxes and services. Nevertheless, the strong dependence of individual tree metabolic rates, such as water use, with size across species may hamper the utility of trait CWM’s as indicators of water fluxes across forest communities. In addition, it is unclear how contextual factors, such as founder effects and management practices, affect the relationships of trait CWM’s to ecosystem fluxes. We used individual tree sap flux and trunk diameter measurements to estimate tree to stand water use and water use efficiency (WUE) in six paired secondary forests and teak plantations in dry tropical Costa Rica (N=108 trees). These two reforestation types are sponsored by national Payments for Ecosystem Services (PES) incentives to landowners. We aimed to answer (1) is convergence in tree water use with size observed across functional groups? (2) what other traits are related to tree and stand water use? (3) is the relationship of trait CWM’s and water use different between reforestation types? and (4) is there a difference in water use and WUE between PES-sponsored secondary forests and teak plantations?
Linear relationships of tree water use to biomass and trunk diameter (R2≥0.72, p<0.0001) varied per functional group, the slope being lower for nitrogen-fixers (ANCOVA: p≤0.01), and were absent in plantation trees (R2≤0.12) due to limited size range. Secondary forest tree water use relationships to height, crown area and sapwood depth (R2≥0.33, p<0.0001) varied by functional group (p≤0.004). We found no relationships between plantation tree water use and traits, except for crown area (R2=0.16, p=0.03), or between tree WUE and traits. Stand water use was linearly related to biomass and basal area (R2≥0.75, p≤0.0004), less to the CWM of sapwood depth (R2=0.48, p=0.01), unaffected by reforestation type (ANCOVA: p≥0.3). Water use was linearly related to CWM’s of wood density, water content, leaf area, thickness, and height in secondary forests (R2≥0.72, p≤0.03) but not in plantations (R2≤0.28). Stand WUE was related to wood water and leaf carbon CWM’s (R2>0.35, p≤0.01) across reforestation types (p>0.07), but to wood carbon only in plantations (R2=0.76, p=0.02). Water use was not different between reforestation types (Paired t-test: p≥0.09) but WUE was higher in plantations (p=0.0008). Our results show functional group and contextual factor effects on trait-function relationships shouldn’t be ignored in ecosystem service assessments and policy.