The water and energy exchange of a shaded coffee plantation in the lower montane cloud forest zone of central Veracruz, Mexico

The water and energy fluxes of a shaded coffee plantation in the lower montane cloud forest (LMCF) zone of central Veracruz, Mexico, were measured over a two-year period (September 2006-August 2008) using the eddy covariance method. Complementary measurements of throughfall and stemflow were made to study rainfall interception. The sum of the observed sensible (H) and latent (λE) heat fluxes was almost 95% of the net radiation (Rn) minus the canopy heat storage fluxes, indicating very good energy balance closure. Monthly means of the mid-day (11:00-15:00 h) Bowen ratio (H/λE) and evaporative fraction (λE/Rn) averaged 0.74 +/- 0.12 and 0.56 +/- 0.05, respectively. Energy partitioning showed distinct seasonal variation, with significantly higher Bowen ratios prevailing during the dry season (0.81 +/- 0.13) compared to the rainy season (0.67 +/- 0.06). The lower evaporation rates during the dry season reflected a combination of lower soil moisture availability and a lower leaf area of the Inga shade trees during this part of the year. Both the eddy covariance, and the throughfall and stemflow measurements showed average wet-canopy evaporation rate to be very low (0.05 mm/h) compared to the corresponding rainfall rate (3.06 mm/h). As a result, and despite the low canopy storage capacity of the coffee plantation (Cm, 0.50 mm), interception was dominated by post-event evaporation of intercepted water rather than by within-event evaporation. Comparing the results for the coffee plantation with interception data from mature and secondary LMCFs in the study area suggests that the conversion of LMCF to shade-coffee may lead to a decrease in interception loss of 8-18% of incident rainfall. This decrease is caused by a three- to seven-fold decrease in Cm due to the lower leaf area and smaller epiphyte biomass of the coffee plantation. Comparing the eddy covariance-based estimate of dry-canopy evaporation for the coffee plantation with sapflow-based estimates of transpiration for the LMCFs did not show any clear differences. Possibly, the potentially lower transpiration of the coffee plantation (due to its lower leaf area) may have been offset by higher soil/litter evaporation.