Leaf traits and foliar CO₂ exchange in a Peruvian tropical montane cloud forest

Abstract

Tropical montane cloud forests (TMCF) are one of the most fascinating, but least understood ecosystems in the world, and the interest in the carbon (C) cycle of TMCFs with regard to carbon sequestration and storage practices has increased rapidly in recent years. One feature that prevails in all TMCFs is a decrease in aboveground net primary productivity (ANPP) and standing biomass and leaf area index (LAI) with increasing altitude, together with the stunted growth form of the trees. This thesis focuses on the input part of the TMCF C-cycle, and investigates the controlling factors on photosynthesis on a leaf, canopy, and ecosystem level in the Kosñipata valley in south east Peru, on the eastern slope of the Andes (13º11’28’’S / 71º35’24’’W). Leaf traits are known to relate to foliar C-exchange, and compared with other altitudinal transect studies of TMCFs, the studied sites had similar altitudinal trends for foliar nitrogen (N) content (though not for phosphorus) and leaf mass per area (LMA), with N content decreasing and LMA increasing with altitude. N concentrations were relatively high and LMA values relatively low, but this observed relationship was consistent with those found in global leaf trait surveys. Examining plant stoichiometry (i.e. N:P ratios), the data suggests that unlike the general hypothesis, the Kosñipata forests are not N limited, except for the study site at 2990 m a.s.l.. At the 2990 m a.s.l. site, which is the focal study site of the thesis, photosynthetic parameters Vcmax (the carboxylation efficiency of the Rubisco protein) and Jmax (the electron transport efficiency) proved to be similar to those found in lowland tropical rainforest leaves when expressed on an area basis and standardised to 25 °C (55.6 ± 2.6 and 106.5 ± 5.2 mmol m-2 s-1, for Vcmax and Jmax, respectively). However, when standardised to the mean ambient TMCF temperature of 12.5 °C, both photosynthetic parameters were much lower than ambient tropical rainforest Vcmax and Jmax values. The TMCF Jmax -Vcmax relationships were steeper than found in other tropical biomes, indicating a possible adaptation to the lower light availability in TMCFs because of frequent cloud cover, or a consequence of little atmospheric evaporative demand, which is also due to the humid conditions in this forest type. Although N-Vcmax relationships were significant (P<0.05), the fit was not very strong and the relationship between nitrogen use efficiency (NUE) and Vcmax indicates that TMCF species can be regarded as a different plant functional type compared with other tropical forest types. Diurnal measurements of net photosynthesis (A), stomatal conductance (gs) and leaf water potential (Yleaf) showed that different TMCF species experienced non-contrasting diurnal patterns of Yleaf and gs in the dry season. The observed patterns suggest that some TMCF species can be classified as isohydric species, while others behave anisohydrically. Additionally, in situ gs was not very responsive to these to the range of experienced photosynthetically active radiation (PAR), vapour pressure deficit (VPD) or soil water content (SWC), leading to the conclusion that in the studied TMCF, drought stress does not play a role in C-uptake. When using the measured photosynthetic parameters for up-scaling C-uptake to stand scale with a Soil-Plant-Atmosphere model, simulated annual gross primary productivity (GPP) was 16.24 ±1.6 T C ha-1 yr-1, which is about half the GPP observed in neotropical lowland rainforests. Analyses of the modelled results showed that GPP in this TMCF is mostly controlled by temperature, PAR and leaf area index (LAI) and when increasing these three factors to values found in tropical lowland forest, GPP increased up to 75%. In addition, the modelled results indicate that hydraulic limitations on TMCF C-uptake are very unlikely under current climatic conditions. The modelled results also showed that increases in radiation as a result of less cloud cover do not translate to straightforward increases of GPP. The cloudy conditions of TMCFs, which reduced incident PAR in TMCFs, should therefore not be regarded simply as a negative control on TMCF GPP. Instead, the increase in fraction of diffuse radiation partially offsets the decrease in GPP following the reduction in PAR. Overall, the results of this study show that leaves of Andean TMCF forests have similar C-uptake capacity to tropical lowland rainforests when standardized to similar temperatures, but that for in situ C-uptake temperature, radiation and LAI are the key controls.