B7 - Drought Effects on Water Consumption and Plant Water Status of Trees in the Natural Forest

Abstract
ENSO-related droughts can have a large influence on growth and vitality of trees in tropical moist forests. As part of the replicated throughfall displacement experiment in Focus 3, this subproject investigates the response of water consumption, hydraulic conductivity and plant water status of various tree species in the species-rich natural forest in the Toro region to applied mild and severe droughts. We test the hypothesis that tree species differences in stem sap flow and vulnerability to cavitation during moist and dry periods are linked to two key attributes, wood density and rooting depths.

Summary
1. Despite high annual rainfall, one of the principal constraints of tree growth in tropical moist, non-seasonal forests is the irregular occurrence of drought on a supra-annual basis. There is evidence that drought-causing ENSO events of increasing strength and duration have developed during the past two decades as a result of global warming. This trend has also been confirmed for Indonesia and Malaysia with Sulawesi being particularly affected.

2. Droughts have numerous implications for forest ecosystems. Drought effects were studied at the physiological, the tree and the stand levels. Many physiological studies on the effects of water shortage on tropical trees have been conducted with seedlings, herbs or shrubs, whereas responses of adult trees in the tropical moist forests were much less investigated. On the stand level, drought severity and frequency have a large influence on the diversity, species composition, dynamics and productivity of tropical forests. Severe ENSO droughts and associated fires were found to increase tree mortality by factors of 12- 30 compared to normal years.

3. Severe embolism was identified as a major primary cause of death in droughted rainforest trees which may differ among the tree species of tropical moist forests. A second important controlling variable of drought resistance and water flux patterns in the stem seems to be wood density which typically is inversely correlated with the growth rate of trees. Species with high wood density are expected to be slow-growing and being less vulnerable to cavitation than species with low wood density. A third trait which should be influential for drought tolerance is root distribution and maximal rooting depth. Deeprooting species may partly escape mild or moderate drought periods by access to deeper water reserves; in this sense, they are drought-avoiding. However, contrary to this widely cited hypothesis, trees with short root systems in a Panamanian moist forest did not suffer more from a drought than canopy trees with deep roots. In conclusion, there is still much uncertainty about how ENSO drought of increasing strength and duration will affect tropical moist forests and what physiological and morphological traits are controlling drought tolerance of adult tropical trees.

4. The main objective of this subproject is to investigate the effects of reduced water availability on water consumption, hydraulic properties, leaf water status and phenology of important rainforest trees in a species-rich submontane moist forest of Central Sulawesi, and to relate drought response to tree morphological attributes. The project will address the following hypotheses:

(a) The leaf area/sapwood area ratio (LA/SA, known as Huber value) varies with changes in rainwater input and soil water content. At a constant evaporative demand, whole tree LA/SA is expected to decrease with decreasing soil water availability because leaves are progressively shed. (b) The rate of reduction of LA/SA is negatively correlated with the rate of eduction in canopy conductance. Drought-induced leaf shedding helps to reduce the risk of cavitation in twigs. (c) Leaf asymmetry can be used as a drought indicator. In contrast, leaf size reduction is thought to be no reliable indicator of trees drought response. In fact, European beech showed an increase, and not a decrease, in leaf size from wetter to drier stands (pers. observation). (d) Stem water content decreases with reduced rainwater input and soil water availability, but it increases with decreasing wood density. (e) Tree species with low wood density have relatively high maximum sap flow
densities, but reduce sap flow greatly under stress and are more sensitive in plant water status to drought. Tree species with a high wood density have less variation in their diurnal water flow rates, have a lower ks and are less vulnerable to cavitation. Thus, species with high wood density are less susceptible to droughts. (f) Drought events induce a stronger competition for water and trees with a deeper root system will show less seasonal variation
in xylem flux rates.

5. This project is part of Focus 3 and its Throughfall Displacement Experiment. It investigates the response of important natural forest tree species to experimental drought. It is complementary to project B4 (Hölscher/Schwendenmann) which studies the water relations of cacao agroforestry. The following variables will be studies before during and after two experimental desiccation periods under the roofs: xylem flux density, leaf water potential, vulnerability to cavitation, and leaf phenology. These parameters will be linked to traits such as wood density, rooting depth and hydraulic architecture.

6. The project is closely linked to other projects in Focus 3. B4 and B7 are complementary in that B4 supplies data on soil hydrology for both agroforest and natural forest, while B7 conducts leaf water potential measurements not only in the forest, but also in the cacao agroforest to support B4. Data will also be transferred to projects C4 (water consumption data and leaf water status), and C5 (leaf water status as stress indicators). The project B7 will obtain data from the projects B1 (local climate), B4 (soil moisture data and water consumption in agrofrest), B5 (soil chemical data), C2 (tree diversity and species identification), C3 (leaf herbivory of trees), C4 (water uptake of roots), and C5 (mycorrhizal infection may correlate with species tolerance to drought).