RELIANCE: nutRiEnt Limitations In A ugaNdan tropiCal forEst
Soil Sciences
Plant Cultivation, Plant Nutrition, Agricultural Technology
Final Report Abstract
Tropical forests are highly productive ecosystems and play a critical role in absorbing anthropogenic carbon dioxide (CO2) from the atmosphere, accounting for up to half of the terrestrial biosphere’s carbon (C) sequestration. For reasons not yet fully reconciled, there are indications that the C sink strength of these forests is slowly declining, thereby decreasing the buffering capacity these forests offer in mitigating global climate change. It is recognized that ecosystem nutrient limitations play a critical regulatory role in plant growth, therein affecting ecosystem C assimilation and specifically net primary production (NPP). However, the direction and magnitude of these limitations are poorly understood, especially on heavily weathered soils in humid tropical forests. To understand how nutrient availability constrains ecosystem processes on highly weathered soils in tropical forests in Africa, we established a large-scale nutrient manipulation experiment (NME) in a highly diverse semi-deciduous tropical forest in northwestern Uganda. Specifically, our objective was to investigate the roles of nitrogen (N), phosphorus (P) and potassium (K) and their interactions have on ecosystem processes across a hierarchy of scales, from microbial biomass to ecosystem NPP. The NME was set up using a factorial experimental design including eight nutrient addition treatments (N, P, K, and in combination as N+P, N+K, P+K and N+P+K), each replicated four times (n = 32). At each of the thirty-two 40 x 40 m plots, we measured the effects of the nutrient additions on NPP (work package 1), on plant-available nutrients in the soil (work package 2), and on nutrient turnover in decomposing leaf litter (work package 3). Through this large-scale NME, we determined that multiple nutrients (co)regulate different ecosystem processes, further substantiating the growing pool of evidence that tropical forests do not follow Liebig’s Law of the Minimum. Instead, resource and nutrient requirements (and their limitations) vary depending on the tree species, tree size (or age), reproductive stage and resource competition. While there were no community-wide tree growth responses to nutrient addition, different species responded differently. For instance, two species (Lasiodiscus mildbraedii and Rinorea ardisiaeflora; accounting for 12% of the total abundance) responded positively to N additions, and a semi-deciduous tree Celtis mildbraedii (15% of the total abundance) grew 300% faster in the K addition treatments during the first dry year of the experiment. For the latter, it is likely K addition helped delay leaf shedding during the dry season, which meant it could maintain photosynthesis during this period and continue to invest in stem growth. For another two tree species (Trichilia prieuriana and Trichilia rubescens), P and K additions inhibited tree growth, which could be related to nutrient imbalances or out-competition by surrounding trees that may require large quantities of P and/or K. Next, biomass accumulation responses to nutrient addition were also dependent on tree sizes. Specifically, in large trees (>30 cm diameter at breast height (DBH)) the biomass increment (growth rate) only increased when all three nutrients (N+P+K) were applied simultaneously. Medium sized trees (10-30 cm DBH), grew significantly faster in all N addition treatments in the second year of the experiment. Lastly, although tree growth among poles and saplings (1-10 cm DBH) was not influenced by any nutrient additions, P addition did reduce tree mortality of saplings (1-5 cm DBH). Leaf litter fall rates are driven by both P and K when trees face strong drought stress, as experienced in the first dry season of the experiment. During this time, it is likely that the addition of both P and K (individually and in combination) reduced leaf shedding by improving water use efficiency in plant leaves. Furthermore, the addition of N reduced fine root biomass by 35% in the first year of the experiment, suggesting an ecosystemscale N limitation. This fast, dramatic reduction in fine root biomass in the N treatments highlights that maintaining a large fine root network is an energy and resource intensive process, and trees will scale back their root network when they have adequate resources available. Specifically, the reductions in fine root biomass correlated strongly with increases in nitrate concentration (NO3), the N form preferred by most trees. Strongest reductions in fine root biomass occurred when the NH4 to NO3 ratio was <1. Nutrient additions caused a cascade of biochemical responses in the soil nutrient availability. First, N addition (as urea) generally (but not always) increased soil nitrification rates, causing NH4 concentrations to decrease in the soil and driving up NO3 concentrations. Second, microbial biomass N decreased in the N addition plots during the dry season. Lastly, P additions increased plant available P by 53%. This large increase, could indicate that the demand for P was not very high. The role nutrients play in litter decomposition is more complex, however. While P addition promoted arthropod activity and accordingly litter fragmentation (evident in increased decomposition in large mesh sized litter bags), the microbial communities responsible for decomposition were not affected by nutrient additions. Correlations between litter decay rates and soil and leaf nitrogen measures however show that N has an inhibitory role in decomposition, particularly lignin decomposition. Asymbiotic N fixation on leaf litter was extremely low in this ecosystem (<1 Kg N ha yr-1) and not directly regulated by nutrient additions. Collectively, these results suggest that all three macronutrients are critical in driving different ecological processes, ranging from decomposition (N and P) to litterfall (P and K) to rooting biomass (N) to (aboveground tree growth (N, P and K).
