No abstract

Question: How do community-weighted means of traits (CWM) and functional dispersion (FDis), a measure of trait variability, change in response to gradients of temperature, precipitation, soil nutrients, and disturbance? Is the decrease in trait similarity between plots continuous or discontinuous? Is species turnover between plots linked to trait turnover? Location: Mount Kilimanjaro, Tanzania, Africa. Accepted ArticleThis article is protected by copyright. All rights reserved.Methods: Sixty plots were established in twelve major vegetation types on Mount Kilimanjaro, covering large gradients of temperature, precipitation, soil nutrients, and anthropogenic disturbance representing the dominant ecosystems in East Africa. Environmental data, plant abundances, and plant traits were recorded for each plot. Trait CWM and FDis were related to environmental factors with partial least squares regressions. Trait similarity between pairs of plots was assessed with a null-model approach.Results: Both CWM and FDis of most traits responded strongly to environmental factors, particularly to precipitation and disturbance. FDis of traits associated with growth and reproduction mostly increased with temperature and precipitation, and decreased with disturbance. Pairwise plot comparisons revealed an inverse relationship of trait similarity with differences in temperature, precipitation, and anthropogenic disturbance, respectively. However, changes in similarity were often discontinuous rather than continuous. Several vegetation types differed strongly in species composition but not in traits. Conclusions:Trait dispersion indicating functional niches increased with productivity and temperature. Conversely, low-productivity conditions were characterized by trait convergence.Discontinuous changes in trait similarity between plots suggested tipping points at which trait expressions change strongly to adjust to environmental conditions. Large sections of the temperature gradient were characterized by species turnover with only minor changes in traits, indicating that the functional composition may be resilient against gradual environmental changes until a tipping point is reached.

In this study, we quantify the impacts of climate and land use on soil N O and CH fluxes from tropical forest, agroforest, arable and savanna ecosystems in Africa. To do so, we measured greenhouse gases (GHG) fluxes from 12 different ecosystems along climate and land-use gradients at Mt. Kilimanjaro, combining long-term in situ chamber and laboratory soil core incubation techniques. Both methods showed similar patterns of GHG exchange. Although there were distinct differences from ecosystem to ecosystem, soils generally functioned as net sources and sinks for N O and CH respectively. N O emissions correlated positively with soil moisture and total soil nitrogen content. CH uptake rates correlated negatively with soil moisture and clay content and positively with SOC. Due to moderate soil moisture contents and the dominance of nitrification in soil N turnover, N O emissions of tropical montane forests were generally low (<1.2 kg N ha year ), and it is likely that ecosystem N losses are driven instead by nitrate leaching (~10 kg N ha year ). Forest soils with well-aerated litter layers were a significant sink for atmospheric CH (up to 4 kg C ha year ) regardless of low mean annual temperatures at higher elevations. Land-use intensification significantly increased the soil N O source strength and significantly decreased the soil CH sink. Compared to decreases in aboveground and belowground carbon stocks enhanced soil non-CO GHG emissions following land-use conversion from tropical forests to homegardens and coffee plantations were only a small factor in the total GHG budget. However, due to lower ecosystem carbon stock changes, enhanced N O emissions significantly contributed to total GHG emissions following conversion of savanna into grassland and particularly maize. Overall, we found that the protection and sustainable management of aboveground and belowground carbon and nitrogen stocks of agroforestry and arable systems is most crucial for mitigating GHG emissions from land-use change.

No abstract

Tropical ecosystems are under increasing pressure of land‐use changes, strongly affecting the carbon cycle. Conversion from natural to agri‐cultural ecosystems is often accompanied by a decrease in the stocks of organic and microbial carbon (Corg, Cmic) as well as changes in microbial activity and litter decomposition. Eleven ecosystems along an elevation gradient on the slopes of Mt. Kilimanjaro were used to investigate impacts of land‐use changes on Corg and Cmic stocks as well as the specific metabolic respiration quotient (qsCO2) in surface soils. Six natural, two semi‐natural and three intensively used agricultural ecosystems were investigated on an elevation gradient from 950 to 3,880 m asl. To estimate the effects of precipitation, rainfall regimes of 3·6 and 20·0 mm were simulated. Corg stocks were controlled by water availability, temperature and net primary production. Agricultural management resulted in decreases of Corg and Cmic stocks by 38% and 76%, respectively. In addition, agricultural systems were characterized by low Cmic:Corg ratios, indicating a decline in available substrate. Enhanced land‐use intensity leads to increased qsCO2 (agricultural > semi‐natural > natural). The traditional homegardens stood out as a sustainable land‐use form with high substrate availability and microbial efficiency. Soil CO2 efflux and qsCO2 generally increased with precipitation level. We conclude that soils of Mt. Kilimanjaro's ecosystems are highly sensitive to land‐use changes and are vulnerable to changes in precipitation, especially at low elevations. Even though qsCO2 was measured under different water contents, it can be used as an indicator of ecosystem disturbances caused by land‐use and management practices. Copyright © 2015 John Wiley & Sons, Ltd.