Fluxes and composition dynamics of atmospheric nitrogen deposition play key roles in better balancing economic development and ecological environment. However, there are some knowledge gaps and difficulties in urban ecosystems, especially for small and medium-sized cities. In this study, both flux and composition (ratio of NH4+-N to NO3--N, RN) of wet-deposited dissolved inorganic nitrogen (DIN, sum of NO3--N and NH4+-N) were estimated and sources were identified at a long-term urban observation station in Tongling, a typical medium-sized city in eastern China during 2010-2016, respectively. Results showed that wet-deposited DIN fluxes were 33.20 and 28.15 kgN ha-1 yr-1 in Tongling city during 2010-2011 and 2015-2016, respectively. Compared to these two periods, both DIN and NO3--N fluxes decreased by 15.2% and 31.8% for a series of NOx abatement measures applied effectively, respectively. At the same time, the NH4+-N flux remained stable and ranged from 19.53 to 20.62 kgN ha-1 yr-1, and the RN increased from 1.7 to 2.2. Seasonally, winds from the southwest and west-southwest with higher frequencies and speeds in spring and summer brought more NH4+-N and DIN wet deposition from an ammonia plant, which could threaten the safety of regional hydrosphere ecosystems. On the whole, the wet-deposited NH4+-N was threatening regional ecosystems of both the hydrosphere and forest. The wet-deposited DIN including NH4+-N in Tongling city stemmed mainly from a combined source of coal combustion and dust from Cu extraction and smelting, ammonia production, and roads. Therefore, production lines should be updated for Cu extraction and smelting industries, thermal power generations and the ammonia plant, old vehicles should be eliminated, and the use of new energy vehicles should be promoted for regional sustainable development and human health in the medium-sized city.
Land use plays a significant role in determining the spatial patterns of water quality in the Johor River Basin (JRB), Malaysia. In the recent years, there have been several occurrences of pollution in these rivers, which has generated concerns over the long-term sustainability of the water resources in the JRB. Specifically, this water resource is a shared commodity between two states, namely, Johor state of Malaysia and Singapore, a neighbouring country adjacent to Malaysia. Prior to this study, few research on the influence of land use configuration on water quality have been conducted in Johor. In addition, it is also unclear how water quality varies under different seasonality in the presence of point sources. In this study, we investigated the influence of land use and point sources from wastewater treatment plants (WWTPs) on the water quality in the JRB. Two statistical techniques - Multivariate Linear Regression (MLR) and Redundancy Analysis (RA) were undertaken to analyse the relationships between river water quality and land use configuration, as well as point sources from WWTPs under different seasonality. Water samples were collected from 49 sites within the JRB from March to December in 2019. Results showed that influence from WWTPs on water quality was greater during the dry season and less significant during the wet season. In particular, point source was highly positively correlated with ammoniacal‑nitrogen (NH3-N). On the other hand, land use influence was greater than point source influence during the wet season. Residential and urban land use were important predictors for nutrients and organic matter (chemical oxygen demand); and forest land use were important sinks for heavy metals but a significant source of manganese.
To understand the variations in the decomposability of tropical peat soil following deforestation for an oil palm plantation, a field incubation experiment was conducted in Sarawak, Malaysia. Peat soils collected from three types of primary forest, namely Mixed Peat Swamp (MPS; Gonystylus-Dactylocladus-Neoscrotechinia association), Alan Batu (ABt; Shorea albida-Gonstylus-Strenonurus association), and Alan Bunga (ABg; Shorea albida association), were packed in polyvinyl chloride pipes and installed in an oil palm plantation. Carbon dioxide (CO2) and methane (CH4) fluxes from soil were monthly measured for 3years. Environmental variables including soil temperature, soil moisture content, and groundwater table were also monitored. The pH, loss on ignition, and total carbon (C) content were similar among the three soils, while total N content was larger in the MPS than in the ABg soils. Based on13C nuclear magnetic resonance (NMR) spectroscopy, C composition of the MPS and ABg soils was characterized by the largest proportion of C present as alkyl C and O-alkyl C, respectively. The C composition of the ABt soil was intermediate between the MPS and ABg soils. The CO2fluxes from the three soils ranged from 78 to 625mgCm-2h-1with a negative correlation to groundwater level. The CH4fluxes ranged from -67 to 653μgCm-2h-1. Both total CO2and CH4fluxes were larger in the order ABg>ABt>MPS (P<0.05). Annual rate of peat decomposition as was estimated from cumulative C loss differed up to 2 times, and the rate constant in exponential decay model was 0.033y-1for the MPS soil and 0.066y-1for the ABg soil. The field incubation results of the three forest peat soils seem to reflect the difference in the labile organic matter content, represented by polysaccharides.
The conversions of forests and grass land to urban and farmland has exerted significant changes on terrestrial ecosystems. However, quantifying how these changes can affect the quality of water resources is still a challenge for hydrologists. Nitrate concentrations can be applied as an indicator to trace the link between land use changes and groundwater quality due to their solubility and easy transport from their source to the groundwater. In this study, 25year records (from 1989 to 2014) of nitrate concentrations are applied to show the impact of land use changes on the quality of groundwater in Northern Kelantan, Malaysia, where large scale deforestation in recent decades has occurred. The results from the integration of time series analysis and geospatial modelling revealed that nitrate (NO3-N) concentrations significantly increased with approximately 8.1% and 3.89% annually in agricultural and residential wells, respectively, over 25years. In 1989 only 1% of the total area had a nitrate value greater than 10mg/L; and this value increased sharply to 48% by 2014. The significant increase in nitrate was only observed in a shallow aquifer with a 3.74% annual nitrate increase. Based on the result of the Autoregressive Integrated Moving Average (ARIMA) model the nitrate contamination is expected to continue to rise by about 2.64% and 3.9% annually until 2030 in agricultural and residential areas. The present study develops techniques for detecting and predicting the impact of land use changes on environmental parameters as an essential step in land and water resource management strategy development.
Tropical peatlands are globally important ecosystems with high C storage and are endangered by anthropogenic disturbances. Microbes in peatlands play an important role in sustaining the functions of peatlands as a C sink, yet their characteristics in these habitats are poorly understood. This research aimed to elucidate the responses of these complex ecosystems to disturbance by exploring greenhouse gas (GHG) emissions, nutrient contents, soil microbial communities and the functional interactions between these components in a primary and secondary peat swamp forest in Peninsular Malaysia. GHG measurements using closed chambers, and peat sampling were carried out in both wet and dry seasons. Microbial community phenotypes and nutrient content were determined using phospholipid fatty acid (PLFA) and inductively-coupled plasma mass spectrometry (ICP-MS) analyses respectively. CO2 emissions in the secondary peat swamp forest were > 50% higher than in the primary forest. CH4 emission rates were ca. 2 mg m-2 h-1 in the primary forest but the secondary forest was a CH4 sink, showing no seasonal variations in GHG emissions. Almost all the nutrient concentrations were significantly lower in the secondary forest, postulated to be due to nutrient leaching via drainage and higher rates of decomposition. Cu and Mo concentrations were negatively correlated with CO2 and CH4 emissions respectively. Microbial community structure was overwhelmingly dominated by bacteria in both forest types, however it was highly sensitive to land-use change and season. Gram-positive and Gram-negative relative abundance were positively correlated with CO2 and CH4 emissions respectively. Drainage related disturbances increased CO2 emissions, by reducing the nutrient content including some with known antimicrobial properties (Cu & Na) and by favouring Gram-positive bacteria over Gram-negative bacteria. These results suggest that the biogeochemistry of secondary peat swamp forest is fundamentally different from that of primary peat swamp forest, and these differences have significant functional impacts on their respective environments.
Tropical primary forests are being destroyed at an alarming rate and converted for other land uses which is expected to greatly influence soil carbon (C) cycling. However, our understanding of how tropical forest conversions affect the accumulation of compounds in soil functional C pools remains unclear. Here, we collected soils from primary forests (PF), secondary forests (SF), oil-palm (OP), and rubber plantations (RP), and assessed the accumulation of plant- and microbial-derived compounds within soil organic carbon (SOC), particulate (POC) and mineral-associated (MAOC) organic C. PF conversion to RP greatly decreased SOC, POC, and MAOC concentrations, whereas conversion to SF increased POC concentrations and decreased MAOC concentrations, and conversion to OP only increased POC concentrations. PF conversion to RP decreased lignin concentrations and increased amino sugar concentrations in SOC pools which increased the stability of SOC, whereas conversion to SF only increased the lignin concentrations in POC, and conversion to OP just increased lignin concentrations in POC and decreased it in MAOC. We observed divergent dynamics of amino sugars (decrease) and lignin (increase) in SOC with increasing SOC. Only lignin concentrations increased in POC with increasing POC and amino sugars concentrations decreased in MAOC with increasing MAOC. Conversion to RP significantly decreased soil enzyme activities and microbial biomasses. Lignin accumulation was associated with microbial properties, whereas amino sugar accumulation was mainly associated with soil nutrients and stoichiometries. These results suggest that the divergent accumulation of plant- and microbial-derived C in SOC was delivered by the distribution and original composition of functional C pools under forest conversions. Forest conversions changed the formation and stabilization processes of SOC in the long run which was associated with converted plantations and management. The important roles of soil nutrients and stoichiometry also provide a natural-based solution to enhance SOC sequestration via nutrient management in tropical forests.
Drought affects vegetation growth to a large extent. Understanding the dynamic changes of vegetation during drought is of great significance for agricultural and ecological management and climate change adaptation. The relations between vegetation and drought have been widely investigated, but how vegetation loss and restoration in response to drought remains unclear. Using the standardized precipitation evapotranspiration index (SPEI) and the normalized difference vegetation index (NDVI) data, this study developed an evaluation framework for exploring the responses of vegetation loss and recovery to meteorological drought, and applied it to the humid subtropical Pearl River basin (PRB) in southern China for estimating the loss and recovery of three vegetation types (forest, grassland, cropland) during drought using the observed NDVI changes. Results indicate that vegetation is more sensitive to drought in high-elevation areas (lag time 8 months). Vegetation loss (especially in cropland) is found to be more sensitive to drought duration than drought severity and peak. No obvious linear relationship between drought intensity and the extent of vegetation loss is found. Regardless of the intensity, drought can cause the largest probability of mild loss of vegetation, followed by moderate loss, and the least probability of severe loss. Large spatial variability in the probability of vegetation loss and recovery time is found over the study domain, with a higher probability (up to 50 %) of drought-induced vegetation loss and a longer recovery time (>7 months) mostly in the high-elevation areas. Further analysis suggests that forest shows higher but cropland shows lower drought resistance than other vegetation types, and grassland requires a shorter recovery time (4.2-month) after loss than forest (5.1-month) and cropland (4.8-month).
Information on temporal and spatial variations in soil greenhouse gas (GHG) fluxes from tropical peat forests is essential to predict the influence of climate change and estimate the effects of land use on global warming and the carbon (C) cycle. To obtain such basic information, soil carbon dioxide (CO2) and methane (CH4) fluxes, together with soil physicochemical properties and environmental variables, were measured at three major forest types in the Maludam National Park, Sarawak, Malaysia, for eight years, and their relationships were analyzed. Annual soil CO2 fluxes ranged from 860 to 1450 g C m⁻2 yr⁻1 without overall significant differences between the three forest sites, while soil CH4 fluxes, 1.2-10.8 g C m⁻2 yr⁻1, differed. Differences in GHG fluxes between dry and rainy seasons were not necessarily significant, corresponding to the extent of seasonal variation in groundwater level (GWL). The lack of significant differences in soil CO2 fluxes between the three sites could be attributed to set-off between the negative and positive effects of the decomposability of soil organic matter as estimated by pyrophosphate solubility index (PSI) and GWL. The impact of El-Niño on annual CO2 flux also varied between the sites. The variation in soil CH4 fluxes from the three sites was enhanced by variations in temperature, GWL, PSI, and soil iron (Fe) content. A positive correlation was observed between the annual CH4 flux and GWL at only one site, and the influence of soil properties was more pronounced at the site with the lowest GWL and the highest PSI. Variation in annual CH4 fluxes was controlled more strongly by temperature where GWL was the highest and GWL and plant growth fluctuations were the least. Inter-annual variations in soil CO2 and CH4 fluxes confirmed the importance of long-term monitoring of these at multiple sites supporting different forest types.
Largely driven by agricultural pressures, biodiversity has experienced great changes globally. Exploring biodiversity responses to agricultural practices associated with agricultural intensification can benefit biodiversity conservation in agricultural landscapes. However, the effects of agricultural practices may also extend to natural habitats. Moreover, agricultural impacts may also vary with geographical region. We analyze biodiversity responses to landscape cropland coverage, cropping frequency, fertiliser and yield, among different land-use types and across geographical regions. We find that species richness and total abundance generally respond negatively to increased landscape cropland coverage. Biodiversity reductions in human land-use types (pasture, plantation forest and cropland) were stronger in tropical than non-tropical regions, which was also true for biodiversity reductions with increasing yield in both human and natural land-use types. Our results underline substantial biodiversity responses to agricultural practices not only in cropland but also in natural habitats, highlighting the fact that biodiversity conservation demands a greater focus on optimizing agricultural management at the landscape scale.
Tree mortality rates appear to be increasing in moist tropical forests (MTFs) with significant carbon cycle consequences. Here, we review the state of knowledge regarding MTF tree mortality, create a conceptual framework with testable hypotheses regarding the drivers, mechanisms and interactions that may underlie increasing MTF mortality rates, and identify the next steps for improved understanding and reduced prediction. Increasing mortality rates are associated with rising temperature and vapor pressure deficit, liana abundance, drought, wind events, fire and, possibly, CO2 fertilization-induced increases in stand thinning or acceleration of trees reaching larger, more vulnerable heights. The majority of these mortality drivers may kill trees in part through carbon starvation and hydraulic failure. The relative importance of each driver is unknown. High species diversity may buffer MTFs against large-scale mortality events, but recent and expected trends in mortality drivers give reason for concern regarding increasing mortality within MTFs. Models of tropical tree mortality are advancing the representation of hydraulics, carbon and demography, but require more empirical knowledge regarding the most common drivers and their subsequent mechanisms. We outline critical datasets and model developments required to test hypotheses regarding the underlying causes of increasing MTF mortality rates, and improve prediction of future mortality under climate change.
What causes individual tree death in tropical forests remains a major gap in our understanding of the biology of tropical trees and leads to significant uncertainty in predicting global carbon cycle dynamics. We measured individual characteristics (diameter at breast height, wood density, growth rate, crown illumination and crown form) and environmental conditions (soil fertility and habitat suitability) for 26 425 trees ≥ 10 cm diameter at breast height belonging to 416 species in a 52-ha plot in Lambir Hills National Park, Malaysia. We used structural equation models to investigate the relationships among the different factors and tree mortality. Crown form (a proxy for mechanical damage and other stresses) and prior growth were the two most important factors related to mortality. The effect of all variables on mortality (except habitat suitability) was substantially greater than expected by chance. Tree death is the result of interactions between factors, including direct and indirect effects. Crown form/damage and prior growth mediated most of the effect of tree size, wood density, fertility and habitat suitability on mortality. Large-scale assessment of crown form or status may result in improved prediction of individual tree death at the landscape scale.
Plant functional traits regulate ecosystem functions but little is known about how co-occurring gradients of land use and edaphic conditions influence their expression. We test how gradients of logging disturbance and soil properties relate to community-weighted mean traits in logged and old-growth tropical forests in Borneo. We studied 32 physical, chemical and physiological traits from 284 tree species in eight 1 ha plots and measured long-term soil nutrient supplies and plant-available nutrients. Logged plots had greater values for traits that drive carbon capture and growth, whilst old-growth forests had greater values for structural and persistence traits. Although disturbance was the primary driver of trait expression, soil nutrients explained a statistically independent axis of variation linked to leaf size and nutrient concentration. Soil characteristics influenced trait expression via nutrient availability, nutrient pools, and pH. Our finding, that traits have dissimilar responses to land use and soil resource availability, provides robust evidence for the need to consider the abiotic context of logging when predicting plant functional diversity across human-modified tropical forests. The detection of two independent axes was facilitated by the measurement of many more functional traits than have been examined in previous studies.
Arbuscular mycorrhizal fungi (AMF) and ectomycorrhizal fungi (EMF) produce contrasting plant-soil feedbacks, but how these feedbacks are constrained by lithology is poorly understood. We investigated the hypothesis that lithological drivers of soil fertility filter plant resource economic strategies in ways that influence the relative fitness of trees with AMF or EMF symbioses in a Bornean rain forest containing species with both mycorrhizal strategies. Using forest inventory data on 1245 tree species, we found that although AMF-hosting trees had greater relative dominance on all soil types, with declining lithological soil fertility EMF-hosting trees became more dominant. Data on 13 leaf traits and wood density for a total of 150 species showed that variation was almost always associated with soil type, whereas for six leaf traits (structural properties; carbon, nitrogen, phosphorus ratios, nitrogen isotopes), variation was also associated with mycorrhizal strategy. EMF-hosting species had slower leaf economics than AMF-hosts, demonstrating the central role of mycorrhizal symbiosis in plant resource economies. At the global scale, climate has been shown to shape forest mycorrhizal composition, but here we show that in communities it depends on soil lithology, suggesting scale-dependent abiotic factors influence feedbacks underlying the relative fitness of different mycorrhizal strategies.
The evolutionary history of the colugo, a gliding arboreal mammal distributed throughout Sundaland, was influenced by the location of and connections between forest habitats. By comparing colugo phylogenetic patterns, species ecology, sample distributions, and times of divergence to those of other Sundaic taxa with different life-history traits and dispersal capabilities, we inferred the probable distribution of paleo-forest corridors and their influence on observed biogeographic patterns. We identified a consistent pattern of early diversification between east and west Bornean lineages in colugos, lesser mouse deer, and Sunda pangolins, but not in greater mouse deer. This deep east-west split within Borneo has not been commonly described in mammals. Colugos on West Borneo diverged from colugos in Peninsular Malaysia and Sumatra in the late Pliocene, however most other mammalian populations distributed across these same geographic regions diverged from a common ancestor more recently in the Pleistocene. Low genetic divergence between colugos on large landmasses and their neighboring satellite islands indicated that past forest distributions were recently much larger than present refugial distributions. Our analysis of colugo evolutionary history reconstructs Borneo as the most likely ancestral area of origin for Sunda colugos, and suggests that forests present during the middle Pliocene within the Sunda Shelf were more evergreen and contiguous, while forests were more fragmented, transient, seasonal, or with lower density canopies in the Pleistocene.
The expansion of oil palm plantations at the expense of tropical forests is causing declines in many species and altering ecosystem functions. Maintaining forest-dependent species and processes in these landscapes may therefore limit the negative impacts of this economically important industry. Protecting riparian vegetation may be one such opportunity; forest buffer strips are commonly protected for hydrological reasons, but can also conserve functionally important taxa and the processes they support.We surveyed leaf litter ant communities within oil palm-dominated landscapes in Sabah, Malaysia, using protein baits. As the scavenging activity of ants influences important ecological characteristics such as nutrient cycling and soil structure, we quantified species-specific rates of bait removal to examine how this process may change across land uses and establish which changes in community structure underlie observed shifts in activity.Riparian reserves had similar ant species richness, community composition and scavenging rates to nearby continuous logged forest. Reserve width and vegetation structure did not affect ant species richness significantly. However, the number of foraging individuals decreased with increasing reserve width, and scavenging rate increased with vegetation complexity.Oil palm ant communities were characterized by significantly lower species richness than logged forest and riparian reserves and also by altered community composition and reduced scavenging rates.Reduced scavenging activity in oil palm was not explained by a reduction in ant species richness, nor by replacement of forest ant species by those with lower per species scavenging rates. There was also no significant effect of land use on the scavenging activity of the forest species that persisted in oil palm. Rather, changes in scavenging activity were best explained by a reduction in the mean rate of bait removal per individual ant across all species in the community.Synthesis and applications. Our results suggest that riparian reserves are comparable to areas of logged forest in terms of ant community composition and ant-mediated scavenging. Hence, in addition to protecting large continuous areas of primary and logged forest, maintaining riparian reserves is a successful strategy for conserving leaf litter ants and their scavenging activities in tropical agricultural landscapes.
Selective logging is the dominant form of human disturbance in tropical forests, driving changes in the abundance of vertebrate and invertebrate populations relative to undisturbed old-growth forests. A key unresolved question is understanding which physiological mechanisms underlie different responses of species and functional groups to selective logging. Regulation of oxidative status is thought to be one major physiological mechanism underlying the capability of species to cope with environmental changes. Using a correlational cross-sectional approach, we compared a number of oxidative status markers among 15 understorey bird species in unlogged and selectively logged forest in Borneo in relation to their feeding guild. We then tested how variation of markers between forest types was associated with that in population abundance. Birds living in logged forests had a higher activity of the antioxidant enzyme superoxide dismutase and a different regulation of the glutathione cycle compared to conspecific birds in unlogged forest. However, neither oxidative damage nor oxidized glutathione differed between forest types. We also found that omnivores and insectivores differed significantly in all markers related to the key cellular antioxidant glutathione irrespective of the forest type. Species with higher levels of certain antioxidant markers in a given type of forest were less abundant in that forest type compared to the other. Our results suggest that there was little long-term effect of logging (last logging rotation occurred ~15 years prior to the study) on the oxidative status of understorey bird species. However, it is unclear if this was owing to plasticity or evolutionary change. Our correlative results also point to a potential negative association between some antioxidants and population abundance irrespective of the forest type.
Studies have shown that food chain length is governed by interactions between species richness, ecosystem size and resource availability. While redundant trophic links may buffer impacts of species loss on food chain length, higher extinction risks associated with predators may result in bottom-heavy food webs with shorter food chains. The lack of consensus in earlier empirical studies relating species richness and food chain length reflects the need to account robustly for the factors described above. In response to this, we conducted an empirical study to elucidate impacts of land-use change on food chain length in tropical forest streams of Southeast Asia. Despite species losses associated with forest loss at our study areas, results from amino acid isotope analyses showed that food chain length was not linked to land use, ecosystem size or resource availability. Correspondingly, species losses did not have a significant effect on occurrence likelihoods of all trophic guilds except herbivores. Impacts of species losses were likely buffered by initial high levels of trophic redundancy, which declined with canopy cover. Declines in trophic redundancy were most drastic amongst invertivorous fishes. Declines in redundancy across trophic guilds were also more pronounced in wider and more resource-rich streams. While our study found limited evidence for immediate land-use impacts on stream food chains, the potential loss of trophic redundancy in the longer term implies increasing vulnerability of streams to future perturbations, as long as land conversion continues unabated.
The assembly of species communities at local scales is thought to be driven by environmental filtering, species interactions and spatial processes such as dispersal limitation. Little is known about how the relative balance of these drivers of community assembly changes along environmental gradients, especially man-made environmental gradients associated with land-use change. Using concurrent camera- and live-trapping, we investigated the local-scale assembly of mammal communities along a gradient of land-use intensity (old-growth forest, logged forest and oil palm plantations) in Borneo. We hypothesised that increasing land-use intensity would lead to an increasing dominance of environmental control over spatial processes in community assembly. Additionally, we hypothesised that competitive interactions among species might reduce in concert with declines in α-diversity (previously documented) along the land-use gradient. To test our first hypothesis, we partitioned community variance into the fractions explained by environmental and spatial variables. To test our second hypothesis, we used probabilistic models of expected species co-occurrence patterns, in particular focussing on the prevalence of spatial avoidance between species. Spatial avoidance might indicate competition, but might also be due to divergent habitat preferences. We found patterns that are consistent with a shift in the fundamental mechanics governing local community assembly. In support of our first hypothesis, the importance of spatial processes (dispersal limitation and fine-scale patterns of home-ranging) appeared to decrease from low to high intensity land-uses, whilst environmental control increased in importance (in particular due to fine-scale habitat structure). Support for our second hypothesis was weak: whilst we found that the prevalence of spatial avoidance decreased along the land-use gradient, in particular between congeneric species pairs most likely to be in competition, few instances of spatial avoidance were detected in any land-use, and most were likely due to divergent habitat preferences. The widespread changes in land-use occurring in the tropics might be altering not just the biodiversity found in landscapes, but also the fundamental mechanics governing the local assembly of communities. A better understanding of these mechanics, for a range of taxa, could underpin more effective conservation and management of threatened tropical landscapes.
Different mechanisms have been proposed to explain how vertical and horizontal heterogeneity in light conditions enhances tree species coexistence in forest ecosystems. The foliage partitioning theory proposes that differentiation in vertical foliage distribution, caused by an interspecific variation in mortality-to-growth ratio, promotes stable coexistence. In contrast, successional niche theory posits that horizontal light heterogeneity, caused by gap dynamics, enhances species coexistence through an interspecific trade-off between growth rate and survival. To distinguish between these theories of species coexistence, we analyzed tree inventory data for 370 species from the 50-ha plot in Pasoh Forest Reserve, Malaysia. We used community-wide Bayesian models to quantify size-dependent growth rate and mortality of every species. We compared the observed size distributions and the projected distributions from size-dependent demographic rates. We found that the observed size distributions were not simply correlated with the rate of population increase but were related to demographic properties such as size growth rate and mortality. Species with low relative abundance of juveniles in size distribution showed high growth rate and low mortality at small tree sizes and low per-capita recruitment rate. Overall, our findings were in accordance with those predicted by foliage partitioning theory.
Tree species in tropical rain forests exhibit a rich panoply of spatial patterns that beg ecological explanation. The analysis of tropical census data typically relies on spatial statistics, which quantify the average aggregation tendency of a species. In this article we develop a cluster-based approach that complements traditional spatial statistics in the exploration and analysis of ecological hypotheses for spatial pattern. We apply this technique to six study species within a fully mapped 50-ha forest census in peninsular Malaysia. For each species we identify the scale(s) of spatial aggregation and the corresponding tree clusters. We study the correlation between cluster locations and abiotic variables such as topography. We find that the distribution of cluster sizes exhibits equilibrium and nonequilibrium behavior depending on species life history. The distribution of tree diameters within clusters also varies according to species life history. At different spatial scales, we find evidence for both niche-based and dispersal-limited processes producing spatial pattern. Our methodology for identifying scales of aggregation and clusters is general; we discuss the method's applicability to spatial problems outside of tropical plant ecology.