Soil fungal communities reflect aspect-driven environmental structuring and vegetation types in a Pannonian forest landscape

In temperate regions, slope aspect is one of the most influential drivers of environmental conditions at landscape level. The effect of aspect on vegetation has been well studied, but virtually nothing is known about how fungal communities are shaped by aspect-driven environmental conditions. I carried out DNA metabarcoding of fungi from soil samples taken in a selected study area of Pannonian forests in northern Hungary to compare richness and community composition of taxonomic and functional groups of fungi between slopes of predominantly southerly vs. northerly aspect. The deep sequence data presented here (i.e. 980 766 quality-filtered sequences) indicate that both niche (environmental filtering) and neutral (stochastic) processes shape fungal community composition at landscape level. Fungal community composition correlated strongly with aspect, with many fungi showing preference for either south-facing or north-facing slopes. Several taxonomic and functional groups showed significant differences in richness between north- and south-facing slopes and strong compositional differences were observed in all functional groups. The effect of aspect on fungal communities likely is mediated through contrasting mesoclimatic conditions, that in turn influence edaphic processes as well as vegetation. The compositional differences observed in fungi are largely consistent with the coenologically described forest types, which indicates the usefulness of these habitat types as a framework to better understand environmental differences that influence fungal community composition at landscape level. Finally, the data presented here provide unprecedented insights into the diversity and landscape-level community dynamics of fungi in the Pannonian forests.     

六种温带森林类型凋落物量长期动态及其环境驱动

阐明凋落物动态及其环境控制机制,可以为森林生态系统生产力及碳汇功能的维持提供重要的数据支持和理论依据。以长白山系余脉张广才岭西坡林龄相近但立地条件不同的4种天然次生林（即硬阔叶林、杨桦林、杂木林和蒙古栎林）和2种人工林（落叶松人工林和红松人工林）为研究对象,对其地上凋落物产量及其组分以及相关环境因子进行了14年（2008-2021年）的连续测定,旨在揭示森林凋落物量及其组分的时空变化（林型间和年际变异）及其环境驱动机制。结果表明：6种森林类型的凋落总量（TL）无显著差异,波动范围为500.5-556.1 g m^-2 a^-1;但其叶凋落量（LL）、繁殖组织凋落量（RT）和其他组织凋落量（OT）均存在显著差异,波动范围依次分别为333.9-391.8 g m^-2 a^-1、8.43-69.93 g m^-2 a^-1和93.4-185.9 g m^-2 a^-1。6种森林类型的TL均存在显著的年际变化;其中LL和OT年际变化的显著性因森林类型而不同,而RT的年际变化不显著。除落叶松人工林外,其余5种森林类型的LL与生长季平均气温、日最低气温均值、土壤10 cm深度处的平均温度、最低温度（Ts_min）和土壤5 cm含水量（Ms）均呈显著正相关。杂木林、硬阔叶林和红松人工林的RT与Ms呈显著负相关;杂木林、杨桦林和硬阔叶林的OT与Ts_min呈显著负相关。样地水平的LL与土壤10 cm处含水量存在显著的正相关关系,而RT和OT则与其呈现显著负相关关系。这些结果表明林龄相似的温带森林地上凋落物总量有趋同趋势,但其通过改变组分分配格局来适应立地条件的变化;土壤湿度和温度变化会引起凋落物量的年际变化,但不同森林类型的凋落物量对环境波动的敏感性不同。     

Species richness and distribution of understorey bryophytes in different forest types in Colombian Amazonia

Abstract

The first bryophyte survey results from Colombian Amazonia are presented. Bryophyte species, differentiated into mosses and liverworts, and further into four life-form classes, were sampled in 0.1-ha plots. These plots were distributed over four landscape units in the middle Caquetá area: floodplains, swamps, terra firme forests and white-sand areas. The total numbers of bryophyte species in the units were 50, 45, 45 and 32, respectively. The plots in floodplains and swamps were richer in moss species than the terra firme and white-sand plots, suggesting that coexistence of many moss species is favoured by high humidity. Moss species with fan life-forms preferred floodplains. On the other hand, liverwort species richness was highest in white-sand plots, which suggests that light incidence controls liverwort species-richness, perhaps more than humidity. All plots from the floodplain of the Caquetá River differed remarkably in species composition (of both mosses and liverworts) from the other landscape units. This may be due to the unique properties of this varzea system where, during yearly flooding events, soil, dead logs and stems are covered with a fresh layer of nutrient-rich fine silt, enhancing the surface for colonization and improving the conditions for productive bryophyte growth compared with elsewhere in the middle Caquetá area.     

Diversity of tropical forest landscape types in Hainan Island, China

A novel landscape classification system was proposed in this study based on landscape ecological theory and the differentiation in climate, topography, soil, vegetation and land use mode. Five basic units (zone, tract, province, region and type) and two assistant units (sub and group) were used in this system. The tropical forest landscape in Hainan Island was regarded as a landscape province, belonging to global tropical forest landscape zone, Asiatic (oriental) tropical forest landscape tract and Chinese tropical forest landscape subtract. Based on the grade system of region, sub-region, type-group and type, this landscape province (Hainan Island) is divided into 6 landscape regions (east moist forest landscape, west semi-arid forest landscape, central-south mountainous forest landscape, tropical evergreen needle-leaved forest landscape, tropical bamboo landscape and tropical plantation landscape), 11 tropical forest landscape sub-regions represented by tropical lowland valley rain forest landscape, 26 tropical forest landscape-type groups represented by tropical lowland valley Dipterocarpaceae forest landscape, and 54 forest landscape types (FLT) represented by lowland valley Vatica mangachapoi forest. Generally, this classification system represents the landscape diversity of the tropical forest in Hainan Island. Further studies are needed to better understand the landscape diversity of Hainan Island.     

Diversity of tropical forest landscape types in Hainan Island, China

A novel landscape classification system was proposed in this study based on landscape ecological theory and the differentiation in climate, topography, soil, vegetation and land use mode. Five basic units (zone, tract, province, region and type) and two assistant units (sub and group) were used in this system. The tropical forest landscape in Hainan Island was regarded as a landscape province, belonging to global tropical forest landscape zone, Asiatic (oriental) tropical forest landscape tract and Chinese tropical forest landscape subtract. Based on the grade system of region, sub-region, type-group and type, this landscape province (Hainan Island) is divided into 6 landscape regions (east moist forest landscape, west semi-arid forest landscape, central-south mountainous forest landscape, tropical evergreen needle-leaved forest landscape, tropical bamboo landscape and tropical plantation landscape), 11 tropical forest landscape sub-regions represented by tropical lowland valley rain forest landscape, 26 tropical forest landscape-type groups represented by tropical lowland valley Dipterocarpaceae forest landscape, and 54 forest landscape types (FLT) represented by lowland valley Vatica mangachapoi forest. Generally, this classification system represents the landscape diversity of the tropical forest in Hainan Island. Further studies are needed to better understand the landscape diversity of Hainan Island.     

Plant functional types and climate in a southern African savanna

Abstract. A system of easily defined plant functional types for a dry savanna is presented and related to climate data. To predict possible changes of vegetation in response to climate change, a system of functional attributes is developed. The attributes are based on both structural and functional characteristics and are chosen to be related to climatic conditions. plant functional types are defined as assemblages of plant species having similar combinations of plant functional attributes. Direct and indirect gradient analyses of a test data set show that the plant functional types as well as the vegetation described from them are strongly associated with total annual precipitation, precipitation of the wettest month and a moisture index (all related to soil moisture) and with temperature of the coldest month.     

Distinct roles of savanna and forest tree species in regeneration under fire suppression in a Brazilian savanna

Questions: Has fire suppression relaxed barriers to the exchange of species between savanna and forest? Do all species or a subset of species participate in this exchange? Would current vegetation structure persist if fire suppression were to cease? Location: A gallery forest edge in the Cerrado region of central Brazil that burned only once in the past 35 years. Methods: Density of tree seedlings, saplings and adults, leaf area index (LAI), tree basal area and diameter were surveyed in 12, 10 m × 70 m transects centred on and perpendicular to the forest–savanna boundary. Community composition was assessed using non‐metric multi‐dimensional scaling (NMDS). Results: Basal area and LAI declined substantially from forest to savanna, with an associated shift in species composition. Savanna tree species were nearly absent in the forest, but accounted for the majority of stems in the savanna. In contrast, forest species comprised 14% of adults and more than one‐third of juveniles in the savanna. Despite the high diversity of trees (85 species) in the forest, five species play a particularly large role in this initial phase of forest expansion. Reintroduction of fire, however, would result in widespread topkill of juveniles and the majority of adult forest trees, thereby interrupting the succession towards forest. Conclusions: After 35 years during which the site burned only once, the savanna still remains dominated by savanna species. Nevertheless, the dominance of forest juveniles in border and savanna tree communities suggests that with a continued policy of fire suppression, the forest will continue to expand.     

A multivariate approach to identify vegetation belts: Gallery forest and its surrounding savanna along the river Kota in north Benin

Abstract

The floristic composition and spatial structure of a West African riverine forest in north-western Benin were studied in order to identify vegetation zones and their arrangement and composition. Data were collected on five continuous belt transects and were analysed using multivariate methods. A total of 133 tree taxa from 34 plant families were identified, with Rubiaceae and Leguminosae-Papilionoideae as the most frequent families. Fifteen homogeneous transect sections were documented. Near-river sections were dominated by water-demanding species (Syzygium guineense, Garcinia ovalifolia, Berlinia grandiflora, Breonardia salicina) (=gallery forest), whereas uphill sections were characterised by savanna species (Terminalia laxiflora, Hymenocardia acida, Detarium microcarpum, Burkea africana) (= savanna). The shift from gallery forest to savanna was reflected by a floristic turnover and the different physiognomy of the vegetation, although diversity values for gallery forest and savanna were very similar. In the Principal Co-ordinates Analysis (PCoA), gallery forest and savanna sections did not overlap. Parallel sections at both sides of the river could be linked by isolines following the topography parallel to the river, resulting in a continuous gallery forest belt (width between 10 and 30 m) flanked by savanna. Tree species typical for gallery forests have wider areas of distribution than savanna species.     

Woody plant diversity in relation to environmental factors in a seasonally dry tropical forest landscape

Questions

Water availability is known to be a first‐order driver of plant diversity; yet water also affects fire regimes and soil fertility, which, in turn, affect plant diversity. We examined how precipitation, fire and soil properties jointly determine woody plant diversity. Specifically, we asked how woody plant diversity varies along a sharp precipitation gradient (about 600–1,800 mm mean annual precipitation [MAP ]within a ~45‐km distance) exhibiting considerable variation in long‐term fire burn frequency and soil fertility, in a southern Indian seasonally dry tropical forest (SDTF ) landscape.

Location

Mudumalai, Western Ghats, India.

Methods

Woody plants ≥1‐cm DBH were enumerated in 19 1‐ha permanent plots spanning a range of tropical vegetation types from dry thorn forest, through dry and moist deciduous forest to semi‐evergreen forest. Burn frequencies were derived from annual fire maps. Six measures of surface soil properties – total exchangeable bases (Ca + Mg + K), organic carbon (OC ), total N, pH , plant available P and micronutrients (Fe + Cu + Zn + Mn) were used in the analyses. Five measures of diversity – species richness, Shannon diversity, the rarefied/extrapolated versions of these two measures, and Fisher's α – were modelled as functions of MAP , annual fire burn frequency and the principal components of soil properties.

Results

Most soil nutrients and OC increased with MAP , except in the wettest sites. Woody productivity increased with MAP , while fire frequency was highest at intermediate values of MAP . Woody plant diversity increased with MAP but decreased with increasing fire frequency, resulting in two local diversity maxima along the MAP gradient – in the semi‐evergreen and dry thorn forest – separated by a low‐diversity central region in dry deciduous forest where fire frequency was highest. Soil variables were, on the whole, less strongly correlated with diversity than MAP .

Conclusions

Although woody plant diversity in this landscape, representative of regional SDTF s, is primarily limited by water availability, our study emphasizes the role of fire as a potentially important second‐order driver that acts to reduce diversity in this landscape.

     

Spatial scaling of ecosystem C and N in a subtropical savanna landscape

Widely occurred woody encroachment in grass‐dominated ecosystems has the potential to influence soil organic carbon (SOC) and total nitrogen (TN) pools at local, regional, and global scales. Evaluation of this potential requires assessment of both pool sizes and their spatial patterns. We quantified SOC and TN, their relationships with soil and vegetation attributes, and their spatial scaling along a catena (hill‐slope) gradient in the southern Great Plains, USA where woody cover has increased substantially over the past 100 years. Quadrat variance analysis revealed spatial variation in SOC and TN at two scales. The larger scale variation (40–45 m) was approximately the distance between centers of woody plant communities and their adjoining herbaceous patches. The smaller scale variation (10 m) appeared to reflect the local influence of shrubs on SOC and TN. Litter, root biomass, shrub, and tree basal area (a proxy for plant age) exhibited not only similar spatial scales, but also strong correlations with SOC and TN, suggesting invasive woody plants alter both the storage and spatial scaling of SOC and TN through ecological processes related primarily to root turnover and, to a lesser extent litter production, as mediated by time of occupancy. Forb and grass biomass were not significantly correlated with SOC and TN suggesting that changes in herbaceous vegetation have not been the driving force for the observed changes in SOC and TN. Because SOC and TN varied at two scales, it would be inappropriate to estimate SOC and TN pools at broad scales by extrapolating from point sampling at fine scales. Sampling designs that capture variation at multiple scales are required to estimate SOC and TN pools at broader scales. Knowledge of spatial scaling and correlations will be necessary to design field sampling protocols to quantify the biogeochemical consequences of woody plant encroachment at broad scales.     

Stream order controls geomorphic heterogeneity and plant distribution in a savanna landscape

We posed the question: does viewing a savanna as a network of streams linked to a matrix of terrestrial hillslopes provide a useful framework to research and understand plant distribution in these landscapes? Our study area, the Phugwane River network, lies in the semi‐arid savanna of Kruger National Park, South Africa. We examined changes in hillslope geomorphology from first‐, third‐ and fifth‐order hillslopes with regression equations. The distribution of geomorphic boundaries was enumerated by moving window analysis and the relationship between geomorphology and plant distribution was explored through ordination. First‐order hillslopes had a simple geomorphology, fewer geomorphic boundaries and a relatively homogeneous plant assemblage. By contrast, fifth‐order hillslopes were more complex in geomorphology, with more boundaries and a relatively heterogeneous vegetation pattern. Stream order classification of a savanna drainage network resulted in landscape units distinguishable by geomorphology, geomorphic boundaries and vegetation pattern. Therefore, the drainage network is a useful template to expose and organize the complexity in savanna landscapes into easily managed and researched units. This perspective should inform a shift from single‐scale phytosociological views of homogeneous vegetation units towards multi‐scale conceptualizations of savannas as water dependent ecosystems.     

Soil and forest structure in the Colombian Amazon

Forests structural differences could result of environmental variations at different scales. Because soils are an important component of plant's environment, it is possible that edaphic and structural variables are associated and that, in consequence, spatial autocorrelation occurs. This paper aims to answer two questions: (1) are structural and edaphic variables associated at local scale in a terra firme forest of Colombian Amazonia? and (2) are these variables regionalized at the scale of work? To answer these questions we analyzed the data of a 6ha plot established in a terra firme forest of the Amacayacu National Park. Structural variables included basal area and density of large trees (diameter > or = 10cm) (Gdos and Ndos), basal area and density of understory individuals (diameter < 10cm) (Gsot and Nsot) and number of species of large trees (sp). Edaphic variables included were pH, organic matter, P, Mg, Ca, K, Al, sand, silt and clay. Structural and edaphic variables were reduced through a principal component analysis (PCA); then, the association between edaphic and structural components from PCA was evaluated by multiple regressions. The existence of regionalization of these variables was studied through isotropic variograms, and autocorrelated variables were spatially mapped. PCA found two significant components for structure, corresponding to the structure of large trees (G, Gdos, Ndos and sp) and of small trees (N, Nsot and Gsot), which explained 43.9% and 36.2% of total variance, respectively. Four components were identified for edaphic variables, which globally explained 81.9% of total variance and basically represent drainage and soil fertility. Regression analyses were significant (p < 0.05) and showed that the structure of both large and small trees is associated with greater sand contents and low soil fertility, though they explained a low proportion of total variability (R2 was 4.9% and 16.5% for the structure of large trees and small tress, respectively). Variables with spatial autocorrelation were the structure of small trees, Al, silt, and sand. Among them, Nsot and sand content showed similar patterns of spatial distribution inside the plot.     

Elephant response to spatial heterogeneity in a savanna landscape of northern Tanzania

Landscape heterogeneity, namely the variation of a landscape property across space and time, can influence the distribution of a species and its abundance. Quantifying landscape heterogeneity is important for the management of semi‐natural areas through predicting species response to landscape changes, such as habitat fragmentation. In this paper, we tested whether the change in spatial heterogeneity of the vegetation cover due to farming expansion affected the distribution of the African elephant in the Tarangire‐Manyara ecosystem, northern Tanzania. Spatial heterogeneity (based on the normalized difference vegetation index) was characterized at multiple spatial scales using the wavelet transform and the intensity‐dominant scale method. Elephant distribution was estimated from time‐series aerial surveys using a kernel density function. The intensity, which relates to the contrast in vegetation cover, quantified the maximum variation in NDVI across multiple spatial scales, whereas the dominant scale, which represents the scale at which this maximum variation occurs, identified the dominant inter‐patches distance, i.e. the size of dominant landscape features. We related the dominant scale of spatial heterogeneity to the probability of elephant occurrence in order to identify: 1) the scale that maximizes elephant occurrence, and 2) its change between 1988 and 2001. Neither the dominant scale and intensity of spatial heterogeneity, nor the probability of the elephant occurrence changed significantly between 1988 and 2001. The spatial scale maximizing elephant occurrence remained constant at 7000 to 8000 m during each wet season. Compared to the findings of a recent, similar study in Zimbabwe, our results suggest that the change in the dominant scale was relatively small in Tarangire‐Manyara ecosystem and well within the critical threshold for elephant persistence. The method is a useful tool for monitoring ecosystems and their properties.     

Termite mounds differ in their importance for herbivores across savanna types,seasons and spatial scales

Herbivores do not forage uniformly across landscapes, but select for patches of higher nutrition and lower predation risk. Macrotermes mounds contain higher concentrations of soil nutrients and support grasses of higher nutritional value than the surrounding savanna matrix, attracting mammalian grazers that preferentially forage on termite mound vegetation. However, little is known about the spatial extent of such termite influence on grazing patterns and how it might differ in time and space. We measured grazing intensity in three African savanna types differing in rainfall and foliar nutrients and predicted that the functional importance of mounds for grazing herbivores would increase as the difference in foliar nutrient levels between mound and savanna matrix grasses increases and the mounds become more attractive. We expected this to occur in nutrient‐poor areas and during the dry season when savanna matrix grass nutrient levels are lower. Tuft use and grass N and P content were measured along transects away from termite mounds, enabling calculation of the spatial extent of termite influence on mammalian grazing. Using termite mound densities estimated from airborne light detection and ranging (LiDAR), we further upscaled field‐based results to determine the percentage of the landscape influenced by termite activity. Grasses in close proximity to termite mounds were preferentially grazed at all sites and in both seasons, but the strength of mound influence varied between savanna types and seasons. In the wet season, mounds had a relatively larger effect on grazers at the landscape scale in the nutrient‐poor, wetter savanna, whereas in the dry season the pattern was reversed with more of the landscape influenced at the nutrient‐rich, driest site. Our results reveal that termite mounds enhance the value of savanna landscapes for herbivores, but that their functional importance varies across savanna types and seasons.     

Seasonal variation in leaf traits between congeneric savanna and forest trees in Central Brazil: implications for forest expansion into savanna

The ecology of forest and savanna trees species will largely determine the structure and dynamics of the forest–savanna boundaries, but little is known about the constraints to leaf trait variation imposed by selective forces and evolutionary history during the process of savanna invasion by forest species. We compared seasonal patterns in leaf traits related to leaf structure, carbon assimilation, water, and nutrient relations in 10 congeneric species pairs, each containing one savanna species and one forest species. All individuals were growing in dystrophic oxisols in a fire-protected savanna of Central Brazil. We tested the hypothesis that forest species would be more constrained by seasonal drought and nutrient-poor soils than their savanna congeners. We also hypothesized that habitat, rather than phylogeny, would explain more of the interspecific variance in leaf traits of the studied species. We found that throughout the year forest trees had higher specific leaf area (SLA) but lower integrated water use efficiency than savanna trees. Forest and savanna species maintained similar values of predawn and midday leaf water potential along the year. Lower values were measured in the dry season. However, this was achieved by a stronger regulation of stomatal conductance and of CO₂ assimilation on an area basis (A _area) in forest trees, particularly toward the end of the dry season. Relative to savanna trees, forest trees maintained similar (P, K, Ca, and Mg) or slightly higher (N) leaf nutrient concentrations. For the majority of traits, more variance was explained by phylogeny, than by habitat of origin, with the exception of SLA, leaf N concentration, and A _area, which were apparently subjected to different selective pressures in the savanna and forest environments. In conclusion, water shortage during extended droughts would be more limiting for forest trees than nutrient-poor soils.     

Large-scale environmental correlates of forest tree distributions in Catalonia (NE Spain)

Aim To explore the environmental correlates of tree species distributions in Catalonia according to the chorological status of the species. Location The study area is the region of Catalonia, in north‐eastern Spain. Methods We used presence‐absence data for 24 species, sampled in random plots distributed throughout forests of Catalonia. A climate model for the Catalonia region provided environmental variables. We used classification tree analysis to explore the environmental correlates of the realized niches of tree species. The predictive accuracy of the models was assessed using the ROC curve approach. Potential distribution maps of tree species were generated for the whole Catalonia region. Results Models were ranked from low to high accuracy for the 24 species. Differences in accuracy among species were related to the chorological status of species. Zonal species, or species at the core of their range (Mediterranean and Sub‐Mediterranean species), were generally well predicted, while extrazonal species, or species at the edge of their range, were predicted only moderately well. Mediterranean species distributions showed good correlations with extreme temperatures and annual precipitation. Main conclusions The above trends confirmed the difficulty of identifying the realized niche of species at the edges of their ranges. In contrast, Mediterranean and Sub‐Mediterranean species, which were at the core of their range, were well‐predicted, confirming the importance of extremes of temperature and annual precipitation as effective surrogates for variables having more direct physiological roles in limiting the ability of plants to survive and grow. Maps of potential tree distributions allowed us to define suitable habitats and to highlight areas where species have been planted outside their natural distribution.     

Influence of environmental and spatial variables on regional distribution of forest plant species in a fragmented and changing landscape

During the past several centuries, forests in Europe and large parts of North America have been subject to extensive forest clearance. The last several decades, however, numerous new forest patches have been established onto former agricultural land. As a result, the present forest area often consists of a mixture of small forest patches of different age, area, habitat quality and connectivity embedded within a hostile agricultural landscape. In these patchy landscapes, distribution patterns of plant species may be affected by both regional and local factors, although the relative importance of both is still poorly understood. In this study, we investigated distribution patterns of 113 forest plant species in a fragmented landscape. Species abundances at the regional scale conformed to a clearly unimodal abundance distribution which we believe to be related to 1) environmental heterogeneity due to succession and 2) inequality in migration rates. Patch incidence was significantly related to life form, which in turn was correlated to seed mass and dispersal mechanism. Multiple logistic regressions showed that presence/absence of 59 species studied was significantly affected by patch connectivity, patch area and age for 35, 30 and 34 species, respectively. The results of this study indicate that distribution patterns of forest plant species are influenced by both local and regional factors. Moreover, they also demonstrate that next to spatial aspects of fragmentation, temporal patterns of landscape change may have far-reaching effects on presence/absence patterns of plant species and therefore should be incorporated in studies dealing with regional population structures of plants.     

Tree distribution on a steep environmental gradient in an arid savanna

Aim The structure of savannas ranges from scrub to woodland over broad geographical gradients. Here we examine the hypothesis that water availability is a major determinant of these structural differences by ascertaining the relationship between water availability and tree growth across a steep moisture gradient. Location The study site is a sub‐tropical savanna, with a mean annual precipitation of 400 mm, located in the Phugwane river basin, Kruger National Park, Limpopo Province, South Africa. Methods We determined plant moisture stress using xylem pressure potentials, stem growth using dendrometer bands, and estimated the water sources available to plants using stable hydrogen and oxygen isotope ratios. The primary objective was to understand tree growth relative to available water along an environmental gradient. Results We found that Philenoptera violacea trees growing close to the stream have lower water stress (least negative xylem pressure potentials) and higher cumulative growth than those growing away from the stream. The stem growth of P. violacea was characterized by steady incremental growth and could not be related to antecedent rainfall. Colophospermum mopane trees experienced higher water stress, yet trees growing adjacent to the stream achieved the highest cumulative growth rates over the study period. The growth of C. mopane could be clearly linked to antecedent rainfall, and most growth was achieved during short growth pulses that followed rainfall events. δ¹⁸O values become progressively more enriched in the heavier isotope with distance from the stream, suggesting that access to groundwater decreases with distance from the stream side. The isotopic data suggest that P. violacea has access to groundwater, but that C. mopane does not. Main conclusions Our results show that water stress increases with distance from the stream side as a result of reduced access to groundwater. Trees without access to deep water adopt an opportunistic growth strategy.