Search

An unresolved issue in the vegetation ecology of the Indian subcontinent is whether its savannas, characterized by relatively open formations of deciduous trees in C4-grass dominated understories, are natural or anthropogenic. Historically, these ecosystems have widely been regarded as anthropogenic-derived, degraded descendants of deciduous forests. Despite recent work showing that modern savannas in the subcontinent fall within established bioclimatic envelopes of extant savannas elsewhere, the debate persists, at least in part because the regions where savannas occur also have a long history of human presence and habitat modification. Here we show for the first time, using multiple proxies for vegetation, climate and disturbances from high-resolution, well-dated lake sediments from Lonar Crater in peninsular India, that neither anthropogenic impact nor fire regime shifts, but monsoon weakening during the past ~ 6.0 kyr cal. BP, drove the expansion of savanna at the expense of forests in peninsular India. Our results provide unambiguous evidence for a climate-induced origin and spread of the modern savannas of peninsular India at around the mid-Holocene. We further propose that this savannization preceded and drove the introduction of agriculture and development of sedentism in this region, rather than vice-versa as has often been assumed.

Many arid and semi‐arid rangelands exhibit distinct spatial patterning of vegetated and bare soil‐dominated patches. The latter potentially represent a grazing‐induced, degraded ecosystem state, but could also arise via mechanisms related to feedbacks between vegetation cover and soil moisture availability that are unrelated to grazing. The degree to which grazing contributes to the formation or maintenance of degraded patches has been widely discussed and modeled, but empirical studies of the role of grazing in their formation, persistence, and reversibility are limited.We report on a long‐term (17 years) grazing removal experiment in a semi‐arid savanna where vegetated patches composed of perennial grasses were interspersed within large (>10 m2) patches of bare soil.Short‐term (3 years) grazing removal did not allow bare patches to become revegetated, whereas following long‐term (17 years) grazing removal, bare soil patches were revegetated by a combination of stoloniferous grasses and tufted bunchgrasses. In the presence of grazers, stoloniferous grasses partially recolonized bare patches, but this did not lead to full recovery or to the establishment of tufted bunchgrasses.These results show that grazers alter both the balance between bare and vegetated patches, as well as the types of grasses dominating both patch types in this semiarid savanna.Synthesis: Large herbivores fundamentally shaped the composition and spatial pattern of the herbaceous layer by maintaining a two‐phase herbaceous mosaic. However, bare patches within this mosaic can recover given herbivore removal over sufficiently long time scales, and hence do not represent a permanently degraded ecosystem state., The stability and spatial pattern of degraded and vegetated patches in dryland ecosystems has been modeled and discussed for decades, with few empirical studies. We present one of the first long‐term experiments that demonstrate these patterns are maintained by large herbivore grazing and can disappear when herbivores are removed.

Bark damage resulting from elephant feeding is common in African savanna trees with subsequent interactions with fire, insects, and other pathogens often resulting in tree mortality. Yet, surprisingly little is known about how savanna trees respond to bark damage. We addressed this by investigating how the inner bark of marula (Sclerocarya birrea), a widespread tree species favoured by elephants, recovers after bark damage. We used a long-term fire experiment in the Kruger National Park to measure bark recovery with and without fire. At 24 months post-damage, mean wound closure was 98, 92, and 72%, respectively, in annual and biennial burns and fire-exclusion treatments. Fire exclusion resulted in higher rates of ant colonization of bark wounds, and such ant colonization resulted in significantly lower bark recovery. We also investigated how ten common savanna tree species respond to bark damage and tested for relationships between bark damage, bark recovery, and bark traits while accounting for phylogeny. We found phylogenetic signal in bark dry matter content, bark N and bark P, but not in bark thickness. Bark recovery and damage was highest in species which had thick moist inner bark and low wood densities (Anacardiaceae), intermediate in species which had moderate inner bark thickness and wood densities (Fabaceae) and lowest in species which had thin inner bark and high wood densities (Combretaceae). Elephants prefer species with thick, moist inner bark, traits that also appear to result in faster recovery rates.

1. Forest–grassland mosaics, with abrupt boundaries between the two vegetation types, occur across the globe. Fire and herbivory are widely considered primary drivers that maintain these mosaics by limiting tree establishment in grasslands, while edaphic factors and frosts are generally considered to be secondary factors that reinforce these effects. However, the relative importance of these drivers likely varies across systems. In particular, although frost is known to occur in many montane tropical mosaics, experimental evidence for its role as a driving factor is limited. 2. We used replicated in situ transplant and warming experiments to examine the role of microclimate (frost and freezing temperatures) and soil in influencing germination and seedling survival of both native forest trees and alien invasive Acacia trees in grasslands of a tropical montane forest–grassland mosaic in the Western Ghats of southern India. 3. Seed germination of both native and alien tree species was higher in grasslands regardless of soil type, indicating that germination was not the limiting stage to tree establishment. However, irrespective of soil type, native seedlings in grasslands incurred high mortality following winter frosts and freezing temperatures relative to native seedlings in adjoining forests where freezing temperatures did not occur. Seedling survival through the tropical winter was thus a primary limitation to native tree establishment in grasslands. In contrast, alien Acacia seedlings in grasslands incurred much lower levels of winter mortality. Experimental night‐time warming in grasslands significantly enhanced over‐winter survival of all tree seedlings, but increases were much greater for alien Acacia than for native tree seedlings. 4. Synthesis. Our results provide evidence for a primary role for frost and freezing temperatures in limiting tree establishment in grasslands of this tropical montane forest–grassland mosaic. Future increases in temperature are likely to release trees from this limitation and favour tree expansion into grasslands, with rates of expansion of non‐native Acacia likely to be much greater than that of native trees. We suggest that studies of frost limitation to plant establishment are needed across a range of tropical ecosystems to re‐evaluate the general importance of frost as a driver of vegetation transitions in the tropics.

Recent work suggests that savanna woody plant species utilise two different strategies based on their defences against herbivory; a low nutrient/high chemical defence strategy and a nutrition paired with mostly architectural defences strategy. The concept that chemical and structural defences can augment each other and do not necessarily trade-off has emanated from this work. In this study, we examine woody plant defence strategies, how these respond to herbivore removal and how they affect plant growth in an East African savanna. At three paired long-term exclosure sites with high browser and mixed-feeder densities at Mpala Ranch, Kenya, we investigated: (a) whether defences employed by the dominant fine- and broad-leaved woody savanna species form defence strategies and if these align with previously proposed strategies, (b) how nine key plant defence traits respond to herbivore removal and (c) how effective the different defence strategies are at protecting against intense herbivory (by measuring plant growth with and without herbivores present). We identified three defence strategies. We found a group (a) with high N, short spines and high N-free secondary metabolites, a group (b) with high N, long spines and low N-free secondary metabolites and a group (c) with moderate N, no spines and low N-free secondary metabolites (most likely defended by unmeasured chemical defences). Structural defences (spine length, branching) were generally found to be induced by herbivory, leaf available N increased or did not respond, and N-free secondary metabolites decreased or did not respond to herbivory. Species with long spines combined with increased “caginess” (dense canopy architecture arising from complex arrangement of numerous woody and spiny axis categories) of branches, maintained the highest growth under intense browsing, compared to species with short spines and high N-free secondary metabolites and species with no spines and low N-free secondary metabolites. Synthesis. At our study site, structural traits (i.e. spines, increased caginess) were the most inducible and effective defences against intense mammalian herbivory. We propose that high levels of variability in the way that nutrient and defence traits combine may contribute to the coexistence of closely related species comprising savanna woody communities.

The paper studies the performance of a free space optical communication link, which consists of a Gaussian Laser beam employing direct-current-biased optical frequency division multiplexing (DCO-OFDM) modulation, to carry information. While optical multi-carrier modulation is capable of alleviating multi-path fading, higher transmit power by virtue of dc-bias, helps the optical beam in partially off-setting signal intensity loss due to beam divergence in free space channel. In order to reduce the overall transmitted power apart from avoiding nonlinearities due to high peak-to-average-power-ratio (PAPR), we utilize a PAPR-dependent dc-bias, to generate a DCO OFDM light beam from a pre-clipped electrical baseband OFDM signal. The radiated optical beam encounters turbulence-induced impairments like beam-spreading, beam-wander, etc., which affect the signal-to-noise ratio (SNR) of the received data symbols. Further signal pre-clipping and optical source linearity generate clipping noise components resulting in compromised receiver performance or reduction of span length, to achieve target bit rate error (BER). In this paper, we derive an analytical model for evaluating the performance of the free-space optical (FSO) link, by modeling various impairments as noise variances and characterizing atmospheric turbulence by refractive index structure parameter under weak- and strong-turbulent conditions. Numerical results are obtained by varying system design parameters in the model, which graphically provide useful insight on the power penalties and various trade-offs involved in operating the link on a longer span, at a higher data rate, and with reduced transmitted power through pre-clipping while ensuring desired signal BER.

International audience; A lack of spatial congruence between carbon storage and biodiversity in intact forests suggests limited cobenefits of carbon-focused policies for conserving tropical biodiversity. However, whether the same applies in tropical human-dominated landscapes (HDLs) is unclear. In India's Western Ghats Biodiversity Hotspot, we found that while HDL forests harbor lower tree diversity and aboveground carbon stocks than relatively intact forests, positive diversity-carbon correlations are more prevalent in HDLs. This is because anthropogenic drivers of species loss in HDLs consistently reduce carbon storing biomass volume (lower basal area), and biomass per unit volume (fewer hard-wood trees). We further show, using a meta-analysis spanning multiple regions, that these patterns apply to tropical HDLs more generally. Thus, while complementary strategies are needed for securing the irreplaceable biodiversity and carbon values of intact forests, ubiquitous tropical HDLs might hold greater potential for synergizing biodiversity conservation and climate change mitigation. K E Y W O R D S basal area, biodiversity conservation, carbon storage, climate change, forest degradation, meta-analysis, tree density, tropical forests, Western Ghats, wood density This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

We investigated flowering phenology, pollinator visitation and visitor community composition in communities of self-incompatible sympatric Primula species in a high-elevation Himalayan ecosystem. Within the tight constraints imposed by short growing seasons in such ecosystems, interactions among co-occurring plants for pollinators may vary from competition to facilitation, depending on the specifics of the system. We found that pollinator community composition changed with elevation in this system: lepidopterans were the dominant visitors at lower elevations (2200–3000 masl), bees (other than bumblebees) dominated at mid-elevations (3000–3800 masl) and bumblebees dominated at higher elevations (3800–4600 masl). However, within an elevation zone, there were no significant differences in pollinators amongst co-occurring Primula species. At a focal study site where multiple Primula species co-occurred, our results showed that even while the overall flowering periods of these species broadly overlapped, the peak flowering periods of different Primula species were temporally segregated. Upon further inferring the nature of interaction amongst co-flowering Primula species, we found that plots with higher Primula diversity (≥ 2 species) and density (80–100 individuals) experienced significantly higher pollinator visitation, compared with plots with single species and low flower densities (40–50 individuals). Our results suggest that in this community of sympatric, self-incompatible Primula species, a broadly aggregated, synchronous floral display of multiple species results in pollinator facilitation by attracting a greater number of pollinator visitors. Within this broadly synchronous display, the temporal segregation of peak flowering period of individual species may reduce competition for pollinators and limit heterospecific pollen transfer.

Colonial encounters with tropical ecosystems were primarily driven by profit-oriented management practices; witness the extensive network of timber and forestry practices that were set up across colonial India. In contrast, the colonial engagement with the montane forest-grassland mosaics of the higher reaches of the Western Ghats in southern India was marked by intensive investment in vegetation management by colonial foresters that yielded no profits. In this archival study, we trace the history of extensive vegetation transformation in this landscape from the early nineteenth to the early twentieth century. We show how the misperception that the grasslands within this mosaic must have resulted from tree felling, fire-setting and buffalo grazing by indigenous communities led colonial foresters into a century-long effort at ‘foresting’ the grasslands, primarily through large-scale planting of exotic tree species. These efforts persisted despite economic losses and ecological evidence that native tree seedlings planted in the grasslands repeatedly failed to establish. These policies continued unabated into the late twentieth century in newly independent India. Today, the once picturesque landscapes of these ancient forest-grassland mosaics are diminished by large-scale plantations of exotic species. Some of these species have become invasive and pose significant threats to the remnant natural grasslands. While this historical narrative is set in the forest-grassland mosaics of southern India, it finds striking parallels in the current day, with grasslands and savannas globally threatened by the misperception that they are ‘degraded ecosystems’ that can be ‘forested’ or converted to other ‘productive’ land uses. We suggest that this case history portends the potential fates of many of earth's threatened tropical grasslands and savannas.

The success of invasive species eradication depends on a variety of factors, including those that initially facilitated the invasion, as well as removal and post-removal protocols. Two factors that appear to influence invasion by, and eradication of, the Neotropical shrub Lantana camara (L.), in southern Indian deciduous forests, are rainfall and removal method. However, their role in influencing eradication success is yet to be quantified, and remains unclear. We conducted an experiment to clarify how rainfall (high vs. low) and removal method (cutting vs. uprooting Lantana) influence re-invasion by Lantana, and native plant recovery. Rainfall influenced both eradication effort and outcomes—drier forest had lower starting levels of invader biomass, requiring less initial eradication effort, as well as lower subsequent Lantana re-invasion (from seed and rootstock) whereas wetter forest typically had greater starting levels of invader biomass, requiring considerably greater initial eradication effort, and greater Lantana re-invasion. However, wetter forest also showed greater native tree and forb recovery. Therefore, the availability of funds, local environmental gradients, and restoration priorities should inform the selection of restoration sites. With regard to removal method, uprooting combined with weeding of germinating Lantana, particularly after the rainy season, minimized overall re-invasion. Therefore, uprooting, followed by regular weeding of germinating Lantana and secondary invaders, is crucial to long-term Lantana eradication success.

Tropical semi-arid grasslands are a widespread and ecologically and economically important terrestrial biome. Here, we use paleoecology to understand woodland–grassland transitions across the mid- to late-Holocene period in the Banni grassland, western India. Multi proxy analyses involving palynology, phytoliths and elemental geochemistry were carried out on two sediment cores retrieved from wetlands (Chachi and Luna), to understand temporal fluctuations in vegetation, moisture availability and other environmental parameters. Based on the results, the Chachi core was divided into two major climatic phases. Phase 1 (4600–2500 cal. yr BP) was characterised by high precipitation and abundance of pollen types and phytolith morphotypes that indicate the presence of woody savanna, and mesic herbaceous taxa. Phase 2 (2500 cal. yr BP to the present) was characterised by lower precipitation, lower abundance of mesic taxa and an increase in grass phytolith abundance. However, the period from ~1000 cal. yr BP to the present was characterised by the increased abundance of leguminous taxa, dryland herbs/shrubs and a decline in grass phytolith abundance. The Luna core (~1000 cal. yr BP to the present) also showed results matching with the Chachi core for this latter period. Overall, moisture availability in the ecosystem appears to have declined since 4600 cal. yr BP, and the vegetation has responded to this. Although the balance between tree, shrub and grass elements has fluctuated, overall, the region has remained as an open ‘grass and shrub savanna’ with sparse woody vegetation throughout this period. Our study provides insights into the vegetation dynamics and environmental settings in a poorly understood tropical arid-grassland ecosystem from Asia during the mid-late-Holocene.

Tropical grasslands and savannas are globally extensive, and are of significant environmental, economic, and ecological importance. These ecosystems are anticipated to be particularly sensitive to future changes in climate, and understanding how these systems have responded to climatic changes in the past can provide us with insights into their potential responses to future global change. In this study, the temporal dynamics of C3-C4 vegetation changes in response to changes in moisture availability, local fire events and changing levels of herbivory in a summer-rainfall region of Western India are reconstructed for the past ~ 4600 cal yr BP. Paleodata such as stable carbon isotope of bulk organic matter (δ13Corg), oxygen isotope from carbonate shells (δ18Oshell), macro-charcoal and herbivore dung fungal spores are reported from the retrieved cores of two wetland sites located in the Banni grasslands of Western India. Results show that vegetation in the Banni was composed mostly of C3 vegetation from ~ 4600 to ~ 2500 cal yr BP, after which there was a decline in C3 vegetation. From the late-Holocene to the present, there was a mix of both C3 and C4 vegetation, with C4 grasses being more abundant in the ecosystem. These shifts were coincident with rainfall changes from more mesic conditions during ~ 4600 to ~ 2500 cal yr BP to more arid conditions towards the present as indicated by δ18Oshell isotope data. The period of increase in C4 vegetation also coincides with a period of increased biotic disturbances in the ecosystem, particularly fire. Given the current scenarios of global warming, recurrent drought events and increased anthropogenic use of similar ecosystems, such studies can provide us insights into potential future trajectories of these ecosystems.

Although studies have shown that mammalian herbivores often limit aboveground carbon storage in savannas, their effects on belowground soil carbon storage remain unclear. Using three sets of long‐term, large herbivore exclosures with paired controls, we asked how almost two decades of herbivore removal from a semiarid savanna in Laikipia, Kenya affected aboveground (woody and grass) and belowground soil carbon sequestration, and determined the major source (C3 vs. C4) of belowground carbon sequestered in soils with and without herbivores present. Large herbivore exclusion, which included a diverse community of grazers, browsers, and mixed‐feeding ungulates, resulted in significant increases in grass cover (~22%), woody basal area (~8 m2/ha), and woody canopy cover (31%), translating to a ~8.5 t/ha increase in aboveground carbon over two decades. Herbivore exclusion also led to a 54% increase (20.5 t/ha) in total soil carbon to 30‐cm depth, with ~71% of this derived from C4 grasses (vs. ~76% with herbivores present) despite substantial increases in woody cover. We attribute this continued high contribution of C4 grasses to soil C sequestration to the reduced offtake of grass biomass with herbivore exclusion together with the facilitative influence of open sparse woody canopies (e.g., Acacia spp.) on grass cover and productivity in this semiarid system.

Bastin et al.’s estimate (Reports, 5 July 2019, p. 76) that tree planting for climate change mitigation could sequester 205 gigatonnes of carbon is approximately five times too large. Their analysis inflated soil organic carbon gains, failed to safeguard against warming from trees at high latitudes and elevations, and considered afforestation of savannas, grasslands, and shrublands to be restoration., Funding: Supported by the Texas A&M Sid Kyle Global Savanna Research Initiative (T.W.B.); Swiss National Science Foundation (20FI20_173691) (N.B.); Centre National pour la Recherche Scientifique CNRS PICS 2018-2020 (RESIGRASS) (E.B.); CNPq (Brazil, 303179/2016-3) (G.D.); CNPq (Brazil) (G.W.F.); CNPq (Brazil, 303988/2018-5) (A.F.); NASA award NNX17AK14G (F.F.); NSF award 1354943 (W.A.H.); Fundação de Amparo à Pesquisa do Estado de Minas Gerais (Brazil, 2016/13232-5) (S.L.S.); the Office of the Royal Society (IC170015) (C.E.R.L.); CNPq (Brazil, 310345/2018-9) (G.E.O.); the Spanish Government (FIROTIC, PGC2018-096569-B-I00) (J.G.P.); the National Research Foundation (ACCESS, 114695) (N.S.); CNPq (Brazil, 303568/2017-8) (F.A.O.S.); NSF awards 1342703 and 1926431 (C.J.S. and D.M.G.); NSF award EAR-1253713 (C.A.E.S.); Deutsche Forschungsgemeinschaft grant 5579 POEM (V.M.T.); and USDA-NIFA Sustainable Agricultural Systems Grant 12726253 (J.W.V.)

Two dominant biomes that occur across the southern Indian peninsula are dry deciduous “forests” and evergreen forests, with the former occurring in drier regions and the latter in wetter regions, sometimes in close proximity to each other. Here we compare stem and leaf traits of trees from multiple sites across these biomes to show that dry deciduous “forest” species have, on average, lower height: diameter ratios, lower specific leaf areas, higher wood densities and higher relative bark thickness, than evergreen forest species. These traits are diagnostic of these dry deciduous “forests” as open, well-lit, drought-, and fire-prone habitats where trees are conservative in their growth strategies and invest heavily in protective bark tissue. These tree traits together with the occurrence of a C4 grass-dominated understory, diverse mammalian grazers, and frequent fires indicate that large tracts of dry deciduous “forests” of southern India are more accurately classified as mesic deciduous “savannas.”

The majority of Earth’s tropical forests have been selectively logged; some on repeated occasions. Selective logging is known to affect forest structure, composition and function in various ways, but how such effects vary with logging frequency and across forest types remains unclear. In the Andaman Archipelago in India, we examined adult and pole-sized trees in baseline (unlogged since 1990s), once-logged (logged between 2007 and 2014) and twice-logged (logged in early 1990s and between 2007 and 2014) evergreen and deciduous forests, and tested whether higher logging frequency was associated with lower canopy cover, tree density, tree diversity, aboveground carbon stocks and divergence in species composition, including increased relative abundances of deciduous species in these forests. While once-logged evergreen and deciduous forests were similar to their respective baselines for most attributes assessed, and twice-logged evergreen forests had 22–24% lower adult tree density and diversity, twice-logged deciduous forests had 17–50% lower canopy cover, pole density, adult species diversity and above-ground carbon stocks, and 12% higher deciduous fractions compared to the baselines. Collectively our results reveal lasting impacts of repeated selective logging, even at relative low intensity, on tree communities and carbon storage. These impacts can potentially be mitigated by reducing logging frequency and retaining unlogged patches in logged landscapes, but management must also incorporate heterogeneity in responses and recovery across different forest types. In the Andaman Islands, all forests may require more than 15–25 years between logging events, with deciduous forests potentially requiring more stringent extraction limits compared to evergreen forests.

Bastin et al . (Reports, 12 May 2017, p. 635) infer forest as more globally extensive than previously estimated using tree cover data. However, their forest definition does not reflect ecosystem function or biotic composition. These structural and climatic definitions inflate forest estimates across the tropics and undermine conservation goals, leading to inappropriate management policies and practices in tropical grassy ecosystems.

The savannahs of Asia remain locally unrecognized as distinctive ecosystems, and continue to be viewed as degraded forests or seasonally dry tropical forests. These colonial-era legacies are problematic, because they fail to recognize the unique diversity of Asian savannahs and the critical roles of fire and herbivory in maintaining ecosystem health and diversity. In this review, we show that: the palaeo-historical evidence suggests that the savannahs of Asia have existed for at least 1 million years, long before widespread landscape modification by humans; savannah regions across Asia have levels of C4grass endemism and diversity that are consistent with area-based expectations for non-Asian savannahs; there are at least three distinct Asian savannah communities, namely deciduous broadleaf savannahs, deciduous fine-leafed and spiny savannahs and evergreen pine savannahs, with distinct functional ecologies consistent with fire- and herbivory-driven community assembly. Via an analysis of savannah climate domains on other continents, we map the potential extent of savannahs across Asia. We find that the climates of African savannahs provide the closest analogues for those of Asian deciduous savannahs, but that Asian pine savannahs occur in climates different to any of the savannahs in the southern continents. Finally, we review major threats to the persistence of savannahs in Asia, including the mismanagement of fire and herbivory, alien woody encroachment, afforestation policies and future climate uncertainty associated with the changing Asian monsoon. Research agendas that target these issues are urgently needed to manage and conserve these ecosystems.This article is part of the themed issue ‘Tropical grassy biomes: linking ecology, human use and conservation’.

Defaunation is causing declines of large-seeded animal-dispersed trees in tropical forests worldwide, but whether and how these declines will affect carbon storage across this biome is unclear. Here we show, using a pan-tropical data set, that simulated declines of large-seeded animal-dispersed trees have contrasting effects on aboveground carbon stocks across Earth's tropical forests. In our simulations, African, American and South Asian forests, which have high proportions of animal-dispersed species, consistently show carbon losses (2–12%), but Southeast Asian and Australian forests, where there are more abiotically dispersed species, show little to no carbon losses or marginal gains (±1%). These patterns result primarily from changes in wood volume, and are underlain by consistent relationships in our empirical data (∼2,100 species), wherein, large-seeded animal-dispersed species are larger as adults than small-seeded animal-dispersed species, but are smaller than abiotically dispersed species. Thus, floristic differences and distinct dispersal mode–seed size–adult size combinations can drive contrasting regional responses to defaunation., Defaunation is linked to the decline of tree species that depend on large animals for seed dispersal, but it is unclear if this affects carbon storage. Here the authors show that defaunation effects on carbon storage vary across continents, driven by relationships between seed dispersal strategies and adult tree size.

The majority of research on savanna vegetation dynamics has focused on the coexistence of woody and herbaceous vegetation. Interactions among woody plants in savannas are relatively poorly understood. We present data from a 10-year longitudinal study of spatially explicit growth patterns of woody vegetation in an East African savanna following exclusion of large herbivores and in the absence of fire. We examined plant spatial patterns and quantified the degree of competition among woody individuals. Woody plants in this semi-arid savanna exhibit strongly clumped spatial distributions at scales of 1 – 5 m. However, analysis of woody plant growth rates relative to their conspecific and heterospecific neighbors revealed evidence for strong competitive interactions at neighborhood scales of up to 5 m for most woody plant species. Thus, woody plants were aggregated in clumps despite significantly decreased growth rates in close proximity to neighbors, indicating that the spatial distribution of woody plants in this region depends on dispersal and establishment processes rather than on competitive, density-dependent mortality. However, our documentation of suppressive effects of woody plants on neighbors also suggests a potentially important role for tree-tree competition in controlling vegetation structure and indicates that the balanced-competition hypothesis may contribute to well-known patterns in maximum tree cover across rainfall gradients in Africa. This article is protected by copyright. All rights reserved.

Savannas are globally important ecosystems of great significance to human economies. In these biomes, which are characterized by the co-dominance of trees and grasses, woody cover is a chief determinant of ecosystem properties. The availability of resources (water, nutrients) and disturbance regimes (fire, herbivory) are thought to be important in regulating woody cover, but perceptions differ on which of these are the primary drivers of savanna structure. Here we show, using data from 854 sites across Africa, that maximum woody cover in savannas receiving a mean annual precipitation (MAP) of less than approximately 650 mm is constrained by, and increases linearly with, MAP. These arid and semi-arid savannas may be considered 'stable' systems in which water constrains woody cover and permits grasses to coexist, while fire, herbivory and soil properties interact to reduce woody cover below the MAP-controlled upper bound. Above a MAP of approximately 650 mm, savannas are 'unstable' systems in which MAP is sufficient for woody canopy closure, and disturbances (fire, herbivory) are required for the coexistence of trees and grass. These results provide insights into the nature of African savannas and suggest that future changes in precipitation may considerably affect their distribution and dynamics.