The Ecophysiological Role of Trees in Dryland Agroecosystems: Implications for Natural Resource Conservation and Sustainable Food Production in Sub-Saharan Africa

Abstract

Agroforestry involves maintaining trees alongside crops and is widely recognised to provide multiple benefits, including improving food security, production efficiency, and soil quality and mitigating climate change. However, in Sub-Saharan Africa, a predominantly dry landscape, various pressures are leading to the removal of trees from farmlands. Evidence from natural dryland systems shows that trees play a central role in regulating the key ecological processes of nutrient and water redistribution, an aspect also invoked in dryland agroecology. In this paper, we synthesise the ecophysiological functioning of trees, focusing on two key processes: water and nutrient redistribution. Additionally, we synthesise the influence of these functions on soil biotic interactions, detailing their ecological significance. Based on available evidence from both natural and agroecosystems, we review the role of tree ecophysiology in sustainable food production in dryland agroecosystems of Sub-Saharan Africa. We provide caveats related to prevalent interpretations and the current understanding of plant resource use in agroecology. Trees in agroforestry systems of Sub-Saharan Africa play a potentially critical role in the ecological intensification of food production. However, there is a lack of data on the roles of tree functions in enhancing crop yields and conserving resources in this region. Although evidence from natural drylands and indirect evidence from dryland agroforests indicate that tree ecophysiological functions may be crucial for ecological intensification of food production in Sub-Saharan Africa, many claims related to agroecosystems are overstated, underscoring the urgent need for focused research. Importantly, large trees on farms need to be conserved. To effectively exploit ecosystem services provided by trees, a key feature of ecological intensification, research tailored to local farm conditions is needed, with a focus on maintaining soil quality, securing long-term productivity, and conserving resources. Balancing agricultural intensification with ecological sustainability remains a challenge, yet it is vital for addressing food security, land degradation, and climate change.

1. Introduction

Agroforestry, a land management system that maintains trees alongside understory crops and pastures on farms and rangelands, is widely acknowledged as a sustainable land-use practice with multiple benefits, including improving food security [1]; resource-use efficiency in food production [2,3]; and soil quality [4,5]. Additionally, trees on farms are proposed to buffer the effects of climate change via improved carbon sequestration, lowering the release of non-CO₂ greenhouse gases from agriculture and restoring the production potential of degraded lands [6]. Studies have shown that trees play a central role in regulating key biogeochemical processes [7,8,9]. This role is also frequently invoked in dryland agroecology. In this paper, we synthesise the role of tree ecophysiology in drylands and review this role in dryland agroecosystems. This role is often cited as having potential for sustainable food production in Sub-Saharan Africa, a heterogeneous landscape with diverse land uses.

A defining feature of drylands is water scarcity. The global extent of drylands accounts for 41% of the Earth’s land area, and more than 2 billion people inhabit these lands [10]. Climate change-related uncertainty in the timing and magnitude of rainfall is likely to negatively impact crop yields. Rapid land-use change will also affect agricultural production and food security [11,12,13]. Sub-Saharan Africa, a mosaic of diverse and contrasting landscapes that supports a high diversity of animal and plant species, is important for global agricultural production; however, producing food in this region sustainably is a challenge due to poor soils and several other socio-economic factors [14,15]. Agricultural production in this region is almost entirely rain-fed [16], with up to 93% of water needs dependent on rainfall [17]. High soil erosion and surface water runoff, which are already extensive in this region, are predicted to exacerbate nutrient and water losses from the system [18,19]. Water demand is also expected to increase, particularly with the need to expand irrigated agricultural production [20]. Currently, the proportion of fertiliser consumed by African farmers is less than 3% of global use, which significantly affects crop yields in nutrient-poor soils. This necessitates finding affordable solutions to increase food production sustainably [15,21]. Consequently, efficient water and nutrient management becomes imperative for improving crop yields, sustaining water needs, and improving soil quality. A significant portion of the discussion in agroforestry science centres on the biophysical role of trees, including their ability to prevent soil erosion and surface water runoff. However, studies reveal that the impact of tree roots on water and nutrient cycling can play an ecosystem engineering role, influencing biotic interactions and soil formation [9]. We present this paper in two parts. First, we present a synthesis of tree ecophysiology, focusing on nutrient and water cycling and the impact on overall nutrient transfers in dryland ecosystems. Second, we review the role of tree ecophysiology in dryland agroforestry systems of Sub-Saharan Africa. We critically analyse the available evidence and examine the current understanding and prevailing discussions related to agroforestry systems regarding resource use, conservation, and sustainable food production, providing caveats as needed.

2. The Role of Trees in Water and Nutrient Cycling in Drylands

Trees are an important component of dryland ecosystems, playing a key role in the ecological functions of these systems due to their size above- and belowground and influencing the cycling of water and nutrients through the physiological processes of transpiration and deep-soil water and nutrient uptake [9,22]. Trees in dryland ecosystems have some of the deepest roots [23], which are central to essential ecosystem services such as pedogenesis, the infiltration of rainwater deeper into the soil, and carbon sequestration [9,24,25]. In this section, we provide a comprehensive synthesis of the ecophysiology of trees and how these functions are central to nutrient and water cycling within the tree–grass–soil system. Furthermore, we illustrate how these functions influence complex biotic interactions.

2.1. Water Redistribution by Trees

Water redistribution within the tree–soil system is a passive process, termed hydraulic redistribution (HR), where water moves passively from wetter ends to drier ends of the soil–plant continuum [26,27]. Figure 1 is a simplified schematic illustrating this process. The upward, downward, and lateral flows of water via tree-mediated HR occur when there is a substantial difference in water potential within the tree or between the tree roots and the soil interface [7,27,28]. Up to 60 species of trees globally have been documented to show HR [7], and its prevalence is more common than previously thought [29,30,31,32].

The results of experiments using stable isotope tracers demonstrate dry-season redistribution of water by dryland savanna trees to understory grasses via tree HR [32]. Furthermore, tree HR is critical for mycorrhizal survival and function [33]. Trees in a seasonally dry ∼30-year-old Douglas-fir (Pseudotsuga menziesii) forest mobilised tightly bound immobile water in the soil, making it available to other plants and soil organisms [34]. A similar process is also highly likely to occur in other seasonally dry systems, as seen in most of Africa. Deep-soil water extraction by trees is expected to be more pronounced in environments where the topsoil undergoes rapid drying, as in regions with coarse soils or unpredictable and variable rainfall [35]. Direct measurements of the magnitudes of HR are yet to be performed and remain a challenge. However, modelled estimates suggest that although HR in water-scarce regions is relatively small (less than 10% of transpiration), it is ecologically significant for the maintenance of functional small roots and mycorrhizal activity [22,33]. Trees increase infiltration rates through downward HR (Figure 1), buffering deep-soil water losses due to environmental fluctuations and making this water a reliable source for deep-rooted plants [9,36,37]. For example, in a Prosopis velutina savanna, tree HR accounted for up to 50% of deep-water recharge and contributed almost equally to transpiration during droughts, extending the growing period and generally improving the water status of the system [38].

Studies indicate that different tree species have different water utilisation capacities, which influence the amount of water that can be retained in the soil [39]. Additionally, changes in the direction and magnitude of HR are regulated by the constantly changing water statuses of the constituent organs, such as leaves, shoots, roots, and soil, influencing the system’s ecohydrology [7,28]. Data indicate that the tree canopy influences water retention by the soil, and the magnitude of this retention increases with aridity [35], likely buffering losses due to evapotranspiration. Furthermore, plant analyses showed that grasses under trees had higher moisture (unpublished data) and nutrient contents (N, P, and other palatability factors for grazers [40,41]) than those in the open. Trees affect the overall water budget of an ecosystem not only through HR but also through the associated gas exchange and transpiration. These processes are maintained without tree mortality during times of water stress, mainly because of HR [35,42,43]. Tree HR also influences soil biotic interactions by sustaining the functions of mycorrhizae and other soil organisms by preventing desiccation [33,44]. Therefore, tree HR in dryland systems is critical to the overall ecohydrology of the ecosystems, influencing nutrient transfers and the associated complex biotic interactions that we describe in the following section.

2.2. The Linkages Between Water Redistribution, Nutrient Transport, and Biotic Interactions

HR has a significant influence on nutrient transport and redistribution in dryland systems. Figure 2 is a schematic that illustrates this process. Nutrient redistribution occurs through several mechanisms, which include the transport of nutrients via mass flow of water (e.g., SO₄²⁻, Ca, and Mg), the uptake of nutrients via diffusion (e.g., PO₄³⁻, NH₄⁺, K, and other metal micronutrients that need water as a medium), and a combination of both these mechanisms [24]. Tree HR has a significant influence on mycorrhizal abundance and activity, which significantly impact the supply of essential plant nutrients [7,45], and increases the overall nutrient capture potential [33]. Direct transfer of water via HR from oak trees to their mycorrhizal symbionts during an imposed drought was shown in an oak savanna [46]. Studies also revealed the influence of HR on organic matter decomposition, resulting in higher aboveground uptake of N [47]. The natural abundance of the stable ¹⁵N isotope in a semi-arid South African savanna of commonly occurring N-fixing and non-N-fixing trees and their under-canopy grasses corresponded with the reported values of mycorrhizal-supplied N (indicating a mycorrhizal source of N) during the dry season [48], the time when HR resulted in water movement to the upper soil layers [32]. This suggests that mycorrhizae were active, even during the driest part of the year, possibly due to dry-season HR.

Mutualistic and symbiotic associations between trees and soil organisms (e.g., rhizobia and mycorrhizae) significantly influence the nutrients supplied to plants [49,50,51]. Up to 90% of terrestrial plants have symbiotic associations with mycorrhizal fungi. These associations are considered to be of evolutionary significance, as studies indicate that plants may have colonised land as a result of these associations [52,53,54]. In natural ecosystems, plants obtain up to 80% of N and 90% of P via mycorrhizal fungi, forming a significant pathway for nutrient capture in plants [55]. The much smaller hyphae and their substantial extent in the soil significantly increase the absorptive area compared to plant roots alone [56]. Evidence of tree’s associations with soil micro-organisms is considerable [57,58,59,60]. However, our understanding of the associated ecophysiological processes and the impacts or magnitudes of these interactions remains inadequate [25,58,60,61], but they are thought to be substantial [55].

Several studies have shown that symbiotic associations between trees and soil organisms are highly significant in overcoming ecosystem nutrient limitations [33,51,54,60,62,63]. These associations are also reported to influence interplant interactions [64]. Furthermore, there is increasing evidence that hyphae of mycorrhizae act as conduits for water and nutrient transfers between plants. In an old-growth pine forest in North America (Oregon, USA), ectomycorrhizal hyphae acted as conduits for water hydraulically lifted by the pine trees, facilitating the supply of both nutrients and water to nearby seedlings [65]. Additionally, carbon isotope labelling of spruce trees in a temperate forest at the tall canopy level showed that carbon assimilates from labelled trees were found in unlabelled neighbouring trees that were supplied through common ectomycorrhizal networks [66]. The natural ¹⁵N abundance values of grasses and trees from a semi-arid savanna in South Africa indicated that both used mycorrhizal-supplied N for most of the year and biologically fixed N during the wet season. In contrast, they did not seem to use mineralised N at all [48], which only became available at the end of the wet season [67]. A significant impact of tree ecophysiology is the linkage between mineralisation and biological N fixation. However, these processes are yet to be disentangled [68]. Research on the process of mineralisation has evolved considerably, moving from the net mineralisation concept, where nitrification was the most limiting step in the mineralisation process, to viewing exoenzyme-driven depolymerisation as the rate-limiting step. This step is highly controlled and influenced by microbial activity, including that of mycorrhizae [69].

There is increasing evidence of linkages between plants and rhizosphere processes, mainly for two reasons: first, associations between plant roots and both endophytic and ectophytic microbes, and second, factors that regulate plant processes also influence rhizosphere processes [45,68]. Studies have also revealed the complex mechanistic processes of N transformations, N use by plants, and the complex linkages between N-fixing soil organisms and plant roots [45,70]. In semi-arid savannas of East Africa, mineralisable N was higher under trees, as were P, K, and Ca [71], indicating the significant influence of trees on the soil characteristics of the under-canopy. Furthermore, under-canopy vegetation has a higher nutrient content [72] and serves as a critical food resource for grazing animals in the dry rangelands of Sub-Saharan Africa [40,41,73]. To summarise, trees have a considerable influence on soil nutrient dynamics, as they move water and nutrients from the physical environment to the biological environment [9,34], which in turn influences the mutualistic associations between trees and soil organisms [44,46,60].

2.3. Nutrient Redistribution by Trees

While the role of trees in nutrient cycling through litterfall and their associations with soil organisms is well established, direct redistribution of nutrients by trees requires further examination. Determining nutrient redistribution by trees has proved challenging. For example, N cycling and soil–plant N dynamics are complex due to internal cycling of N by plants [45,74,75]. Furthermore, these dynamics are impacted by plant–rhizosphere interactions [57,70]. There is evidence of direct uptake of amino acids in the soil, organic forms of soil N, and water-soluble organic N by plants [45,58,69,76]. However, in both savanna ecology and agroecology, much of the research focuses on plant uptake of inorganic soil nitrogen.

Most interpretations of deep-soil nutrient cycling are based on studies of root activity patterns or distributions [61,77]. However, there is limited evidence of direct uptake and redistribution of soil nutrients by plants. Using a ¹⁵N stable isotope tracer, we previously showed deep-soil redistribution of N by semi-arid savanna trees to under-canopy grasses throughout the year, independent of HR [48]. Furthermore, this capacity was seen in both N fixers and non-N fixers. Similarly, natural variation in ⁸⁷Sr/⁸⁶Sr isotope data (used as a surrogate for nutrients like Ca) revealed plant-mediated movement of deep-soil P, Ca, and Mg to the topsoil that could not be explained by any mechanism other than plant uplift from deeper layers of the soil [78]. There is increasing evidence of tree nutrient (re)distribution from the deep soil to the topsoil, making these otherwise unavailable nutrients available to shallow-rooted plants and the microbiota [48,78,79].

3. Trees in Agroecosystems of Sub-Saharan Africa

Agroforestry is a central tenet of ecological intensification of agriculture and conservation agriculture [1,80,81]. However, adoption rates remain low in Sub-Saharan Africa [82,83], and existing on-farm trees are being removed due to various pressures [84,85,86]. Key resource-conservation functions of agroforestry include (a) water conservation by reducing surface and subsurface runoff and improving plant water use; (b) nutrient conservation by preventing leaching losses; and (c) nutrient addition through crop residues, mulch, and litterfall.

Prudent resource management is essential to raise yields in the water- and nutrient-limited soils of Sub-Saharan Africa. Research shows that yield improvements are unlikely without fertiliser use [87,88]. Thus, strategies to reduce inputs while increasing outputs are a major agronomic focus [3,79,89,90]. Tree-mediated improvements in resource-use efficiency have been proposed, making agroforestry central to ecological intensification and conservation agriculture [1,3,83,91]. However, there is no consensus on trees’ role in improving this efficiency [91,92,93]. Agroforestry, where trees or shrubs co-occur with crops, aims to optimise these resources and maximise yields [3,79,90,94]. However, resource use is complex due to interactions among plants and soil, making it difficult to disentangle mechanisms [95,96,97,98,99,100]. Understanding these interactions is vital yet challenging in the diverse, resource-limited agroecosystems of Sub-Saharan Africa.

3.1. Why Is the Focus on Sub-Saharan Africa Important?

Approximately 66% of Sub-Saharan Africa is desert or dryland, and 45% of Africa’s human population inhabits this region [101]. The most prevalent livelihoods in this region are agriculture and livestock farming [102], and economies are tightly linked to agricultural growth [15,102]. African rangelands contribute 33–41% of global ruminant meat production [103], making this region a significant land resource for animal production [104]. It accounts for 52% of remaining global arable land [15], making it an important region for global agricultural production. However, about 65% of this region’s arable land and 30% of grazing land face high levels of soil degradation [105,106,107,108]. Agricultural expansion into previously unexploited regions is expected, including Miombo and Guinean savanna woodlands [109]. Additionally, this region has shown the highest growth in agricultural production globally since 2000, but this increase has been driven by land expansion rather than yield gains [15]. Forests, woodlands, and scrublands in East Africa have declined sharply due to conversion to agricultural lands [110].

Data shows that agricultural expansion is a major driver of soil degradation [111]. For instance, in the Sahel, up to 30% of human-induced degradation is attributed to expansion, resulting in croplands that are largely treeless [112]. A survey of six nations (1980s–2014) revealed land-use intensity has reached permanent cropping (including silvopastoral use), requiring high fertiliser inputs and disappearance of fallow periods [113]. In eastern Tanzania, the Miombo woodland area decreased by 50% (1964–1996), while the cropland area increased by nearly 600%, largely due to charcoal harvesting and shifting cultivation [114]. Elsewhere in the Miombo, large areas were cleared for tobacco [115]. A matter of concern is that agroforestry adoption remains marginal, with only a 2% increase reported in Sub-Saharan Africa [93,116], leaving monocultures widespread and degradation high [93,117,118].

A global analysis shows tree cover of 10–20% on farms in Sub-Saharan Africa, mostly in smallholder agroforestry [119]. However, these drylands risk becoming inadequate for sustained production of tree-based goods (e.g., charcoal, fuelwood, and browse) [120,121]. Rising demand for timber, fuelwood, and charcoal is the main threat to on-farm trees [84,85]. In South Africa, communal rangelands that are severely degraded from fuelwood extraction can no longer produce adequate biomass [122]. Furthermore, woody biomass is predicted to be exhausted by 2027 in these rangelands unless there is a 15% reduction in its consumption [121]. The International Energy Agency estimates that by 2040, 1.8 billion Africans will still depend on solid biomass (mainly wood and dung) for domestic energy [123]. Estimates suggest that Africa already produces 51% of global charcoal [124]. Demand for bioenergy is predicted to increase substantially [84]. Recent biofuel crop investments, spurred by trade policies in North America and Europe, could further divert farmland to industrial biofuel production [123]. Global fuelwood demand is expected to rise sharply, approaching industrial wood prices [85]. Earlier reforestation and energy-plantation policies were scaled back in the 1990s, as low-cost fuelwood remained the preferred option over farm forestry interventions [84]. Under these conditions, substantial efforts are needed to promote ecological intensification in Sub-Saharan Africa, an approach that is rarely addressed in this region [79].

3.2. The Role of Tree Ecophysiology in Agroforestry Systems of Sub-Saharan Africa

The nutrient and water management practices of agriculturists in Sub-Saharan Africa are diverse. Farmers use water management structures such as Zaï planting pits in the Sahel, planting basins in southern Africa, and in situ water harvesting systems (e.g., open and tied ridges) to reduce runoff, trap rainwater, enhance infiltration, trap organic matter, and reduce nutrient losses via leaching [107,125,126,127,128]. Conservation practices also include maintaining mulch to improve infiltration [125,126,127] and nutrient management strategies like intercropping with trees, rotational cropping, natural fallows (leaving the land uncultivated for a period), and improved fallows (planting woody/herbaceous species to replenish soil nutrients) [129,130,131]. A major form of nutrient management in Sub-Saharan Africa is adding biomass (either trees or crop residues) [132,133]. However, experimental field trials show complex relationships between biomass additions and N recovery by crop plants; crop plants seem to mainly rely on soil N, and tree and crop components differ in N uptake [132], indicating resource partitioning.

In the preceding sections, we highlighted the central role of trees in water and nutrient cycling in drylands. In the following sections, we examine whether these functions have potential for ecological intensification of food production and identify constraints when harnessing the ecophysiological functions of trees in these agroecosystems.

3.2.1. Influence of Tree Ecophysiology on Resource Use

The main objective in agriculture is to increase crop yields while improving resource-use efficiency and minimising external inputs [79,81,99]. Research shows that interspersed trees can boost crop yields [134,135]. In Sub-Saharan Africa, maize (Zea mays) yields on degraded, N-limited soils were found to be higher under agroforestry than under monoculture, and modelling suggests agroforestry maintains yields even in dry years, unlike conventional monocropping [6]. However, tree–crop interactions are complex; yield gains depend on the crop and tree species [90,135,136,137], as well as management practices [134,135]. For example, taro (Colocasia esculenta) performed better with shade than millet (Pennisetum glaucum) [135], while sorghum (Sorghum bicolor) yielded more (by 56%) under Vitellaria paradoxa than under Parkia biglobosa [134]. Elsewhere, intercropping sorghum and cowpea (Vigna unguiculata) with Acacia saligna trees showed no yield benefit, and pruning had no effect, indicating that tree shade was not an influencing factor [138]. Furthermore, ¹⁵N tracer studies revealed that A. saligna trees took up a higher proportion of the tracer than sorghum, indicating competition [139]. Tree–crop–fertiliser interactions also vary with agroforestry tree species [90]. In Ethiopia and Rwanda’s Rift Valley, Faidherbia albida + wheat (Triticum aestivum var. aestivum) improved yields and N/P-use efficiency, while Acacia tortilis + maize and Grevillea robusta + maize (Zea mays) produced lower yields [90]. Fertiliser inputs and tree–crop combinations must therefore be tailored locally to improve yields.

Direct measurements of tree–crop interactions in agroforestry remain scarce, and differing methods complicate comparisons of resource-use efficiency. Some studies report increased rain-use efficiency under agroforestry, based on calculated water use or productivity [125,126,127,138,140]. Despite the frequent emphasis on root overlap and competition, evidence suggests that agroforestry may lead to higher water-use efficiency, although the magnitude of this effect remains unclear. It was inferred that this improvement occurred due to subsoil water becoming available to shallow-rooted crop plants via presumed HR or potential water-use complementarity (trees use deeper soil water, while crops use water from the topsoil) and reduced surface water runoff [126,127]. Additionally, supplemental irrigation often had little effect on yields in agroforestry [134], suggesting improved water-use efficiency. Modelling estimates show that during droughts, crops used 85% of the annual rainfall, with only 15% lost [127,140]. Spatial water-budget modelling in West African parklands found that intermediate tree cover enhances groundwater recharge [141], aligning with the evidence of HR-mediated deep infiltration in savannas [36,38]. A few agroforestry tree and shrub species showed HR during dry seasons or droughts, and modelling simulations suggested the possibility that trees have a significant influence on dry-season water and nutrient dynamics [142]. The use of N-fixing trees such as Sesbania sesban during fallows shortens fallow periods and boosts subsequent maize yields [2,5,143,144]. However, evidence related to the role of trees in resource conservation is mixed, and the mechanisms remain unclear. Further research is needed on how HR or deep-soil nutrient redistribution can enhance sustainability in dryland agroforestry.

A characteristic of nutrient management in African agriculture is the use of tree cuttings and residues to fertilise depleted soils [87,129]. While such inputs can provide adequate nutrients for understory crops, except P [132,133], there is competition from livestock feeding, fuelwood collection, and other socio-economic demands [87,130]. Moreover, this practice is mostly successful in areas with high biomass productivity [145]. The biomass must also undergo mineralisation before nutrients become plant-available, a process influenced by soil conditions, biomass quality, and decomposer diversity [146]. Compared to mineral fertilisers, biomass inputs may result in nutrient immobilisation (e.g., of P) and slower release rates [87,147]. This often necessitates fallow periods, yet such periods are disappearing or becoming increasingly rare [113,148]. A big challenge is the removal of mature trees, which undermines the presence of perennial root systems, arguably one of the most valuable features of agroforestry. Establishing new trees to harness their ecophysiological functions can take 10–20 years [149]. Although there is some evidence of resource conservation linked to tree ecophysiology, it remains limited and is often based on assumptions rather than empirical data. However, indirect evidence suggests that trees have significant potential for resource conservation. Targeted research on tree–crop interactions is urgently needed to support sustainable resource use.

3.2.2. Influence of Tree Ecophysiology on Nutrient and Water Cycling

Few studies have examined nutrient uptake, nutrient recovery, or leaching losses in tree–crop mixtures [139,150]. Agroforestry research has largely focused on nitrogen (N) use, often showing N-use complementarity in these systems [130]. Some studies, however, report competition for N [139], though N leaching losses were lower when intercropped with trees due to extensive perennial root networks [136]. Overall, data on nutrient uptake and transfer are limited, making a comprehensive understanding of nutrient redistribution challenging, perhaps due to the difficulty in measuring the magnitudes of water and nutrient transfers.

The role of soil organisms in N cycling in agroecosystems has been largely overlooked [68,151,152]. Although an inoculum is applied to some non-promiscuous legumes, its effectiveness is debated [87]. However, studies indicate that tree–mycorrhizal associations can significantly enhance crop nutrient acquisition [68,151,152], consequently improving nutrient-use efficiency [153]. For instance, mycorrhizal infection can facilitate N exchange among co-occurring pasture plants [154]. A ¹⁵N tracer study demonstrated transfer of labelled ¹⁵N from N-fixing mycorrhizal plants to non-N-fixing, nonmycorrhizal plants [155]. Tree roots provide an extensive and deeper area for mycorrhizal colonisation, contributing to perennial and persistent mycorrhizal functioning [56,151,156] and serving as important sources of inocula [157,158].

Associations with rhizobia act as biofertilisers by fixing atmospheric N (unlike mycorrhizae, which improve access to soil nutrients) [56]. Intercropping with legumes, a popular management strategy for replenishing lost soil N, remains central to conservation agriculture [80], though adoption in Sub-Saharan Africa remains low [80,87]. Observations from natural savannas reveal opportunistic nodulation, where N fixation occurs mainly under nutrient-limited or competitive conditions [45,159]. Thus, N transfer in mixed legume–crop systems may not always occur [87,139]. For example, Faidherbia albida, although leguminous, showed no nodulation [149], yet crop yields and fertiliser-use efficiency were higher when it was retained on farms [90,92]. In addition, only a subset of legumes nodulate [160]. In our South African studies, no nodulation was observed; the natural abundance of ¹⁵N indicated mycorrhizal-derived N [48], suggesting trees may contribute to nutrient cycling even without nodulation, perhaps due to mycorrhizal associations [9,45]. Few studies have directly assessed soil resource uptake in dryland systems [125,139,142,150]. This knowledge gap extends beyond agroecosystems to natural drylands as well.

Vitellaria globosa and Parkia globosa, which are commonly used agroforestry trees in semi-arid West Africa, exhibited HR during the dry season [142]. Another shrub, Guiera senegalensis, also redistributed water under experimental drought conditions [161]. A ²H tracer experiment confirmed HR benefits: millet grown with G. senegalensis yielded 900% more than without shrubs, indicating they supplied deep water [161]. Most HR studies rely on soil or root analyses; studies on nutrient redistribution are even rarer. Water relations between trees and the understory in wild savanna areas vary with the seasons, the magnitude of rainfall, and grazing pressure [71,162,163]. Shade effects also shift seasonally, with the dry season exhibiting a positive influence on the understory [164]. Measurements of the natural abundance of ²H₂O revealed a change in the water sources of trees and grasses in response to changes in rainfall and the soil water status, indicating a shift in water relations from competition to partitioning of water [32]. Given the dynamic interactions among plants, soil, and water, current data are insufficient to evaluate HR’s impact on yields. A modelling exercise suggests that high yields can be sustained under agroforestry, even during dry periods [6]. Focused research is needed to determine whether tree ecophysiological functions can be harnessed to sustainably boost food production in Sub-Saharan Africa.

3.2.3. Influence of Tree Ecophysiology on Soil Quality

The largest positive impact of the presence of trees is an increase in soil quality. Trees improve soil structure, create macropores that enhance infiltration, reduce runoff and erosion, and prevent physicochemical degradation (e.g., nutrient depletion and soil organic carbon loss) [1,6,61,165]. Consequently, agroforestry’s potential to improve soil quality is widely recognised [166]. However, the role of HR and nutrient redistribution remains unclear and understudied [167]. In Figure 3, we illustrate tree–soil interactions in a simplified schematic. In previous sections, we described how tree water and nutrient redistribution affect soil processes (see also Figure 1 and Figure 2). For example, soils under Faidherbia albida canopies have higher organic C, N, and mineralisation rates than soils outside the canopies [168,169]. Furthermore, microbial biomass and biomarkers of major microbial groups were higher under the tree canopy, indicating higher microbial activity [169].

Tree removal from dryland agroecosystems causes cascading effects, altering soil pH [170], disrupting soil microbiota and their interactions [54,171], and ultimately degrading fertility and soil health. Trees have been used to restore soil quality in parts of Sub-Saharan Africa, increasing soil organic carbon, total N, and P [172,173,174]. However, these benefits result from the presence of large trees on farms, reflecting long-term accumulation rather than short-term effects. Declines in on-farm tree abundance are reported across Sub-Saharan Africa [116,175], especially in the eastern and southern regions [176]. After tree removal, it can take decades for new trees to mature and provide facilitative effects (e.g., 20 years for F. albida) [149]. Conserving remaining trees is therefore vital, yet motivations to plant new ones are often weak [82,177,178,179], as seen in the low agroforestry adoption and the ongoing removal of large trees when immediate needs arise [93,94,179].

3.2.4. Influence of Tree Ecophysiology to Improve Climate Resilience

The effects of climate change are omnipresent, and an increase in the extent of African drylands is expected [180]. Climate change is also predicted to make African drylands highly vulnerable [106,181,182,183]. Increasing temperatures and changes in precipitation patterns are key drivers that will negatively impact food production in African drylands [11,12,13,182]. Drastic reductions in rainfall have been reported in various parts of Africa [16]. Although research efforts are presently insufficient [11,182], there is a consensus that the vulnerability of the agricultural sector in Sub-Saharan Africa will increase due to unpredictable changes in rainfall and increases in pests and diseases [12,13,184,185]. Furthermore, the impact of climate change on infectious diseases in crop plants has also been shown, although to a lesser extent than the impacts on livestock and wildlife [186,187]. The spread of pathogens from wild plants to economically important agricultural plants is increasing due to a rise in emerging infectious diseases likely resulting from expanding host ranges or changes in pathogenesis [187]. Changes in plant phenology and flower/tassel initiation are expected in crop plants, and reductions in yields have been reported [12,188,189,190]. The severe changes in rainfall over the last several decades have led to the overuse of marginal lands, resulting in further degradation of land that was already less productive [16]. Furthermore, the transition to irrigated agriculture in this region is highly unsustainable [191,192].

Agroforestry has been proposed to mitigate and buffer the effects of climate change [3,6,173,193,194]. Environmental changes strongly influence tree ecophysiological functions. Therefore, whether tree ecophysiological functions can buffer the effects of climate change in dryland agroecosystems of Sub-Saharan Africa remains to be determined. As the immediate results of climate change are increased CO₂, increased temperature, and changes in climatic patterns impacting water regimes, what are the potential implications for trees? Figure 4 is a simplified schematic illustrating the possible impacts of climate change on tree ecophysiology in dryland ecosystems.

The carbon sequestration potential of trees in agroecosystems is proposed as a means to mitigate the effects of climate change [6,89] and forms a large part of the carbon offset market when opportunities are available [110,195]. However, tree responses to increasing temperatures and CO₂ are not straightforward, and different functional groups (e.g., evergreen vs. deciduous trees) respond differently to these changes [196]. Trees also influence the CO₂ dynamics in the soil through litterfall, which is influenced by leaf traits and leaf chemistry [197]. Furthermore, litter decomposition and other soil processes are influenced by the availability of N and other nutrients in the soil [198]. Large-scale vegetation die-offs and increased pest infestations have occurred as a result of climate change [42,199]. Experimental evidence from boreal forest trees indicates that tree responses to increases in CO₂ and temperature are contingent on water and nutrient availability and do not have a direct relationship (high CO₂ ≠ higher growth) [200]. Similarly, synthesis of data from climate change-related droughts in different ecosystems across Europe indicates the role of water in influencing tree responses [201]. Such information is not available for dryland tropical trees; however, it is likely that these trees will also exhibit similar responses to water and nutrient limitations, as the fundamental processes of photosynthesis and nutrient uptake pathways are unlikely to be drastically different, albeit with adaptations that have evolved to survive local climatic conditions.

Results from Free Carbon Dioxide Enrichment (FACE) sites show linkages between physiological controls on tree growth (e.g., rubisco-controlled carbon fixation) and increased concentrations of CO₂ in the environment. Moreover, there seem to be species-specific responses to increases in CO₂ [202]. Recent remote sensing analyses indicate an increase in tree crown size rather than tree densities in Africa [203], which aligns with experiments at the FACE sites, which show increases in wood volume, the leaf area index, canopy size, and fine root density [202,204]. This was also shown in a greenhouse study using Acacia karoo and Acacia nilotica in South Africa [205]. However, the conclusion that an increased root carbohydrate content in seedlings indicates a higher potential for re-sprouting, leading to bush encroachment [205], is premature. Data from large-scale CO₂ enrichment sites that have been monitored for longer than 10 years show an increase in photosynthesis in the short term but a decrease in photosynthesis in the long term, termed photosynthetic downregulation, when plants have been exposed to enriched CO₂ for a longer period [202].

Tree mortality is an important factor to consider in the context of climate change. Figure 5 illustrates the cascading effects of tree mortality in dryland agroecosystems. Heat-induced tree mortality has been documented in different parts of Sub-Saharan Africa [206,207,208]. However, differences in cavitation tolerance, as a mechanism to withstand water limitations brought about by drought, could be used in dryland agroecosystems, but cavitation tolerance also has a threshold before impacting tree survival. There is little clarity about the effects of an increase in atmospheric N on N cycling in general. Experimental evidence suggests that the complex linkages of mycorrhizae, which regulate nutrient exchange and transport, may be affected via disruption of the nutrient exchange mechanisms [45,209]. However, a recent meta-analysis shows that arbuscular mycorrhizae can greatly increase the resilience of trees to climate change [210].

In agroforestry, research often highlights how the multifunctionality of trees enhances climate resilience by improving food and livelihood security [83]. However, the claims of enhanced water use by trees [3,185] are not fully supported by the evidence. While the effects of climate change on carbon sequestration by trees are established in the short term, the longer-term effects on farms are difficult to predict due to the limited available data and the complex interlinkages between the environment and plant physiology. Furthermore, only established trees will sequester carbon, which might also be offset by agricultural practices that result in farms being carbon sources rather than sinks. Therefore, the contribution of trees to carbon sequestration is complex and contingent on factors such as tree age, tree type, and the agricultural practices in operation.

3.3. Tree Ecophysiological Function and Agroecological Sustainability

Terrestrial ecosystems comprise a diverse array of plant species that coexist through complex ecological interactions. Agroforestry emulates this diversity by integrating trees with crops, yet the mechanisms governing interplant relationships in these systems are insufficiently understood. Classical ecological theory proposes two principal mechanisms for plant coexistence: habitat and resource differentiation [211], which are both grounded in the Gaussian principle of niche differentiation [211,212]. Habitat differentiation posits that species occupy distinct habitat niches, while resource partitioning suggests that species utilise different components of limiting resources (e.g., water and nutrients), enabling coexistence. However, these frameworks do not adequately explain instances where plants share space and resources without evident competition, a phenomenon frequently observed in agroforestry.

Despite claims that agroforestry enhances water and nutrient efficiency, improves soil health, boosts biodiversity, sequesters carbon, and supports food security and livelihoods, a disconnect persists between theory and empirical evidence [1,3,83,89,91,93,94]. This gap stems largely from generalised, often untested assumptions and limited direct measurements. For instance, a meta-analysis found increased water infiltration and soil moisture under agroforestry across diverse climates, suggesting improved resource availability [83]. While the study asserted that agroforestry enhances crop yields without compromising ecosystem services, it offered little explanation of the underlying mechanisms. Similarly, although the importance of selecting tree species to minimise crop competition was emphasised, the evidence was restricted to shading effects. Another study challenged the presumed benefits of trees for soil fertility, citing Faidherbia albida as an example and underscoring the scarcity of empirical data [149]. What remains poorly understood yet potentially significant for the ecological intensification of food production in Sub-Saharan Africa is how tree ecophysiological processes, particularly water and nutrient redistribution, influence biogeochemical cycling and resource conservation. Limited data and methodological challenges in measuring these complex interactions continue to obscure the role of tree ecophysiology in ecological intensification and conservation agriculture [167].

In dryland agroecology, discourse often focuses on competition and the need for resource-use complementarity between trees and crops [1,127,149,212,213,214,215]. A common assumption is that deep-rooted trees access nutrients only from lower soil layers while shallow-rooted crops rely solely on topsoil nutrients, thereby reducing competition [139,150,152,216]. However, root systems exhibit significant phenotypic and physiological plasticity, with seasonal shifts in function [7,25,32,217,218,219,220]. Evidence of resource partitioning and facilitation via water and nutrient redistribution exists but remains sparse and largely indirect.

Extrapolating ecosystem services from natural savannas to agroecosystems is problematic because tree ecophysiological functions may differ in these systems. Trees are often described as “irrigators” and “nutrient pumps” [1,167], yet these labels are based on limited data from natural systems. The phenomenon of HR is regulated by pronounced differences in water potentials, often during droughts or seasonal aridity, as an adaptive response to water stress that mitigates the risk of hydraulic cavitation within the plant vascular system, as detailed in Section 2.1 [7,22,28,32,142,161,221,222,223,224,225,226]. However, how HR may change under irrigated conditions remains unclear, especially since irrigation can alter root architecture, inducing shallow-rootedness in trees [219].

Focused research is urgently needed to understand the role of HR in enhancing environmental sustainability. Bayala and Prieto (2020) briefly outline future research directions for the role of HR in dryland agroecosystems [167]. Some studies suggest that trees contribute to water conservation and groundwater recharge in dryland agroecosystems [36,43], and intermediate tree cover has been linked to increased recharge in West African plantations [141]. While crop yield improvements have been reported, they depend on numerous factors (see Section 3.2.1 and Section 3.2.2 for details). Sole reliance on organic nutrients is insufficient for achieving yield gains in Sub-Saharan Africa; mineral supplementation is often necessary [87,88]. Field trials show that strategic plant mixtures can improve fertiliser efficiency, but recommendations must be locally tailored [90]. Agroforestry is already being used in parts of Sub-Saharan Africa to restore soil quality [194]. However, models optimising species combinations and management practices, such as ideal tree–crop pairings, remain underdeveloped [93,94].

Tree HR, nutrient transfers, and soil biotic interactions are closely linked, yet leveraging these processes for sustainable food production is challenging. Experimental evidence shows that trees can redistribute nutrients to shallow-rooted understory plants, particularly nitrogen, though the extent of these transfers is uncertain [48,78]. Evidence, though sparse, suggests that trees impact nutrient cycling, particularly cycling of N in agroecosystems [139]. Nitrogen uptake is an active process [48,227,228,229], and HR can affect nutrient uptake via passive flow or mass flow (Figure 2). Interactions between trees and soil microbiota, such as mycorrhizae and nitrogen-fixing organisms, significantly impact nutrient cycling (as detailed in Section 2.2). Mycorrhizae mediate nutrient transfers in natural ecosystems [33,46,66] and have been reported to perform this function in some agroforestry systems [155], offering potential for ecological intensification [153,230]. However, agricultural intensification can disrupt the functional diversity of soil microbial communities, and natural selection may result in less beneficial strains [56]. A transition toward low-input, sustainable agroecosystems will require deeper investigations into plant–rhizosphere interactions [70,152]. Furthermore, the agricultural practices and management strategies employed will profoundly influence these ecological processes, shaping the future of sustainable food production in Sub-Saharan Africa [56,81,87].

4. Conclusions

Studies in Sub-Saharan African agroecosystems show that agroforestry trees play a critical role in conserving resources by reducing surface water runoff, soil erosion, nutrient loss, and leaching. However, the influence of tree ecophysiology, despite being acknowledged often, remains understudied. Evidence from natural dryland systems demonstrates the significant influence of tree ecophysiology in regulating water and nutrient cycling, as well as sustaining the complex biotic interactions that support plant productivity. However, even in these systems, the underlying mechanisms are not fully understood. Indirect evidence from agroforestry systems also suggests that leveraging these processes may have significant potential for sustainable food production. For this, large trees need to be conserved on farms. Mariotte et al. [230] propose a conceptual framework that integrates insights from both natural and managed ecosystems to promote ecologically sustainable food production, with particular emphasis on plant–soil feedback loops that improve nutrient retention, cycling, and acquisition [230]. This framework is especially relevant to Sub-Saharan Africa, where ecological intensification has received limited attention [79]. In theory, processes such as HR and nutrient transfer by trees offer great potential for ecological intensification in dryland agroforestry. In practice, however, their application is highly context-dependent and shaped by factors such as the soil type, tree–crop combinations, fertiliser use, and farm management practices. Given this region’s diverse and heterogeneous agroecosystems, broad policy frameworks are unlikely to be effective at local or sub-regional scales [129,231]. To effectively exploit the ecosystem services provided by trees, a key feature of ecological intensification, research tailored to local farm conditions is needed, with a focus on maintaining soil quality, securing long-term productivity, and conserving resources. Balancing agricultural intensification with ecological sustainability remains a challenge, yet it is vital for addressing food security, land degradation, and the effects of climate change.