Chapter 7
Trees potentially out-compete grasses by growing taller and thereby shading grasses growing beneath their canopies, but soil moisture is the limiting resource in savannas, not sunlight. Water enables both the capture of carbon dioxide from the atmosphere and the uptake of mineral nutrients from the soil. Competition for access to water and nutrients operates mostly underground, out of sight and is thus easily overlooked. Once established, trees can build up supporting stems to overtop grasses and raise their leaf canopies beyond the flame zone. However, grasses can activate faster than trees in accessing soil moisture when conditions permit. Competition is effective mainly during the establishment phase for tree seedlings, which must somehow claim rooting space amid the grass roots.
Plants derive carbon, the structural constituent of their biomass, from atmospheric carbon dioxide. Mineral nutrient resources are obtained primarily from the products of decaying organic matter in the soil. Nitrogen, taken up in the form of nitrate or ammonia, is an essential constituent of proteins. Potassium is needed for ionic regulation. Magnesium is a constituent of chlorophyll. Calcium is a component of cell walls. Phosphorus, in the form of phosphates, fuels cell metabolism. Sulphate is needed for sulphur-containing amino acids. Legumes in the subfamilies Mimosoideae and Papillionoideae overcome nitrogen deficiencies by gaining atmospheric nitrogen fixed by bacteria housed in root nodules. Phosphorus uptake is facilitated by associations between roots and mycorrhizal fungi in the soil.1 These processes are energised by solar radiation captured by chlorophyll molecules located in chloroplasts in leaves.
Organic matter decomposition is brought about by soil organisms. Termites and other invertebrates play the leading role in breaking down plant litter. Bacteria and fungi complete the mineralisation of the organic matter. Their activity is dependent on sufficient moisture and ceases at some stage during the dry season. This results in a pulse of nutrient release following the first adequate rains initiating the wet season.2,3 Most of the available nutrients are generated in the top 20 cm of the soil.4 Water availability determines when growth can occur, while soil nutrient availability can restrict the rate of growth achieved while moisture is sufficient. The low concentration of carbon dioxide in the atmosphere constrains the growth rate achieved when soil nutrient supplies are adequate.5
In this chapter, I will outline how trees and grasses grow and compete in the context of the seasonal restriction in rainfall and its erratic distribution even within the wet season. Large herbivores set back plant growth through their consumption of leaves, stems and sometimes roots, but also benefit plants by dispersing seeds contained in fruits and grass inflorescences (flowering heads).
Growth of Woody Plants
Woody plants increase in standing biomass by annual increments to stems and branches bearing leaves. Savanna trees typically produce their new leaves shortly before the start of the rains by drawing on moisture stored in their stems or accessed from deeper soil levels.6,7 Miombo woodland trees flush their new leaves especially early, 4–8 weeks before the first rain showers initiate the wet season.8 These leaves are tinted with red or orange colours indicating high anthocyanin contents, which may act as a sunscreen (Figure 7.1). The leaf display is followed by shoot extension, then stem expansion and lastly root growth, during the course of the wet season.9 The initial growth of new leaves at the start of the wet season is generated from stored carbohydrate and mineral reserves. Heights reached by savanna trees rarely exceed 20 m, and most remain under 10 m. Shrubs typically do not exceed 3–5 m in height and are characteristically multi-stemmed from their base. Savanna trees typically exhibit spreading canopies, in contrast to the upward growth towards light shown by forest trees.
Figure 7.1
Pre-rain flush of red leaves on (A) Brachystegia in Zambia and (B) Julbernadia in Zimbabwe.
Most of the woody plants in African savannas are deciduous, shedding their leaves at some stage during the dry season.10,11 This forestalls the moisture loss that they would have incurred had leaves been retained.12 The thorny acacias typical of dry/eutrophic savannas generally drop their leaves quite early in the dry season, while the broad-leaved trees found in moist/dystrophic woodlands on sandy soils tend to retain their leaves somewhat later into the dry season. Nutrients are mostly withdrawn before the leaves are shed. Evergreen or semi-evergreen trees and shrubs that retain their leaves through the dry season tend to be found in bottomlands where soil moisture persists longer and less fire penetrates as a result of the trampled grass cover. Termite mounds also provide refuge sites from fires. Evergreen leaves are tougher than leaves of deciduous species in order to resist wilting when water is deficient. Only a tiny remnant of the foliage produced by savanna trees remains available to herbivores through the dry season (Figure 7.2). By the end of the dry season, savannas become bare of tree leaves and bereft of shade (Figure 7.3). Conditions may become baking hot and dry before the first rains.
Figure 7.2
Seasonal variation in tree foliage retained within browsing height in mixed broad-leaved woodland in the Nylsvley Nature Reserve, subdivided by plant types. PSpin, spinescent acacias; PDec, deciduous bushwillows and wild raisins; PEvg, evergreen monkey oranges and karee; UDec, wild seringa and other deciduous species with foliage defended by tannins; UEvg, evergreen shrubs with high tannin contents in their leaves. Note that most acacias lose their leaves quite early in the dry season, while the broad-leaved legumes (unpalatable deciduous) produce new leaves earliest preceding the start of the wet season.
Figure 7.3
Dry season aspects. (A) Bushwillow trees bare of leaves, Kruger NP; (B) mopane trees bare of leaves, Luangwa Valley, Zambia; (C) bared soil in central Kruger by end of dry season, September 2016; (D) eastern short grass plains in the dry season, Serengeti, Tanzania; (E) thatch grass grassland grazed down to stubble in Maasai Mara NR, February 2017; (F) bared soil and leafless trees, Chobe riverfront, northern Botswana; (G) bare ground and trees during the 1983 drought, Kruger NP; (H) leafless mopane and bare soil during drought, near Kruger NP.
Because trees commence their seasonal growth of leaves and shoots ahead of the rains, their annual production of these tissues is not determined by the current season’s rainfall.13 However, leaves may be shed earlier in drier years, and there may be carryover effects on the commencement of growth in the following wet season, mediated by soil water reserves but buffered by moisture stored within stems.
Rooting Patterns
Rooting patterns are a fundamental indication of niche distinctions among tree species related to resource capture, but are exposed only after considerable effort in digging.14,15,16 Some trees extend shallow roots outwards well beyond their leaf canopies, competing with grass roots for both moisture and mineral nutrients. Others extend taproots downwards several metres to access water that persists longer at greater depths, particularly in sandy soils. Most acacias have well-developed lateral root systems, coupled with taproots in some cases. Umbrella thorn has contorted surface roots enabling it to draw water from a wide area. Leadwood trees have exceptionally large taproots and relatively small lateral roots. Red bushwillow (Combretum apiculatum), a small tree widely prevalent on shallow sandy or rocky soils in southern Africa, has mainly an extensive lateral root system. Russet bushwillow (C. hereroense), commonly found low down on catenas where soils are more clayey, has several descending roots as well as abundant lateral roots. Silver clusterleaf (Terminalia sericea), common in sandy soils both on crests and further down bordering seep zones, has roots extending outwards from the central rootstock. Marula trees are comparatively deep-rooted. Shepherd’s tree, evergreen despite being prevalent in dry environments, has a large taproot but lacks lateral roots. It exhibits the deepest root depth recorded for any tree: 68 m.17 Second in line for rooting depth is giraffe thorn, with a taproot penetrating down to 60 m. Trees associated with deep sandy soils, such as Zimbabwe teak and msasa, also exhibit a comparatively deep root distribution.18 A teak seedling can produce a taproot extending 1.5 m deep within one year while the above-ground shoot at that stage remains under 15 cm tall.19 In contrast, wild seringa has predominantly shallow roots. Shrubs growing on sandy sediments such as sand camwood (Baphia massaiensis) and Zambezi jessebush (Combretum celastroides) are also relatively shallow-rooted, with 95 percent of their roots less than a metre deep. Nevertheless, they have proportionately more biomass below ground than most trees, and some have swollen lignotubers facilitating stem regeneration after fires. Trees with shallow roots are subject to greater mortality during severe droughts.16 On firmer, more clay-rich soils, roots of both trees and shrubs are restricted to shallower rooting depths.20,21 Mopane trees have mostly shallow roots because they grow in arid regions where little moisture penetrates deeper.22 Trees with deep taproots still depend mostly on water in the topsoil layer for nutrient capture and thus do not evade competition with grasses during the wet season.23
Fruits
Savanna trees generally start producing fruits when around 20 percent of their potential height and reach maximum production once they attain 67 percent or more of maximum size.24 The predominance of leguminous trees means that fruits mainly take the form of dry pods in both fine-leaved and broad-leaved savannas. Nevertheless, certain acacias drop their ripened pods early in the dry season and these are consumed by various ungulates, including elephants,25 dispersing the seeds contained far and wide. Seeds that have not passed through the gut of an ungulate germinate less well. Marula produce abundant crops of plum-like drupes dropped on the ground and so available for consumption late in the wet season (Figure 7.4A). Antelope spit out their hard seeds while ruminating in the shade some distance away. Elephants travel long distances to feed on marula fruits and play a major role in seed dispersal. Various species of monkey orange (Strychnos spp.) produce large hard-shelled fruits with fleshy interiors late in the dry season when few other fruits are available (Figure 7.4E). These are sought out by primates (including humans) as well as ungulates. However, elephants seem not to eat them, perhaps because their powerful jaws would crush the seeds, releasing toxic alkaloids. Baobab trees (Figure 7.4F) produce huge pods containing seeds buried in a dry pulp, which are shed during the dry season. Humans are again among the primates consuming them. Small berries are produced by jackalberry, bird plum (Berchemia spp.), raisin bush (Grewia spp.), gwarrie (Euclea spp.) and snow-berry (Flueggia virosa) shrubs and consumed by birds, primates and, along with leaves, by ungulates.
Figure 7.4
Savanna fruits eaten by humans as well as baboons and antelope. (A) Drupes from widespread marula tree, South Africa, dropped when ripe; (B) mobola plums, Zambia, featuring in miombo woodlands; (C) mahobohobo fruits, Zambia, also featuring in miombo woodlands; (D) sour plums (Ximenia caffra), South Africa; (E) monkey orange fruits with a hard shell and soft pulp inside, South Africa, ripening during the dry season; (F) baobab pods, gathered by people as well as eaten by baboons.
(photo: Sarah Venter)
The pods produced by members of the Caesalpinoideae break open to scatter their seeds, which are dispersed only a few metres beyond the parent trees. Nevertheless, miombo woodlands also contain trees producing a rich variety of succulent fruits. These include drupes of mobola plum (Parinari curatellifolia), mahobohobo or sugar plum (Uapaca spp.) and large waterberry (Syzigium guineaense), which all typically ripen late in the dry season (Figure 7.4B,C).
Species in the Combretaceae produce dry winged pods that are not attractive to ungulates, although they are occasionally eaten. They are evidently adapted for wind dispersal over short distances. Other large fruits remain puzzling with regard to the role of herbivores in seed dispersal. Sausage trees produce huge woody pods sometimes eaten by hippos, giraffes and black rhinos, but not by much else. Wild gardenias (Gardenia volkensii) bear hard round fruits the size of a cricket ball supposedly eaten occasionally by baboons and monkeys as well as antelope.
Resistance to Fire
Juvenile woody plants are vulnerable to having all of their top growth burnt by recurrent fires while they remain below 2–3 m in height.26 Because most savanna species can resprout from buds in the root bole or stem, they are rarely killed by fires once they have built up sufficient root reserves to restore the incinerated parts. Saplings of savanna trees progressively develop a thickened bark providing protection for their woody stems, and resistance to fire depends also on stem thickness relative to height and bark moisture content.27 Vulnerability to top-kill decreases sharply with increasing height in the range 1–4 m. Shrubs gain some protection against losing all of their above-ground biomass by having multiple stems. Woody seedlings are vulnerable to being killed before they have established sufficient underground reserves. Beyond this stage, fires only temporarily set back the growth of trees and shrubs unless burns are both frequent and especially hot. This occurs mainly where there is little grazing pressure on the grass layer. Nevertheless, even quite large trees can be killed if bark damage allows fire to burn into their trunks. Trees typical of miombo woodlands regenerate via profuse stem coppicing after being burnt back.
Trees with evergreen or semi-evergreen foliage suffer a greater setback than deciduous species when their potentially long-lasting leaves get burnt. These species occupy sites where they are less exposed to fire because of the lack of much grass cover, flanking rivers and on termite mounds. The shade cast by the spreading canopies of savanna trees favours grass types less likely to burn, conferring some protection against fires once established. Woody plant species typical of forests or thickets are much more vulnerable to mortality from fire than fire-adapted savanna trees and shrubs.28
Plant Anti-herbivore Defences
Plants can have much of their energy-gathering potential removed through consumption of their leaves by herbivores. They can inhibit consumption by incorporating secondary chemicals not involved in leaf metabolism in their foliage. For trees, these anti-herbivore defences commonly take the form of carbon-based secondary metabolites like tannins or other phenolics, although terpenoids, alkaloids and toxic amino acids may also come into play.29 Trees growing on nutrient-deficient soils potentially have a surplus supply of carbon, which continues to be captured while stomata remain open to draw up scarce mineral nutrients.30 Little extra cost may be entailed in elaborating the surplus carbon into chemical deterrents. Biochemically, tannins form complexes with proteins, thereby blocking the enzymatic action of these proteins. The prime adaptive value of plant tannins may be to protect leaves from fungal or bacterial pathogens. They also restrict rates of decomposition of fallen leaves, thereby helping retain nutrients against leaching. However, tannins also interfere with the digestion of leaves by large herbivores, both by complexing with plant proteins and by interfering with the activity of cellulolytic enzymes secreted by gut microbes. Herbaceous plants growing under shade canopies of trees or grasses where carbon capture is restricted tend to be defended against herbivory by nitrogen-based compounds like alkaloids or non-protein amino acids.
The anthocyanins responsible for the red spring colours of the leaves of miombo woodland trees may be precursors to the formation of tannins in these leaves. Woody plants associated with dystrophic savannas generally have higher tannin or phenolic contents in their leaves than those growing in more fertile soils.29,31 Plants growing in fertile soils have less surplus carbon relative to nutrients so that producing carbon-based chemical deterrents would be at the cost of growth. Thus, few of the acacias prevalent in dry/eutrophic savannas contain much in the way of tannins. They rely mainly on structural defences to restrict leaf losses (Figure 7.5).32 However, structural defences in the form of thorns, spines or prickles on stems are not effective against insect herbivores.
Smaller antelopes are less restricted by thorns than larger ones because they can nibble leaves between the thorns. Giraffes overcome the effects of spines by stripping leaves from branch tips of acacias with their lower incisor teeth. Juvenile trees may deter browsing by developing a cage structure coupling prominent spines with short internodes, protecting inner leaves (Figure 7.5H).33,34 Spinescent trees and shrubs experience less leaf loss than plants lacking such defences.35 Many acacia species produce longer, thicker or denser spines when exposed to browsing than when protected from browsers.36 Shrubs with evergreen leaves attracting browsing during the dry season produce especially formidable spines (Figure 7.5E).
Figure 7.5
Plant structural defences against browsers. (A) Conspicuously white straight spines of sweet thorn; (B) prominent straight spines of ant-gall thorn (A. drepanolobium); (C) mixed straight and hooked spines of umbrella thorn; (D) hooked entangling spines formed by knob thorn; (E) formidable branched thorns of num-num (Carissa bispinosa), an evergreen shrub; (F) sharp branchlet tips of sickle-bush (Dichrostachys cinerea); (G) spiny knobs formed on stem of young knob thorn; (H) prickly cage formed by sweet thorn spines.
The chemicals that serve as deterrents against large mammals may not be effective against insect herbivores, and vice versa. At Nylsvley, some of the tree species with high tannin contents in their leaves were subject to defoliating outbreaks of caterpillars. Those containing aromatic oils showed little leaf damage from insects, but were readily consumed by ungulates.29 None of the acacias incurred much leaf loss to insect herbivores, despite being most nutritious in terms of leaf protein content.
The chemical and structural defences presented by trees and shrubs generally do not prevent browsing. Nevertheless, they adequately restrict rates of leaf loss and impose metabolic costs on herbivores that influence the diet choice of the latter.37 At Nylsvley, all of the 60 woody species present in the study enclosure were eaten by kudus at some stage. Even those containing high tannin levels were nibbled when little else remained to eat. However, certain forbs and geoxyles were never eaten, suggesting that they contain more potent toxins, like the monofluoracetatic acid present in gifblaar (‘poison-leaf’ Dichapetalum cymosum; Figure 6A.5D), which acts as a heart poison by disrupting energy metabolism. Cattle and goats die after eating just a few leaves, but wild ungulates seem somehow to know to avoid it, even hand-reared impalas.
Grass Growth
Grasses retain living biomass underground in the form of root tissues and buds. Thus, while their above-ground stems and leaves are removed by dry season fires, these parts grow back during the start of the following wet season. Hence their growth patterns are adapted to cope with seasonality in rainfall and recurrent fires. The primary buds (meristems) that generate above-ground growth are low down in the crown, protected from fire and herbivory (Figure 7.6). Single or multiple stems or tillers extend upwards from the crown with attached leaves. Tufted or bunch grasses are formed by clusters of tillers, each functioning largely independently. Eventually stems rise beyond the last leaf to produce inflorescences in the form of clusters of tiny flowers and later seeds. Some grasses extend their stems laterally along the ground in the form of stolons, with roots developing where nodes make contact with the soil. Roots underground typically constitute over 60 percent of the peak biomass of grasses, with the finest roots concentrated in the top 10 cm of the soil. During the course of the year, most of the grass leaves above-ground not consumed by herbivores get transformed into standing dead tissues or ‘necromass’ (Figure 7.7).
Figure 7.6
Structure of a grass plant.
Reproduced with permission from the University of Missouri.
Figure 7.7
(A) Seasonal dynamics of grass parts above-ground partitioned between green and brown (derived from Grunow et al. (1980) Journal of Ecology 68:877–889); (B) grassland on fertile clay soils in Maasai Mara Reserve, Kenya, subject to grazing by abundant wild herbivores.
(derived from Boutton et al. (1988a) African Journal of Ecology 26:89–101)
Grasses generate above-ground biomass during the course of the wet season either by extending tillers or by producing new tillers. Individual leaves rarely last longer than a few weeks, while tillers typically persist for 1–2 years. This means that there is ongoing turnover of the above-ground parts, so that the amount of biomass accumulated by the end of the wet season is less than the total biomass actually produced. Stems remaining from the preceding growing season hamper the growth of new tillers by shading them. Accordingly, bunch grasses depend on regular fires or grazing to enable their annual growth. More than 80 percent of the root biomass of grasses can turn over annually in tropical grasslands as some fine roots extend their growth towards sources of mineral nutrients, while others die back.38
Annual growth of the above-ground parts of grasses is directly controlled by rainfall in savannas. Grasses grow while the soil remains sufficiently moist so that growth may be pulsed in response to rainfall events during the course of the wet season. Grasses growing on sandy soils derived from granite retain green leaves longer into the dry season than grasses growing on clayey basaltic soils.39 There is an almost linear relationship between the annual accumulation of above-ground biomass by grasses and the wet season rainfall total below a MAR of around 1000 mm.40 Carbohydrates and other nutrients stored in the crown and roots enable the above-ground tissues to be regenerated at the start of each wet season.41
C4 Versus C3 Photosynthesis
Photosynthesis takes place in the chloroplasts present within leaves, which absorb infrared radiation and reflect green wavelengths to energise carbon fixation. Tropical grasses have evolved a special mechanism to concentrate carbon dioxide at the sites where rubisco, the enzyme catalysing photosynthesis, is lodged. Carbon capture is mediated via a four-carbon molecule rather than simply by the three-carbon molecule used by temperate grasses and most other plants.42 The C4 mechanism effectively acts as a pump drawing in carbon dioxide to the sites in bundle sheath cells where photosynthesis takes place. This enhances the efficiency of carbon gained relative to the water and nutrient resources needed, compared with grasses employing the C3 pathway, despite the extra energy required. It also helps overcome the limitation on growth rates posed by low atmospheric concentrations of carbon dioxide. Accordingly, tropical C4 grasses are able to grow rapidly in hot and dry climates during periods while soil moisture remains adequately available. While C4 grasses can grow faster than C3 trees and shrubs while water remains available, they close their stomata sooner than woody plants when soils become dry. The evolution of this biochemical mechanism among tropical grasses was a major factor in the expansion of savanna grasslands during the Miocene, as will be explained in Chapter 9.
Some succulents have evolved the C4 photosynthesis pathway, with a difference. They open their stomata at night, incorporate CO2 in a four-carbon molecule, then continue with photosynthesis during daylight using the captured carbon.43 This is called the CAM (crassulacean acid metabolism) pathway, because it was originally identified in the family Crassulaceae. Rather than increasing growth rate, this biochemical adaptation, along with moisture retained in leaves or stems, enables succulents to continue growing during prolonged periods without rainfall.
Grass Adaptations to Grazing
Grasses may either deter herbivory by their high fibre contents, governed largely by stem:leaf ratios, or restrict tissue losses by keeping their leaves close to the ground. If ungrazed, or unburnt, erect bunch grasses shade out shorter grasses. Locally concentrated grazing can enable the spread of low-growing grasses to form grazing lawns (Figure 6A.4F–I).44 Some grass species have the capacity to grow either erect or decumbent.
Mega-grazers, in the form of white rhinos and hippos, are the archetypal lawn-mowers, using their broad lips to crop grasses evenly.45 Wildebeest equipped with broader muzzles than other ruminants can also maintain grazing lawns.46 Smaller grazers like impalas and gazelles selectively nibble individual leaves, thereby exerting a more detrimental impact on grass plants than larger herbivores. However, they can contribute to maintaining grazing lawns once these have been formed.47 Lawn-forming grasses tend to predominate on uplands where soils are shallower and around termitaria.48 Grazing concentrated on more palatable grass species can lead to the replacement of these grasses by less nutritious types of grasses, which are called ‘increasers’ in the range management literature.
The extent of the grazing lawns cultivated by white rhinos facilitated my wanderings through the acacia parkland of the Hluhluwe-iMfolozi Park. Traversing the taller grasslands in the wetter north of the park, sopping wet with dew on summer mornings, loaded with ticks and potentially concealing dangerous animals, was much more inhibiting. Similar experiences must have been shared by ancestral humans, considered in Part IV of this book.
How Grasses Overrule Trees
Grass roots are vastly more abundant than tree roots in the upper 20 cm of savanna soils, except under tree canopies.49 Rainfall seldom penetrates much deeper than this, especially in clay-rich soils,50 and nutrient release occurs mostly in the upper topsoil. Numerous experiments show how woody seedlings grow more slowly and survive less well in the presence of grasses than they do following grass removal. 28,51,52,53 This shows how grasses out-compete trees for access to water in the topsoil zone. Tree seedlings need to gain access to sufficient soil moisture both during the season of their germination and through the next growing season in order to become established.51 To do so, they must occupy a gap within the dense mat of shallow roots established by grasses. Gaps can arise where grass tufts have died during droughts or following severe fires, or where some animal has dug up the soil to get at a bulb. Savanna grasslands have more gaps than montane or Highveld grasslands, as a consequence of their more erratic rainfall.54 Within Highveld grasslands, deeper penetration by tree roots can be blocked by the hardpan layer typically present beneath the topsoil.55,56 Tree roots are relatively more abundant than grass roots only beyond a depth of around 80 cm, i.e. in fairly deep soils. Savanna tree seedlings must succeed in extending their taproots to deeper levels in the soil, where moisture reserves remain, before growing much in height, in order to evade intense competition from grass roots.
Juvenile trees can potentially access soil water not taken up by grasses by exploiting windows in time. By putting out their new leaves before the first rains, woody plants can draw on soil moisture provided by the early rains, producing a pulse of nutrient release, before the grasses commence growth.57,58 Juvenile trees can also retain their leaves longer into the dry season, thereby extending their growth after grasses have mostly become dormant.59,60 By coupling spatial partitioning in rooting depth with temporal separation in growth, woody plants may bypass the superior rates of extraction of soil water by C4 grasses, especially in drier savannas, when circumstances allow.
However, the difference in rooting depth between trees and grasses diminishes with increasing rainfall.61 Grasses generally have shallower roots than trees, but trees become less deep-rooted with increasing rainfall.62 In wetter savannas, water saturation in the topsoil can be more than sufficient for the needs of both trees and grasses through much of the wet season. In drier savannas, established trees can gain water from deeper soil levels by hydraulic lift, although this mechanism also benefits grasses.63,64 Earlier leaf flush may benefit trees primarily via access to the pulse of mineral nutrients released by the first soil wetting, especially in infertile savannas where nutrients are especially precious.
The grass cover restricts the establishment of juvenile trees further by promoting repeated fires, which are more intense in moist savannas because of the greater grass biomass produced, and thus more likely to incinerate small trees. A multi-year interval in time between fires is needed for juvenile trees to escape the flame zone.
Tree establishment can thus depend on chance opportunities in space for water abstraction coupled with windows in time, governed by the seasonal hydrology of soil moisture. Water remains available for longer at greater depths, especially in places where soils are sandy and in lowlands where soils are deeper, unless water logging occurs. The tree cover thus depends on landscape position and on soil texture and depth in addition to the local rainfall.
Despite variation in the annual growth produced by trees and grasses, one fundamental reality remains. Within the savanna biome, tree canopies generally cover less than 40 percent of the land surface, so that the grasses growing between and beneath them capture more solar radiation and hence potentially produce more plant biomass than all of the trees. Trees get prime attention because of the capital accumulation presented by their prominent stems and branches, but these are constituted mostly by accumulated deadwood rather than metabolically active biomass. The grasses operate more nimbly, particularly underground, in response to erratic rainfall and pulsed releases of mineral nutrients. They are better adapted to resist the destructive effects of fires as well as defoliation by large herbivores.
Overview
Grasses restrict the presence of trees in savannas through being superior competitors for soil moisture, thereby retarding the growth of juvenile trees and holding the latter in the flame zone for longer. Grasses with the C4 photosynthetic pathway can grow rapidly in hot and periodically dry environments while water remains available, overcoming limitations on growth posed by low atmospheric levels of carbon dioxide. Trees may coexist by tapping into water at deeper depths, thereby extending their growth periods. Earlier seasonal growth by trees enables them to capture the pulse of mineral nutrients released by the first rains before grasses start growing. Juvenile trees need sufficient time to extend taproots and establish space for their surface roots amid the mat of grass roots. At a later stage, tree growth switches from gaining height towards spreading the leaf canopy, which ameliorates fire intensity in the shade cast. Between-tree competition contributes to maintaining an open woody canopy in drier savannas, while allowing ample space for grasses to coexist beneath and between trees. Trees tend to get excluded where soils are shallow and hardpans restrict water infiltration. This limitation contributes to the treeless grasslands prevalent in the South African Highveld and on upland surfaces elsewhere in Africa. Woodlands or even forests develop in lowlands and elsewhere where soils are deeper and retain moisture longer, except where seasonal waterlogging restricts rooting depth. The annual production of grass biomass above-ground is controlled largely by the wet season rainfall total, while the annual production of leaves and shoots by woody plants is affected little by the current year’s rainfall.
The trees that predominate in South American savannas are mainly evergreen,12 and the eucalypts that typify Australian savannas are also mostly evergreen. The evergreen habitat is believed to aid nutrient conservation in regions with extremely infertile soils, avoiding nutrients lost when leaves are shed. It could also be advantageous in dry regions where rainfall and hence opportunities for plant growth are highly unpredictable, as in much of Australia. In tropical Asia, most of the trees growing in savanna-like woodlands are deciduous, but shed their leaves quite late during a brief dry season. Fires generally have a lesser influence on the tree cover than in Africa, perhaps because more rainfall is received during the dry season.
Several of the grasses prevalent in South America cerrado exhibit the C3 pathway. Relative to African grasses, South American grasses have slower growth rates and appear less efficient in dry matter produced per unit amount of nitrogen or phosphorus absorbed.65 They are generally less tolerant of defoliation than African grasses and less digestible.66 Some introduced African grasses have become aggressive invaders in South America, including signal grasses (Brachiaria spp.), red thatching grass (Hyparrhenia rufa) and Guinea grass, probably due to the fertilisation provided by agriculture. Blue buffalo grass has spread widely in Australia, but needs fertiliser applications to be maintained. Native Australian grasses tend to be particularly low in nitrogen content, although readily digestible.65 Tussock-forming Mitchell grasses (Astrebla spp.) form the nearest approach to sweetveld.
However, we need to look beyond immediate or annual growth and consider changes in the populations of the individual plant species contributing to the vegetation cover over multi-annual periods. The comparative life-history patterns of trees and grasses in the context of recurrent fires will be the subject of the next chapter.
SUGGESTED FURTHER READING
Gibson, DJ. (2009) Grasses and Grassland Ecology. Oxford University Press, Oxford.
McNaughton, SJ. (1983) Serengeti grassland ecology: the role of composite environmental factors and contingency in community organization. Ecological Monographs 53:291–320.
O’Connor, TG; Bredenkamp, GJ. (1997) Grassland. In Cowling, RM, et al. (eds) Vegetation of Southern Africa. Cambridge University Press, Cambridge, pp. 215–257.
REFERENCES
1.Stevens, BM, et al. (2018) Mycorrhizal symbioses influence the trophic structure of the Serengeti. Journal of Ecology 106:536–546.
2.Augustine, DJ; McNaughton, SJ. (2004) Temporal asynchrony in soil nutrient dynamics and plant production in a semiarid ecosystem. Ecosystems 7:829–840.
3.Higgins, SI, et al. (2015) Feedback of trees on nitrogen mineralization to restrict the advance of trees in C4 savannahs. Biology Letters 11.
4.Wigley, BJ, et al. (2013) What do ecologists miss by not digging deep enough? Insights and methodological guidelines for assessing soil fertility status in ecological studies. Acta Oecologica – International Journal of Ecology 51:17–27.
5.Kgope, BS, et al. (2010) Growth responses of African savanna trees implicate atmospheric [CO2] as a driver of past and current changes in savanna tree cover. Austral Ecology 35:451–463.
6.Higgins, SI, et al. (2011) Is there a temporal niche separation in the leaf phenology of savanna trees and grasses? Journal of Biogeography 38:2165–2175.
7.February, EC; Higgins, SI. (2016) Rapid leaf deployment strategies in a deciduous savanna. PLoS One 11:e0157833.
8.Frost, P. (1996) The ecology of miombo woodlands. In Campbell, B (ed.) The Miombo in Transition: Woodlands and Welfare in Africa. Centre for International Forestry Research, Bogor, India, pp. 11–57.
9.Rutherford, MC. (1983) Growth rates, biomass and distribution of selected woody plant roots in Burkea africana–Ochna pulchra savanna. Vegetatio 52:45–63.
10.Boaler, S. (1966) Ecology of a miombo site, Lupa North Forest Reserve, Tanzania: II. Plant communities and seasonal variation in the vegetation. The Journal of Ecology 54:465–479.
11.Ryan, CM, et al. (2017) Pre-rain green-up is ubiquitous across southern tropical Africa: implications for temporal niche separation and model representation. New Phytologist 213:625–633.
12.Eamus, D. (1999) Ecophysiological traits of deciduous and evergreen woody species in the seasonally dry tropics. Trends in Ecology & Evolution 14:11–16.
13.Rutherford, MC; Panagos, MD. (1982) Seasonal woody plant shoot growth in Burkea africana–Ochna pulchra savanna. South African Journal of Botany 1:104–116.
14.Cole, MM; Brown, R. (1976) The vegetation of the Ghanzi area of western Botswana. Journal of Biogeography 3:169–196.
15.O’Donnell, FC, et al. (2015) A quantitative description of the interspecies diversity of belowground structure in savanna woody plants. Ecosphere 6:1–15.
16.Zhou, Y, et al. (2020) Rooting depth as a key woody functional trait in savannas. New Phytologist 227:1350–1361.
17.Canadell, J, et al. (1996) Maximum rooting depth of vegetation types at the global scale. Oecologia 108:583–595.
18.Holdo, RM; Timberlake, J. (2008) Rooting depth and above-ground community composition in Kalahari sand woodlands in western Zimbabwe. Journal of Tropical Ecology 24:169–176.
19.Seymour, CL. (2008) Grass, rainfall and herbivores as determinants of Acacia erioloba (Meyer) recruitment in an African savanna. Plant Ecology 197:131–138.
20.Seghieri, J. (1995) The rooting patterns of woody and herbaceous plants in a savanna; are they complementary or in competition? African Journal of Ecology 33:358–365.
21.Case, MF, et al. (2020) Root–niche separation between savanna trees and grasses is greater on sandier soils. Journal of Ecology 108:2298–2308.
22.Smit, G; Rethman, N. (1998) Root biomass, depth distribution and relations with leaf biomass of Colophospermum mopane. South African Journal of Botany 64:38–43.
23.Verweij, RJT, et al. (2011) Water sourcing by trees in a mesic savanna: responses to severing deep and shallow roots. Environmental and Experimental Botany 74:229–236.
24.Shackleton, CM. (1997) The prediction of woody productivity in the savanna biome, South Africa. PhD thesis, University of the Witwatersrand, Johannesburg.
25.Dudley, JP. (2000) Seed dispersal by elephants in semiarid woodland habitats of Hwange National Park, Zimbabwe. Biotropica 32:556–561.
26.Bond, WJ; Keeley, JE. (2005) Fire as a global ‘herbivore’: the ecology and evolution of flammable ecosystems. Trends in Ecology & Evolution 20:387–394.
27.Higgins, SI, et al. (2012) Which traits determine shifts in the abundance of tree species in a fire-prone savanna? Journal of Ecology 100:1400–1410.
28.Bond, WJ, et al. (2017) Demographic bottlenecks and savanna tree abundance. In Cromsigt JPG, et al. (eds) Conserving Africa’s Mega-Diversity in the Anthropocene. Cambridge University Press, Cambridge, pp. 161–188.
29.Owen-Smith, N. (1993) Woody plants, browsers and tannins in southern African savannas. South African Journal of Science 89:505–510.
30.Coley, PD, et al. (1985) Resource availability and plant antiherbivore defense. Science 230:895–899.
31.Wigley, BJ, et al. (2018) Defence strategies in African savanna trees. Oecologia 187:797–809.
32.Cooper, SM; Owen-Smith, N. (1986) Effects of plant spinescence on large mammalian herbivores. Oecologia 68:446–455.
33.Charles‐Dominique, T, et al. (2017) The architectural design of trees protects them against large herbivores. Functional Ecology 31:1710–1717.
34.Wigley, BJ, et al. (2019) A thorny issue: woody plant defence and growth in an East African savanna. Journal of Ecology 107:1839–1851.
35.Skarpe, C, et al. (2000) Browsing in a heterogeneous savanna. Ecography 23:632–640.
36.Wigley, BJ, et al. (2015) Mammal browsers and rainfall affect acacia leaf nutrient content, defense, and growth in South African savannas. Biotropica 47:190–200.
37.Owen-Smith, N; Cooper, SM. (1987) Palatability of woody plants to browsing ruminants in a South African savanna. Ecology 68:319–331.
38.Lauenroth, WK; Gill, R. (2003). Turnover of root systems. In de Kroon, H; Visser, EJW (eds) Root Ecology. Springer, Berlin, pp. 61–89.
39.Mutanga, O, et al. (2004) Explaining grass–nutrient patterns in a savanna rangeland of southern Africa. Journal of Biogeography 31:819–829.
40.Rutherford, MC. (1980) Annual plant production–precipitation relations in arid and semi-arid regions. South African Journal of Science 76:53–57.
41.O’Connor, TG. (1994) Composition and population responses of an African savanna grassland to rainfall and grazing. Journal of Applied Ecology 31:155–171.
42.Osborne, CP; Freckleton, RP. (2009) Ecological selection pressures for C-4 photosynthesis in the grasses. Proceedings of the Royal Society B – Biological Sciences 276:1753–1760.
43.Ehleringer, JR; Monson, RK. (1993) Evolutionary and ecological aspects of photosynthetic pathway variation. Annual Review of Ecology and Systematics 24:411–439.
44.Hempson, GP, et al. (2015) Ecology of grazing lawns in Africa. Biological Reviews 90:979–994.
45.Olivier, RCD; Laurie, WA. (1974) Habitat utilization by hippopotamus in the Mara River. African Journal of Ecology 12:249–271.
46.McNaughton, SJ. (1984) Grazing lawns: animals in herds, plant form, and coevolution. The American Naturalist 124:863–886.
47.Waldram, MS, et al. (2008) Ecological engineering by a mega-grazer: white rhino impacts on a South African savanna. Ecosystems 11:101–112.
48.Van der Plas, F, et al. (2013) Functional traits of trees on and off termite mounds: understanding the origin of biotically‐driven heterogeneity in savannas. Journal of Vegetation Science 24:227–238.
49.Hipondoka, MHT, et al. (2003) Vertical distribution of grass and tree roots in arid ecosystems of Southern Africa: niche differentiation or competition? Journal of Arid Environments 54:319–325.
50.Knoop, WT; Walker, BH. (1985) Interactions of woody and herbaceous vegetation in a southern African savanna. The Journal of Ecology 9:235–253.
51.Cramer, MD, et al. (2012) Belowground competitive suppression of seedling growth by grass in an African savanna. Plant Ecology 213:1655–1666.
52.Stevens, N, et al. (2018) Transplant experiments point to fire regime as limiting savanna tree distribution. Frontiers in Ecology and Evolution 6:137.
53.Morrison, TA, et al. (2019) Grass competition overwhelms effects of herbivores and precipitation on early tree establishment in Serengeti. Journal of Ecology 107:216–228.
54.Wakeling, JL, et al. (2015) Grass competition and the savanna-grassland ‘treeline’: a question of root gaps? South African Journal of Botany 101:91–97.
55.Tinley, K. (1982) The influence of soil moisture balance on ecosystem patterns in southern Africa. In Huntley, BJ; Walker, BH (eds) Ecology of Tropical Savannas. Springer, Berlin, pp. 175–192.
56.O’Connor, TG; Bredenkamp, GJ. (1997) Grassland. In Cowling, RM, et al. (eds) Vegetation of Southern Africa. Cambridge University Press, Cambridge, pp. 215–257.
57.Guan, K, et al. (2014) Terrestrial hydrological controls on land surface phenology of African savannas and woodlands. Journal of Geophysical Research: Biogeosciences 119:1652–1669.
58.February, EC, et al. (2019) Physiological traits of savanna woody species: adaptations to resource availability. In Scogings, PF; Sankaran, M (eds) Savanna Woody Plants and Large Herbivores. Wiley, Oxford, pp. 309–329.
59.Novellie, P. (1989) Tree size as a factor influencing leaf emergence and leaf fall in Acacia nigrescens and Combretum apiculatum in the Kruger National Park. Koedoe 32:95–99.
60.Shackleton, CM. (1999) Rainfall and topo‐edaphic influences on woody community phenology in South African savannas. Global Ecology and Biogeography 8:125–136.
61.Schenk, HJ; Jackson, RB. (2002) Rooting depths, lateral root spreads and below-ground/above-ground allometries of plants in water-limited ecosystems. Journal of Ecology 90:480–494.
62.Holdo, RM, et al. (2018) Rooting depth varies differentially in trees and grasses as a function of mean annual rainfall in an African savanna. Oecologia 186:269–280.
63.Dohn, J, et al. (2013) Tree effects on grass growth in savannas: competition, facilitation and the stress‐gradient hypothesis. Journal of Ecology 101:202–209.
64.Priyadarshini, KVR, et al. (2016) Seasonality of hydraulic redistribution by trees to grasses and changes in their water‐source use that change tree–grass interactions. Ecohydrology 9:218–228.
65.Fisher, MJ, et al. (1996) Grasslands in the well-watered tropical lowlands. In Hodgson, J; Illius, AW (eds) The Ecology and Management of Grazing Systems. CAB International, Wallingford, pp. 393–425.
66.Simoes, M; Baruch, Z. (1991) Responses to simulated herbivory and water stress in two tropical C 4 grasses. Oecologia 88:173–180.