Norman Owen-Smith

Effects of plant spinescence on large mammalian herbivores

SummaryPlant thorns and spines had these effects on the feeding behaviour of the three species of browsing ungulate that we studied, kudu, impala and domestic goats: (i) bite sizes were restricted, in most cases to single leaves or leaf clusters; (ii) hooked thorns retarded biting rates; (iii) the acceptability of those plant species offering small leaf size in conjunction with prickles was lower, at least for the kudus, than those of other palatable plant species; (iv) the inhibitory effect of prickles on feeding was much less for the smaller impalas and goats than for the larger kudus; (v) from certain hook-thorned species the kudus bit off shoot ends despite their prickles; (vi) for certain straight-thorned species the kudus compensated partially for the slow eating rates obtained by extending their feeding durations per encounter. Most spinescent species were similar in their acceptability to the ungulates to unarmed palatable species, despite higher crude protein contents in their foliage than the latter. Such structural features furthermore reduce the tissue losses incurred by plants per encounter by a large ungulate herbivore, by restricting the eating rates that the animals obtain. In this way prickles function to restrict foliage losses to large herbivores below the levels that might otherwise occur.

Effects of Severe Defoliation on the Long-Term Resistance to Insect Attack and on Leaf Chemistry in Six Woody Species of the Southern African Savanna

Severe defoliation as occurs in insect outbreaks can alter the chemistry and food value of woody-plant leaves for insects for several years after defoliation ceases. Two hypotheses, the active-defense hypothesis (AD) and the hypothesis of a carbon/nutrient balance (CNB), attempt to explain the mechanism of such responses to defoliation. We tested these hypotheses by studying the responses to severe defoliation by six southern African savanna woody species. Manual defoliation of the three fast-growing deciduous species we studied (Grewia flavescens, Acacia tortilis, Dichrostachys cinerea) resulted in increased resistance to insect attack in the next year. Concomitantly, nitrogen (N) and phosphorus (P) concentrations in leaves decreased and leaf total phenol and condensed tannin concentrations increased. The three slowly growing species that we studied (Burkea africana, Ochna pulchra, Euclea natalensis) responded differently to manual defoliation. In the year following defoliation, leaf resistance to insect attack declined, concentrations of total phenols and condensed tannin in leaves decreased, and concentrations of N and P increased. These opposite responses are consistent with the expectations of CNB rather than AD. We also found that defoliation of B. africana by caterpillars of one of its major defoliators, the saturniid moth Cirina forda, resulted in a greater loss of resistance than did manual defoliation of B. africana. This result further suggests that the AD does not apply to B. africana. We conclude that inherently slow growth constrains the long-term responses of woody plants to severe defoliation. Most of our results indicate that the CNB is a useful working model of woody-plant responses to defoliation. However, CNBs current emphasis on evergreenness as a control over the way slowly growing woody species respond to severe defoliation cannot explain our results. Thus, we suggest that a low photosynthetic capacity is also an important control over these responses.

Condensed tannins deter feeding by browsing ruminants in a South African savanna

SummaryThe palatability of 14 species of woody plant was assessed for three species of browsing ruminant, namely kudus, impalas and goats. Results show that palatability was most clearly related to leaf contents of condensed tannins. The effect was a threshold one, with all plants containing more than 5% condensed tannins being rejected as food during the wet season period. In contrast palatability was not influenced by concentrations of protein-precipitating polyphenols, and only weakly related to contents of nitrogen, phosphorus, cations, fibre components and other secondary metabolites. Insect herbivory shows a different pattern. These findings support the hypotheses that (i) condensed tannins function to protect plant cell walls against microbial attack; (ii) hydrolyzable tannins function to inactivate the digestive enzymes of insect herbivores. Large mammalian herbivores are influenced by condensed tannins due to their dependance upon microbial fermentation of plant cell walls for part of their energy needs.

Predator–prey size relationships in an African large-mammal food web

1. Size relationships are central in structuring trophic linkages within food webs, leading to suggestions that the dietary niche of smaller carnivores is nested within that of larger species. However, past analyses have not taken into account the differing selection shown by carnivores for specific size ranges of prey, nor the extent to which the greater carcass mass of larger prey outweighs the greater numerical representation of smaller prey species in the predator diet. Furthermore, the top-down impact that predation has on prey abundance cannot be assessed simply in terms of the number of predator species involved. 2. Records of found carcasses and cause of death assembled over 46 years in the Kruger National Park, South Africa, corrected for under-recording of smaller species, enabled a definitive assessment of size relationships between large mammalian carnivores and their ungulate prey. Five carnivore species were considered, including lion (Panthera leo), leopard (Panthera pardus), cheetah (Acinonyx jubatus), African wild dog (Lycaon pictus) and spotted hyena (Crocuta crocuta), and 22 herbivore prey species larger than 10 kg in adult body mass. 3. These carnivores selectively favoured prey species approximately half to twice their mass, within a total prey size range from an order of magnitude below to an order of magnitude above the body mass of the predator. The three smallest carnivores, i.e. leopard, cheetah and wild dog, showed high similarity in prey species favoured. Despite overlap in prey size range, each carnivore showed a distinct dietary preference. 4. Almost all mortality was through the agency of a predator for ungulate species up to the size of a giraffe (800-1200 kg). Ungulates larger than twice the mass of the predator contributed substantially to the dietary intake of lions, despite the low proportional mortality inflicted by predation on these species. Only for megaherbivores substantially exceeding 1000 kg in adult body mass did predation become a negligible cause of mortality. 5. Hence, the relative size of predators and prey had a pervasive structuring influence on biomass fluxes within this large-mammal food web. Nevertheless, the large carnivore assemblage was dominated overwhelmingly by the largest predator, which contributed the major share of animals killed across a wide size range.

Empirical evidence of density-dependence in populations of large herbivores

Density‐dependence is a key concept in population dynamics. Here, we review how body mass and demographic parameters vary with population density in large herbivores. The demographic parameters we consider are age‐ and sex‐specific reproduction, survival and dispersal. As population density increases, the body mass of large herbivores typically declines, affecting individual performance traits such as age of first reproduction and juvenile survival. We documented density‐dependent variations in reproductive rates for many species from the Arctic to subtropical zones, both with and without predation. At high density, a trade‐off between growth and reproduction delays the age of primiparity and often increases the costs of reproduction, decreasing both survival and future reproductive success of adult females. Density‐dependent preweaning juvenile survival occurs more often in polytocous than monotocous species, while the effects of density on post‐weaning juvenile survival are independent of litter size. Responses of adult survival to density are much less marked than for juvenile survival, and may be exaggerated by density‐dependent changes in age structure. The role of density‐dependent dispersal in population dynamics remains uncertain, because very few studies have examined it. For sexually dimorphic species, we found little support for higher sensitivity to increasing density in the life history traits of males compared to females, except for young age classes. It remains unclear whether males of dimorphic species are sensitive to male density, female density or a combination of both. Eberhardts model predicting a sequential effect of density on demographic parameters (from juvenile survival to adult survival) was supported by 9 of 10 case studies. In addition, population density at birth can also lead to cohort effects, including a direct effect on juvenile survival and longterm effects on average cohort performance as adults. Density effects typically interact with weather, increasing in strength in years of harsh weather. For some species, the synchronization between plant phenology and reproductive cycle is a key process in population dynamics. The timing of late gestation as a function of plant phenology determines whether density‐dependence influences juvenile survival or adult female reproduction. The detection of density‐dependence can be made difficult by nonlinear relationships with density, high sampling variability, lagged responses to density changes, changes in population age structure, and temporal variation in the main factors limiting population growth. The negative feedbacks of population size on individual performance, and hence on life history traits, are thus only expected in particular ecological contexts and are most often restricted to certain age‐specific demographic traits.

Demography of a large herbivore, the greater kudu Tragelaphus strepsiceros, in relation to rainfall

(1) The study was conducted from 1974 to 1984 in two areas of the Kruger National Park differing in mean rainfall. Age-class specific survival rates were estimated by individual registration of the kudus present in each female-young social unit, using natural markings. (2) The kudu populations in both study areas increased in 1974-78 then declined in 1978-83. Because of high annual variability in juvenile recruitment, no stable age distribution was attained.

Palatability of Woody Plants to Browsing Ruminants in a South African Savanna

We recorded food selection by free-ranging kudus (Tragelaphus strepsiceros), impalas (Aepyceros melampus), and domestic goats in the Nylsvley Nature Reserve in South Africa. The predominant vegetation is a deciduous wooded savanna growing on infertile sandy soils; patches of Acacia savanna occur on the more fertile sites of former human settlements. Woody species fall into two basic categories of acceptability to the animals: (1) species favored year-round; and (2) species generally rejected, except during certain periods. Among the latter, some species increase in acceptability during the dry season; others are favored temporarily while new leaves predominate. Some species remain low in acceptability year-round. We propose a palatability classification of woody species, considering features of leaf retention (deciduous vs. evergreen) and spinescence, as well as seasonal variations in acceptability. Estimated annual foliage losses to browsing ungulates were 1-3% for unpalatable deciduous species, compared with 10-30% for most palatable unarmed deciduous and evergreen species. Unpalatable deciduous species dominate the woody foliage biomass on the infertile soils, while palatable but spinescent species make up most of the leaf biomass of trees and shrubs on the more fertile sites. These patterns seem widely typical of African savanna. Plants known to have chemical defenses against vertebrate herbivory are prominent on nutrient-deficient soils, while those with structural defenses are prominent on fertile soils.

Foliage acceptability to browsing ruminants in relation to seasonal changes in the leaf chemistry of woody plants in a South African savanna

SummaryWe investigated seasonal changes in food selection by hand-reared kudus and impalas in savanna vegetation in northern Transvaal, South Africa. The acceptability of the leaves of woody plants to these animals was compared with leaf concentrations of nutrients, fibre components and old leaf phenophases. No consistently significant correlation was found between acceptability and any single chemical factor. Based on an a priori palatability classification, discriminant function analysis separated relatively palatable species from unpalatable species in terms of a linear combination of protein and condensed tannin concentrations. The high acceptability of certain otherwise unpalatable species during the new leaf phenophase was related to elevation of protein levels relative to condensed tannin contents. Species were added to the diet during the dry season approximately in the order of their relative protein-condensed tannin difference.

Foraging Responses of Kudus to Seasonal Changes in Food Resources: Elasticity in Constraints

Foraging behavior provides the mechanistic link between species perfor- mance and food resources in the environment. I document the foraging responses of a browsing ruminant, the kudu (Tragelaphus strepsiceros), to changes in food abundance and quality over the seasonal cycle and interpret the results in the context of optimal foraging theory. The study was carried out on hand-reared but free-ranging animals in a savanna region with summer wet season and winter dry season in Transvaal, South Africa. During the dry season, kudus expanded their diet to include evergreen and unpalatable deciduous woody species neglected during the wet season. The acceptability rating of woody plants was correlated with food value expressed in terms of a protein-condensed tannin index, relative to handling time during ingestion or digestion. However, dietary expansion alone was inadequate to satisfy the daily energy requirement during this period. The animals also increased the fraction of trees of palatable species that was accepted for feeding, extended feeding duration at feeding stations, and increased encounter rate with evergreen trees still retaining leaves. Both total time active and proportion of active time spent foraging increased over the dry season. Digestive capacity was apparently increased to accommodate a higher daily food intake, which compensated for reduced diet quality. Only in September at the end of the dry season, when little foliage remained, were compensatory adjustments inadequate to meet energy requirements. The functional response relating consumption rate to seasonal changes in food abundance showed little variation, despite a decline in foliage biomass by more than an order of magnitude. The kudus were neither energy maximizers, nor time minimizers, but rather targeted on their energy requirements with least overall cost. Findings demonstrate that the foraging time and digestive capacity constraints assumed in optimal diet models are somewhat elastic. Dietary predictions obtained using average parameter settings may be misleading, because constraints may become effective only under extreme conditions. Further research is needed to establish the costs associated with stretching physiological constraints towards their upper tolerance limits.

Establishing the causes of the roan antelope decline in the Kruger National Park, South Africa

Abstract Between 1986 and 1993, the roan antelope Hippotragus equinus population in the Kruger National Park, South Africa, declined from about 450 to ca.45 animals. Data from long term monitoring of herbivore numbers, sex, age and herd structure and of vegetation condition, supported by spatial and demographic modelling and predator sighting records, were used to evaluate potential causes of the population decrease, including: (1) competition from zebra and wildebeest, which moved into the roan’s range after the introduction of waterpoints; (2) increased predation following the influx of zebra and wildebeest; (3) habitat deterioration due to an extended period of low rainfall; (4) outbreaks of the disease anthrax; and (5) stress associated with immunisation against anthrax. Evidence did not support the latter two causes. Drought alone was inadequate as an explanation, although progressive desiccation over a sequence of low rainfall years, and resultant grassland deterioration, may have been contributory. Grassland recovery may have been suppressed by grazing pressure from zebra and wildebeest. The population crash was associated with an increase in adult mortality, but little apparent change in calf survival, suggesting that nutritional factors were not the prime cause. Increased predation on adult roan due to a build-up in lion numbers, following the zebra and wildebeest influx, seemed to be the proximate cause of the initial sharp decline in the roan population. However, the ultimate cause was the provision of numerous artificial waterpoints in the roan range, which attracted the influx of zebra and wildebeest, particularly during drought conditions. Following the closure of waterpoints in a section of the roan range, the roan herds in the vicinity are recovering. Deficiencies in current monitoring programmes in the Kruger Park are identified.