Conservation Evidence strives to be as useful to conservationists as possible. Please take our survey to help the team improve our resource.

Providing evidence to improve practice

Individual study: The effect of excluding vertebrate and invertebrate grazers on growth of Acacia xanthophloea seedlings in Amboseli National Park, Kenya

Published source details

Western D. & Maitumo D. (2004) Woodland loss and restoration in a savannah park: a 20 year experiment. African Journal of Ecology, 42, 111-121


Large scale loss of acacia savannah woodland, dominated by Acacia xanthophloea, has been recorded since at least 1961 in Amboseli National Park, Kenya. Several theories were proposed for the woodland loss and associated habitat changes, ranging from livestock overgrazing, game over-browsing, pathogens/pests, increasing soil salinity and climate change. Each theory assumed a different cause and by inference, remedy.

Of particular concern was collapse of elephant Loxodonto africana migration caused by heavy poaching in the 1970s in areas beyond the park boundaries. Elephants have a complex social structure that is sensitive to external abrupt changes e.g. if the alpha matriarch is taken out through poaching the group may alter dispersal behaviour, such as abandoning migration across dangerous territories. This led to a sharp rise in elepant numbers within the park resulting in increased grazing and browsing pressure. This was the main theory advanced for loss and lack of regeneration of the acia woodland and an experiment, summarised here, was set up to look at the effect of both invertebrate and vertebrate grazers on A.xanthophloea seedling growth. A second experiment was subsequently undertaken to assess A.xanthophloea seedling growth when soley elephants were excluded (see Case 244).

Study site: The study site was located in the centre of Amboseli National Park, Kenya, in an area between Ol Tukai Lodge and David Western's research house. The site was one of eighteen previously studied areas where investigations into causes of woodland decline had been undertaken in 1969 (Western & Van Praet 1973).

At this site tree canopy cover in 1969 was estimated at 17%. At this time 8% of mature trees (average height 17 m) were dead or dying. There was also no evidence of any Acacia xanthopholea recruitment since about the 1950s, based on stem diameter measurements and aerial photographs dating back to this period (Western 1973).

A series of experiments in two phases (see Case 244 for second phase) were designed to establish the cause of the decline in acacia woodland, and to establish mangement techniques to enhance woodland restoration.

Establishment of study plots: In 1981, when this was initiated, all mature acacia trees had died in the study area and no successful recruitment had occurred for about 30 years. However, acacia seedlings were observed and an experiment was designed to investigate the potential problem of grazing, by both invertebrate and vertebrate grazers, upon seedling growth. In the study area, 32 plots were randomly selected. Plot sizes varied from 0.0025 to 0.4 ha, depending on seedlings density, which averaged 92 seedlings/ha.

Phase 1 experiments: The first phase of the study was conducted in 1981 to September 1984. Forty A.xanthophloea seedlings were tagged (to allow identification) in each study plot. Twenty were protected from large herbivores and the other 20 were left exposed. Both the protected and exposed seedlings were randomly assigned to four experimental treatments with five replicates. The treatments were aimed to detect:

i) Effect of invertebrate browsers - A light atomised mist spray of Roxion insecticide was applied every two weeks to deter invertebrate browsers.

ii) Effect of vertebrate browsers - Protection of the twenty seedlings in each plot from vertebrate browsers was undertaken by installation of a large 1.6 m wide x 1.3 m deep game barrier to exclude elephants Loxodonto africana, and an inner 1.5 m 5-strand wire fence to exclude other large ungulates including impala Aepyceros melampus, zebra Equus burchelli, wildebeest Connochaetes taurinus and buffalo Syncerus caffer. The remaining twenty seedlings outside the exclosure were exposed to large herbivores.

iii) Effect of of plant competitors - Above-ground plant competition was eliminated by regularly clipping the herb-layer to ground level within a metre radius of the experimental seedlings.

iv) Interactions between these three factors - Effects of varying the three treatments above on acacia seedlings growth. These observations continued twelve months after terminating the three above treatments.

The assumption was recruitment in the exclosure and the absence of recruitment in the control plot would highlight grazers as the main reason for lack of seedling recruitment and discount other theories such as climate change, soil salinity and pathogens as holding back woodland regeneration.

Seedling measurements: Height, stem diameter, leaf length, leafiness (an index of leaf area) and signs of herbivory were recorded monthly. Large herbivore browsing was distinguished from invertebrate browsing by the presence of stripped-off whole leaves and distinctly nibbled shoots.

The growth of Acacia seedlings under the different experimental treatments and inside vs. outside a large herbivore exclosure are shown in Tables 1 and 2, and Figures 1 and 2 (reproduced by kind permission of the authors, D.Western & D.Maitumo), results are summarised below.

Effect of invertebrates: No caterpillars or other invertebrate herbivores were observed on the sprayed plants, and leaf vigour was not affected by insecticide spraying. Seedling growth was slowed by invertebrate grazing in the early stages but these plants recovered.

Seedling growth inside & outside the exclosures: By October 1983, Acacia seedlings inside the exclosure were growing vigorously, the tallest plants had reached 2.8 m and were beginning to overshadow slower growing seedlings. In 1984 seedlings inside the exclosure were too tall to warrant further insect or plant competitor treatment so only stem diameter and height measurements of seedlings protected from and exposed to herbivores were monitored to see if they reached maturity (see Tables 1 and 2, attached).

Conclusions: The herbivory hypothesis predicted that the Acacia seedlings would grow in the absence of browsers but not in their presence. The experiment was consistent with this as seedlings grew rapidly when large vertebrate browsers were excluded but were stunted in the unprotected controls. To assess if elephants were the main cause of seedling failure as opposed to other large grazers, a second experiment was undertaken (see Case 344).

Note: If using or referring to this published study, please read and quote the original paper. Do not quote as a case as this is for previously unpublished work only.