Impact of sustained drought on a semi‐aridColophospermum mopanesavanna

Abstract

Vegetation change on different substrate types was measured over 15 years in semi-arid Mopani Veld. Baseline data were collected in the terminal year (1982) of a decade of above-average rainfall, whilst the period of monitoring was characterised by below-average rainfall, with severe drought years occurring in 1988/1989 and 1991/1992. Grazing pressure was, on average, alleviated over the period of recording so that changes reflect primarily the effect of rainfall. Long-lived perennial grasses had declined markedly, annual and micro-perennial grasses had increased markedly, and the proportion of forbs had increased. Despite 7-17 years since the replacement of livestock grazing with wildlife at lower stocking rates, there was no evidence of an improvement in sward or soil condition, owing ostensibly to the effect of drought. The extent of change in botanical composition and in sward condition was influenced by substrate type. The rate of growth of the woody component had been slow: the average per annum increase in the radius of trees was 0.67 mm. For the dominant species, Colophospermum mopane, only individuals of less than 1 m in height consistently showed an increase in size. The basal area of dead standing trees had increased threefold.Complete mortality or substantial topkill of shrub species, in particular of Grewia spp., Ximenia americana, and Combretum apiculatum, and of adults of the latter, had occurred owing to drought. Shrub mortality was most pronounced on colluvial soils and sandveld. All species exhibited partial dieback of the crown. Overall, shrub recruitment has been limited except for that of H. mopane on colluvial soils but not on other substrate types, which is therefore likely to increase further its dominance on this substrate at the expense of species richness. Grazing pressure on desirable perennial grasses and browsing pressure on palatable woody plants has probably increased because of their decline, even if animal abundance has not increased, which will have a negative effect on their re-establishment. Acting in opposition, dieback of trees and shrubs has probably created a window of opportunity for the recovery of perennial grasses, but it would likely require an extended run of above-average years to reverse the changes in herbaceous vegetation, for which the availability of seed of perennial grasses is likely to be a major constraint.

Determinants of vegetation type patch dieback in a semi-arid area, Tutume Sub-District, Botswana

“…The extent of dieback of different species varied on the two soil types and that was similar to what has been reported by O'Connor (1999).…”

Impacts of high utilisation pressure on biodiversity components inColophospermum mopanesavanna

“…Although its herbaceous layer cover was relatively sparse and mainly annual, this sparse understorey with relatively little perennial grass is typical of many areas of C. mopane savanna (Werger andCoetzee 1978, Klintenberg andVerlinden 2008). However, there appears to be much variation (Siebert et al 2003, Poilecot andGaidet 2010) where perennial herbaceous species are often replaced by annuals following drought (O'Connor 1998, Buitenwerf et al 2011). The drought conditions in the few years preceding this study may have contributed to the relatively high proportion of annual plants on LU, although the drought conditions appeared to not have been broken at the time of sampling of vegetation.…”

Development and Parameterization of a Rain- and Fire-driven Model for Exploring Elephant Effects in African Savannas

“…The lowest published estimate of sapling survival, 0.78, is for P. angolensis [25]. The same authors give P. angolensis shrub survival as 0.948-0.969 [25], but again we adopt the value for C. mopane of s 6 , s 7 =0.994 [77]. For mature trees, we assumed greater tolerance of shortterm fluctuations in rainfall, only experiencing increased mortality in droughts of greater than 2 years of duration [86]:…”

“…We use s 1 =0.95 in this model. For sapling survival we use s i =0.99, 2≤i≤5, after an estimate based on 9-year C. mopane drought mortality [77]. The lowest published estimate of sapling survival, 0.78, is for P. angolensis [25].…”

“…P. angolensis adult survival has been measured as 0.944 and adult tree survival in Lamto, Cote d'Ivoire was modeled as 0.985 [45]. Survival estimates for C. mopane vary between 0.99 [13,27] and zero observed adult mortality throughout a 9-year drought [77]. We set adult tree mortality in a full-scale drought (x=1) to m r 8À9 ¼ 0:05 [86].…”