Personal tools
You are here: Home eNewsletter Archives 2017 october2017 Even no clouds have their silver lining – finding the good (grazing) in the lowveld’s recent drought
Research Publications


OUTPUTS 2006-2017

Log in

Forgot your password?

NRF logo



Even no clouds have their silver lining – finding the good (grazing) in the lowveld’s recent drought

By Dr Dave Thompson, SAEON Ndlovu Node
mail.jpg facebook.jpg

The summers of 2014/15 and 2015/16 were some of the driest and hottest on record across the subcontinent, including South Africa - with the central and eastern regions having borne the brunt of unusual back-to-back El Niño conditions.

This periodic phenomenon, linked to above-normal sea-surface temperatures in the Pacific Ocean, impacts global weather patterns.

In South Africa, El Niño is typically experienced as below-normal rainfall and above-normal temperatures during summer. According to AgriSA, 2015 was the driest year in South Africa in over a century, although some reports indicate that 2016 was drier still. Further, the summer of 2015/16 saw 36 South African Weather Service stations record their highest ever maximum temperatures.

Immediate impacts

The impacts of drought, defined by unusually low rainfall and the subsequent shortage of water, are dire. These effects are magnified when coupled with excessive heat, as any available water is quickly lost through evaporation and the physiological water demands for survival by plants and animals are increased.

From the human perspective, rain-fed agriculture (Fig. 1) and subsistence farming are negatively impacted first, with reduced or failed yields. Plant growth, particularly the production of biomass by grass in non-cultivated lands falters.

0201.jpg 0202.jpg

Figure 1. Rain-fed agriculture such as maize typically shows the first impacts of drought. Over the past three decades South Africa has produced approximately 10 million tonnes of maize annually; production dropped to just four million tonnes during the previous droughts of 1982-84 and 1991/92. Photo courtesy of

Figure 2. As drought conditions intensify, water resources and the production of forage in rangeland and commercial grazing systems become limiting. Under such conditions livestock mortality is inevitable. Photo courtesy of                                                                                                                                       

Thereafter, irrigated crops fail as water reserves are depleted, and livestock in both rangeland and commercial grazing systems eventually succumb to either hunger or thirst (Fig. 2). These anthropocentric impacts are measured in terms of livelihoods, health, and economy.

It is therefore no surprise that, in describing drought, poet Julia Ward wrote: ‘Sinister dryness is gorging all around me. Gorgeous has gone’.

So where then, amidst this sinister dryness - this suffering, is the silver lining? For that, we need to take an ecological perspective.

Much of north-eastern South Africa comprises savannas. Seasonal and yearly fluctuations in water availability and temperature are typical here, as are occasional climate extremes (e.g. droughts, floods, heat waves). Dominant land-uses are game farming, communal rangeland and conservation - including the Kruger National Park (KNP).

The drought of 2014-16 is arguably the worst experienced by the lowveld and KNP in recent history, although similarly dry conditions occurred in 1982-84 and 1991-92. In all cases the annualised rainfall approached only half of the long-term average, which for central and southern KNP is 500-550 mm (Fig. 3).

0203.jpg 0204.jpg

Figure 3. Cumulative rainfall for 2015/16 (green line) compared with the previous worst drought in KNP history (1991/92, red line) and the park-wide long-term average (blue line). These two drought events, as well as the drought of 1982-84 (not shown), saw annual rainfall totals fall short of half of the long-term average. Courtesy of SANParks.                                                                                                                                                                                                                                                                                                                         

Figure 4. Extreme heat contributed to the severity of the 2014-2016 drought. By February 2016, following back-to-back years with below-average rainfall, the number of days where maximum temperatures exceeded 40C in Phalaborwa, sited adjacent to north-central KNP, was far more than any summer since 1960/61. Similar patterns of increased maximum temperatures and an increase in the number of days experiencing extreme conditions were noted elsewhere in KNP, and from across South Africa. Courtesy of Dr Tony Swemmer, SAEON.

KNP was especially hard hit by the most recent drought because of an unprecedented number of days where maximum temperatures exceeded 40C (Fig. 4), by rivers having reduced flow, and by higher-than-previous densities of certain animal species, especially the grass-dependent grazers. The immediate impacts of this ‘perfect storm’ of circumstances in KNP were painfully obvious to those watching: the herbaceous vegetation comprising mainly the grass forage which sustains the system was first reduced, then lost altogether (Fig. 5), resulting in high grazer die-off.

0205.jpg 0206.jpg

Figure 5. Below-average rainfall and searing temperatures in the back-to-back summers of 2014-2016 effectively halted plant, especially grass growth in the savannas of the Kruger National Park. Supporting high densities of grass-dependent grazers, the landscape was quickly reduced to bare soil. Photo courtesy of Dr Izak Smit, SANParks.                                                                                                                                                                                                           

Figure 6. A conceptual framework of ecosystem response to climate extremes such as drought. Transient impacts with rapid recovery (B) are expected when only the growth response of individuals is reduced. Large impacts and prolonged recovery (C), or even a state change (D), are expected when extreme response thresholds are crossed, resulting in shifts in species composition due to species reordering or local species loss. Modified from Smith (2011), Journal of Ecology 99: 656-663.

Longer-term effects of drought and ecosystem recovery

Lessons learned from previous severe climate events have aided savanna ecologists in broadly understanding what to expect during droughts. However, due to their infrequent occurrence, extreme droughts tend to be studied opportunistically and after-the-fact. Consequently, their longer-term impacts, and the process and rate of ecosystem recovery once the rain returns to normal, are poorly understood.

Lags exist in the recovery of the vegetation, which in turn impact on herbivore distribution and numbers, and other critical savanna processes such as fire, erosion and nutrient cycling. Scientists have many theories about the process and eventual outcomes (Fig. 6) - those longer-term effects, which can be positive or negative, of extreme drought in savannas, but have little data and have had little opportunity to test these ideas.

The dissipation of the 2014-2016 El Niño conditions and near-normal rainfall across much of KNP in the summer of 2016/17 (Fig. 7) gave a group of researchers just such an opportunity. From 2006 a research collaboration* between several US universities, the University of KwaZulu-Natal and the SAEON Ndlovu Node has been assessing the response of the vegetation of central KNP to manipulations of fire frequency and grazing, both key drivers of the plant community.


Figure 7. Near-average rainfall across north-eastern South Africa, including KNP, during the summer of 2016/2017 ended the drought experienced since the end of 2014. This provided a multi-national group of researchers working in KNP the opportunity to study post-drought recovery patterns and processes in this savanna system. Map courtesy of Manstrat Agricultural Intelligence Solutions.

Multiple scientific publications from this work have significantly advanced the understanding of fire-vegetation and herbivore-vegetation dynamics in savanna systems. Included here is the unanticipated finding that after a decade of herbivore exclusion under different burning treatments, the vegetation community remained mostly unchanged (Fig. 8).


Figure 8. Previous research into the drivers of vegetation in central KNP unexpectedly found that manipulations of fire and grazing had little effect on the vegetation community, likely because the dominant grass in the system - Bothriochloa radicans (inset), is seldom grazed and responds positively to fire. These existing data, collected during ‘normal’ pre-drought conditions, provide the benchmark for evaluating drought response and recovery. Figure from Koerner et al. (2014), Ecology 95: 808-816.

This the researchers attribute to traits of a dominant grass in the system, Bothriochloa radicans (Stinking grass or Stinkgras). The longevity of this grass, combined with its resistance to grazing and positive response to fire permits it to dominate, seemingly indefinitely.

Bothriochloa is unpalatable and offers little in the way of valuable forage for grazers, leaving conservation managers grappling with the challenge of improving the forage value of these savannas for wildlife, but with no tools to do so. At the same time, the research team was speculating that a disturbance larger than either fire or grazing may be the answer…

Unbeknown to the researchers, their decade-long data set collected under variable but ‘normal’ climate conditions would become the pre-drought data against which to test ideas regarding two distinctly positive longer-term drought effects - a shift in the vegetation from dominance by unpalatable to palatable grasses, and increased plant species diversity. Underpinning this is the thinking that during drought conditions animals are forced to consume all available forage, whether palatable or not.

Further, the return of the vegetation occurs under conditions of reduced herbivore numbers. Working on forecasts that the drought would end during summer 2016/2017, the consortium motivated for and secured last-minute funds via the US-NSF RAPID funding mechanism to conduct a resampling campaign of the permanent vegetation plots established a decade earlier in Satara, one of the areas of KNP hardest-hit by the drought (Fig. 9). The RAPID program specifically supports monitoring at established long-term research sites where further data collection has severe urgency, including quick-response research following natural or anthropogenic disasters.


Figure 9. The 2015/2016 drought in Southern Africa (left) based on NOAA AVHRR satellite data from the STAR programme NESDIS (modified from the RIASCO Action Plan for Southern Africa: Response Plan for the El Niño-induced Drought in Southern Africa, 2016). The black circle shows the study location in KNP. Satara experienced a moderate drought in 2014/15 (defined by exceeding the 15th percentile), followed by an extreme drought in 2016 (exceeding the 5th percentile) based on an estimated probability function calculated from 20 years (1984-2004) of annual precipitation for the study site (right).

Intensive vegetation sampling during the 2016/17 growing season revealed that an extreme drought may be a potential landscape-scale game changer; with recovery within one year of the drought providing a reset to a more productive savanna system. Some areas, barren just months earlier, now supported grass growth at levels which would have been considered normal prior to the drought (Fig. 10).


Figure 10. Production of grass each growing season - termed the primary productivity, typically fluctuates with annual rainfall. For all purposes, this productivity ceased during the recent drought, with no forage available for grazers (inset). However, the system proved highly resilient, with productivity at some sites returning to near-normal levels in 2016/17.

The previous dominant grass, now nearly absent from the system at some sites, had in many instances been replaced by Urochloa mosambicensis (Bushveld signal grass) - a palatable pioneer species of comparatively higher forage value (Fig. 11). Further, plant richness was higher than recorded previously, with the complete removal of grass during drought conditions providing the physical space and reduced competition needed for the recruitment and establishment of less commonly seen annual, and especially forb, species.


Figure 11. Drought variously but negatively impacted the unpalatable and low-forage quality grass Bothriocloa. At many sites a new dominant - Urochloa (inset), emerged. This switch in dominance provides for a more palatable system offering higher forage value, at least during immediate drought recovery.

With plans to repeat their vegetation surveys during the summers of 2017/18 and beyond, the researchers are in an unprecedented position to map the trajectory of the savanna as time since drought accumulates, as plants begin competing for resources, and grazer numbers again increase. Of particular interest is documenting the process of change and the eventual fate of this newly transformed and rejuvenated savanna - will the current high forage quality-high diversity system persist indefinitely as a permanent and positive state change, or will the species reordering be transient, with rapid or prolonged return of the vegetation to pre-drought conditions?

Understanding the impacts of climate extremes

Climate models predict an intensification and increased frequency of climate extremes, a phenomenon already evident in global weather patterns. A pressing scientific need therefore exists to better understand the impacts of such extreme events in terrestrial, especially arid and semi-arid, ecosystems.

This is true not only in protected areas like KNP, but also in ranching and rangeland contexts where livelihoods depend on the productivity of these savannas and the animals they support. Based on the RCP8.5 or ‘business as usual’ pathway, which sees atmospheric CO2 concentrations approach 1 000 ppm by the end of the century (403 ppm at October 03, 2017), summer in the South African lowveld is forecast to be 2oC hotter by 2065, and 3oC hotter by 2095.

For the same near-future and far-future horizons, annual summer rainfall is likely to have decreased by 10-20 mm (Fig. 12). It’s a sobering thought that, as temperatures rise and rainfall decreases, what is currently considered a climate ‘extreme’ will become the climate ‘normal’.

Any understanding of drought effects on species, communities and ecosystem functioning that can be gained now will likely prove crucial in decades to come.


Figure 12. Summer (Dec-Feb) mean temperature (oC, left column) and rainfall (mm, right column) change projected for near- (top row) and far-future (bottom row) periods, relative to present (1976-2005), under conditions of the RCP 8.5 (‘business as usual’) pathway. Modified from the South African Weather Service Climate Change Reference Atlas (2017). 

* Prof. Deron Burkepile (University of California, Santa Barbara), Prof. Scott Collins (University of New Mexico), Prof. Kevin Kirkman (University of KwaZulu-Natal), Prof. Sally Koerner (University of North Carolina, Greensboro), Prof. Alan Knapp (Colorado State University), Dr Nathan Lemoine (Colorado State University), Prof. Melinda Smith (Colorado State University), Dr Dave Thompson (SAEON) and Dr Kevin Wilcox (University of Oklahoma).

Document Actions