Changes in species composition, herbaceous biomass yield, soil seedbank and soil properties of semi-arid grassland subjected to different levels of rainfall interception and resting period

Climate change is expected to result in rainfall reduction in South African semi-arid and arid rangelands and may have severe ecological consequences. This rainfall reduction may cause irreversible and substantial socio-economic changes if resilient based management strategies are not developed to assist farm managers in their ability to sustain desirable ecosystem services by incorporating adaptive management strategies. Therefore, it is important to understand how these rangelands will respond to predicted impacts of climate in particular that of rainfall reduction. The current study was conducted to determine the impact of rainfall interception on soil water content and soil properties, annual primary production vegetation composition and soil seedbank storage. To achieve this, rainout shelters that reduced rainfall at several levels (15%, 30% and 60%) and ambient rainfall (0%) interceptions were used. Effects of resting periods (70 and 90 days) were investigated (following simulated grazing) with the rainfall interception treatment to determine the main effect and its interaction with these treatments. The data were collected over three seasons (with the exception of winter) in 2016/17 and 2017/18 and used in four separate studies. In the first study, the effects of rainfall interception and resting period on residual soil water content and selected soil chemical properties of the semi-arid grassland were evaluated. Results from this study showed that reducing annual rainfall by 60% could negatively affect residual soil water content, thus influencing regrowth potential after defoliation by grazing or cutting natural grasslands. The negative effect of rainfall interception on residual soil water content was higher during summer than autumn. Rainfall interception led to higher extractable residual soil cations (calcium, potassium, magnesium, sodium, and phosphorus) and anions (sulphate and nitrate) partly because of the interception in leaching and partly owing to reduced uptake of soil nutrients by plants. Rainfall interception, however, did not significantly affect soil pH, total N, and total C, indicating the minimal impact of climate change manifested in terms of rainfall interception on these parameters. Generally residual soil water content was higher at a deeper soil layer than the top layer, suggesting that plants with creeping ground cover and tap root systems would have competitive advantage to dominate the vegetation at 15% and 60% rainfall interception (RI), which was characterized by low residual soil water content, rather than 0% or 30% RI. Under 60% RI scenarios, however, the plants might not be able to grow fast and switch to reproductive stage. In the second study annual net primary productivity (ANPP) was evaluated of various functional groups as affected by different intensities of rainfall interception in the semi-arid grasslands of the province of Gauteng in South Africa. It was concluded that reduced rainfall had a significant negative impact on ANPP at 60%. Reduced rainfall significantly (p = 0.0188) affected grass ANPP more than forbs ANPP. Whereas grass ANPP decreased with the reduction in rainfall from 2017 to 2018, that of forbs increased. Various grass species responded in different ways to the reduced water supply, because those tropical grass species that were equipped with better ability to fix carbon dioxide by closing their stomata were likely to be affected less by rainfall interception than temperate and other tropical grass species adapted to waterlogged areas. It was found that grass species such as Eragrostis curvula required less water to produce herbage than Digitaria eriantha. Livestock farmers who rely on natural grazing land should promote the growth of grasses such as E. curvula to produce enough grass during periods of low to moderate intensity rainfall interception, because such grass species are more favourable to those conditions and able to produce more biomass per unit of rainfall. This can be archived by reseeding the rangelands with such grass species. In the third and fourth study, seasonal herbaceous structure and the response of biomass production to rainfall interception and resting period were studied. In this study, it was concluded that rainfall interception associated with climate change posed a threat to the country’s rangeland productivity as it would lead to a decreased forage biomass yield. It was found that increaser grasses and forbs dominated under high rainfall interceptions. Although the grassland productivity had shown resilience to a small interception in rainfall (up to 15%), a severe reduction would lead to changes in the herbaceous layer, which might lead to rangelands being dominated by forbs rather than grasses beyond a 30% interception in rainfall. Within the grass species, increasers are more tolerant of rainfall interception than decreasers, and thus are likely to increase with the decrease in rainfall beyond 15%. Resting the rangelands for 90 days or more compared with 70 days would improve rangeland biomass productivity during spring and summer, whereas there was no benefit to resting for 90 days compared with 70 days during the autumn harvest. Further studies need to be conducted to quantify the impact of long-term rainfall interception on the growth pattern of key species, soil vegetation cover, soil nitrogen and carbon dynamics, soils seedbank and soil microbiota diversity to accurately model and predict the impact of reduced rainfall on the country’s rangelands to come up with climate change adaptation strategies. Finally, in the fifth study, the effects of rainfall interception and resting period on soil seed banks (SSBs), were determined. The results showed that rainfall interception had a negative effect on a grass SSB, whereas it improved the forb SSB. Rainfall interception affected the SSB of the most palatable species such as Themeda triandra and D. eriantha negatively. These species were fewer in the SSB compared at 60% RI, because they did not produce seed. Therefore rainfall interception might lead to grasslands being dominated by forbs, resulting in larger bare patches that expose the land to degradation through soil erosion. Forbs seemed to be suppressed by grasses at lower rainfall interception treatments. However, the authors recommend that more studies of this nature should be conducted, taking into consideration the effect of time lags between SSB collections. This calls for alternative interventions that could be considered if the droughts became more persistent, for example reseeding of rangelands with naturally occurring grasses in a specific area for example E. curvula. Key words: decreaser, increaser, forbs, rainfall reduction, resting period Abstract Climate change is expected to result in rainfall reduction in South African semi-arid and arid rangelands and may have severe ecological consequences. This rainfall reduction may cause irreversible and substantial socio-economic changes if resilient based management strategies are not developed to assist farm managers in their ability to sustain desirable ecosystem services by incorporating adaptive management strategies. Therefore, it is important to understand how these rangelands will respond to predicted impacts of climate in particular that of rainfall reduction. The current study was conducted to determine the impact of rainfall interception on soil water content and soil properties, annual primary production vegetation composition and soil seedbank storage. To achieve this, rainout shelters that reduced rainfall at several levels (15%, 30% and 60%) and ambient rainfall (0%) interceptions were used. Effects of resting periods (70 and 90 days) were investigated (following simulated grazing) with the rainfall interception treatment to determine the main effect and its interaction with these treatments. The data were collected over three seasons (with the exception of winter) in 2016/17 and 2017/18 and used in four separate studies. In the first study, the effects of rainfall interception and resting period on residual soil water content and selected soil chemical properties of the semi-arid grassland were evaluated. Results from this study showed that reducing annual rainfall by 60% could negatively affect residual soil water content, thus influencing regrowth potential after defoliation by grazing or cutting natural grasslands. The negative effect of rainfall interception on residual soil water content was higher during summer than autumn. Rainfall interception led to higher extractable residual soil cations (calcium, potassium, magnesium, sodium, and phosphorus) and anions (sulphate and nitrate) partly because of the interception in leaching and partly owing to reduced uptake of soil nutrients by plants. Rainfall interception, however, did not significantly affect soil pH, total N, and total C, indicating the minimal impact of climate change manifested in terms of rainfall interception on these parameters. Generally residual soil water content was higher at a deeper soil layer than the top layer, suggesting that plants with creeping ground cover and tap root systems would have competitive advantage to dominate the vegetation at 15% and 60% rainfall interception (RI), which was characterized by low residual soil water content, rather than 0% or 30% RI. Under 60% RI scenarios, however, the plants might not be able to grow fast and switch to reproductive stage. In the second study annual net primary productivity (ANPP) was evaluated of various functional groups as affected by different intensities of rainfall interception in the semi-arid grasslands of the province of Gauteng in South Africa. It was concluded that reduced rainfall had a significant negative impact on ANPP at 60%. Reduced rainfall significantly (p = 0.0188) affected grass ANPP more than forbs ANPP. Whereas grass ANPP decreased with the reduction in rainfall from 2017 to 2018, that of forbs increased. Various grass species responded in different ways to the reduced water supply, because those tropical grass species that were equipped with better ability to fix carbon dioxide by closing their stomata were likely to be affected less by rainfall interception than temperate and other tropical grass species adapted to waterlogged areas. It was found that grass species such as Eragrostis curvula required less water to produce herbage than Digitaria eriantha. Livestock farmers who rely on natural grazing land should promote the growth of grasses such as E. curvula to produce enough grass during periods of low to moderate intensity rainfall interception, because such grass species are more favourable to those conditions and able to produce more biomass per unit of rainfall. This can be archived by reseeding the rangelands with such grass species. In the third and fourth study, seasonal herbaceous structure and the response of biomass production to rainfall interception and resting period were studied. In this study, it was concluded that rainfall interception associated with climate change posed a threat to the country’s rangeland productivity as it would lead to a decreased forage biomass yield. It was found that increaser grasses and forbs dominated under high rainfall interceptions. Although the grassland productivity had shown resilience to a small interception in rainfall (up to 15%), a severe reduction would lead to changes in the herbaceous layer, which might lead to rangelands being dominated by forbs rather than grasses beyond a 30% interception in rainfall. Within the grass species, increasers are more tolerant of rainfall interception than decreasers, and thus are likely to increase with the decrease in rainfall beyond 15%. Resting the rangelands for 90 days or more compared with 70 days would improve rangeland biomass productivity during spring and summer, whereas there was no benefit to resting for 90 days compared with 70 days during the autumn harvest. Further studies need to be conducted to quantify the impact of long-term rainfall interception on the growth pattern of key species, soil vegetation cover, soil nitrogen and carbon dynamics, soils seedbank and soil microbiota diversity to accurately model and predict the impact of reduced rainfall on the country’s rangelands to come up with climate change adaptation strategies. Finally, in the fifth study, the effects of rainfall interception and resting period on soil seed banks (SSBs), were determined. The results showed that rainfall interception had a negative effect on a grass SSB, whereas it improved the forb SSB. Rainfall interception affected the SSB of the most palatable species such as Themeda triandra and D. eriantha negatively. These species were fewer in the SSB compared at 60% RI, because they did not produce seed. Therefore rainfall interception might lead to grasslands being dominated by forbs, resulting in larger bare patches that expose the land to degradation through soil erosion. Forbs seemed to be suppressed by grasses at lower rainfall interception treatments. However, the authors recommend that more studies of this nature should be conducted, taking into consideration the effect of time lags between SSB collections. This calls for alternative interventions that could be considered if the droughts became more persistent, for example reseeding of rangelands with naturally occurring grasses in a specific area for example E. curvula