The dynamics of certain grass species used during the restoration of degraded semi-arid rangelands

Abstract

Addressing desertification in South Africa requires the restoration of degraded land, particularly degraded pasture. Enhanced (coated) and untreated seed of four palatable, perennial grass species and one semi-perennial grass species were broadcasted into part of degraded pastures following a rip plough cultivation method. Different grass mixtures and ratios were used. Quantitative information for the success of the restoration methods as well as the sustainability and change of the oversown grass population over time was determined in four different study areas (Tweefontein, Totiuskraal, Kromspruit and Davidskatnagel) during 1995 to 1999 and 1995 to 2000 for Totiuskraal. The composition of the soil seedbank of each study area for the summer (1998) was determined as an indication of the representation of the oversown grasses in the seedbank and to monitor the sustainability of the community. The quantitative data were in the format of frequency and density surveys. Density data was subdivided in the number of seedlings, vegetative and reproductive growth stages of each grass species per square meter. Seed viability was determined. Data analysis included the analysis of variance (STATISTICA2), multi variate analysis (DECORANA; Hill, 1979) and the use of ISPD (Integrated System for Plant Dynamics; Bosch et al., 1992). In all the study areas the untreated seed had the highest germination rate in the laboratory, but the enhanced seed had the highest seedling survival rate under natural conditions in the field, except Chloris gayana at the Tweefontein study area. The environmental, climatological and managerial conditions favoured specific grass mixtures at each study area at the end of the first year. • C. gayana (untreated seed), Anthephora pubescens (enhanced seed), Digitaria eriantha (enhanced seed) and Panicum maximum (enhanced seed) were appropriate after one year for the 25% sandy-clay-loam sand of the Tweefontein study area with its average short-term rainfall of 596 mm/yr. (1995 - 1999). C. gayana is a nursing plant because it stabilizes the eroding

soil and thus contribute to the establishment of other perennial species. Only D. eriantha and C. gayana formed sustainable populations in the first season after the application of the restoration technique. These two species provided soil stability in the first two years of the application and contributed to improved environmental conditions for A. pubescens and P. maximum. • The 29% clay of the Totiuskraal study area near Potchefstroom combined with the average short-term rainfall of 773 mm/yr. was ideal for a grass mixture of C. ciliaris and A. pubescens (untreated seed). • P. maximum and C. ciliaris were suited for the 23% sandy-clay-loam soil and 517 mm/yr. (short-term average) of Kromspruit. • Davidskatnagel's 29% sandy-clay-loam soil and 316 mm/yr. average short-term rainfall favoured a D. eriantha and P. maximum enhanced grass mixture. • The frequency of the oversown grass species at Tweefontein, except for the frequencies of D. eriantha decreased. The density of D. eriantha increased, but the other species decreased. • At the Totiuskraal study area the density of A. pubescens (enhanced) increased over a five-year period. A. pubescens's (u) frequency at first increased and then decreased. The frequency and density of C. ciliaris (untreated) first increased as well, and then decreased, while the frequency and density of C. ciliaris (u) stabilised. • Some of the species at Kromspruit increased in frequency (D. eri u, C. gay e + 50%, etc.). The other ratios and treatments first increased and then decreased. The overall number of species increased in density, except D. eriantha, C. ciliaris and P. maximum. • The frequencies of the Davidskatnagel's oversown species decreased, except for D. eriantha, which increased. The grass species richness of each study area increased for the period 1995 to 1999 with between eight and 14 species. This increase hints at a more stable and resilient grass community. There was significant variation in the seed densities of the soil seedbanks of the different sub-plots, and although the soil seedbanks are rich in seed, none of the oversown seed appeared in the soil seedbanks three years after the restoration application, except for A. pubescens and C. ciliaris.

The rip plough cultivation method at the Tweefontein study area influenced the frequencies of C. plurinodis, Eragrostis racemosa, Aristida congesta, Melinis repens and Elionurus muticus negatively. The frequencies of Setaria sphacelata, Eragrostis curvula, Themeda triandra, P. maximum, Cynodon dactylon, Eragrostis chloromelas and C. ciliaris increased because of the rip plough cultivation method. Based on the edaphic and climatological information and the grass species favoured through the rip plough cultivation method, the following species are recommended for restoration purposes: S. sphacelata, E. curvula, T. triandra, P. maximum, E. chloromelas, C. ciliaris, D. eriantha and A. pubescens. Chloris virgata, Urochloa mosambiscensis, C. dactylon and Brachiaria eruciformis's frequencies decreased after the rip plough cultivation method at Kromspruit. The density of some species, however, increased, e.g. D. eriantha, P. maximum, Eragrostis . rigidior, Setaria nigrirostris, Aristida congesta and Bothriochloa insculpta. As mentioned for Tweefontein, the above mentioned species were suitable for the edaphic and climatological characteristics of the specific area. D. eriantha, P. maximum and B. insculpta could be an appropriate seed mixture for restoration in areas of similar environmental conditions (p. ii, paragraph 2). B. eruciformis, Setaria iocladis, C. virgata, E. curvula and Sporobolus africanus's frequencies decreased where the rip plough cultivation method was applied at Davidskatnagel. Species frequencies that increased due to the rip plough cultivation method (i.e. species adapted to disturbance), was A. congesta and E. rigidior. In all the cases the species composition and thus the field condition of the study areas improved over the five seasons after the restoration application. At Tweefontein and Davidskatnagel where D. eriantha was oversown, the grass communities became D. eriantha dominant. In the case of Totiuskraal, though C. ciliaris and A. pubescens dominated the grass community at the end of the study period (2000). The latter two species prefer more sandy soil as is the case for Totiuskraal. At Kromspruit the dominant species were the annuals C. dactylon and C. virgata, but the perennial grass that became dominant was D. eriantha. In conclusion, the composition of unpalatable or annual grass communities can improve over a short period (approximately five years) through the combined use of rip plough cultivation method and oversowing of perennial climax grass species.