SOUTH AFRICA'S ARID AND SEMIARID RANGELANDS: WHY ARE THEY CHANGING AND CAN THEY BE RESTORED? S U Z A N N E J. MILTON and W. R I C H A R D J. D E A N FitzPatrick Institute, University of Cape Town, Rondebosch 7700, South Africa

Abstract. Since the mid-19th century settled livestock ranching has been the major form of land use in South Africa, occupying 68% of the land surface. Decreases in livestock densities and ranch numbers during the past century imply that carrying capacities for domestic herbivores are falling. Differences in carbon isotope signals with soil depth and abrupt shifts in dominant plant species across ranch boundaries reveal that southern African rangelands are changing. Case studies suggest ways to control altered grassland composition, bush encroachment in arid savanna, and dominance by toxic and halophytic shrubs in arid shrublands. But climatic and biological factors constrain rates of passive recovery, and guidelines for active restoration are poor and techniques costly. Moreover, conservation of remaining good rangeland is seldom enforced, and economic considerations usually outweigh the land user's desire to sustain diversity and productivity.

1. Introduction

Ranching occupies 68% of South Africa's land surface (Directorate Agricultural Information, 1991). Although it generates only 2% of the gross domestic product and supports little more than 1 million people, ranching supplies most of the local demand for meat and wool. Desertification, or a functionally irreversible decline in rangeland productivity, therefore has serious social and economic implications for this country. As the theme of this conference is desertification in developed countries, we will not discuss the problems of the underdeveloped 15% of South Africa (Boonzaier et al., 1990; Hatch and Grossman, 1993), where 30% of the nation's population subsists on communal grazing land (Directorate Agricultural Information, 1991). Most of South Africa's 706,000 km 2 of rangeland is owned by commercial ranchers. The present review focuses on this sector, which is "developed" in its access to education, transport, finance, and markets. To understand desertification in the developed parts of South Africa we review changes in land use and the evidence for declining productivity on commercial rangeland at various spatial scales. We then explore the biological processes behind decreases in carrying capacity for livestock and discuss constraints on rangeland rehabilitation. Finally, we discuss the social and economic factors that motivate landowners to overexploit their rangelands. Environmental Monitoring and Assessment 37: 245-264, 1995. (~) 1995 Kluwer Academic Publishers. Printed in the Netherlands.

246

SUZANNE J. MILTON AND W. RICHARD J, DEAN

Mean annual rainfall isohyets (mm)

(a)

200 mm 400 mm 600 mm >600 mm

Fig. la. Figs. l(a)-(c). Rainfallisohyets(a), vegetation(b) and distribution6f livestock(c) in South Africa.

Vegetation types

(b)

Savanna Grassland Succulent Karoo Karoo Forest Macchia

Fig. lb. 2. Changes in L a n d Use The indigenous grazing and browsing ungulates of South Africa's semiarid Karoo and savanna regions (Figures la and lb) were nomadic. Eland (Taurotragus oryx), gemsbok (Oryx gazella), hartebeest (Alcelaphus buselaphus) and springbok (Anti-

SOUTH AFRICA'S ARID AND SEMIARID RANGELANDS

Livestock and regions of South Africa

247

(c)

SHEEP GOATS CATTLE

CApE__ Fig. lc.

dorcas marsupialis) moved seasonally or opportunistically to greener pastures (Skead, 1980). The native hunters and pastoralists of the region followed the game or moved with their herds of sheep (Smith, 1986). Seasonal shortages of surface water necessitated a similar lifestyle among the European ranchers (trekboers) who colonized the Cape Colony of South Africa in the 18th century (Talbot, 1961; Penn, 1986). Trekking between summer and winter pastures continued after formal alienation of land in the 1820s. But the Land Settlement Act of 1912, together with the state-assisted introduction of wind-driven waterpumps and fencing, ended the era of nomadic ranching, even in the most arid regions (Talbot, 1961). By this time cattle, sheep, and goats (Figure lc) had largely replaced the indigenous ungulate populations, which had been reduced by efficient hunting and fencing (Skead, 1980).

3. Evidence for Declining Productivity Evidence for declining secondary productivity and changes in vegetation composition has been accumulating since the early 1900s. We discuss this evidence at national, regional, ranch, and patch scales. This distinction is necessary because desertification may be a patchy rather than a homogeneous process (Dean et al., 1995). Therefore, changes at the ranch scale cannot be inferred from productivity trends at the national scale.

248

SUZANNE J. MILTON AND W. RICHARD J. DEAN 35

HUMA PN OPULATION~

30 25 03 Z

20

...I ..d

15

O

CATTLE SHEEP GOATS

~,

.,&.~

/~

10

,

~. ""

" .4

. . . . . . .

5

~ . . , . E ] . EIB .... ..~....Er.~ ...... ~..-.t~[].~.~

SMALL STOCK (LSU) 0

1 ~,~00

1

I

I

|

I

I

I

I

1910

1920

1930

1940

1950

1960

1970

1980

Fig. 2. Trends in human and livestock populations in South Africa. Livestock populations are expressed as large stock units (LSU). (Data from Fuggle, 1983; Directorate Agricultural Information, 1991.)

3.1.

NATIONAL

TRENDS

Government enquiries into drought and rangeland desertification began in the early 1900s (Union of South Africa, 1914, 1923) and were followed a generation later by other investigations (Kokot, 1948; Union of South Africa, 1951). All concluded that annual rainfall showed no negative trend and that the observed drying up of rivers and rangelands was the result of overgrazing and soil erosion. Generating a lucrative market for wool, the industrial revolution in Europe led to a rapid increase in sheep ranching in South Africa between 1840 and 1850. Small stock (sheep and goat) populations reached a peak of 50 million in 1930 shortly after the fencing and water provision on ranches, but fell to 35 million in 1933 and appears to have stabilized at this level (Downing 1978; Macdonald, 1989). The human population in South Africa has grown by 2-3% annually since 1950 (Fuggle, 1983), whereas total livestock populations (large stock units - LSUs - equivalent to one bovid weighing 420 kg (Dean and Macdonald, 1994) have remained stable or decreased during this period (Figure 2). Ranch numbers peaked in 1950 as a result of subdivision of deceased estates*, but many of the units were subeconomic (Union of South Africa, 1951). We infer from the decrease in ranch numbers and increase in average ranch size (Figure 3a) that rangeland productivity declined on a national scale thereafter. The change in ranch sizes was greater in the arid grasslands and shrublands of the Cape and * A problem peculiar to South Africa was caused by subdividing ranches when an estate was left jointly to a rancher's sons. Each section of the original ranch was then independently ranched by the inheritor, often with disastrous results for the owner and the land. The process of fragmentation continued when the inheritor died and the subsection was again subdivided.

249

SOUTHAFRICA'SARID AND SEMIARIDRANGELANDS

"-O

1.2

120 110

if)

o r" LU "r" 0 Z rr

1.0

90

0.9

uJ N

x

ol E

0.8

60

< LU r'r
95% of the grass biomass because perennial grass species were absent from the seed bank (Moore, 1989). Clearing of Karoo shrubland resulted in abundant seedling recruitment, but

258

SUZANNE J. MILTON AND W. RICHARD J. DEAN

the seedling composition was once again determined by localized seed availability (Milton, 1995). 5.3. RESEEDING Large-scale reseeding trials have rarely been well monitored. In some cases, competition from established plants precluded establishment of new seedlings, whereas in others, changes in soil and microclimate appeared to limit recruitment. Without intervention, less palatable grasses can remain dominant in grassland for >20 years. Addition of seeds of perennials results in limited improvement, but rapid restoration usually requires plowing, fertilizing, and resowing (Edwards, 1981). Reseeding into undisturbed Karoo vegetation results in little or no recruitment of shrubs. When existing vegetation was cleared or where the soil was cultivated to a depth of 30 ram, seedling survival improved (Joubert and van Breda, 1976; Milton, 1994). Mud-pelleting of Karoo grass and shrub seed reportedly improved its germination in the field relative to uncoated seeds (Van Breda, 1939). Ripping, bushcutting, and reseeding on Karoo ranches with carrying capacities

South Africa's arid and semiarid rangelands: Why are they changing and can they be restored?

Since the mid-19th century settled livestock ranching has been the major form of land use in South Africa, occupying 68% of the land surface. Decrease...
1MB Sizes 0 Downloads 0 Views