Woody Plant Encroachment in Grasslands and Savannas

Woody plant encroachment has been widespread in grassland and savanna ecosystems of North and South America, Australia, Africa, and southeast Asia over the past century (Table 1). This encroachment, typically by unpalatable trees and shrubs, has gone to completion on some landscapes and is in progress on others. It jeopardizes grassland biodiversity and threatens the sustainability of pastoral, subsistence, and commercial livestock grazing (Rappole et al, 1986; Noble, 1997). As such, it may adversely impact-20% of the world's population (Turner et al, 1990). The proximate causes of displacement of perennial grasses by woody plants are subjects of debate. Land-use practices such as heavy grazing and reductions in fire frequency have often been implicated. However, climate change, historic atmospheric C02 enrichment, and exotic species introductions are potentially important contributing factors (Idso, 1995; Archer et al, 1995; Polley, 1997; Polley et al, 1997). Current trends in atmospheric C02 enrichment may exacerbate the shifts from grass to woody plant domination, especially where the invasive trees/shrubs are capable of symbiotic N2 fixation. Expansion of woody plants into grasslands may also be favored by recent increases in atmospheric N deposition (Kochy, 1999). In addition to influencing vegetation composition, changes in each of these factors would have the potential to alter the storage and dynamics of C and N in savanna ecosystems. The net outcome of such interactions over the recent past is poorly understood and has not been well documented.

Although woody plant encroachment has long been a concern of land managers in grassland and savanna regions (e.g., Fisher, 1950, 1977), research on this problem has been primarily "applied" and focused on the effects of woody plants on grass production and the development of chemical or mechanical methods to reduce the abundance of established trees and shrubs. Despite the long-standing recognition of woody plant encroachment as a worldwide dryland management problem, little is known of the rates and dynamics of the phenomenon or its impact on fundamental ecological processes related to energy flow, nutrient cycling, and biodiversity. Grassland/savanna systems account for 30-35% of global terrestrial net primary production (Field et al, 1998). Hence, when woody species increase in abundance and transform shrublands into woodlands, grasslands into savannas, or savannas into shrublands and woodlands, the potential to alter C and N sequestration and cycling at regional and global scales may be significant. Consequently, this type of land cover change has the potential to contribute significantly to the terrestrial global carbon sink (cf. Ciais et al, 1995). Savanna landforms may have a larger impact on the global carbon cycle than previously appreciated (Hall et al, 1995; Ojima et al, 1993; Scholes and Hall, 1996; Scholes and Bailey, 1996; Scholes and van der Merwe, 1996). Indeed, recent assessments suggest that savanna ecosystems have among the highest potential C gain and loss rates of the world's biomes (ORNL, 1998). In addition, emissions of radiatively active

TABLE 1 Survey of Studies Describing or Quantifying Woody Plant Encroachment into Grassland, Tree/Shrub Proliferation in Savannas, and Tree Encroachment into Shrubland.



South Dakota

Reichard and Hamilton, 1997

Burrows et al., 1998

Arnold, 1950

Abrams, 1986

Bock and Bock, 1984

Rogers, 1982

Burrows et ai, 1990

Bahre, 1991

Bragg and Hulbert, 1976

Progulske, 1974

Robinson, 1965

Cook et al., 1996

Bahre and Shelton, 1993

Briggs and Gibson, 1992

Tieszen and Archer, 1990

Tieszen and Archer, 1990

Cunningham and Walker, 1973

Brown, 1950

Knapp and Seastedt, 1986

Tieszen and Pfau, 1995

Wall, 1999

Gardiner and Gardiner, 1996

Brown et al., 1997

Knight et al., 1994


West, 1988

Grice, 1996

Cooper, 1960

Loehle etai, 1996

Ansley ef1995

Young et al, 1979

Grice, 1997

Covington and Moore, 1994

Owensby etai, 1973

Archer et al., 1988


Harrington et ai, 1979

Glendening, 1952


Archer, 1989

Acocks, 1964

Harrington and Hodgkinson, 1986

Humphrey and Mehrhoff, 1958

Grimm, 1983

Boutton etai, 1998

Ambrose and Sikes, 1991

Hodgkin, 1984

Hastings and Turner, 1965

Johnston et al., 1996

Bogusch, 1952

Ben-Shaher, 1991

Lonsdale, 1993

Johnsen, 1962


Bray, 1901

Bews, 1917

Noble, 1997

Kenney et al., 1986

Arno and Gruell, 1986

Bruce etai, 1995

Bond et al, 1994

Panetta and McKee, 1997

Martin, 1975

Arno et al., 1995

Ellis and Schuster, 1968

Priedel, 1987


Martin and Turner, 1977


Foster, 1917

Grossman and Gandar, 1989

Archibold and Wilson, 1980

McClaran and McPherson, 1995

Johnson, 1994

Inglis, 1964

Höchberg et al, 1994

Brown, 1994

McPherson et til., 1993

Steinauer and Bragg, 1987

Johnston, 1963

Holmes and Cowling, 1997

Kôchy, 1999

Miller, 1921

Steuteret«/., 1990

McKinney, 1996

Jeltsch etai., 1997


Reynolds and Glendening, 1949

New Mexico

McPherson etai., 1988

Le Roux, 1997

Adamoli et til., 1990

Savage and Swetnam, 1990

Branscomb, 1958

Nelson and Beres, 1987

Menaut et ill., 1990

Biicher, 1982

Smith and Schmutz, 1975

Buffington and Herbei, 1965

Smeins et al., 1974

Norton-Griffiths, 1979

Biicher 1987


Connin ct ill., 1997

Scanlan and Archer, 1991

O'Connor and Roux, 1995

Distel and Boo, 1996

Bossard, 1991

Dick-Peddie, 1993

Weltzin et al., 1997

Palmer and van Rooyen, 1998

Dussart etai, 1998

Bossard and Reimanek, 1994

Gibbens et al, 1992

Wondzell and Ludwig, 1995

Ramsay and Rose Innes, 1963

San José and E:arinas, 1983

Callaway and Davis, 1993

Hennessy et ill., 1983


Reid and Ellis, 1995

San José and Farinas, 1991

McBride and Heady, 1968

McCraw, 1985

Madany and West, 1983

Ringose et al., 1996

San José el al., 1991

Hobbs and Mooney, 1986

York and Dick-Peddie, 1969

Yorks et ill., 1992

Sabiiti, 1988

San José and Montes, 1997

Vivrette and Muller, 1977



Schwartz et al., 1996

San ¡osé et ai, 1998

Williams««/., 1987

Blackburn and Tueller, 1970

Rummell, 1951

Scott, 1966

Schofield and Bucher, 1986

Young and Evans, 1981

North Dakota


Shantz and Turner, 1958



Potter and Green, 1964

Fisher et al., 1987

Skarpe, 1990

Backéus, 1992

Baker and Weisberg, 1997


Regional Assessments

Skarpe, 1991

Binggeli, 1996

Mast et al., 1997

Knapp and Soule, 1996

Glendening and Paulsen, 1955

Thomas and Pratt, 1967

Walker et aL, 1981

Mast et ill., 1998

Knapp and Soule, 1998

Gruell, 1983

Trollope, 1982

Skarpe, 1992

Veble N and Lorenz, 1991

Miller and Rose, 1995

Hart and Lavcock, 1996

Van Vegten, 1983


Miller and Halpern, 1998

Humphrey, 1938

West, 1947

Anderson and Holte, 1981

Miller and Rose, 1999

Humphrey, 1987


Burkhardt and Tisdale, 1976

Skovlin and Thomas, 1995

Johnson, 1987

Booth and Barker, 1981

Zimmerman and Neunschwander,

Soule and Knapp, 1999

Leopold, 1951

Bowman and Panton, 1995



Milchunas and Lauenroth, 1993



Engle et ill., 1996

Miller andWigand, 1994

Brown and Carter, 1998

Wang et al., 1993

Snook, 1985

McClaran and McPherson, 1995

Burrows et al., 1985

Note. Documentation includes historical observations, long-term monitoring, repeat ground or aerial photography, stable carbon isotope analysis, dendrochronology, and, in some cases, simulation modeling. The focus is on arid and semi-arid "rangelands." Hence, studies documenting tree/shrub invasion of abandoned agricultural fields (cf., Smith, 1975; Johnston et at., 1996; De Steven, 1991) or regeneration following forest clearing are not included here. Studies discussing or reviewing causes or consequences of woody encroachment into grasslands/savannas (cf, Humphrey, 1953; Fisher, 1977; Smeins, 1983; Rappole et al, 1986; Grover and Musick, 1990; Schlesinger et til., 1990; Archer, 1994; Archer et til., 1995; Idso, 1995; Polley et al., 1997) are also excluded. Some papers in the list reference other papers which have documented woody plant increases in historical times (Backeus, 1992; Noble, 1997). An updated and more extensive version of this table, including a list of woody genera, can be found at http://cnrit.taimi.edu/rlem/faculty/archer/.

Desert Scrub

Grassland or Shrub-Steppe

Woodland t

Grassland or Savanna

Grassland, Pasture, Savanna, Heathland

Forest or Woodland




Drought Grazing

Drought Loss of Fire/Browsers Grazing




Tree Clearing Wood Harvesting Browsing



FIGURE 2 Xerifkation/desertifkation (West, 1986) and deforestation have received much attention. Although increases in woody plant abundance in drylands are geographically widespread and well documented (Table 1), little is known of the ecological consequences of this vegetation change (adapted from Archer and Stokes, 2000).

trace gases, NOx and aerosols from savanna fires may contribute significantly to global emissions and influence climate and atmospheric chemistry (Crutzen and Andreae, 1990; Hao el al., 1990; Crutzen and Goldammer, 1993).

Desertification has long been a topic of concern to land managers and ecologists (Moat and Hutchinson, 1995; Arnalds and Archer, 2000). More recently, changes in the storage and dynamics of C and N in the terrestrial biosphere have been evaluated with respect to deforestation, intensive agricultural practices, succession on abandoned agricultural lands, and afforestation / reforestation (Fig. 2) (Houghton et al., 1987; Post, 1993). Increased abundance of woody plants in drylands has the potential to alter land surface-atmosphere interactions and atmospheric chemistry by affecting biophysical processes, and C and N storage and dynamics (e.g., Schlesinger etui, 1990; Graetz, 1991; Bonan, 1997). Even so, its significance has yet to be thoroughly evaluated or quantified. Here, we review results from a case study of a subtropical dryland landscape which has been undergoing a transformation from grassland to savanna to woodland. Some of our recent work has explored the implications of this change in vegetation on the hy-drological cycle (Brown and Archer, 1990; Midwood et al, 1998; Boutton et al., 1999). Here, our emphasis is on the rates of change in soil and plant carbon and nitrogen pools and fluxes.

Was this article helpful?

0 0

Post a comment