Most dry land ecosystems in Africa are characterized by distinctive vegetation types including grassland, savanna, woodland, scrubland, and desert (White, 1983; Friis and Vollesen, 2005). The vegetation structure and composition of these ecosystems have undergone dynamic changes over time, and ecosystem degradation resulted from bush encroachment has been documented to be one of the recent expressions of those dynamic changes (Bester and Reed, 1997; Briske et al., 2003; Reynolds et al., 2007; Eldridge et al., 2011). According to Ward (2005) encroachment of bush can be defined as “ the aggression of destructive and unwanted woody plants, which causes a disproportional increase in bushes and a decline in biological diversity, leading to a decline in grazing capacity of both wild and domestic herbivores in semi-arid rangeland ecosystems” . The encroaching woody species include both indigenous and exotic woody species, and thus bush encroachment actually includes two kinds: (1) the proliferation of indigenous woody plant species and (2) the invasion of exotic/alien woody species in savannas and grassland ecosystems (Brown and Archer, 1989; Ward, 2005). It should be stressed that the second kind of bush encroachment, i.e., invasion of exotic/alien woody species, is a widespread phenomenon over most of the African savannas and grasslands (Hoffman and O’ Connor, 1999; Oba et al., 2000; Smit, 2002; Ward, 2005; Gemedo et al., 2006; O’ Connor et al., 2014).

Researchers have persistently attempted to explore those driving factors triggering encroachment of woody plants in the semi-arid environments (Archer et al., 1995; van Auken, 2000; Briske et al., 2003; Ward, 2005). Among many factors, continuous heavy grazing, frequent fires, declines in soil moisture and nutrient, and climate change associated global warming have been causally linked with encroachment of bushes or woody plants. However, it is yet difficult to attribute a single factor or a set of factors as the cause of the observed bush encroachment simply because most factors are spatially correlated and some of those factors are scale (both space and time) dependent (Walker, 1971; Archer et al., 1995; Scholes and Archer, 1997; Ward, 2005; van Auken, 2009; Archer, 2010; Tessema et al., 2012). Of course, encroachment of woody plants has been tied to the degradation of semi-arid rangelands, and thus often perceived negatively. However, recent studies suggest that encroaching woody species sometimes had positive effects on the structure and function of some ecosystems (Smit, 2002; Gemedo et al., 2006; Maestre et al., 2009; O’ Connor et al., 2014). To explain the locally-observed bush encroachment, three models were proposed in earlier times: (1) Walter’ s two-layer model (Walter, 1939), (2) Moir’ s one-layer model (Moir, 1954), and (3) state-and-transition system (Briske et al., 2003; Sankaran et al., 2004; Vetter, 2005). As longer-term data were accumulated from larger geographic areas and studies were further deepened, more comprehensive theoretical schemes have emerged. The following three theoretical schemes are well regarded: (1) equilibrium rangeland system, (2) disequilibrium rangelands system, (3) non-equilibrium rangeland system. This paper attempts to provide a comprehensive review of published works primarily on the observed bush encroachment and the associated degradation of rangelands in the dryland ecosystems in Africa. The debated causes and the proposed mechanisms are critically reviewed and the practical implications for developing mitigation and adaptation strategies are also discussed.

Encroachment of bush, i.e., the spread of indigenous woody species and/or incursion of woody species introduced from other ecosystems into semi-arid savannas and grassland ecosystems, has been widely reported phenomena (Prins and Van der Jeugd, 1992; Gibbens et al., 2005; Van Auken, 2009). For example, proliferation of woody plants was documented as a well understood threat in many semi-arid savannas during the twentieth century and has later came out as one of the major recognized threats of rangelands (Hoffman and O’ Connor, 1999). Previous studies indicated that semi-arid savannas and grassland ecosystems globally have been changed by bush encroachment in the past 50 years (Ward, 2005). Again, encroachment of bush refers to the spread of indigenous or/and exotic woody plants in savannas and grassland ecosystems (Bews, 1917; Van Auken, 2009).

The phrase “ bush encroachment” is similar with other commonly-used vocabularies such as thickening of woody plants (Van Auken, 2009), woody re-growth (Eldridge et al., 2003), invasion of woody weed (Ayres et al., 2001), xerification (Archer, 2010), and invasion of shrubs (Noble, 1997). Other authors defined encroachment of bush from the perspectives of grazing and browsing values of encroaching woody plant species and productivity. For example, encroachment or incursion of bush could be defined as the destruction or overthrow of pleasant grass species by bushes and/or shrubs, which are often indigestible to herbivores (Ward, 2005). The phenomenon of bush encroachment, which was described in different terminologies, was reportedly widespread in arid and semi-arid savannas and grasslands in most parts of the world such as in North and South Americas, Africa, south-east Asia, and Australia (Prins and Van der Jeugd, 1992; Eldridge et al., 2011). It was demonstrated to have caused herbaceous vegetation loss in many semi-arid grasslands and savannas (Oba et al., 2000; Abule et al., 2005; O’ Connor et al., 2014).

Furthermore, invasion by bush has been tied to rangeland degradation and thus often perceived negatively (Gemedo et al., 2006). The well-known negative connotation was given by Ward (2005) and the connotation defined encroachment of bush as “ the aggression of destructive and unwanted woody plants, which causes a disproportional increase in bushes and a decline in biological diversity, leading to a decline in grazing capacity of both wild and domestic herbivores in semi-arid rangeland ecosystems” . However, recent studies suggest that encroaching woody species sometimes had positive effects on the structure and function of ecosystems (Smit, 2002; Gemedo et al., 2006; Maestre et al., 2009; O’ Connor et al., 2014) and these woody species may also be indispensable resources such as fodder for both domestic and wild herbivores, sources of firewood, medicine, mulch, and more (Smit, 2002; Maestre et al., 2009; Eldridge et al., 2011). In addition, some published works suggested that shrubs could enhance resource stocks such as belowground carbon and nitrogen or aboveground net primary productivity (Maestre et al., 2009).

The balance between the negative and positive effects of bush encroachment may have been juggled by the “ degree” of bush encroachment. White (1980) reported that bush encroachment becomes a problem (i.e., negative) when trees, shrubs and bush covers exceed 30%. More recent works found that an increase of 10% woody cover would lead to a decline of about 7% of the grazing potential and fodder resources in semi-arid rangelands (Oba et al., 2000). Price and Morgan (2008) found a significant reduction in the number of species and also in the emerging seeds of soils in areas with more than 15% coverage by an encroaching shrub in comparison with open areas.

Studies indicated that degradation of land become a single major constraint in most rangeland ecosystems in dry lands of Africa (Ward, 2005; Gemedo et al., 2006; O’ Connor et al., 2014). The key drivers of ecological changes in arid and semi-arid rangeland environments are those leading to degraded condition for primary productivity (Reich et al., 2000). The processes leading to the degraded conditions for primary productivity are called “ land degradation” . The worst-case scenario for land degradation in dry land ecosystems often termed as “ desertification” , which can be regarded as an appearance of a continual decrease in the capability of a dry land environment to offer various goods and services (Reich et al., 2000). UNCCD et al. (2009) defined “ land” as a terrestrial biological productive system based on its primary productivity service, and land degradation is simply loss of provision of its primary productivity service of the land. Therefore, degradation of land can be described as “ the decline or complete damage of the biological productivity and/or socioeconomic benefit of land in general” .

Some authors defined rangeland degradation as general declines in biotic and abiotic qualities including losses of topsoil, changes in flora/fauna composition, and continuous reduction of primary productivity (Barrow, 1991). Others defined rangeland degradation as measurable declines in aboveground and belowground biomass production, in carbon storage and soil quality, and also in water infiltration and holding capacities (Dregne, 2002; Richardson et al., 2005). A third group of authors examined rangeland degradation primarily from socioeconomic perspectives. They viewed rangeland degradation as processes resulting in the capacity reduction of the ecosystems to support primary productivity and livestock production (Illius and O’ Connor, 1999).

Bush encroachment is explained to be a common problem of all dry lands globally (Bews, 1917; Archer, 1990; Hoffman and O’ Connor, 1999; Oba et al., 2000; Smit, 2002; Gibbens et al., 2005; Gemedo et al., 2006; O’ Connor, 2014), and the problem is the most serious in Africa. As aforementioned, bush encroachment has been tied to rangeland degradation and thus often perceived negatively (Gemedo et al., 2006). The causal linkages between encroachment of bush and degradation of rangeland have been the foci of scientific researches around the globe, especially in Africa, simply because bush encroachment poses serious difficulties to the socioeconomic development of dry lands (Vetter, 2005). Encroachment by bush is recognized as a major factor that aggravates degradation of rangeland in dry land areas of the world (Belay et al., 2013). For instance, more than 70% of rangelands have already experienced severe degradation (Dregne and Choun, 1992). Moreover, the swift incursion of woody plants into freely grazed areas shares a main responsibility to the reported rangeland degradation (Smit, 2002; Ward, 2005; Archer, 2010). Depending on the degree of bush encroachment and the characteristics of encroaching plants, encroachment of bush could adversely change the soil biological, physical, and chemical characteristics of the local biodiversity of rangeland ecosystems (Biederman and Boutton, 2009).

However, encroaching woody species sometimes had positive effects on the structure and function of various ecosystems (Smit, 2002; Gemedo et al., 2006; Maestre et al., 2009; O’ Connor et al., 2014) and these woody species may also be indispensable resources (Smit, 2002; Maestre et al., 2009; Eldridge et al., 2011). For example, bush encroachment-resulted woody plants were reported to have significantly contributed to livestock fodder, particularly in drought periods (Moleele and Perkins, 1998; Gemedo et al., 2006) and deliver a variety of products and amenities for the pastoral people in many parts of Africa (Abule et al., 2007). For example, leguminous woody plants encroaching the rangelands can improve soil quality through nitrogen fixation (Kgosikoma et al., 2012). Although bush encroachment has been negatively perceived from pastoralism point of view (Illius and O’ Connor, 1999; Gemedo et al., 2006; Van Auken, 2009), the positive ecological effects and the positive economic effects are also abundantly reported (Di Tomaso, 2000; Moleele et al., 2002; Parizek et al., 2002; Smith and Johnson, 2004; Maestre et al., 2009; Eldridge et al., 2011; Ratajczak et al., 2012). The nature of the effect depends on whether the encroaching woody species are trees or shrubs and on whether the encroaching woody species are indigenous or invasive. For instance, Angassa and Baars (2000) indicated that the total score of rangeland condition (a proxy for rangeland quality) was lower under bush encroached rangeland than non-encroached rangelands in Borana of southern parts of Ethiopia. It is now apparent that a straightforward description of bush encroachment could not be supported with critical explanation of available information. If a wider concept of land degradation is given to bush encroachment, its effects (negative or positive) on the structure and function of various ecosystems can be more fairly assessed.

Both indigenous and exotic woody species are known encroachers of the semi-arid rangelands. Encroachment by woody species of the genus Acacia, Dichrostachys, Prosopis, Balanites, and Grewiain arid and semi-arid rangelands was widely reported from North America, South America, Asia, and Australia (Prins and Van der Jeugd, 1992; Ayres et al., 2001; Smit, 2002; Gibbens et al., 2005; Ward, 2005; Solomon et al., 2007), and encroachment by woody plant species (indigenous and exotic) was reported to have occurred in many parts of Africa (Hoffman and O’ Connor, 1999; Smit, 2002; Ward, 2005; Gemedo et al., 2006; Russell and Ward, 2014; Ward et al., 2014). A host of other woody species was also reported to have encroached into rangelands and savannas, like Maytenus, Euclea, and Acalypha species. Table 1 includes the most common native woody species of trees and shrubs collected from published literatures to indicate how the grasslands and savannah rangelands are severely encroached by plants of various woody species in the semi-arid ecosystems.

Table 1

Table 1

Table 1 Most common encroached indigenous woody species in the semi-arid rangelands Family Species Regions Sources Fabaceae Acacia senegal, A. nubica, A. melifera, A. nilotica, A. tortilis, A. seyal, A. karroo, A. hebeclada, A. etbaica, A. erubescens, A. sieberiana, A. erioloba, A. brevispica, A. bussei, A. reficiens Africa, Australia, North America O’ Connor (1991); Ayres et al. (2001); Smit (2003); Ward (2005); Gemedo-Dalle et al. (2006); Abule et al. (2007); Tefera et al. (2007); Russell and Ward (2014); Ward et al. (2014) Dichrostachys cinerea Africa Hottman and O’ Connor (1999) Ormocarpum mimosoides Africa Tefera et al. (2007) Prosopis africana, P. tamarungo Africa Australia Bester and Reed (1997); Ayres et al. (2001) Prosopis cineraria, P. farcta, P. koelziana Asia Gibbens et al. (2005) Tiliaceae Grewia rewiaflava, G. bicolor, G. tenax, G. villosa, G. tembensis Africa Gemedo-Dalle et al. (2006); Eldridge et al. (2011) Balanitaceae Balanites aegyptiaca Africa Abule et al. (2007) Capparaceae Cadaba rotundifolia Africa Richter et al. (2001) Capparis tomentosa Africa Australia Prins and Van der Jeugd (1992) Anacardiaceae Rhus natalensis Africa Gemedo-Dalle et al. (2006) Lannea floccosa Africa Tefera et al. (2007) Combretaceae Terminalia brownie, T. prunioides, T. sericea Africa Abule et al. (2007); O’ Connor et al. (2014) Combretum molle Africa Tefera et al. (2007) Celastraceae Maytenusspp. Africa Tefera et al. (2007) Euphorbiaceae Acalypha fruticosa Africa Smit (2003) Ebenaceae Euclea crispa, E. divinorum E. shimperi Africa Asia Russell and Ward (2014) Bignoniaceae Rhigozum trichotomum Africa O’ Connor et al. (2014); Ward et al. (2014) Asteraceae Tarchonanthus camphoratus Africa O’ Connor et al. (2014); Ward et al. (2014) Pentzia incana Africa Ward et al. (2014) Burseraceae Commiphora africana Africa Tefera et al. (2007) Rhamnaceae Ziziphus mucronata, Ziziphussp. Africa Smit (2003); O’ Connor et al. (2014) Aloaceae Aloe schimperiana Africa Tefera et al. (2007) Rubiaceae Gardenia volkensii Africa Tefera et al. (2007)

Table 1 Most common encroached indigenous woody species in the semi-arid rangelands

The main genus of exotic plants of woody species encroaching rangelands in North and South Americas, Africa and Australia are Acacia, Prosopis, Opuntia, Lanthana and Larrea(Abule et al., 2005; Gemedo et al., 2006; Van Auken, 2009; Eldridge et al., 2011; O’ Connor et al., 2014). Some of the common undesirable noxious (exotic or alien) woody plant species of these genus and others are indicated in Table 2. It should be mentioned that although indigenous bush encroachment or spreading generally poses great threats to the existing ecological structure and function in semi-arid rangelands, exotic bush encroachment or invasion is associated with greater threats to the existing ecological structure and function.

Table 2

Table 2

Table 2 Most common encroached exotic woody species in the semi-arid rangelands Family Species Regions Sources Fabaceae Acacia mearnsii, A. nubica, A. drepanolobium, A. horrida, A. saligna, A. xanthoploea Africa, South America, Mexico Gemedo-Dalle et al. (2006); O’ Connor et al. (2014); Russell and Ward (2014) Prosopis glandulosa, P. velutina, P. torreyana, P. juliflora North America, Africa, South America, Mexico Archer (1990); Richter et al. (2001); Van Auken (2009) Senna didemobotrya Africa Gemedo-Dalle et al. (2006) Verbenaceae Lanthana camara, L. trifolia Africa Abule et al. (2007) Cactaceae Opuntia stricta North America, Africa, South America Grover and Musick (1990); Schmutz et al. (1991) Zygophyllacceae Larrea tridentata North America Van Auken (2009) Asteraceae Gutierreziasp. North America Van Auken (2009) Euphorbiaceae Euphorbia tirucalli Africa Tefera et al. (2007) Calitropis procera Africa Nott and Savage (1985); Zewdu and Oustelet (2007) Apocynaceae Cryptostegia grandiflora Africa, Australia Ayres et al. (2001); Eldridge et al. (2003); Abule et al. (2007) Asparagaceae Dracaena afromontana Africa Abule et al. (2007) Solanaceae Solanum incanum, S. somalense Africa Richter et al. (2001); Zewdu and Oustelet (2007)

Table 2 Most common encroached exotic woody species in the semi-arid rangelands

To explain the locally-observed bush encroachment, researchers proposed two models in earlier times: (1) the two-layer model of Walter and the one-layer model of Moir. As studies were later deepened, more comprehensive theories have emerged and four theories were put forward: (1)

state-and-transition, (2) equilibrium, (3) disequilibrium, and (4) non-equilibrium rangeland theories.

In the aforementioned mechanisms on bush encroachment, various driving factors were discussed (Table 3). Among them several are reported to be quite active in triggering the current encroachment of indigenous and exotic woody plant into grasslands and savannas in arid and semi-arid areas around the globe (Prins and Van der Jeugd, 1992; Smit, 2002; Van Auken, 2000). It should be particularly noted that combined effects of several driving factors and interacting effects of some driving factors were attributed to the bush encroachment, but the quantitative definition of the combined effects and the underlying processes of the interacting effects are still issues under heated debates (Van Auken, 2009; O’ Connor et al., 2014). Following is our brief review on the reported driving factors of bush encroachment.

Table 3

Table 3

Table 3 Drivers of bush encroachment reported in published studies Driver of bush encroachment Reference Heavy grazing pressure Bothma (2002); Sankaran et al. (2004) Ward (2005); Abule et al. (2007); Van Auken(2009) Reduced mobility of animals and break down of traditional land use Oba et al. (2000); Abule et al. (2007) Change in fire frequency and intensity Oba et al. (2000); Bothma (2002); Van Auken (2009) Absences and misuse of fire Belay et al. (2013); O’ Connor et al. (2014) Raising atmospheric CO2 concentration Buitenwerf et al. (2012) Climate change Van Auken (2000, 2009) Expansion of cultivation Abule et al. (2007) Nitrogen deposition Kö chy and Wilson (2001) Nutrient availability and rainfall patterns Sankaran et al. (2004); Ward (2005) Soil moisture stress Van Auken (2009) Soil nutrient levels Ward (2005) Kind of herbivory Van Auken (2009); O’ Connor et al. (2014) Weakening of the grass stratum by drought Clarke et al. (2000); Gemedo-Dalle et al. (2006) Spread of seed by livestock Abule et al. (2005); Van Auken (2009) Belay et al. (2013)

Table 3 Drivers of bush encroachment reported in published studies

Encroachment of bushes must unambiguously be viewed as societal problem at national, regional, and global levels and just not as isolated problem to pastoralist community alone. This indicates that government, pastoralists, concerned stockholders, and the private sector need to act together to search for solution. Different coping strategies were suggested by different studies. Moreover, pastoralists and ecologists agreed that disappearance of consistent fire has facilitated the spreading of woody species (Kgosikoma et al., 2012). Consequently, depending on the availability of fuel biomass, burning should be an essential component of dryland savanna ecosystems. When the significant of fire is understood, establishment of institutions is essential to develop scientifically sounding burning regimes of savanna ecosystems (Fatunbi and Dube, 2008). Scientifically, proper burning as a control strategy of bush encroachment requires the acquaintance of upcoming climatic situations and the knowledge of minimizing its adverse impacts (e.g., air pollution and carbon emission).

The main emphasis must be placed on the development of strategies that can prevent the grasslands from bush encroachment. For instance, a community-based investigation in Borana of southern Ethiopia the perception and attitude of pastoralists were reported to have provided an insight into the management and control of bush encroachment (Gemedo et al., 2006). Since pastoralists are the most important stakeholders of rangelands in Africa, their perceptions may hold the key to the success in controlling bush encroachment. Moreover, pastoralists believe that a ban on fire is one of the factors causing bush encroachment and re-introduction of fire and consolidation of customary rangeland management approaches are strongly favoured. In brief, the success in controlling bush encroachment depends on our better understanding of underlying causes and also on our willingness to take an integrated approach. Therefore, future investigation on bush encroachment must be inter- and intra-disciplinary oriented through involvement of various partnerships.

Bush encroachment (including proliferation of indigenous woody plant species and invasion of exotic/alien woody species) is one of the most widespread phenomena in semi-arid rangelands of Africa. Unfortunately, its exact causes are not fully understood yet and the understanding requires an integrated approach to expanding the relevant scientific knowledge and to absorbing the indigenous ecological knowledge. Accordingly, a comprehensive use of the indigenous and customary ecological understanding and an effective expansion of the relevant scientific knowledge are so crucial to protecting the rangelands. Even if a single driving factor is often reported to be responsible for the observed bush encroachment, most reports concluded that couplings and interactions among different driving factors are more often at work in determining the transitions of arid and semi-arid grasslands and savannas to bush lands and woodlands. A number of studies reported both the synergies and trade-offs between bush encroachments and rangeland degradation in the semi-arid ecosystems. Rangeland degradation has a much wider connotation than bush encroachment. However, if a wider concept of degradation of rangeland is applied to encroachment of bush, it can envisioned that there are a range of positive, negative, and neutral effects of bush encroachment.

The authors would like to acknowledge the information used in this review paper.