There are relatively few data on the resistance of arid lands to invasion of non-native species. One study (Fox and Fox 1986) found little difference among plant communities in the percentage of non-native species present (with the exception of high invasion rates in Mediterranean-climate regions), suggesting that arid communities are not likely to differ from any others in their susceptibility to invasion. At one Chihuahuan Desert research area, non-native plants make up about 3% of the species present (8 of 270 species; J. Anderson, Jornada Long-Term Ecological Research program, personal communication, 1994), a number typical of those reported by Fox and Fox. The pool of potential invaders or new species will likely be smaller for more arid systems, owing to the relatively low proportion of potential colonists able to cope with the physiological demands of desert life, and perhaps to the low likelihood of desert species being transported as weeds or agricultural pests. Abu-Irmaileh (1994) compared germination of native and non-native plants in osmotic and water-stressed conditions, and found that in general non-native specics were unable to germinate under such desert conditions. As in other habitats, many non-native plants arc able to establish only after direct disturbance. On the other hand, Bromus tectorum and some other non-native plants are able to invade and persist in undisturbed North American dry grasslands (e.g. Brandt and Rickard 1994). Where these invaders are successful, they may have dramatic negative effects on the diversity and productivity of native plants (e.g. Anable et at. 1992).
In general, invasive plants of desert regions are most successful and conspicuous in riparian or wetland systems (Loope et al. 1988; Australian National Parks and Wildlife Service 1991). The most dramatic example, perhaps, is that of Tamarix, an Asian genus of trees that is known for invading desert riparian habitat in North America, Australia and elsewhere. Its high évapotranspiration can have strong impacts on local water tables, and its presence in riparian channels often spurs the accumulation of stable sediments, thus altering stream morphology and function (Graf 1978). The dense shade, heavy accumulations of salt in leaf litter, and other properties of tamarisk result in an elimination of most native plants and animals from the site (e.g. Loope et al. 1988). Bermuda-grass (Cynodon dactyl on) has invaded desert riparian areas and can have an effect on the entire disturbance cycle and successional dynamics by stabilizing the sediment bed in the streams, which normally have frequent flood-initiated disturbances of sediment and biota (Dudley and Grimm 1994).
In Australia, plant invasions of arid and semi-arid regions are often characterized by the dominance of a single invader over vast areas (Australian National Parks and Wildlife Service 1991). There, invaders of varying growth form, from trees and shrubs to annua! plants, appear to occupy space and pre-empt moisture and nutrients that would otherwise be available to native vegetation with different morphology and phenology. Both annual and perennial grass invaders are known to alter fire regimes in semi-arid systems, often causing further reduction of native vegetation (D'Antonio and Vitousek 1992).
There arc numerous cases of the active elimination of vertebrate species from arid ecosystems. In many cases, these actions have targeted carnivores and predators, owing to the risk of livestock losses and danger to human life (e.g. extinction of the Mexican wolf in the wilds of the US Southwest, recounted by Brown (1983), and rattlesnake "roundups" in the same region (Warwick 1990), but we found no published study of response of native rodent or lagomorph populations. Other hunted animals include burrowing mammals and those herbivores viewed as competitors for forage (prairie dogs, Australian marsupials). Brown and Heske (1990) and Heske et al. (1993) have demonstrated that the elimination of small rodents (in particular, kangaroo rats) from experimental plots in the US Chihuahuan desert can result in a decrease in shrub abundance and an increase in the abundance of at least some perennial grasses. Interestingly, a non-native grass showed the most dramatic increase. A reduction of soil disturbance by the rodents appeared to be the primary factor facilitating the change in vegetation.
In other cases the decline of native mammals has been less deliberate but no less influential. Perhaps the most dramatic example is that of Australia, where the introduction of Europeans and their livestock has been associated with the reduction or extinction of a diverse group of medium-sized marsupials. Hypotheses for these negative effects range from direct competition for forage with livestock and direct impacts of predation by introduced carnivores. to indirect effects of removal of cover and decreasing patchiness of the landscape due to widespread fire suppression (Burbidge and McKenzie
1989; Morton 1990; Short and Turner 1994). Dawson and Ellis (1994) described significant overlap in the diets of kangaroos and livestock, with interference (and impact on grazing-sensitive shrubs) most severe during drought.
Was this article helpful?