Posteradication Flora And Fauna Dynamics

TNC recognized that there were many potential outcomes of the sheep eradication program (Schuyler 1993). The fundamental and most desirable one was that native biodiversity would be maintained or improved, but they were also aware that other outcomes could be undesirable or have unknown consequences. These included an increase in the pig population because of increased food and cover, increased fire frequency and extent resulting from increased vegetation biomass, and an increase in non-native plant species after being released from grazing pressure (Schuyler 1993). In 1984 a relatively small project was initiated to monitor vegetation change, primarily in grasslands. This was augmented with a broader photo-monitoring program. From 1990 to 1998 TNC implemented a conservation program that combined more extensive monitoring with field experiments to try to gain a better understanding of the dynamics of the pig population, rare plants, vegetation, selected vertebrate communities, and fire.


Annual counts and systematic hunting were conducted from 19901999 to collect data on the density and population structure of the pigs. Estimates of their island-wide abundance ranged from 800-5400, with pronounced crashes following years with low rainfall and poor mast crops (Figure 18.3A; R. Klinger unpublished data). A population model indicated that over the 150 years the pigs had been on SCI they were characterized by large fluctuations in population growth, with a long-term geometric mean X = 1 (Figure 18.3B; R. Klinger unpublished data).

Rooting occurred in all vegetation types, and in years when pig numbers were high an estimated 8-10% of the island was rooted. Impacts on rare plants could be severe. Populations of Arabis hoffmannii (Munz) Rollins and Thysanocarpus conchuliferus Greene, two of the rarest endemic plants on SCI, were either destroyed or seriously reduced by pig rooting. At the community level, shrub cover and density were reduced where rooting was frequent and intense, which favored both native and non-native annual forbs and non-native annual grasses.


A rapid increase in vegetation cover occurred after a 5-year drought ended in 1991 (Klinger et al. 2002). Cover, density, and recruitment of trees and shrubs increased across the island, especially in chaparral and pine forests (Figure 18.4; Klinger, unpublished data). These were largely native species, including many endemics. However, non-native species comprised a significant proportion of the herbaceous layer in many vegetation types. Mean cover of non-native grasses and forbs was 87% in grasslands, 46% in oak woodlands, and 26% in coastal scrub. Only in chaparral and pine forests did non-native species comprise <10% of the ground cover (Klinger, unpublished data). Ironically, the richness and cover of native herbaceous species were greatest in sites where sheep grazing had been the most intense (Klinger et al. 2002).

















0 20 40 60 80 100 120 140 160 Years

FIGURE 18.3 (A) Density estimates of feral pigs on Santa Cruz Island, California. (B) Stochastic population model of variation in finite population growth rate (Lambda) for the 150 years feral pigs were on Santa Cruz Island (Klinger, unpublished data).

The distribution of 77% of the island's 43 endemic species increased between 1991 and 1996 (Klinger et al. 2002). Of the five endemic species that were monitored most intensively, the abundance of Berberis pinnata Munz remained unchanged, Malacothamnus fasciculatus (B.L. Rob.) Kearney and the remaining populations of Arabis hoffmannii increased in population size, and Dudleya nesiotica Moran and Thysanocarpus




1991 1992 1993 1994 1995


1991 1992 1993 1994 1995


• Chaparral X Pine

FIGURE 18.4 There was a significant increase in density (# stems/60 m2) of trees and shrubs in chaparral (solid line) and Bishop pine forest (dashed line) communities on Santa Cruz Island, California, in the decade after feral sheep were eradicated from the island. The decrease in density in pine forests in 1995 was due to sheep remaining on the east end of the island re-invading the eastern stand of pines (Klinger, unpublished data).

conchuliferus had serious declines. The decline in Dudleya nesiotica was strongly correlated with an increase in cover of alien grasses and forbs and a related buildup of organic litter. Besides the populations that had been rooted by pigs, the disappearance of six other populations of Thy-sanocarpus conchuliferus appeared to be associated with an increase in cover of alien annual grass (Klinger et al. 2002).

The most explosive increase by a single plant species was the perennial non-native forb fennel (Foeniculum vulgare Mill.) (Brenton and Klinger 1994). Fennel had occurred on the island since at least the 1880s, but was never particularly dense except in a few localized areas (Beatty and Licari 1992, Junak et al. 1995). Between 1991 and 1992 the percentage of plots with fennel more than doubled (12% to 27%) and mean fennel cover increased from 10% to over 25% (Figure 18.5). In 1995 it occurred in 32% of the plots and mean cover was >38%. By 1998 cover was 50%, including some areas that were virtual monocultures (Figure 18.5).


TNC and the National Park Service conducted a fire management and research program on SCI from 1990-1999. The program is summarized in Klinger et al. (2004) and Klinger and Messer (2001), but an increase in fire frequency because of greater vegetation biomass following sheep eradication did not occur.

1991 1992 1993 1994 1995
1991 1992 1993 1994 1995

FIGURE 18.5 Primarily because of removal of cattle from Santa Cruz Island, the distribution and cover of the perennial non-native forb fennel (Foeniculum vulgare) doubled in a year and continued to increase. By 1998 mean cover was 50% and it occurred across 15% of the island (Klinger, unpublished data).


As anticipated, there was a strong response at both species and community levels to removal of grazers. Some of these responses were not surprising, such as the increase in vegetation biomass. But others, such as changes in abundance of some species, were not expected. More important than responses by individual species though were the complex interactions that presented unanticipated management challenges.

More times than not these interactions involved multiple interacting species, both native and non-native, as well as abiotic factors such as rainfall (Klinger and Messer 2001, Klinger et al. 2002). Though there were a number of examples of these interactions, probably the most useful to examine from the perspective of ecosystem engineering is the expansion of fennel and the subsequent efforts to manage it (Brenton and Klinger 1994, Brenton and Klinger 2002, Ogden and Rejmanek 2005).

Several lines of evidence indicated it was actually the removal of the cattle more than the sheep that led to the increased distribution and abundance of fennel (Beatty 1991, Klinger et al. 2002). But while removal of cattle grazing was certainly a critical influence, Brenton and Klinger (1994) proposed three other contributing factors without which they believed fennel's rapid expansion would not have occurred. These were the species being pre-adapted to the Mediterranean-type climate, its phenology and growth form allowing it to out-compete other herbaceous species for light and moisture, and a 5-6 year drought ending in a series of 3 wet years. Removal of grazing pressure probably allowed fennel to begin storing nutrients in its deep, fleshy taproot. But its phenology, prolific seed production, and tall stature (Klinger 2000) likely gave it an important advantage over potential competitors. This was translated into rapid growth and spread, both vegetative and from seed, once the drought ended.

Questions arose whether pigs were enhancing the spread of fennel, either through dispersing their seeds or by disturbing the soil. Pig abundance was high in the fennel, possibly because the tall, dense stands provided them with more shade, cover, and food than grasslands and coastal scrub. But while there was a positive correlation between fennel cover and rooting intensity (Klinger, unpublished data), this did not necessarily mean pig disturbance created conditions favoring its spread. Fennel cover is greater in loose soils (Klinger, unpublished data), and it may have been that more intense rooting simply occurred in denser patches of fennel because it was easier for pigs to root in them. And while fennel seeds were often found on the hides of pig kills, they were also found on the pelage of rodents and feathers of mist-netted birds. So while pigs may have enhanced the spread of fennel, many native animals did as well.

The structure of the fennel stands had quite different effects on native plants and animals. Species richness of native grasses and forbs had a negative relationship with fennel cover (Figure 18.6A). Not all herbaceous species were shaded out though; the understory of the stands was comprised primarily of non-native annual grasses and forbs (Brenton and Klinger 1994, 2002; Ogden and Rejmanek 2005). In contrast, abundance of native vertebrates, including several endemic species, was

30 25

30 25

Chaparral Coast Scrub Fennel Grassland

Chaparral Coast Scrub Fennel Grassland i-1-r

FIGURE 18.6 The spread of fennel between 1991 and 1998 had different effects on native plant and wildlife species on Santa Cruz Island, California: (A) the relationship between the number of native herbaceous species and fennel cover; (B) density of the endemic mouse Peromyscus maniculatus insularis in five vegetation types; and (C) mean avian species richness-plot in six vegetation types (Klinger, unpublished data).

greater than in some other vegetation types (Figure 18.6B, C). Presumably this was because of the structural complexity of the stands and higher levels of food.

+1 0

Post a comment