The study of community interactions is challenging for both plant and animal ecologists, but animal populations offer additional difficulties. Many species of animals are secretive, mobile, and long lived. Relative abundance can be quite variable over time and influenced by many factors, both biotic and abiotic. Marking individuals is rife with potential biases and accurate estimation of animal densities is problematic, even though methods are improving greatly (e.g., Buckland et al. 1993).
These logistical and methodological problems exist even in the study of small mammals (rodents), which have small home ranges and are abundant relative to other mammalian species (Dueser et al. 1989). Even some of the most frequently cited studies on competition in small mammal communities may have serious statistical flaws, which led Dueser et al. (1989:111) to conclude that we are less certain about the role of competition in rodent communities than is generally believed.
One of the primary problems with past studies is the lack of replication (Box 2.2). Dueser et al. (1989) reviewed 25 North American field experiments of competition that included both treatment and control plots. They reported that interspecific competition was prevalent, but that detection of competition was affected by experimental protocols. For example, competition was more evident between enclosed populations than between populations on open grids and, more disturbingly, competition was more evident in unreplicated than replicated studies. The evidence on competition in rodent communities is thus substantial, but ambiguous, which makes the generality of this evidence "unknowable but suspect" (Dueser et al. 1989:117). Dueser et al. (1989:123) emphasized
Box 2.2 Replication
Replication is the application of a treatment, or set of treatments, to more than one experimental unit. Replication demonstrates that observed trends are consistent, thereby reducing the possibility that a trend has occurred by chance. It increases the precision of estimates and it provides an estimate of experimental error, which is needed for appropriate statistical analyses.
In many cases, replication is precluded by ethical or logistical concerns. For example, it may be infeasible to conduct a replicated experiment on the behavior of large carnivores in a field setting, or on the effects of acid rain on trophic dynamics in large lakes. In such cases, inferences are restricted to the observed populations and considerable caution should be exercised when interpreting statements about broader populations (in these cases, carnivores or lakes not included in the study). The application of inferential statistics to data resulting from these studies is forced to rely on inappropriate error terms (i.e., sampling error is assumed to represent experimental error). As a result, these studies rely on pseudorepli-cation (Hurlbert 1984). The reader is encouraged to understand the causes and consequences of pseudoreplication with a thorough review of Hurlbert (1984) and Ramsey and Schafer (1997).
the need for increased statistical power in studies of community interactions by: (1) increasing replication; (2) reducing within-treatment variation; (3) using appropriate test statistics, such as one-tailed tests and repeated-measures analysis of variance; and (4) testing for large effect sizes by selecting an abundant species or by implementing the simultaneous removal and addition treatments that maximize between-treatment differences in density.
Management objectives frequently include unnatural conditions or situations. "Natural" levels of production (e.g., of red meat or wood) often do not meet societal demands. Similarly, "natural" levels of resource
Figure 2.5 (cont.) dynamics. Hare and lynx populations, interacting with their environment and each other, show cyclic highs and lows about every 10 years. Photos by Stephen DeStefano.
quality often are deemed insufficient (e.g., size of antlers or individual trees). Thus, managers focus on the creation of "unnatural" conditions to favor greater quantity or quality of various resources derived from ecosystems.
Many of the techniques designed to increase the quantity or quality of resources for human use rely on alterations in environmental conditions as a mechanism to alter species interactions and, therefore, change species composition. For example, removal of forest canopies generates large changes in several environmental parameters: at the soil surface, light and wind speed increase and diurnal temperature fluctuations become more pronounced. These and other changes in the physical environment tend to favor recruitment of species that are short lived, fast growing, and high in photosynthetic tissue at the expense of species that are long lived, slow growing, and high in structural tissue. The resulting suite of "early successional" species provides excellent habitat for some game species (e.g., white-tailed deer) compared with dense, closed-canopied forests. Conversely, many plant and animal species are found primarily in old forest stands with closed canopies, and the distribution and abundance of these species will likely be affected negatively by these management approaches (Figure 2.6).
Interference is a critically important interaction on most sites, and, in many cases, interference between plants can be manipulated to achieve management objectives. Effective manipulation of interference to meet objectives requires knowledge of two factors: the limiting resource(s) and the influence of the environment on the interaction. Considerable volumes of literature identify resource limitations and illustrate plant-environment relationships at an appropriate depth for the management of some systems. In other systems, relevant experiments should be conducted in order to identify and elucidate these relationships.
In most arid and semi-arid regions, water is the most limiting resource during much of the year, which implies that manipulation of other resources will have little or no impact on interference (and, hence, community structure). In contrast, light commonly constrains plant growth beneath forest canopies; thus, increasing the amount of light reaching understory plants is the most effective strategy for increasing the productivity of understory plants. These examples are simple and
obvious, yet they are representative of many sites. On other sites, plant productivity and diversity are constrained by one or more additional factors. Identifying the factors that impose the greatest constraints on the growth and survival of specific plants is a necessary first step toward manipulating the outcome of interference.
The ability of plants to interfere with neighbors is dependent on the environment (Keddy 1989; Grace and Tilman 1990). For example, the ability of many plants with the C3 photosynthetic pathway to interfere with the growth of C4 plants is enhanced under conditions of cool temperatures and low light (Williams et al. 1999). Similarly, recruitment of woody plants within a stand of grass is enhanced by a series of years with above-average precipitation or by defoliation of grasses by herbivores (McPherson 1997; Scholes and Archer 1997). Recognition of environmental effects on competitive interactions may allow manipulation of the environment to favor one plant or group of plants over others.
Herbivory represents an interaction and, in some cases, a management technique. Native herbivores associated with all wildland ecosystems remove an enormous amount of biomass each year and may constrain the establishment and growth of some species (Evans and Seastedt 1995). Thus, protection from these herbivores may be required to sustain populations of some plants. In effect, protection of desired plants from herbivores indicates that the primary constraint on plant establishment has been identified; protection represents an attempt to overcome this constraint. This strategy is particularly effective if plants require protection from herbivores only until they reach some minimum size or age. If native herbivores have been reduced via anthropogenic activities, reintroduction may represent a minimal-risk, low-cost strategy for reducing their forage plants (e.g., prairie dogs and woody plants; Weltzin et al. 1997; McPherson and Weltzin 2000).
Nonnative herbivores, notably livestock, have been introduced into most terrestrial ecosystems. The primary goals of livestock introduction are to convert biomass to a form suitable for human consumption and to produce other animal products (e.g., leather). Considerable volumes of literature address the consequences of livestock grazing and provide guidelines for maintaining the sustainability of this activity (e.g., Vallentine 1990; Heitschmidt and Stuth 1991; McPherson and Weltzin 2000). In general, livestock grazing requires minimal cultural input and represents a sustainable use of many areas if soils are inherently capable of recovering from livestock-induced compaction and if removal of nutrients does not exceed inputs from natural sources (e.g., precipitation, nitrogen-fixing organisms) (Figure 2.7).
Livestock can be manipulated to influence plant-plant interactions. Most large ungulates preferentially defoliate herbaceous plants, thereby enhancing the establishment and growth of woody plants. In contrast, goats tend to select woody plants and, hence, facilitate the spread of herbs. Managers can exploit these relationships to encourage desirable life forms and discourage undesirable ones. Historically, stocking rates of livestock on most natural areas greatly exceeded "sustainable" limits and thus impacted many populations of native flora and fauna. If applied judiciously, however, livestock grazing can be a useful method of manipulating vegetation in some areas (Box 2.3).
Several management activities represent attempts to overcome seed limitations on a site. As with herbivory, any attempt to increase the seed supply implies that seed dispersal is a primary constraint on establishment of desirable species. Constraints on seed availability can be overcome with a variety of techniques, ranging from direct sowing or planting to the attraction of animals likely to deposit the desired seeds.
Direct sowing is used by landscape architects, silviculturists, and revegetation specialists. Seeds of desirable species are frequently sown into areas after plants are removed by a disturbance (e.g., fire, overstory removal, road construction). The objective in these cases is to enhance the recruitment of desirable species by giving them a "head start" on undesirable species.
A less direct means of dispersing seeds involves feeding the seeds to domestic herbivores which are then released into target areas. Obviously, seeds must be resistant to deterioration or digestion by the animal. Seeds are defecated in feces, which provide a nutrient-rich, interference-free environment for germination and early establishment. This technique has been used with livestock to restore tropical dry forest in Guanacaste National Park, Costa Rica (Janzen 1986).
The historical consequences of livestock grazing continue to influence contemporary management decisions. Many vegetation managers refuse to consider livestock an appropriate tool for vegetation management because of historical "transgressions." For example, livestock are excluded from most national parks in the western United States: livestock are not native to these systems, and their presence appears to be inconsistent with the restoration and maintenance of preColumbian plant communities.
Grazing by livestock has contributed to extensive and widespread soil erosion and undesirable vegetation change in many of the world's ecosystems. Nonnative grazing animals caused, and continue to cause, reduced biological diversity. The livestock industry causes much economic and ecological harm with few societal benefits, and it appears to survive in many areas primarily because it offers a unique and colorful livelihood. However, livestock grazing is a useful tool for some site- and objective-specific goals of vegetation management, including the maintenance of biological diversity.
In the southwestern United States, nonnative annual grasses carpet the Sonoran Desert during years with above-average winter precipitation (Abbott and McPherson 1999). The associated increase in fine fuel enhances fire occurrence and spread, and most native plants are poorly adapted to fire. Long-lived succulents such as the giant saguaro cactus are particularly vulnerable. Opportunistic, short-term grazing by livestock may reduce the fine fuel load and prevent fires in these systems. Impacts of livestock grazing on this ecosystem presumably are minor relative to the long-term detrimental effects of fire (Abbott and McPherson 1999).
Finally, a very indirect means of overcoming constraints on seed dispersal relies on passive assistance from native animals. Again, seeds are resistant to deterioration or digestion, and they are defecated in desirable locations. Specifically, birds are attracted with artificial perches that serve as recruitment foci for woody plants (McClanahan and Wolfe 1993; Robinson and Handel 1993). The effectiveness of perches is enhanced by providing seeds of desirable plants in bird feeders. As with the application of much other ecological research, this strategy is not always successful; perching structures typically increase seed dispersal, but may not overcome other constraints on woody plant recruitment (Holl 1998).
Was this article helpful?