Gap Analysis for Plant Species
Walter Fertig, William A. Reiners, and Ronald L. Hartman
Department of Botany, University of Wyoming, Laramie
GAP projects have made an important contribution towards modeling distributions of terrestrial vertebrates but rarely have addressed other organisms. Vascular plants, in particular, have been treated primarily as components of vegetation rather than as individual species. As with vertebrates, there are also significant conservation gaps between distributions of plant species and protected areas. A GAP approach to plant species (Hartman and Reiners 1997) may be overdue in light of recent studies suggesting that 12.5% of the world’s flora is in danger of extinction (Walter and Gillett 1997).
Performing gap analyses for plants presents different challenges than for vertebrates. Theoretically, sedentary organisms like plants should be easier to map and model than motile organisms. In practice, plant distributions may be more sensitive to vagaries of local climate, topography, and substrate than vertebrates and require finer-grained environmental data for predictive modeling. State vascular plant floras also typically consist of thousands of species, compared with hundreds of species in similar terrestrial vertebrate faunas. Due to their high species richness, it is far less practical and more laborious to model distributions for entire floras than vertebrate faunas.
An alternative to attempting a gap analysis of an entire state flora is to randomly select a subsample of representative plant taxa. But on what basis does one select representative species? In Wyoming we have developed a computerized database (using Microsoft Access) to stratify the state’s 2,752 native and introduced vascular plant species and varieties (Dorn 1992; Fertig unpublished data) in order to randomly select subsets of taxa for predictive modeling purposes. To create this database, each taxon was ranked according to four criteria: geographic distribution pattern, in-state abundance, growth form, and major biome affinity.
Geographic distribution patterns reflect Wyoming’s context within the global range of each taxon. Geographic pattern classes included exotic (non-native to Wyoming), widespread (found commonly throughout the state and North America), sparse (numerically uncommon and restricted to widely scattered, specialized habitats in the state, but wide-ranging outside Wyoming), peripheral (at the edge of its range in Wyoming), disjunct (geographically isolated in Wyoming from its main continuous range), or endemic (restricted to a small geographic area only in Wyoming or 1-2 adjacent states). These distribution patterns were determined from state range maps produced by the University of Wyoming’s Rocky Mountain Herbarium and the literature.
In-state abundance was derived from Nature Conservancy Heritage ranks using a modified 5-point scale ranging from critically rare (1) to widespread and abundant (5) (Fertig 1997). State ranks take into account population size, trend, and number of occurrences. Each plant was also categorized by its typical mature growth form (tree, shrub, perennial forb, annual forb, perennial graminoid, annual graminoid, or fern-like). Lastly, biome affinity was based on the chief biogeographic region occupied by each taxon in North America as determined by the literature. Major biomes in the Wyoming flora include the Great Plains, eastern deciduous forest, Rocky Mountain forest, intermountain desert steppe, arctic/alpine, and wetlands (Barbour and Billings 1998).
The value of the database comes in its ability to readily sort groups of taxa with similar abundance, growth form, and distribution patterns. A total of 1260 such groups are theoretically possible, although only 262 are actually represented in the Wyoming flora (for example, there are no widespread, alpine/arctic tree species ranked S1 in the state). This stratification process has allowed us to randomly and proportionally sample species from selected groups for GAP modeling. It has also facilitated the study of interesting ecological and biogeographic patterns in the state flora (Figure 1).
 Figure 1. Geographic distribution patterns of the Wyoming flora. See text for description of distribution pattern types. |
For the purpose of GAP modeling in Wyoming, we have randomly selected 200 plant taxa from our stratified database to represent the state flora. Exotics and ubiquitous species (those ranked S4S5 or S5) were purposefully excluded from the modeling pool because they are not of preservation concern or because their ranges are too large to benefit from predictive modeling. Species were also eliminated if they were too uncommon within Wyoming and adjacent states to meet a minimum sample size of 25 locations required for our distributional modeling needs. With the removal of these species, the selection pool consisted of 2016 taxa in 212 groups. Small groups of similar composition were then combined and the relative proportion of each group to the total flora was calculated to completely stratify the selection pool. Two hundred taxa, with at least one species representing each combined group, were then randomly chosen.
Development of this database has facilitated gap analysis by stratifying the flora into manageable, yet representative, subsamples. Customized sortings are possible depending on the criteria, classes, and weightings one might wish to impose on the stratification process. We have also been able to quantify the flora of Wyoming and identify patterns in plant distributions related to growth form, state abundance, geographical distribution, and biome affinity. Similar databases can be developed in other states provided that adequate information is available on the status and distribution of the flora.
Literature Cited
Barbour, M.G., and W.D. Billings 1998. North American terrestrial vegetation, second edition. Cambridge University Press.
Dorn, R.D. 1992. Vascular plants of Wyoming, second edition. Mountain West Publishing, Cheyenne, Wyoming.
Fertig, W. 1997. Wyoming plant and animal species of special concern. Wyoming Natural Diversity Database, Laramie, Wyoming.
Hartman, R.L., and W.A. Reiners. 1997. Predicting plant species distributions in Wyoming: A GAP pilot project. Gap Analysis Bulletin 6:60.
Walter, K.S., and H.J. Gillett. 1997. The 1997 IUCN red list of threatened plants. IUCN, Gland, Switzerland, and Cambridge, United Kingdom.