Contributions from the field of population biology hold promise for understanding and managing invasiveness; invasive species also offer excellent opportunities to study basic processes in population biology. Life history studies and demographic models may be valuable for examining the introduction of invasive species and identifying life history stages where management will be most effective. Evolutionary processes may be key features in determining whether invasive species establish and spread. Studies of genetic diversity and evolutionary changes should be useful for 0066-4162/01/1215-0305$14.00 305 Annu. Rev. Ecol. Syst. 2001.32:305-332. Downloaded from www.annualreviews.org by NORTH CAROLINA STATE UNIVERSITY on 09/26/12. For personal use only. 306 SAKAI ET AL.understanding the potential for colonization and establishment, geographic patterns of invasion and range expansion, lag times, and the potential for evolutionary responses to novel environments, including management practices. The consequences of biological invasions permit study of basic evolutionary processes, as invaders often evolve rapidly in response to novel abiotic and biotic conditions, and native species evolve in response to the invasion.

No abstract

No abstract

No abstract

Although the destruction of tropical rain forests receives much attention, tropical dry forests are in general far more threatened and endangered. Eliminating grazing ungulates is often considered a key first step toward protecting these ecosystems, but few studies have investigated the long‐term effects of this technique. We examined the effects of ungulate exclusion from a 2.3‐ha native dry‐forest preserve on the island of Hawaii by comparing its present flora to the flora of an adjacent area subjected to continuous grazing since the preserve was fenced over 40 years ago. Relative to this adjacent area, the fenced preserve contained a more diverse flora with substantially greater coverage of native overstory and understory species. Until recently, however, regeneration of native canopy trees within the preserve appears to have been thwarted by a dominant herbaceous cover of alien fountain grass (  Pennisetum setaceum) and predation by alien rodent species. Our results indicate that although ungulate exclusion may be a necessary and critical first step, it is not sufficient to adequately preserve and maintain Hawaii's remaining tropical dry forest remnants. Our recent efforts to control the dominant alien species within the fenced preserve suggest that this practice may facilitate both the regeneration of native species and the colonization and potential invasion of new alien plants. Comparisons of seedlings of the dominant native canopy tree Diospyros sandwicensis growing in sites both dominated by and free of fountain grass suggested that fountain grass inhibits Diospyros seedling growth and photosynthesis but may increase survival if seedlings are protected from ungulates.

The transition from biotic to wind pollination and the consequencesof this transition for the evolution of dioecious breeding systems wereinvestigated in Schiedea and Alsinidendron(Caryophyllaceae: Alsinoideae), genera endemic to the Hawaiian Islands. The potential for wind pollination was studied for five species ofSchiedea using a wind tunnel. Morphological correlates of windpollination for these species were then used to infer the presence orabsence of wind pollination in the remaining Schiedea species. Hermaphroditic Alsinidendron and Schiedea species,which occur in mesic to wet forests, or less commonly in dry habitats,show little or no indication of wind pollination. These species had lowpollen:ovule ratios, large relative pollen size, diffuse inflorescences,substantial nectar production in several cases, and appear to bebiotically pollinated or autogamous. Sexually dimorphic species, whichall occur in dry habitats, are wind pollinated, based on wind tunnelresults or morphological adaptations indicating the potential for windpollination. These adaptations include high pollen:ovule ratios, smallpollen size, moderately to highly condensed inflorescences, and reducednectaries and nectar production. Shifts to wind pollination anddimorphism are strongly correlated in Schiedea, suggesting theclose functional relationship of the twophenomena.

Because of its presumed high levels of dioecy (separate male and female plants), study of the native Hawaiian angiosperm flora has been important in development of many hypotheses about conditions favoring the evolution of dioecy. The importance of ecological correlates with dioecy has proven difficult to assess, however, because of lack of data on the origins of dioecy in the Hawaiian Islands. Clearly, these correlations are of greater interest in taxa where dioecy evolved in the Hawaiian Islands (autochthonous evolution of dioecy) than in taxa that are the result of dioecious colonists with subsequent speciation in the Hawaiian Islands. Because the Hawaiian flora is small and extremely isolated, colonists can be identified and their breeding systems hypothesized, thus allowing inferences on the origins of dioecy. Using current taxonomic information, the incidence of dioecy in native Hawaiian angiosperm species is 14.7%, lower than previous estimates, but still the highest of any known flora worldwide. Ten percent of the colonists were sexually dimorphic (dioecious, gynodioecious, polygamodioecious, and subdioecious), and over half (55.2%) of current dimorphic species are in lineages arising from dimorphic colonists, showing that dimorphism is high in part because colonists were dimorphic. Autochthonous evolution of dimorphism occurred in at least 12 lineages (e.g., hermaphroditic colonists of Bidens (Asteraceae), the Hawaiian Alsinoideae (Caryophyllaceae), and Hedyotis (Rubiaceae) led to species—rich lineages that include many dimorphic species). One—third (31.8%) of current dimorphic species are in lineages arising from monomorphic colonists. Dioecy in the Hawaiian Islands is a result of both dimorphic colonists as well as evolution of dioecy in Hawaiian lineages from hermaphroditic colonists. The high incidence of dimorphism is not because dimorphic colonists evolved more species per colonist than monomorphic colonists. Detailed studies of individual lineages are critical to elucidate causal factors in the evolution of dioecy.

The Hawaiian flora, because of its great isolation, high levels of endemism, known lineages, and high rates of endangerment, offers unique opportunities to explore patterns of endangerment related to phylogeny, ecological and life history traits, and geographic patterns. Nine percent of the native flora of 1159 taxa are already extinct, and 52.5% are at risk (extinct, endangered, vulnerable, or rare). Risk is strongly associated with limited geographic distribution at several scales: endemic taxa (native only to the Hawaiian Islands) are at far greater risk than indigenous taxa (with both Hawaiian and extra-Hawaiian ranges); single-island endemics are more at risk than multi-island endemics; small islands have the highest proportion of endemic taxa at risk; and endemics with more limited habitat distributions (elevation, community type) are more at risk. Historic population density is a strong predictor of risk, and taxa with low historic population densities are at greatest risk with rapid anthropogenic changes. Among the major islands, Maui Nui has the highest percent of taxa that are extinct. Kaua'i has the lowest percent of extinct taxa and the highest proportion of single-island endemic taxa that are rare. Endemic taxa at risk are associated with distributions in shrublands, forests, bogs, and cliff habitats. Endemic taxa with distributions in low elevation dry habitats have the highest proportion of taxa at risk, but the greatest absolute numbers of taxa at risk have distributions in mesic lowland and montane forests, and in wet montane forests. The life history patterns associated with risk are complicated, and inclusion of the effects of evolutionary relationships (lineages) changes some of these patterns. Species level analyses without respect to lineage shows risk associated with monomorphic (hermaphroditic) breeding systems and bird pollination because of the large number of hermaphroditic, bird-pollinated species in the Campanulaceae. Analyses incorporating the effect of lineage greatly reduce the impact of large lineages and result in an association of risk with insect pollination, and no effect of breeding system. There is no association of lineage size and the percent of taxa at risk within the lineage; endemic taxa from lineages with large radiations are at no greater risk than endemic single-taxon lineages. The percentages of taxa at risk at the family level in the Hawaiian Islands and worldwide (excluding Hawaiian taxa) are positively correlated, although flowering plant families in the Hawaiian Islands have a much greater proportion of taxa at risk. Some of the approaches described here may be useful to predict geographical and biological patterns of endangerment in island and island-like ecosystems under increasing pressures of endangerment and extinction.