* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Long-term dynamics of biotic and abiotic resistance to exotic species
Survey
Document related concepts
Storage effect wikipedia , lookup
Molecular ecology wikipedia , lookup
Unified neutral theory of biodiversity wikipedia , lookup
Habitat conservation wikipedia , lookup
Occupancy–abundance relationship wikipedia , lookup
Biodiversity action plan wikipedia , lookup
Latitudinal gradients in species diversity wikipedia , lookup
Ecological fitting wikipedia , lookup
Reconciliation ecology wikipedia , lookup
Island restoration wikipedia , lookup
Introduced species wikipedia , lookup
Transcript
Ecological Applications, 21(6), 2011, pp. 2105–2118 Ó 2011 by the Ecological Society of America Long-term dynamics of biotic and abiotic resistance to exotic species invasion in restored vernal pool plant communities SHARON K. COLLINGE,1,2,4 CHRIS RAY,1 AND FRITZ GERHARDT3 1 2 Department of Ecology and Evolutionary Biology, 334 UCB, N122 Ramaley, University of Colorado, Boulder, Colorado 80309-0334 USA Environmental Studies Program, University of Colorado, Boulder, Colorado 80309-0397 USA 3 Beck Pond LLC, 394 Beck Pond Road, Newark, Vermont 05871 USA Abstract. Invasion of native ecosystems by exotic species can seriously threaten native biodiversity, alter ecosystem function, and inhibit conservation. Moreover, restoration of native plant communities is often impeded by competition from exotic species. Exotic species invasion may be limited by unfavorable abiotic conditions and by competition with native species, but the relative importance of biotic and abiotic factors remains controversial and may vary during the invasion process. We used a long-term experiment involving restored vernal pool plant communities to characterize the temporal dynamics of exotic species invasion, and to evaluate the relative support for biotic and abiotic factors affecting invasion resistance. Experimental pools (n ¼ 256) were divided among controls and several seeding treatments. In most treatments, native vernal pool species were initially more abundant than exotic species, and pools that initially received more native seeds exhibited lower frequencies of exotic species over time. However, even densely seeded pools were eventually dominated by exotic species, following extreme climatic events that reduced both native and exotic plant densities across the study site. By the sixth year of the experiment, most pools supported more exotics than native vernal pool species, regardless of seeding treatment or pool depth. Although deeper pools were less invaded by exotic species, two exotics (Hordeum marinum and Lolium multiflorum) were able to colonize deeper pools as soon as the cover of native species was reduced by climatic extremes. Based on an information-theoretic analysis, the best model of invasion resistance included a nonlinear effect of seeding treatment and both linear and nonlinear effects of pool depth. Pool depth received more support as a predictor of invasion resistance, but seeding intensity was also strongly supported in multivariate models of invasion, and was the best predictor of resistance to invasion by H. marinum and L. multiflorum. We conclude that extreme climatic events can facilitate exotic species invasions by both reducing abiotic constraints and weakening biotic resistance to invasion. Key words: abiotic constraints; biotic resistance; California; ephemeral wetlands; exotic species; grasslands; Hordeum marinum; invasion; Lolium multiflorum; vernal pools. INTRODUCTION Invasion of native ecosystems by exotic species is a leading cause of the loss of native biodiversity (Wilcove et al. 1998, Lockwood et al. 2007). Moreover, the conservation and restoration of native plant communities is often impeded by exotic species invasion (Suding et al. 2004). Ecological research regarding exotic species invasion has focused primarily on the roles of resource availability, productivity, disturbance, and native species diversity in determining the vulnerability of native communities (Fox and Fox 1986, Crawley 1987, Rejmánek 1989, Ewel et al. 1999, Mack et al. 2000). These factors may be categorized into two distinct classes of mechanisms: abiotic constraints, which include physManuscript received 30 May 2010; revised 23 December 2010; accepted 19 January 2011. Corresponding Editor: R. A. Hufbauer. 4 E-mail: [email protected] ical conditions that limit performance of exotics; and biotic resistance, whereby native communities compete with exotic species for limiting resources. Abiotic constraints are important when exotic species lack the morphological and physiological traits required to tolerate the abiotic conditions found in less invaded communities (Fox and Fox 1986, Rejmánek 1989). Many studies have shown that exotic species richness and abundance generally decrease with increasing abiotic stress (Forcella and Harvey 1983, D’Antonio 1993, King and Grace 2000, Harrison et al. 2001, Holway et al. 2002, Williamson and Harrison 2002, Gerhardt and Collinge 2003, Seabloom and van der Valk 2003, 2007). Alternatively, exotic species may be excluded from less invaded communities by competition from native species assemblages. As reviewed in Mack et al. (2000), numerous studies have addressed the roles of species and functional diversity, species identity, and productivity in determining community invasibility. Despite many theoretical and empirical studies of exotic species invasion, there is still no 2105 2106 SHARON K. COLLINGE ET AL. consensus regarding the relative importance of biotic and abiotic constraints in determining the invasion resistance of native communities. Although abiotic constraints and biotic resistance have typically been considered independently, they may act synergistically and they may also be influential at different points in time during the invasion process (D’Antonio 1993, Dietz and Edwards 2006). Dietz and Edwards (2006) suggested that invasions may occur in distinct phases: a primary phase, in which exotics increase rapidly in response to disturbances and/or resource enrichment, and a secondary phase, in which exotics respond to new ecological circumstances. In one of the few long-term studies of invasion, Wiser et al. (1998) found that the invasion process was controlled initially by dispersal limitation and later by other biotic and abiotic factors, especially those related to resource abundance and species richness. Abiotic constraints and biotic resistance may also interact in complex ways. For example, the strength of competition may differ among sites along abiotic gradients (Grime 1973, 1979; but see Wiens 1977), or the relative impact of abiotic constraints and biotic interactions may vary along an environmental gradient (Connell 1961, Gurevitch 1986, Hacker and Bertness 1999, Greiner La Peyre et al. 2001). The few studies that have examined interactions between abiotic and biotic controls on invasibility in plant communities have shown that the outcomes of competitive interactions among native and exotic species depend on abiotic conditions (e.g., David 1999), and that the relative importance of abiotic constraints and biotic resistance may vary among sites and among native communities depending on the life stage of the invading species (e.g., D’Antonio 1993). Our long-term study of restored vernal pool plant communities in California, USA, has allowed us to investigate the temporal dynamics of invasion and to evaluate the relative importance of biotic resistance and abiotic constraints governing invasion since 2002. Vernal pools are spatially discrete, ephemeral wetlands that occur in a variety of ecosystems. In Mediterranean climates, distinct wet and dry seasons lead to winter and spring inundation of pools, followed by complete drying in summer (Keeley and Zedler 1998, Grillas et al. 2004, Deil 2005). In California, these seasonal wetlands are characteristically underlain by an impervious hardpan or claypan (Holland and Jain 1988). Relative to other habitats, vernal pools harbor a disproportionate number of taxa characteristic of the California flora (Stebbins 1976, Zedler 1987, Keeley and Zedler 1998). The extreme variation in physical conditions within these pools, combined with the relative isolation of these patchily distributed habitats, has resulted in the development of unique communities of plants and animals. For example, California’s vernal pools harbor .60 endemic species of plants and invertebrates (Holland 1976, Stone 1990, Witham 1998), representing an important component of California’s native biolog- Ecological Applications Vol. 21, No. 6 ical diversity (Jokerst 1990, Pollak and Kan 1998). Due to destruction of vernal pools from agricultural and residential development, many species restricted to vernal pools in California are currently categorized as rare or endangered (Witham 1998). Thus, protection and restoration of native plant and animal populations and communities are critical aspects of vernal pool conservation within the region. In previous research, we compared patterns and mechanisms of plant community development and exotic species invasions in naturally occurring and restored vernal pools (Gerhardt and Collinge 2003, 2007, Collinge and Ray 2009). Because most exotic and native species in these pools are annuals, studying the year-to-year variation in species abundance and distribution can reveal the relative importance of factors that directly influence population dynamics. We observed that, in general, exotic species were strongly and negatively affected by the relatively harsh abiotic conditions associated with pool inundation during the winter wet phase. Most exotic species were unable to tolerate extensive periods of inundation, and therefore, their abundances in the field were negatively correlated with pool depth (Gerhardt and Collinge 2003). Greenhouse experiments in which water depth was experimentally controlled confirmed that inundation significantly affected most measures of exotic plant performance and played the dominant role in determining invasibility in vernal pools (Gerhardt and Collinge 2007). Both field and greenhouse studies suggested that there were relatively slight, though detectable, influences of native species cover on exotic species occurrence. In field surveys of naturally occurring and restored pools, the abundances of individual exotic species were more often related to water depth than to the cover of native species (Gerhardt and Collinge 2003). These studies offered meaningful insights into short-term patterns and processes of exotic species invasion, as they focused predominantly on recently reconstructed vernal pools (Gerhardt and Collinge 2003) and community dynamics within a single growing season (Gerhardt and Collinge 2007). We now have data from a long-term experiment in vernal pool restoration, including seven years of data on plant community dynamics in pools that were initially subject to different native plant seeding treatments (Collinge and Ray 2009). This seven-year time series allows us to more fully assess the temporal dynamics of invasions under varying abiotic and biotic conditions, during a critical phase in community development. In established communities, there may be less opportunity to observe the relative importance of biotic and abiotic factors because some constraints have been alleviated by the very changes (e.g., species turnover) that result in community establishment. Our data address the need for information on how exotic species invasion proceeds in restored communities, for practical applications in ecological restoration. September 2011 EXOTIC SPECIES IN RESTORED VERNAL POOLS In the present study, we addressed five questions regarding exotic species invasions in restored vernal pools: (1) Is there evidence for biotic resistance to exotic species invasions? (2) Are restoration trajectories divergent for experimental pools that were initially seeded with vernal pool species vs. those that were not seeded? (3) Are exotic species invasions constrained by abiotic factors, such as pool depth? (4) Do the two dominant exotic species in this system, Lolium multiflorum and Hordeum marinum, differ from other exotic species in their responses to pool depth and seeding treatment? (5) Are long-term dynamics well predicted by results from short-term experiments? In addition to answering these qualitative questions, we used information theory to quantify the relative contribution of biotic and abiotic factors affecting invasion. Previous studies of this system suggest that biotic resistance to exotic species should be related to S, the number of native seeds sown in a pool, while abiotic constraints on invasion are largely related to D, a pool’s maximum depth of inundation (Gerhardt and Collinge 2003, 2007, Collinge and Ray 2009). Using S and D as predictor variables, we developed a set of candidate models to explain invasion in each pool. We then used an information criterion to calculate the weighted average support for each predictor across models, providing a quantitative result that can be used as prior information in subsequent analyses of biotic vs. abiotic controls on invasion in this system. METHODS Study site We conducted this study at Travis Air Force Base (AFB), Solano County, California, USA. Travis AFB is located in the Sacramento Valley near Fairfield, California (38815 0 0000 N, 122800 0 0000 W; 6 m elevation). This area receives ;50 cm of rain per year, concentrated in the winter wet season from December to April. There is typically no measurable precipitation from May to November. Approximately 100 naturally occurring vernal pools occur in the 15-ha study area, embedded in a grassland matrix recently used for grazing and cultivation agriculture (BioSystems Analysis 1994). The site was leveled over 50 years ago, using simple tools to eliminate the basin-andmound topography typical of vernal pool habitats. This treatment did not destroy the hardpan clay layer that supports vernal pools in this region. We have identified 76 plant species in vernal pools at this study site during the past ten years (S. K. Collinge, unpublished manuscript). Nearly half (35 of the 76 species, or 46%) of these species are ‘‘exotics’’ not native to California, including Bromus hordeaceus, Centaurea solstitialis, Cotula coronopifolia, Erodium botrys, Hordeum marinum, and Lolium multiflorum (nomenclature follows Hickman [1993]). About one-third (26 of the 76, or 34%) are ‘‘vernal pool’’ or ‘‘native’’ species, considered to be native indicators or associates of vernal pools (Keeler-Wolf et al. 1998), 2107 including Deschampsia danthonioides, Downingia concolor, Eryngium vaseyi, Lasthenia conjugens, Layia chrysanthemoides, Plagiobothrys stipitatus, Pleuropogon californicus, and Veronica peregrina. The remaining 20% (15 of the 76) are ‘‘other’’ species native to California, but not necessarily restricted to vernal pool habitats; e.g., species described as ‘‘generalists’’ by Keeler-Wolf et al. (1998). In early December 1999, we constructed 256 ‘‘experimental’’ pools in a regular grid bisected by an aircraft landing strip. Many of these natural vernal pools line the sides of this small landing strip. Experimental pools were constructed on the same soils as natural pools in this area (Antioch-San Ysidro complex and San Ysidro sandy loam), were similar in elevation and topography (0–2% slopes), and were located in close proximity (150 m) to naturally occurring pools. Each experimental pool was located ;10 m from the nearest experimental pool, a typical distance between naturally occurring pools at this study site. Experimental pools were constructed in three size classes (25, 50, and 100 m2; n ¼ 85 each), to bracket the median area (78 m2) of natural pools within the study site. Using seeds collected from natural pools at the site, each experimental pool received one of five seeding treatments, as detailed in Collinge and Ray (2009). (1) ‘‘LACO-1x’’ pools received 100 seeds of L. conjugens in 1999. (2) ‘‘LACO-3x’’ pools received 100 seeds of L. conjugens in each year 1999–2001 (300 seeds total). (3) ‘‘Group-A’’ pools received 300 seeds in 1999 (100 seeds each of Group-A species L. conjugens, D. danthonioides, and E. vaseyi ), followed by 300 seeds in 2000 (100 seeds each of Group-B species L. conjugens, L. chrysanthemoides, and P. stipitatus). (4) ‘‘Group-B’’ pools received Group-B species in 1999 followed by Group-A species in 2000. Within each pool, a single 0.50 3 0.50 m plot was subjected to a seeding or control treatment, and each treatment plot was permanently marked for long-term observation. Seeds were mixed with a tablespoon of fine sand and sown over lightly raked plots. (5) Control pools were subject to a similar disturbance involving raking and ‘‘sowing’’ of seed-free sand. Of the 256 constructed pools, 249 received the assigned treatment and were analyzed for this study. Field data collection Because seasonal inundation is a defining property of vernal pools, we measured pool flooding during three wet seasons (limited financial resources precluded measurements in additional years). At weekly intervals during December–April, we measured water depth at the center of each pool’s marked treatment plot during two years early in the study (1999–2000 and 2001–2002) and one year late in the study (2008–2009). From these data, we calculated MiY, the maximum water depth recorded within pool i during the wet season ending in year Y, and Di, the mean maximum water depth in pool i. To avoid a temporal bias in Di, we first averaged Mi2000 and Mi2002 to obtain a single set of depths measured ‘‘early’’ 2108 Ecological Applications Vol. 21, No. 6 SHARON K. COLLINGE ET AL. in the study that were then averaged with the single set of depths measured many years later. In equation form, Di ¼ 0.5(Mearly þ M2009), where Mearly ¼ 0.5(M2000 þ M2002). Because each plot was positioned within a relatively deep portion of its pool, plot and pool depth have been closely related. For example, weekly measurements in 2001–2002 demonstrated no significant difference between maximum water depth at the plot center vs. at the pool center (paired t test, t ¼ 0.283, df ¼ 236, P ¼ 0.778). We measured plant occurrence in experimental pools during the flowering phase (April–May) each year during 2000–2008. We placed a square frame (0.50 3 0.50 m ¼ 0.25 m2) divided into 100 subquadrats (each 5 3 5 cm) over the treatment plot in each pool and recorded the frequency (number of subquadrats out of 100 in which the species occurred) of each species present in each year. Plants were identified to species using Hickman (1993), and a voucher collection is maintained in the Collinge Laboratory at the University of Colorado, Boulder, Colorado, USA. Data analyses Most of our analyses focused on the relative cover of two species classes: exotic species and species native to California vernal pools. We characterized the ‘‘cover’’ of each species class as the sum of the frequencies of all species in that class. We compared the cover of each species class across years and seeding treatments using 95% confidence intervals on mean cover, and using box plots showing medians, quartiles, range, and outliers. In box plots, notches on the sides of each box represent bounds on the median, such that there is ‘‘strong evidence’’ that two medians differ if their notches do not overlap (Chambers et al. 1983:62). The initial (2002) difference in exotic species cover between pools sown with 0–100 seeds and those sown with 300–600 seeds was determined using Welch’s two-sample t test. Theories of resource competition predict that the abundance of one species will decline with the abundance of a competitor, such that interspecific competition results in a negative correlation in the abundances of species competing within replicate habitats. Similarly, we sought evidence for interclass competition between native and exotic species by analyzing the relative cover of each class within pools. Because we did not expect strong competition in sparsely vegetated pools, we used quantile regression to detect and model a key aspect of competition. We used the 95th regression quantile to model an upper bound on the cover of native species in our plots as a function of increasing exotic species cover, interpreting a negative slope of regression as evidence for competition. Interannual variation in this model was investigated using 95% confidence bounds on each annual slope and intercept of regression. Using the same regression framework, we also investigated the converse effect of native species cover on the upper bound of exotic cover. We quantified the relative support for biotic and abiotic controls on invasion using an informationtheoretic approach (Burnham and Anderson 2002). Our response variable (y) was the temporal (2002– 2008) average cover of exotic species (e), alone or in combination with cover of native vernal pool species (v). We considered three indices of invasion: y ¼ e, y ¼ e v, and y ¼ e/(e þ v). The latter two indices include useful but potentially confounding information on the relative cover of native species. In each case, direct effects of seeding on the cover of native species may vie with indirect effects (like biotic resistance) to complicate the interpretation of any relationship between invasion and explanatory variables. Here, we focused on y ¼ e, while also summarizing results for the other two indices. Assuming that the success of exotic species in this system should be affected by the number of native seeds sown (S ) and the maximum depth of inundation (D) within each pool i, we developed a set of candidate models of the form yi ¼ f(Si, Di ). (We used a similar approach to model the frequency of the dominant exotic species.) Di was the temporally weighted average of maximum water depths, and Si was coded as the number of native seeds sown within the treatment plot of pool i, divided by 100. Candidate models included linear and/or nonlinear effects of either (or both) S and D. The evidence in support of each fitted model was compared using an information criterion (AICc). AICc discounts model likelihood by the number of model parameters and is appropriate for ranking non-nested models (Burnham and Anderson 2002). The model with lowest AICc is best supported by the data, and the positive difference between AICc values (DAICc) was used as a metric of relative support. Following Burnham and Anderson (2002), we assumed that models with DAICc , 2 were well supported by the data, while models with DAICc . 4 were poorly supported. DAICc values were used to generate an Akaike weight for each model, a scaled metric of relative support that can be summed across models to compare the relative support for individual predictors (Burnham and Anderson 2002), such as S and D. Analyses and plots were coded in R version 2.6.2 (R Development Core Team 2008). RESULTS Although vernal pool species were initially more abundant than exotic species in these restored pools, exotics dominated this system by 2007 (Fig. 1). ‘‘Other’’ species comprised ,10% of plant cover in each year, and were omitted from analyses. Biotic effects The initial dominance of vernal pool species was driven by results from seeded pools. Pools sown with more native seeds and more native species generally exhibited lower frequencies of exotic species than pools that were unseeded or received seeds of only one native species (Fig. 2). For Group-A and Group-B pools, these September 2011 EXOTIC SPECIES IN RESTORED VERNAL POOLS FIG. 1. Temporal progression, 2002–2008, of vernal pool species cover (dark gray) and exotic species cover (light gray) in permanent study plots within a large number of experimental vernal pools (n ¼ 119–249). The study area was on the Travis Air Force Base, California, USA. The cover of each species class within each pool was the percentage of sub-quadrats in which the species occurred (out of 100 sub-quadrats in a 100cm2 sampling quadrat), summed over all species in the class. Dark horizontal lines represent medians, boxes contain upper (75%) and lower (25%) quartiles, and dashed whiskers contain the range of the data excepting outliers. Outliers (lying outside the box by more than 1.5 times the interquartile range) are represented by open circles. effects of seeding were significant in early years of the experiment (Fig. 2a, b). In contrast, initial cover of exotic species was higher in pools sown with few or no native seeds (Fig. 2d, e). Specifically, mean cover of 2109 exotics in 2002 averaged 89.25 in control and LACO-1x pools, and 70.90 in other pools (Welch’s two-sample t test, t ¼ 2.199, df ¼ 224.583, P ¼ 0.014). However, even densely seeded Group-A and Group-B pools were dominated by exotic species in 2008, just seven years after the final seeding treatment (Fig. 2a, b). In both seeded and control pools, the combined cover of exotic and vernal pool species was limited. The scatter plots in Fig. 3 reveal the cover achieved by every community, from those comprised of only vernal pool species (points near the vertical axis in each graph), through mixed communities (central points), to those comprised of only exotic species (points near the horizontal axis). The lack of points in the upper right quadrant of each graph demonstrates a lack of communities with high frequencies of both exotic and vernal pool species. Seeded and control pools exhibited similar temporal patterns in community structure. Statistical evidence for this similarity derived from our models of maximum cover. When fitted separately to data from seeded vs. control pools, there was extensive overlap in confidence intervals for model parameters; e.g., the 95% CI on the slope (or intercept) of the quantile regression for seeded pools overlapped the fitted slope (or intercept) of the quantile regression for control pools in almost every year. For this reason, we report only on models of the combined data (dashed lines in Fig. 3). In each year, the slope of the 95th regression quantile relating the cover of native to exotic species was significantly negative (Table 1). Slopes were significantly more negative during 2004–2005, when FIG. 2. Invasion trajectories, 2002–2008, following different seeding treatments in experimental vernal pools. The cover of each species class within each pool was the percentage of sub-quadrats in which the species occurred (out of 100 sub-quadrats in a 100cm2 sampling quadrat), summed over all species in the class. Mean cover was calculated by averaging cover over all study pools within a treatment. Vertical lines represent 95% confidence intervals on mean cover. See Methods: Study site for seeding treatment descriptions. 2110 Ecological Applications Vol. 21, No. 6 SHARON K. COLLINGE ET AL. FIG. 3. Time series of community composition in experimental vernal pools. Scatter plots relate the total cover of native species to the total cover of exotic species in each pool. The cover of each species class within each pool was the percentage of sub-quadrats in which the species occurred (out of 100 sub-quadrats in a 100-cm2 sampling quadrat), summed over all species in the class. Total species cover was as limited in pools seeded with native species (open circles) as in unseeded control pools (solid circles). Upper bounds on total species cover were estimated using 95th regression quantiles (dashed lines), as summarized in Table 1. TABLE 1. Results of regression for the 95th quantile of each annual relationship between the cover of vernal pool species and exotic species (Fig. 3). Intercept Slope Year Fitted value Lower 95% CI Upper 95% CI Fitted value Lower 95% CI Upper 95% CI n 2002 2003 2004 2005 2006 2007 2008 287.85 318.25 260.45 250.00 198.00 225.42 142.04 275.05 309.42 248.84 232.95 184.06 202.32 135.77 339.01 331.60 277.73 260.83 204.94 239.04 173.66 0.8305 0.8375 1.3495 1.1905 0.9565 0.9015 0.6040 1.0394 0.9025 1.4527 1.2480 0.9800 1.0494 0.7673 0.6502 0.6547 1.2765 1.0951 0.9306 0.6450 0.5891 246 249 233 248 244 119 236 Note: Each regression models an upper bound on the summed frequencies of species in the community. Variation in sample size (n) represents variation in sampling effort (2007 only) and occasional lack of access to certain sampling plots. September 2011 EXOTIC SPECIES IN RESTORED VERNAL POOLS 2111 communities with relatively high cover of exotic species were lost (Fig. 3c, d). Abiotic effects Results from quantile regression of native cover on exotic cover included variation in intercepts among years (Table 1). Significant reductions in intercept values suggested three phases of decline in the resources available to this community: 95% confidence intervals did not overlap for sequential pairs of intercept values estimated for 2003–2004, 2005–2006, and 2007–2008 (Table 1). This pattern of decline was robust to a reversal of response and predictor values. Intercepts for 95th quantile regression models of maximum exotic cover fell from 246 to 175 between 2002 and 2008 (95% CI ¼ 219–271 for 2002, and 163–207 for 2008). Results summarized in Fig. 3 and Table 1 indicate that the upper bound on total species cover in these communities declined over time in the following sequence: In 2002, maximum cover was similar in all communities, regardless of whether native or exotic species dominated each treatment plot (the slope of the 95th regression quantile was not significantly different from1.0; Table 1). By 2004, maximum cover had declined in all communities, but especially in those dominated by exotic species (the intercept had declined significantly and the slope was significantly below 1.0). In 2006, 2007, and especially 2008, maximum cover declined mainly in pools dominated by native species (intercepts declined significantly in 2006 and 2008, and slopes rose above 1.0). Between 2006 and 2008, many communities appeared to ‘‘leap’’ from the vertical axis toward the horizontal axis in Fig. 3, representing a shift from the dominance of native species to the dominance of exotics. The dramatic shift in community composition after 2005 corresponded with two extreme climatic events. Extreme flooding occurred during the growing season of 2006 (Fig. 4a), when winter (October–April) precipitation in nearby Fairfield, California (;5 km from our study site) was 104.32 cm, more than two standard deviations above the long-term mean (for 1950–2009, 55.97 6 22.33 cm, mean 6 SD; available online).5 High precipitation occurred both early (December) and late (March) in the growing season of 2006 (Fig. 4b), and was followed by declines in both native and (especially) exotic species cover, as measured in April (Figs. 1 and 2). The density of points near the vertical axis in Fig. 3e indicates that exotic species cover was very low in many pools during 2006. Subsequent drought during the growing season of 2007 (Fig. 4a) corresponded with continued depression of native species cover, as the performance of vernal pool species was hampered by a total lack of precipitation in January and March of this year (Fig. 4b). In contrast with native species, exotic 5 hhttp://www.wrcc.dri.edui FIG. 4. Precipitation recorded at Fairfield, California (5 km from the study site). (a) Precipitation during the wet season (1 October–30 April) of 2006 was much higher than the longterm (1950–2008) mean, while precipitation during the wet season of 2007 was well below the mean. (b) Excess precipitation during 2006 occurred first and foremost in December, and the drought of 2007 began in January. species cover began to rebound during the drought year of 2007 (Figs. 1, 2, and 3f ). Pool depth accounted for much of the variation among pools in the average cover of exotic species, and there was strong evidence for nonlinearity in this relationship (Fig. 5a). The best model was quadratic, but a saturating model garnered similar support (DAICc ¼ 0.60) and a logistic model was nearly supported (DAICc ¼ 2.01 for a model with fitted upper and lower asymptotes). A linear model was not supported (DAICc ¼ 16.28). Quadratic curves were also best for modeling effects of pool depth on invasion data from each individual year (not shown), and for modeling other 2112 SHARON K. COLLINGE ET AL. Ecological Applications Vol. 21, No. 6 FIG. 5. Effects of inundation and sowing native seed on the invasion of vernal pool plant communities by exotic species. The mean cover of exotic species for each pool is the summed frequencies of exotic species within the study plot, averaged over the period 2002–2008. Frequency was the percentage of sub-quadrats in which the species occurred, out of 100 sub-quadrats in a 100cm2 quadrat. (a) The best relationship between exotic species cover and water depth was quadratic (shown) or saturating (DAICc ¼ 0.60). (b) Stronger biotic resistance was offered by pools that received more seeds of vernal pool species. This relationship appeared quadratic, but a linear decline in invasion index was also well supported (DAICc ¼ 0.39 and 6.21 for linear and null models, respectively). (c) The best model of invasion included a linear effect of pool depth and quadratic effects of pool depth and seeding treatment, but a similar model with linear effect of seeding garnered similar support (see Table 2). The linear fit between observed and modeled data (solid line) rises along the diagonal (dashed line). indices of invasion. For both y ¼ e/(e þ v) and y ¼ e v, quadratic models garnered support similar to the best logistic models (DAICc ranges were 0.77–1.45 for quadratic models, 4.74–5.92 for saturating models, and 22.41–24.53 for linear models). Across all indices of invasion, quadratic models reached their minima (lowest average indices of invasion) at annual maximum water depths averaging 22–25 cm. Relative support for biotic vs. abiotic resistance Exotic species cover was lower in pools that received more seeds of vernal pool species (Fig. 5b). The best model of this treatment effect was quadratic with a peak at zero (unseeded control pools), but a linear effect of seeding was also supported (DAICc ¼ 0.39). Results were similar for all three indices of invasion (quadratic models garnered best support and DAICc ranged 0.39– 1.77 for linear models, 6.21–19.52 for null models). Our model selection analysis suggested strong support for effects of both pool depth (D) and number of native seeds sown (S ) on invasion resistance in these vernal pools. Again, results were qualitatively identical and quantitatively similar among invasion indices. The best model of exotic species cover (y ¼ e) included a negative linear effect of pool depth and nonlinear effects of both seeding treatment and pool depth (y ¼ 155.85 0.52S2 10.16D þ 0.21D2). This same model also received highest support in analyses of y ¼ e v and y ¼ e/(e þ v), where its relative support was even higher (w ranged from 0.55 to 0.62) than in the analysis reported in Table 2 (w ¼ 0.46). A model with similar support included the same predictors excepting a linear, rather than nonlinear, September 2011 EXOTIC SPECIES IN RESTORED VERNAL POOLS 2113 TABLE 2. Relative support for alternative models of invasion resistance. Model y y y y y y y y y ; ; ; ; ; ; ; ; ; b0 b0 þ b1S b0 þ b1S2 b0 þ b1D b0 þ b1S þ b2D b0 þ b1D þ b2D2 b0 þ b1S þ b2D þ b3D2 b0 þ b1S2 þ b2D þ b3D2 b0 þ b1S þ b2S2 þ b3D þ b4D2 k LL AICc DAICc w 2 3 3 3 4 4 5 5 6 1317.10 1313.16 1312.97 1210.90 1202.28 1201.73 1190.96 1190.74 1190.73 2638.24 2632.42 2632.03 2427.90 2412.73 2411.62 2392.17 2391.73 2393.80 246.51 240.69 240.30 36.17 21.00 19.89 0.44 0.00 2.07 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.3719 0.4634 0.1646 Notes: The mean cover of exotic species, y ¼ e, is modeled as a linear or nonlinear function of seeding treatment (S ) and maximum water depth (D) in each pool (see Methods: Study site for seeding treatment descriptions). Each bx represents a fitted intercept or coefficient of regression. For each model, n ¼ 249. Relative support (Akaike weight, w) for each model is derived from an information criterion (AICc), which balances a metric of model fit (LL ¼ log-likelihood) against the number of fitted parameters (k). Better models have higher values of w and lower values of DAICc, the difference between AICc for a given model and the lowest AICc. The best model is indicated in boldface. effect of seeding (Table 2). The only other model with some support was the full model; however, its added parameter did little to improve model fit (cf. loglikelihood values in Table 2) and garnered a penalty (DAICc ¼ 2.07 ’ 2) almost exactly as high as expected for the addition of an uninformative parameter (Burnham and Anderson 2002). Using Akaike weights to compare support for individual predictors, linear and nonlinear effects of water depth were relatively strong, followed by nonlinear effects of seeding (Table 3). The order of relative support for each predictor was identical across indices of invasion, but the absolute support for S2 was elevated in analyses of y ¼ e v and y ¼ e/(e þ v) due to lower model uncertainty in those analyses. Our candidate models, chosen a priori based on previously observed effects of S and D, did not equalize the number of models (m) containing each variable, and considering all possible models is not sensible in this study. However, equal representation is easily achieved by omitting the three models in the middle of Table 2, and is nearly achieved by omitting the fourth and seventh models. In the first case, m ¼ 3 for each predictor and the support for each is unchanged (to four decimal places) relative to wp values in Table 3. In the second case, m ¼ 3 for all but D (m ¼ 4), and the only appreciable change is a rise in relative support for S2 (wp ¼ 0.16, 0.63, 0.63, and 0.63 for S, S2, D, and D2, respectively). Only three of our candidate models were strictly alternative (the second, third, and fourth models in Table 2), with no overlap in predictor variables. Among these univariate models, effects of seeding were overwhelmed by effects of water depth (wp ’ 0.00, 0.00, and 1.00 for S, S2, and D, respectively). Although pool depth reduced invasion by many exotic species, two exotic species were able to invade pools of any depth as soon as competition from vernal pool species was alleviated. Hordeum marinum and Lolium multiflorum were the dominant exotics within this system, combining to account for over 29% of total species cover across years, and over 51% of total species cover in 2008 (Fig. 6a). The rise in frequency of these two species followed the reduction in total species cover after the extreme flooding in 2006, and the continued reduction in vernal pool species cover after the extreme drought in 2007. Both H. marinum and L. multiflorum also exhibited relatively high tolerance for inundation in this experiment. By 2008, both species were most frequent in pools of intermediate depth (Fig. 6b, c), in direct contrast with the general effect of pool depth on exotic species invasion (Fig. 5a). Despite substantial scatter in the data, there was strong support for a quadratic effect of pool depth on the frequency of both species (DAICc ¼ 18.34 for linear and 17.90 for null models of L. multiflorum data; DAICc ¼ 3.21 for linear and 9.46 for null models of H. marinum data). Although we did not find support for typical effects of pool depth on invasion by the dominant exotic species in this system, we did find support for effects of biotic resistance to invasion by these species. Mean frequencies of L. multiflorum and H. marinum during 2002–2008 declined with the number of native seeds sown in a pool. The best model of L. multiflorum frequency ( f ) was quadratic ( f ¼ 40.10 0.19S2), with a peak at S ¼ 0. There was also support for a simple linear decline in frequency with number of seeds sown (DAICc ¼ 0.83), and relatively low support for the null model (DAICc ¼ 2.14). For H. marinum, the best model was linear ( f ¼ TABLE 3. Relative support for linear and nonlinear predictors of exotic species invasion, given the set of invasion models considered in Table 2. Predictor wp m S S2 D D2 0.54 0.63 1.00 1.00 4 3 6 4 Notes: Abbreviations are: S, seeding treatment; and D, maximum water depth in pool. The support for a given predictor (wp) is the sum of support (w in Table 2) for all models that incorporate the given predictor. Relative values of wp were insensitive to the apparent inequality in the number of models (m) in which different predictors appear. 2114 Ecological Applications Vol. 21, No. 6 SHARON K. COLLINGE ET AL. FIG. 6. Trends in cover for the two dominant exotic species in this system. (a) For each year, the mean frequencies of Lolium multiflorum (dark gray) and Hordeum marinum (light gray) are shown relative to the summed frequencies of all species occurring in sampled pools. Frequency was the percentage of sub-quadrats in which the species occurred, out of 100 sub-quadrats in a 100-cm2 quadrat. Each vertical line represents one standard error. Both (b) L. multiflorum and (c) H. marinum achieved higher frequency in pools of intermediate depth. The quadratic models (solid curves) shown garnered far more support than null models: DAICc values for null models were 17.90 in panel (b) and 9.46 in panel (c). 12.21–1.00S ), but there was also support for a quadratic decline in frequency with number of seeds sown (DAICc ¼ 0.86), and little support for the null model (DAICc ¼ 9.67). For both species, similar treatment effects were evident in every year of the study, but appeared strongest in 2007. The frequency of L. multiflorum in 2007 averaged 36% lower in the more heavily seeded Group-A and Group-B pools than in control pools. The best model for L. multiflorum in 2007 was f ¼ 55.68– 0.59S2 (DAICc was 1.62 for a negative linear effect of seeding, and 6.66 for the null model). The frequency of H. marinum in 2007 averaged 78% lower in Group-A and Group-B pools than in controls. The best model for H. marinum in 2007 was f ¼ 32.45–4.10S (DAICc was 0.62 for a quadratic decline, and 12.29 for the null model). DISCUSSION In summary, these long-term data from experimental plant communities provide new insights into the roles of biotic resistance, abiotic constraints, and disturbance events in determining the long-term dynamics of biological invasion. We found strong evidence for pronounced effects of both biotic resistance and abiotic constraints on exotic species invasions. Initially the trajectories of native and exotic species in plots seeded with native species differed from those in unseeded plots. However, the occurrence of extreme flooding followed by extreme drought dramatically impacted both native and exotic species, ultimately allowing exotic species to dominate nearly every plot. As detailed in the following paragraphs, these results have important implications for our understanding of biological invasions and also suggest general guidelines for the conservation of native September 2011 EXOTIC SPECIES IN RESTORED VERNAL POOLS plant communities, including the biologically rich plant communities associated with vernal pools. Temporal dynamics of invasion In control pools and pools seeded lightly with just one native species, exotic species tended to dominate the community throughout the experiment (Fig. 2d, e). In pools seeded more heavily with native species, we observed a temporal shift from dominance by native species to dominance by exotic species (Fig. 2a–c). This shift was facilitated by two years of extreme conditions that appeared to release exotic species from competition with native species in a two-step process. First, extreme flooding reduced the frequency of all species, but especially exotic species. Then, drought inhibited the rebound of vernal pool species, which, unlike most exotics, require substantial water resources for growth and seed production (Bauder 1989, Zedler 1990, Collinge et al. 2003). Thus, ironically, the importance of biotic resistance and abiotic constraints became most evident when each was relaxed. The rapid transition to communities dominated primarily by exotic species followed immediately after a relaxation in biotic resistance caused by flood-induced decline in both native and exotic species in. The drought of 2007, representing a relaxation in abiotic constraints on exotic species, allowed exotics to rebound quickly from the floods of 2006, while native species lagged or continued to decline (Fig. 1). Competition was evident in every year of this experiment, resulting in a strong inverse relationship between native and exotic species cover in the most productive pools (Table 1). We characterized the upper bound on joint cover of native and exotic species using a linear relationship (Fig. 3). A linear bound on joint cover (above which resource competition precludes growth) is appropriate if the strength of competition between individuals is similar within and between native and exotic species classes. In some years, however, maximum cover appears nonlinear, with higher maxima in pools dominated by either native or exotic species, and lower maxima in pools that contain a mixture of native and exotic species (see especially Fig. 3c–f ). This nonlinearity may signify that competition is more intense between native and exotic species classes than within each class (Gilpin and Ayala 1973). Heightened competition in pools with mixed assemblages, relative to pools dominated by either natives or exotics, may have several important effects on community dynamics, as follows. Under typical abiotic conditions, pools dominated by native species should be relatively difficult for exotic species to invade, as observed during the first five years of this experiment. Despite a general decline in the cover of native species during 2002–2006, exotics failed to replace natives prior to major climatic disturbances. Conversely, it may be difficult to restore native communities in pools overrun by exotics. In addition, the rate of transition from native to exotic species (or conversely) should be affected by 2115 the relative intensity of competition between species classes. Heightened competition between native and exotic species should hasten the outcome favored by current abiotic conditions, whether the outcome is invasion or restoration. Thus, even fleeting abiotic conditions may facilitate transitions like invasion. This prediction appears to be supported by the rapid invasion of exotics observed in this system during 2007–2008. Each of these predictions depends on the shape (linear or curvilinear) but not the position of the isocline describing maximum cover, so these predictions should not be affected by the observed fluctuation in maxima among years (Table 1, Fig. 3). Fluctuations in maximum productivity can also provide opportunities for invasion, and such fluctuations were clearly observed during the course of this study. Maximum cover in pools dominated by native species varied by more than a factor of two between 2003 and 2008, and evidenced a series of declines (Table 1). The most productive pools observed in this study were those dominated by native species, suggesting that native species can make more efficient use of abiotic resources in this system, under certain conditions. Maximum cover in pools dominated by exotic species also declined by nearly a factor of two between 2003 and 2004, and did not recover during the remainder of this study (Fig. 3). This pattern is most likely explained by climatic events that reduced the productivity of exotic species without eliminating their competitive advantage. The importance of climatic and other abiotic conditions affecting the outcome of competition need not overshadow the sometimes corresponding effects of biotic conditions. We have demonstrated strong effects of both abiotic ( pool depth) and biotic (seeding treatment) conditions on time-averaged indices of invasion and especially strong effects of biotic resistance to invasion by the dominant exotics in this system. Our temporal data reiterate the importance of biotic resistance to invasion. The floods of 2006 and drought of 2007 combined to produce a sequence of abiotic as well as biotic effects that hastened exotic species invasion, as follows: median (Fig. 1) and mean (Fig. 2) exotic species cover fell significantly between 2005 and 2006 (95% confidence intervals overlapped in only one treatment; Fig. 2a), and flooding is the only available explanation for this dramatic decline. Exotic cover then rebounded during the 2007 drought (Figs. 1 and 2), in a pattern at first suggesting mainly abiotic effects on invasion dynamics. In contrast, the average cover of native species rose and fell in 2007–2008 in a pattern that could be explained as lagged evidence of competition with exotic species. (A one-year lag in native response would be expected if competition affected seed production in these annual species.) Although reduced competition from exotic species may explain the slight rise in native cover in all seeded pools during 2007 (Fig. 2a–d), native cover was understandably outpaced during this drought year by rising exotic cover. In every treatment, 2116 SHARON K. COLLINGE ET AL. exotic cover in 2007 was significantly higher than in 2006 (95% confidence intervals did not overlap; Fig. 2), a pattern not seen in native cover under any treatment. The subsequent decline in native cover across all treatments in 2008 (Figs. 1–3) cannot be explained by the fairly typical amount of precipitation received in that year (Fig. 4a). This decline did, however, coincide with often unprecedented competition from exotic species (Fig. 2c–e). Similarly, details of the trajectory followed by exotic species suggest that invasion was facilitated by a relaxation in biotic resistance. Notably, the 2007 rise in exotic cover was increasingly dramatic in treatments with lower seeding intensity and most dramatic in control pools (Fig. 2), precisely where we would expect weaker competition from native species. The rapid replacement of native cover with exotic cover between 2006 and 2008, after four years of more gradual change, also suggests that invasion was facilitated by a positive feedback process or nonlinear, threshold dynamics. There were no significant interannual changes in median or mean native cover prior to the significant declines of 2007–2008, and no significant interannual increases in exotic cover prior to those of 2006–2008 (Figs. 1 and 2). Although many pools remained free of exotics during the first five years of the experiment (points against the vertical axis in Fig. 3a–e), exotic species were able to invade almost every one of these during just the last two years (Fig. 3g). Furthermore, although maximum productivity was significantly lower in 2007–2008 than earlier in this experiment (Fig. 3 and Table 1), exotics were able to rebound from their nadir in 2006 to achieve complete dominance by 2008. The dynamics of invasion in this field experiment were driven largely by two exotic species. Two exotic grasses, Lolium multiflorum and Hordeum marinum, were relatively unaffected by flooding in 2006, and appeared to thrive during the drought of 2007 (Fig. 6a). Previous field observations showed that H. marinum tended to occur in deeper pools and its abundance was positively associated with pool depth (Gerhardt and Collinge 2003). In contrast, L. multiflorum occurred primarily at pool edges and its abundance was negatively associated with pool depth. In the same field study, the abundances of both species were also negatively associated with native plant cover. In a greenhouse study, survival of both species was unaffected by inundation, but growth and reproduction decreased with increasing water depth (Gerhardt and Collinge 2007). However, both species were also strongly influenced by interactions with neighboring plants, as both growth and reproduction decreased with increasing densities of native neighbors. In agreement with these previous observations, both species appeared to benefit dramatically from reduced biotic resistance following the extreme climatic events in our long-term field study. In addition, our large data set allowed the detection of important nonlinearities in the effects of seeding and inundation on both species. Both species reached higher frequencies in pools of interme- Ecological Applications Vol. 21, No. 6 diate depth (Fig. 6b, c), allowing them to invade a broad spectrum of pools in this system. The pattern of exotic species invasion documented here was presaged by a greenhouse experiment demonstrating the potential for L. multiflorum and H. marinum to tolerate inundation (Gerhardt and Collinge 2007). The same experiment demonstrated negative effects of competition with native species on population growth in these two exotics. One conclusion of that short-term study was that abiotic resistance generally eclipsed biotic resistance to invasion in this system, and that ‘‘biotic resistance reduces invasibility only in more benign environments’’ (Gerhardt and Collinge 2007:922). The current results from our long-term field study suggest a different conclusion: Biotic resistance was overcome more easily in shallower pools, which presumably represent more benign environments for exotic species. The harshest environments (deepest pools) were invaded only after biotic resistance was removed and a key abiotic constraint (pool flooding) was relaxed by a drought. For example, L. multiflorum had become well established (frequency . 40%) in only eight pools deeper than 15 cm by 2005, 0 by 2006 (the year of extreme flooding), 20 by 2007 (the drought year), and 43 by 2008. Thus, this long-term field study has revealed the formerly cryptic potential for key effects of biotic resistance, as well as revealing temporal features (the sequence and confluence of events) that may allow widespread invasion of vernal pool communities. Lessons for conservation and restoration The success of many ecosystem restoration efforts is compromised by invasion of exotic species (D’Antonio and Meyerson 2002, Suding et al. 2004, Funk et al. 2008), so our efforts to clarify factors affecting invasion of restored vernal pools may be useful to future conservation and restoration efforts. We observed that exotic species cover varied inversely with the number of native seeds initially sown, but this effect lasted only while abiotic conditions were ‘‘typical.’’ Extreme flooding followed by extreme drought allowed invasion by exotics regardless of initial seeding treatment. These findings suggest that efforts to maintain and restore native biodiversity are likely to be most successful under a range of abiotic conditions typical for native ecosystems, and management interventions may be necessary following climatic extremes. Our results may also appear to suggest that more diverse communities offer more biotic resistance, in agreement with many other theoretical and empirical studies. However, density and diversity varied in tandem across our seeding treatments, with one exception. Although our study was not designed to evaluate the relative effects of density and diversity on biotic resistance, it did address effects of density in the case of one species, Lasthenia conjugens. Pools sown with 300 seeds of this species offered more biotic resistance than those sown with 100 seeds (Fig. 2c, d). This effect was September 2011 EXOTIC SPECIES IN RESTORED VERNAL POOLS likely driven by the initially higher abundance of L. conjugens in more heavily seeded pools (Collinge and Ray 2009). Although vernal pools are unique in many respects, the lessons learned from our long-term study of invasibility in this system have important implications for our understanding of the forces shaping biological invasions in general. In particular, the unique characteristics of these systems allowed us to detect the importance of climatic extremes in altering the abiotic and biotic environments. Such disturbances have been shown to play important roles in facilitating exotic species invasion in a variety of ecosystems (Hobbs and Huenneke 1992, Davis et al. 2000). In addition, our study allowed us to detect both the cryptic effects of biotic resistance and how the invasion process was altered by changes in abiotic conditions and overall ecosystem productivity. Both abiotic constraints and productivity have been shown to be important determinants of the success of biological invasions (see reviews in Ewel et al. 1999, Mack et al. 2000), but this long-term study demonstrated one of the mechanisms whereby this process occurs, principally by reducing the productivity and presumably the competitive edge of the native plant community. Many environmental factors that act as abiotic constraints on invasions, such as precipitation patterns and the resulting levels of inundation, are not obviously amenable to control by humans. In addition, climatic variability is likely to increase as a consequence of global climate change (IPCC 2007, Füssel 2009), leading a number of authors to propose that biological invasions will become more frequent (Mooney and Hobbs 2000 and chapters therein, Hellmann et al. 2008, Dukes et al. 2009). Our study contributes a relevant example of how extreme climatic events can enhance exotic species invasions. ACKNOWLEDGMENTS We are indebted to two anonymous reviewers for many helpful comments on a previous version of this manuscript. We especially thank Jaymee Marty for assistance with all aspects of this project, and we are especially grateful for the excellent field assistance of C. Battaglia, C. Christian, J. Compton, W. Johnson, T. Levy, and C. Wise. We thank M. Anderson, R. Holmes, and K. Neurer for facilitating our research at Travis AFB, and Y. Linhart, K. Rice, E. Warne, and D. Wright for their insights into vernal pool ecology and restoration. This project was funded by the Air Force Center for Environmental Excellence, the U.S. Fish and Wildlife Service, CH2M Hill, and a grant from the U.S. National Science Foundation (Long Term Research in Environmental Biology, DEB-0744520). Permission to collect seeds of L. conjugens was granted by the U.S. Fish and Wildlife Service to S. K. Collinge (permit number TE828382-3). LITERATURE CITED Bauder, E. T. 1989. Drought stress and competition effects on the local distribution of Pogogyne abramsii. Ecology 70:1083–1089. BioSystems Analysis. 1994. Vernal pool resources at Travis Air Force Base, Solano County, California. Weston Incorporated, Tiburon, California, USA. 2117 Burnham, K. P., and D. R. Anderson. 2002. Model selection and multi-model inference. Springer-Verlag, New York, New York, USA. Chambers, J. M., W. S. Cleveland, B. Kleiner, and P. A. Tukey. 1983. Graphical methods for data analysis. Wadsworth and Brooks/Cole, Duxbury Press, Boston, Massachusetts, USA. Collinge, S. K., and C. Ray. 2009. Transient patterns in the assembly of vernal pool plant communities. Ecology 90:3313–3323. Collinge, S. K., C. A. Wise, and B. Weaver. 2003. Germination, early growth, and flowering of a vernal pool annual in response to soil moisture and salinity. Madroño 50:83–93. Connell, J. H. 1961. The influence of interspecific competition and other factors on the distribution of the barnacle Chthamalus stellatus. Ecology 42:710–723. Crawley, M. J. 1987. What makes a community invasible? Symposium of the British Ecological Society 26:429–453. D’Antonio, C. M. 1993. Mechanisms controlling invasion of coastal plant communities by the alien succulent Carpobrotus edulis. Ecology 74:83–95. D’Antonio, C. M., and L. A. Meyerson. 2002. Exotic plant species as problems and solutions in ecological restoration: a synthesis. Restoration Ecology 10:703–713. David, P. G. 1999. Response of exotics to restored hydroperiod at Dupuis Reserve, Florida. Restoration Ecology 7:407–410. Davis, M. A., J. P. Grime, and K. Thompson. 2000. Fluctuating resources in plant communities: a general theory of invasibility. Journal of Ecology 88:528–534. Deil, U. 2005. A review on habitats, plant traits and vegetation of ephemeral wetlands: a global perspective. Phytocoenologia 35:533–706. Dietz, H., and P. J. Edwards. 2006. Recognition that causal processes change during plant invasion helps explain conflicts in evidence. Ecology 87:1359–1376. Dukes, J. S., et al. 2009. Responses of insect pests, pathogens and invasive species to climate change in the forests of northeastern North America: What can we predict? Canadian Journal of Forest Research 39:231–248. Ewel, J. J., et al. 1999. Deliberate introductions of species: research needs. Bioscience 49:619–630. Forcella, F., and S. J. Harvey. 1983. Eurasian weed infestation in western Montana in relation to vegetation and disturbance. Madroño 30:102–109. Fox, M. D., and B. J. Fox. 1986. The susceptibility of natural communities to invasion. Pages 57–66 in R. H. Groves, and J. J. Burdon, editors. Ecology of biological invasions. Cambridge University Press, Cambridge, UK. Funk, J. L., E. E. Cleland, K. N. Suding, and E. S. Zavaleta. 2008. Restoration through reassembly: Plant traits and invasion resistance. Trends in Ecology and Evolution 23:695–703. Füssel, H.-M. 2009. An updated assessment of the risks from climate change based on research published since the IPCC Fourth Assessment Report. Climatic Change 97:469–482. Gerhardt, F., and S. K. Collinge. 2003. Exotic plant invasions of vernal pools in the Central Valley of California, USA. Journal of Biogeography 30:1043–1052. Gerhardt, F., and S. K. Collinge. 2007. Abiotic constraints eclipse biotic resistance in determining invasibility along experimental vernal pool gradients. Ecological Applications 17:922–933. Gilpin, M. E., and F. J. Ayala. 1973. Global models of growth and competition. Proceedings of the National Academy of Sciences USA 70:3590–3593. Greiner La Peyre, M. K., J. B. Grace, and I. A. Mendelssohn. 2001. The importance of competition in regulating plant species abundance along a salinity gradient. Ecology 82:62– 69. Grillas, P., P. Gauthier, N. Yavercovski, and C. Perennou, editors. 2004. Mediterranean temporary pools. I. Issues 2118 SHARON K. COLLINGE ET AL. relating to conservation, functioning and management. Station Biologique de la Tour du Valat, Arles, France. Grime, J. P. 1973. Competitive exclusion in herbaceous vegetation. Nature 242:344–347. Grime, J. P. 1979. Plant strategies and vegetation processes. John Wiley and Sons, London, UK. Gurevitch, J. 1986. Competition and the local distribution of the grass Stipa neomexicana. Ecology 67:46–57. Hacker, S. D., and M. D. Bertness. 1999. Experimental evidence for factors maintaining plant species diversity in a New England salt marsh. Ecology 80:2064–2073. Harrison, S., K. Rice, and J. Maron. 2001. Habitat patchiness promotes invasion by alien grasses on serpentine soil. Biological Conservation 100:45–53. Hellmann, J. J., J. E. Byers, B. G. Bierwagen, and J. S. Dukes. 2008. Five potential consequences of climate change for invasive species. Conservation Biology 22:534–543. Hickman, J. C., editor. 1993. The Jepson manual: higher plants of California. University of California Press, Berkeley, California, USA. Hobbs, R. J., and L. F. Huenneke, editors. 1992. Disturbance, diversity, and invasions: implications for conservation. Conservation Biology 6:324–337. Holland, R. F. 1976. The vegetation of vernal pools: a survey. Pages 11–15 in S. Jain, editor. Vernal pools: their ecology and conservation. Institute of Ecology Publication, Number 9. University of California-Davis, Davis, California, USA. Holland, R. F., and S. Jain. 1988. Vernal pools. Pages 515–533 in M. G. Barbour, and J. Major, editors. Terrestrial vegetation of California. California Native Plant Society, Sacramento, California, USA. Holway, D. A., A. V. Suarez, and T. J. Case. 2002. Role of abiotic factors in governing susceptibility to invasions: A test with Argentine ants. Ecology 83:1610–1619. IPCC [Intergovernmental Panel on Climate Change]. 2007. Climate change 2007: the physical science basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change[S. Solomon, , D. Qin, M. Manning, Z. Chen, M. Marquis, K. B. Averyt, M. Tignor, and H. L. Miller, editors]. Cambridge University Press, Cambridge, UK. Jokerst, J. D. 1990. Floristic analysis of volcanic mudflow vernal pools. Pages 1–30 in D. H. Ikeda and R. A. Schlising, editors. Vernal pool plants: their habitat and biology. Studies from the Herbarium, Number 8. California State University, Chico, California, USA. Keeler-Wolf, T., D. Elam, K. Lewis, and S. Flint. 1998. California vernal pool assessment preliminary report. California Department of Fish and Game, Sacremento, California, USA. hhttp://www.dfg.ca.gov/biogeodata/wetlands/ pdfs/VernalPoolAssessmentPreliminaryReport.pdfi Keeley, J. E., and P. H. Zedler. 1998. Characterization and global distribution of vernal pools. Pages 1–14 in C. W. Witham, editor. Ecology, conservation, and management of vernal pool ecosystems. California Native Plant Society, Sacramento, California, USA. King, S. E., and J. B. Grace. 2000. The effects of soil flooding on the establishment of cogongrass (Imperata cylindrica), a nonindigenous invader of the southeastern United States. Wetlands 20:300–306. Ecological Applications Vol. 21, No. 6 Lockwood, J., M. Hoopes, and M. Marchetti. 2007. Invasion ecology. Blackwell, Malden, Massachusetts, USA. Mack, R. N., D. Simberloff, W. M. Lonsdale, H. Evans, M. Clout, and F. A. Bazzaz. 2000. Biotic invasions: causes, epidemiology, global consequences, and control. Ecological Applications 10:689–710. Mooney, H. A., and R. J. Hobbs. 2000. Invasive species in a changing world. Island Press, Washington, D.C., USA. Pollak, O, and T. Kan. 1998. The use of prescribed fire to control invasive exotic weeds at Jepson Prairie Preserve. Pages 241–249 in C. W. Witham, editor. Ecology, conservation, and management of vernal pool ecosystems. California Native Plant Society, Sacramento, California, USA. R Development Core Team. 2008. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. hhttp://www. r-project.org/i Rejmánek, M. 1989. Invasibility of plant communities. Pages 369–388 in J. A. Drake, editor. Biological invasions: a global perspective. John Wiley and Sons, London, UK. Seabloom, E. W., and A. G. van der Valk. 2003. Plant diversity, composition, and invasion of restored and natural prairie pothole wetlands: Implications for restoration. Wetlands 23:1–12. Stebbins, G. L. 1976. Ecological islands and vernal pools of California. Pages 1–4 in S. Jain, editor. Vernal pools: Their ecology and conservation. Institute of Ecology Publication, Number 9. University of California-Davis, Davis, California, USA. Stone, R. D. 1990. California’s endemic vernal pool plants: Some factors influencing their rarity and endangerment. Pages 89–107 in D. H. Ikeda, and R. A. Schlising, editors. Vernal pool plants: their habitat and biology. Studies from the Herbarium, Number 8. California State University, Chico, California, USA. Suding, K. N., K. L. Gross, and G. R. Houseman. 2004. Alternative states and positive feedbacks in restoration ecology. Trends in Ecology and Evolution 19:46–53. Wiens, J. A. 1977. On competition and variable environments. American Scientist 65:590–597. Wilcove, D. S., D. Rothstein, J. Dubow, A. Phillips, and E. Losos. 1998. Quantifying threats to imperiled species in the United States. BioScience 48:607–615. Williamson, J., and S. Harrison. 2002. Biotic and abiotic limits to the spread of exotic revegetation species. Ecological Applications 12:40–51. Wiser, S. K., R. B. Allen, P. W. Clinton, and K. H. Platt. 1998. Community structure and forest invasion by an exotic herb over 23 years. Ecology 79:2071–2081. Witham, C. W., editor. 1998. Ecology, conservation, and management of vernal pool ecosystems. California Native Plant Society, Sacramento, California, USA. Zedler, P. H. 1987. The ecology of southern California vernal pools: a community profile. Biological report 85 (7.11). US Fish and Wildlife Service, Washington, D.C., USA. Zedler, P. H. 1990. Life histories of vernal pool vascular plants. Pages 123–146 in D. H. Ikeda, and R. Schlising, editors. Vernal pool plants: their habitat and biology. Studies from the Herbarium, Number 8. California State University, Chico, California, USA.