Cattle waste reduces plant diversity in vernal pool mesocosms

In California, much of the remaining vernal pool habitat is used for cattle grazing. Some studies suggest that grazing helps promote native plant diversity on grasslands, but the impact of grazing on plants that reside in pool basins is largely unknown. We investigated how one aspect of cattle grazing, the deposition of waste, affects these plant species by adding dung and urine to mesocosms lined with vernal pool soil. As a result of dung input, orthophosphate, conductivity, and turbidity increased in our mesocosms while dissolved oxygen decreased. Such changes in water quality are consistent with a shift toward a eutrophic state. Algal biomass and percent-cover also increased in dung-treated mesocosms. When the mesocosms dried, vascular plant species richness and percent-cover in dung-treated mesocosms were reduced by up to 54% and 87%, respectively. We attribute this to light attenuation by algal mats that flourished in the nutrient-enriched water. We also found that dung input caused significant, but weak, shifts in the composition of the vascular plant community. We conclude that cattle grazing may be detrimental to plant communities in vernal pools via increased nutrient loading, which promotes algal growth. Any beneficial effects of grazing may thus be limited to the surrounding grassland. Studies that examine the regional-scale impacts of grazing on vernal pool grasslands should separately consider the impacts to local-scale (i.e., within-pool) plant diversity, as most of the threatened and endangered plant species of California vernal pools reside primarily in pool basins.     

Multiscale resistant kernel surfaces derived from inferred gene flow: An application with vernal pool breeding salamanders

The importance of assessing spatial data at multiple scales when modelling species–environment relationships has been highlighted by several empirical studies. However, no landscape genetics studies have optimized landscape resistance surfaces by evaluating relevant spatial predictors at multiple spatial scales. Here, we model multiscale/layer landscape resistance surfaces to estimate resistance to inferred gene flow for two vernal pool breeding salamander species, spotted (Ambystoma maculatum) and marbled (A. opacum) salamanders. Multiscale resistance surface models outperformed spatial layers modelled at their original spatial scale. A resistance surface with forest land cover at a 500‐m Gaussian kernel bandwidth and normalized vegetation index at a 100‐m Gaussian kernel bandwidth was the top optimized resistance surface for A. maculatum, while a resistance surface with traffic rate and topographic curvature, both at a 500‐m Gaussian kernel bandwidth, was the top optimized resistance surface for A. opacum. Species‐specific resistant kernels were fit at all vernal pools in our study area with the optimized multiscale/layer resistance surface controlling kernel spread. Vernal pools were then evaluated and scored based on surrounding upland habitat (local score) and connectivity with other vernal pools on the landscape, with resistant kernels driving vernal pool connectivity scores. As expected, vernal pools that scored highest were in areas within forested habitats and with high vernal pool densities and low species‐specific landscape resistance. Our findings highlight the success of using a novel analytical approach in a multiscale framework with applications beyond vernal pool amphibian conservation.     

Loss of biodiversity and hydrologic function in seasonal wetlands persists over 10 years of livestock grazing removal

Ecological restoration provides a means to increase biodiversity in ecosystems degraded by natural and human‐induced changes. In some systems, disturbances such as grazing can be key factors in the successful restoration of biodiversity and ecological function, but few studies have addressed this experimentally, especially over long time periods and at landscape scales. In this study, we excluded livestock grazing from plots within a grassland landscape containing vernal pools in the Central Valley of California for 10 years and compared vernal pool hydrology and plant community composition with areas grazed under an historic regime. In all 10 years, the relative cover of native plant species remained between 5 and 20% higher in the grazed versus ungrazed plots. This effect was particularly prominent on the pool edges, though evidence of invasion into the pool basins was evident later in the study. Native species richness was lower in the ungrazed plots with 10–20% fewer native species found in ungrazed versus grazed plots in all years except the first year of treatment. Ungrazed pools held water for a shorter period of time than pools grazed under an historic regime. By the ninth year of the study, ungrazed pools took up to 2 weeks longer to fill and dried down 1–2 weeks sooner at the end of the rainy season compared to grazed pools. The results of this study confirm that livestock grazing plays a key role in maintaining biodiversity and ecosystem function in vernal pools.     

Invertebrate community composition differs between restored and natural vernal pools

The loss of freshwater wetlands worldwide has underscored the importance of restoration to enhance biodiversity and functional objectives. While aquatic invertebrate communities within restored perennial freshwaters are well studied, few studies have occurred in the greatly reduced habitat of seasonal wetlands, such as vernal pools. California vernal pools have experienced high habitat loss and support many threatened or endangered invertebrate species. We compared 90 natural and 90 restored vernal pools of different ages across 10 sites throughout California and Southern Oregon using the Sars' method. Large branchiopod abundance, total invertebrate abundance, class richness, and community composition were assessed between pool types (natural vs. restored) and along environmental gradients (e.g. site, pool depth, surface area, age since restoration). Large branchiopod and total invertebrate abundance were 215 and 274% higher in natural pools than restored pools, but class richness was not different. Community composition was significantly different and driven by greater abundances of vernal pool fairy shrimp, San Diego fairy shrimp, Ostracoda, Cladocera, and Copepoda in natural pools. Few environmental or habitat variables explained patterns in richness or abundance. Our work demonstrates that restored pools, even those decades old, are different than natural pools. Future mitigation and monitoring guidelines for restored vernal pools should include quantitative evaluations for aquatic invertebrates. Restored pools are not adequate compensation for lost natural pools because they do not have the same ecological functions and values.     

Response of a Native Perennial Grass Stand to Disturbance in California's Coast Range Grassland

To assess the potential for enhancing an existing stand of native perennial grasses on a California Coast Range Grassland site, we experimentally manipulated the seasonal timing and presence of grazing for 3 years (1994 through 1996) and of autumn burning for 2 years (1994 and 1995) and measured species cover for 6 years (1993 through 1998). We subjected the species matrix to classification (TWINSPAN) and ordination (CCA) and tested the ordination site scores as well as diversity indices with linear mixed effects models. Four distinct plant community groups emerged from the classification. Two of these were dominated by annual grasses and two by perennial grasses. No treatment effects were observed on diversity. For composition, temporal and spatial random effects were important mixed effects model parameters, as was the fixed effect covariate, pre‐treatment CCA site score, indicating the importance of random environmental variation and initial starting conditions. Incorporation of these random effects and initial condition terms made for more powerful tests of the fixed effects, grazing season, and burning. We found no significant burning effects. Grazing removal imparted a shift in plant community from more annual‐dominated toward more perennial‐dominated vegetation. Individual perennial grass species responded differently according to genus and species. Nassella spp. increased gradually over time regardless of grazing treatment. Nassella pulchra (purple needlegrass) increase was greatest under spring grazing and N. lepida (foothill needlegrass) was greatest with grazing removal. Danthonia californica (California oatgrass) had little response over time under seasonal grazing treatments, but increased with grazing removal. Under relatively mesic weather conditions it appears that grazing removal from Coast Range Grasslands with existing native perennial grass populations can increase their cover. However if N. pulchra is the sole existing population, spring season‐restricted grazing should be equally effective at enhancing cover of the native grass species.     

California vernal pool endemic responses to hydroperiod,plant thatch,and nutrients

Many endemic large branchiopods inhabit ephemeral freshwater ecosystems, including California vernal pools. Hydroperiod, inundation length, has been well studied in these systems that cycle between aquatic and terrestrial phases, but species’ responses to other ecological processes are still poorly known. For example, temporal (plant thatch from the previous terrestrial phase) and spatial (nutrient runoff) factors may have strong effects on emergence and population densities during the aquatic phase. We examined the effects of hydroperiod stability, thatch, and nutrients on the emergence and density of 4 vernal pool endemic species: Branchinecta lynchi (Anostraca), Linderiella occidentalis (Anostraca), Lepidurus packardi (Notostraca), and Cyzicus californicus (Spinicaudata). A full factorial mesocosm experiment was conducted which measured species densities, along with water quality variables. Hydroperiod and thatch differentially affected 3 of the 4 species based on emergence timing and life cycle. Treatments had effects on many water quality variables, and these variables were correlated with densities. These results highlight how hydroperiod stability along with other processes can affect large branchiopod species in temporary freshwater ecosystems. California vernal pools are a greatly reduced habitat rich in endemic and endangered species (including Branchinecta lynchi and Lepidurus packardi), and therefore, these results have implications for conservation and management.     

Comparing amphibian habitat quality and functional success among natural,restored, and created vernal pools

The fact that several vernal pool restoration and creation attempts in eastern Pennsylvania and New Jersey have been paired with conservation of natural pools in the same area provided a valuable research opportunity to compare amphibian habitat quality between project sites and natural reference pools. To measure desired outcomes, we used successful reproduction and metamorphosis of two vernal pool indicator species, the wood frog and spotted salamander. Although many previous studies indicate that restored and created pools rarely replace function lost in the destruction of natural pools, success of vernal pool indicator species was not necessarily related to pool type in this study. Results indicate a strong correlation between reproductive success for both species and vernal pool size (i.e. mean depth and volume), regardless of pool type. Although overall survival rates of wood frog larvae were significantly higher in natural pools with hydroperiods between 12 and 35 weeks, wood frogs were also successful in one restored and one created vernal pool. Salamander survival rates were highest in two natural and two created pools, which had in common both greater volumes and higher proportions of forest land cover in the surrounding 1,000 m. The documented success of vernal pool indicator species in two well‐established created pools demonstrates that pool creation can sometimes restore communities and ecological functions lost, especially when nearby natural pools are degraded or destroyed.     

Grazing effects on plant cover,soil and microclimate in fragmented woodlands in south‐western Australia: implications for restoration

This study investigated the impacts of livestock grazing on native plant species cover, litter cover, soil surface condition, surface soil physical and chemical properties, surface soil hydrology, and near ground and soil microclimate in remnant Eucalyptus salmonophloia F. Muell woodlands. Vegetation and soil surveys were undertaken in three woodlands with a history of regular grazing and in three woodlands with a history of little or no grazing. Livestock grazing was associated with a decline in native perennial cover and an increase in exotic annual cover, reduced litter cover, reduced soil cryptogam cover, loss of surface soil microtopography, increased erosion, changes in the concentrations of soil nutrients, degradation of surface soil structure, reduced soil water infiltration rates and changes in near ground and soil microclimate. The results suggest that livestock grazing changes woodland conditions and disrupts the resource regulatory processes that maintain the natural biological array in E. salmonophloia woodlands. Consequently the conditions and resources in many remnant woodlands may be above or below critical thresholds for many species. The implications of these findings for restoration of plant species diversity and community structure are discussed. Simply removing livestock from degraded woodlands is unlikely to result in the restoration of plant species diversity and community structure. Restoration will require strategies that capture resources, increase their retention and improve microclimate.     

Changes in vernal pool edaphic settings through mitigation at the project and landscape scale

Vernal pool mitigation is a highly controversial process that has been frequently criticized for its inability to adequately replicate the ecosystem functions of the original intact wetlands. We analyzed past mitigation practices in two rapidly growing counties in California's Great Central Valley to determine if mitigation procedures are re-arranging the vernal pool landscape by substituting more common or less ecologically significant pool types (as defined by soil type and geomorphology) for rarer or ecologically richer pool types. Results indicate that most development projects impacting vernal pools conduct at least a portion of their mitigation requirements at a site with similar edaphic settings. However, when examined at a landscape-scale across all development projects, the more common edaphic settings such as Northern Hardpan and Low Terrace pools are increasing while more rare types such as Northern Claypan and Volcanic Mudflow pools are decreasing. Results also show that Drainageway pools, a less-specialized pool type with generally lower species richness, are becoming more common through mitigation. These results are confirmed by an analysis of landscape diversity, which showed that overall landscape diversity was lower at mitigation sites than at project sites. Despite these results, the ecological significance of vernal pool mitigation practices remains unclear for several reasons. The lack of maps showing exact locations of vernal pools at project sites make it difficult to precisely determine vernal pool acreage and distribution among edaphic settings. Additionally, more research is needed to determine precise relationships between edaphic settings and species distributions and the effects of mitigation area management practices on species distribution and persistence.     

Livestock exclusion alters plant species composition in fen meadows

Questions

Our study evaluated how species composition and plant traits that indicate functioning condition in fens responded to grazing cessation over time in an arid ecosystem of the western US. The specific questions addressed were: (i) how does livestock exclusion influence species composition in fens; (ii) is grazing cessation associated with shifts in species functional traits that indicate fen condition; and (iii) what is the pattern of response to livestock exclusion over time?

Location

Plumas National Forest, California, US.

Methods

We studied paired fenced and unfenced study sites in two fens to examine the effects of livestock exclusion. Parallel transects were established at each site, and plant species and ground cover were repeatedly surveyed, once prior to and multiple times following treatment, using 0.01 m² frequency frames. We used NMDS to analyse species composition, RLQ and fourth‐corner analysis to evaluate species functional traits and environmental variables, and linear mixed effects models to examine differences in responses between fenced and unfenced study sites over time.

Results

After fencing, we observed unexpected shifts in species composition and plant functional traits. Grazed sites were associated with peat‐forming obligate wetland, moss and sedge species, while fenced sites were characterized by non‐peat‐forming facultative upland, and upland forb, grass and early seral species. Species composition also varied between sites and sample years.

Conclusions

We found that livestock exclusion strongly affects plant species composition in fens, including promoting species with functional traits that indicate a loss of functioning condition, such as ruderal and upland species. Possible explanations for these observed shifts include: (1) biomass accumulation in the absence of herbivory, (2) competitive exclusion in fenced sites, (3) succession, (4) the abiotic conditions of our study sites, particularly hydrology and nutrient status, and (5) interactions among these factors. We conclude that degradation of fen wetlands caused by livestock grazing in the arid western US may not be reversed by excluding livestock alone.     

A Long‐Term Comparison of Hydrology and Plant Community Composition in Constructed Versus Naturally Occurring Vernal Pools

Successful restoration of ephemeral wetlands worldwide is particularly challenging, given the often‐precise relationship between hydrological features and plant community dynamics. Using a long‐term experiment in vernal pool restoration, we compare hydrological and vegetative characteristics of constructed pools with those of adjacent, naturally occurring reference pools. Although constructed and reference pools were similar in maximum water depth and duration of inundation at the beginning of our experiment in 2000, constructed pools were shallower and inundated for shorter periods by 2009. Native vernal pool species were able to establish populations in many constructed pools, and seeding sped their establishment. Comparing seeded plots in constructed pools with unseeded plots in reference pools, we found no significant difference in the cover of seeded species, native species, or exotic species in most years. In recent years, however, native species have declined in both constructed and reference pools. Finally, the cover of native vernal pool species was positively and non‐linearly associated with both water depth and seeding treatment. We conclude that the establishment of appropriate hydrological conditions was necessary, but not sufficient to promote successful performance of vernal pool species in constructed pools. Constructed pools with hydrologic conditions similar to those of reference pools were more likely to support populations of native vernal pool plant species, but only seeded pools were similar to reference pools in abundance of native cover. Most importantly, hydrological conditions in experimental pools have worsened since their construction, which may hamper persistence of native species in this restoration effort.     

The effects of an unpredictable precipitation regime on vernal pool hydrology

1. Vernal pools are small precipitation‐fed temporary wetlands once common in California. They are known for their numerous narrowly endemic plant and animal species, many of which are endangered. These pools experience the typical wet season/dry season regime of Mediterranean climates. Their hydrological characteristics are determined by a complex interaction between the highly variable climate and topographic relief. 2. Hypotheses regarding the effects on ponding of total precipitation, storm intensity and pattern were examined using long‐term weather records combined with two decades of data on the length and depth of inundation in 10 individual pools. Similarly, data on pool landscape position and microtopography allowed examination of the interactions between topography and rainfall amount and pattern. 3. The total amount of precipitation and length of inundation were strongly correlated. Landscape position affected ponding duration, with collector pools holding water longer than headwater pools. Basin microtopography interacted with climatic variability to determine the nature and extent of within‐basin microhabitats sufficiently different in hydrological and/or soil conditions to support or exclude individual species. The effect on hydroperiod of precipitation concentrated in a few months rather than spread more evenly over the season depended on total precipitation. 4. Changes in climate, the mound‐and‐depression landscape or pool microtopography could have profound impacts on the hydrology of individual pools as well as the array of hydrological conditions in the system. Given the individualistic responses of the numerous endemic species supported by vernal pools, any of these environmental changes could diminish their sustainability and increase the risk of species extinction. Conservation, restoration and management decisions should take these factors into account.     

Introduced and native herbivores have different effects on plant composition in low productivity ecosystems

Questions

Understanding how livestock grazing alters plant composition in low productivity environments is critical to managing livestock sustainably alongside native and introduced wild herbivore populations. We asked four questions: (1) does recent livestock and rabbit grazing reduce some plant attributes more strongly than others; (2) does grazing by introduced herbivores (i.e. livestock and rabbits) affect plants more strongly than native herbivores (i.e. kangaroos); (3) do the effects of recent livestock grazing differ from the legacy effects of livestock grazing; and (4) does the probability of occurrence of exotic plants increase with increasing net primary productivity (NPP)?

Location

South‐eastern Australia.

Methods

We measured the recent grazing activity of co‐occurring livestock (cattle, sheep, goats), rabbits and kangaroos by counting faecal pellets; historic grazing activity by measuring livestock tracks; and derived NPP from satellite imagery. We used a hierarchical GLMM to simultaneously model the presence or absence (i.e. probability of occurrence) of all plant species as a function of their attributes (growth form, lifespan and origin) to assess their average response to recent grazing, historic grazing and productivity in a broad‐scale regional study.

Results

Recent and historic livestock grazing, rabbit grazing and increasing NPP reduced the average probability of occurrence of plant species, although responses varied among plant attributes. Both recent and historic livestock grazing strongly reduced the average probability of occurrence of native species, and forbs and geophytes, but differed in their relative effects on other growth forms. Recent livestock grazing, rabbit grazing and NPP had similar effects, strongly reducing native species and forbs, geophytes, shrubs and sub‐shrubs. The overall effects of recent kangaroo grazing were relatively weak, with no clear trends for any given plant attribute.

Conclusion

Our results highlight the complex nature of grazing by introduced herbivores compared with native herbivores on different plant attributes. Land managers need to be aware that domestic European livestock, rabbits and other free‐ranging introduced livestock such as goats have detrimental impacts on native plant communities. Our results also show that kangaroo grazing has a relatively benign effect on plant occurrence.     

Recent changes in the plant composition of wetlands in the Jura Mountains

Aim

To assess vegetation changes in montane fens and wet meadows and their causes over 38 years.

Location

Wetlands, Jura Mountains (Switzerland and France).

Methods

Plots were inventoried in 1974 and re‐located in 2012 (quasi‐permanent plots) on the basis of sketches to assess changes in plant communities. The 110 plots belonged to five phytosociological alliances, two in oligotrophic fens (Caricion davallianae, Caricion fuscae) and three in wet meadows (Calthion, Molinion, Filipendulion). Changes between surveys were assessed with NMDS, and changes in species richness, Simpson diversity, species cover and frequency and the causes of these changes were evaluated by comparing ecological indicator values.

Results

Changes in species composition varied between alliances, with a general trend towards more nutrient‐rich flora with less light at ground level. Species diversity declined, with a marked decreasing trend for the typical species of each alliance. These species were partly replaced by species belonging to nitrophilous and mesophilous grasslands. However, no trend towards drier conditions was detected in these wetlands. The largest changes, with an important colonization by nitrophilous species, occurred in the Swiss sites, where grazing was banned 25 years ago. As a result of floral shifts, many plots previously belonging to fens or wet mesotrophic meadows shifted to an alliance of the wet meadows, generally Filipendulion. Moreover, communities showed a slight trend towards more thermophilous flora.

Conclusions

The investigated wetlands in the Jura Mountains have suffered mainly from eutrophication due to land‐use abandonment and N deposition, with a loss of typical species. Areas with constant land use (grazing or mowing) showed less marked changes in species composition. The most important action to conserve these wetlands is to maintain or reintroduce the traditional practices of extensive mowing and livestock grazing in the wetlands, especially in areas where they were abandoned 25 years ago. This previous land‐use change was intended to improve fen conservation, but it was obviously the wrong measure for conservation purposes.     

Slow recovery of arbuscular mycorrhizal fungi and plant community after fungicide application: An eight‐year experiment

Aim

In our previous study, we found strong effects of fungicide application on diversity and composition of grassland plant community. Here, we evaluated the recovery of the plant community and arbuscular mycorrhizal fungi (AMF ) infectivity after fungicide application and the effects of grazing management on the recovery.

Location

Northern Bohemia, Czech Republic.

Methods

We recorded plant species composition and AMF infectivity in permanent plots in dry grassland over a period of 5 years after termination of fungicide application and grazing introduction.

Results

The negative effect of fungicide on plant species composition, diversity, AMF infectivity and cover of forbs still persisted 5 years after the last fungicide application. The cover of graminoids decreased, and their cover reached the level before fungicide application. While grazing had no effect on plant species recovery, it led to recovery of AMF infectivity.

Conclusion

Although graminoids lost their dominance after termination of fungicide application and grazing led to the recovery of AMF infectivity, the dry grassland plant community was not completely restored. The forbs were not able to recolonize the site. Their absence might be caused by dispersal limitation or changes in restored AMF community composition. Direct seed sowing may thus be used to support the plant recovery.

     

Grazing and productivity alter individual grass size dynamics in semi-arid woodlands

The spatial arrangement of perennial vegetation is critical for ecosystem function in drylands. While much is known about how vegetation patches respond to grazing and abiotic conditions, the size dynamics of individual plants is mostly limited to theoretical studies. We measured the size distribution (mean, variance, skewness) and density of individual grasses, and grass species composition at 451 sites spanning a range of grazing intensities across three broad vegetation communities in semi-arid eastern Australia. We assessed the relative role of grazing by livestock (cattle and sheep), native (kangaroos) and introduced (rabbits) free ranging herbivores, and several environmental measures (productivity, diversity, composition and groundstorey plant cover) on the size distribution and density of individual grasses. We found mean grass size and density were more sensitive to shifts in grazing intensity and environmental conditions than size variance or the frequency of the smallest individuals (skewness), and shifts were mostly driven by site productivity and cattle and kangaroo grazing. Sheep grazing only reduced mean grass size, and rabbit grazing had no consistent effects. Importantly, we found that site productivity and species composition altered the impacts of grazing on grass density and size distribution. For example, increasing cattle grazing led to larger grasses in low productivity sites. It also led to larger, denser, more variable-sized grasses among grass species from sites with finer soil texture. Increasing kangaroo grazing led to smaller, denser individuals among grass species from sites with coarse soil texture. At high diversity sites kangaroo grazing led to denser, more homogenised grass sizes with a lower frequency of small individuals. Understanding the in situ response of individual plant sizes gives us insights into the processes driving shifts in perennial vegetation patchiness, improving our ability to predict how the spatial arrangement of ecosystems might change under global change scenarios.     

Predator–prey relationships within natural,restored, and created vernal pools

We performed a multiyear monitoring study to compare amphibian habitat quality among four natural, four restored, and six created pools. We used successful reproduction and metamorphosis of two vernal pool indicator species, the wood frog and spotted salamander, to represent desired outcomes. Ordination techniques were used to identify the aspects of habitat quality that were most correlated with desired outcomes. Previously published results indicated that pool depth, volume, and hydroperiod were among the best predictors of success, regardless of pool type. Observations in the first few years of monitoring also suggested that pools with longer hydroperiods had a greater abundance of aquatic predators of eggs and larvae of indicator species. This follow‐up study further explores and compares predator–prey relationships among pool types. We quantified within‐pool predator and prey abundance and diversity and collected another year of data on the reproductive success of indicator species. Our results confirmed that mean predator abundance was eight times higher in pools with longer hydroperiods. We documented a 96% decrease in wood frog survival rates in a semi‐permanent, natural pool following a 41% decrease in overhead canopy cover and an increase in green frog abundance. At the same time, wood frog reproductive success increased in nearby restored pools with lower predator abundance. Pools with the highest mean survival rates for the two indicator species combined were short‐ or long‐cycle pools (i.e. hydroperiod of 12–35 weeks) with low predator abundance (i.e. <1 organism L^?1) and greater proportions of arthropod prey relative to other food items.     

Slow recovery in desert perennial vegetation following prolonged human disturbance

Questions: How long may it take for desert perennial vegetation to recover from prolonged human disturbance and how do different plant community variables (i.e. diversity, density and cover) change during the recovery process? Location: Sonoran Desert, Arizona, USA. Methods: Since protection from grazing from 1907 onwards, plant diversity, density and cover of perennial species were monitored intermittently on ten 10 m × 10 m permanent plots on Tumamoc Hill, Tucson, Arizona, USA. Results: The study shows an exceptionally slow recovery of perennial vegetation from prolonged heavy grazing and other human impacts. Since protection, overall species richness and habitat heterogeneity at the study site continued to increase until the 1960s when diversity, density and cover had been stabilized. During the same period, overall plant density and cover also increased. Species turnover increased gradually with time but no significant relation between any of the three community variables and precipitation or Palmer Drought Severity Index (PDSI) was detected. Conclusions: It took more than 50 yr for the perennial vegetation to recover from prolonged human disturbance. The increases in plant species richness, density, and cover of the perennial vegetation were mostly due to the increase of herbaceous species, especially palatable species. The lack of a clear relationship between environment (e.g. precipitation) and community variables suggests that site history and plant life history must be taken into account in examining the nature of vegetation recovery processes after disturbance.