Abstract. The grassy meadow on the top of Marys Peak, in the Oregon Coast Range, is being invaded, primarily along the margins, by Abies procera. To examine vegetation and environmental changes across the forest‐meadow transition and to evaluate factors affecting tree invasion, belt transects were established at 10 sites. Correlations of environmental variables with ordination axes from Detrended Correspondence Analysis suggest that the sites were distributed along a moisture gradient. Although the sites varied considerably in species composition and environment, tree invasion of the meadow was occurring at all sites. Reduction in abundance of the dense herbaceous vegetation of the meadow is required before the small seedlings of Abies procera can become established. Almost all tree invasion of the meadow occurs in the narrow forest‐meadow ecocline because trees at the edge of the forest reduce the cover of herbaceous plants in the adjacent meadow. On some sites, above average snow accumulations further reduce vigor of meadow vegetation and increase tree establishment. On the driest sites wet summers facilitate tree establishment. Infrequent fires, which remove the fire sensitive Abies procera, are probably required to reverse the slow but persistent trend of forest expansion and ultimately maintain the meadow.

Abstract:We evaluated the floristic condition of freshwater palustrine wetlands dominated by wet meadow, emergent marsh, aquatic vegetation, or open water within the rapidly urbanizing area of Portland, Oregon. USA by (1~ characterizing plant species richness (presence/absencel and composition of naturally occurring wetlands (NOWs) and mitigation wetlands (MWs) and (2) identifying relationships between floristic characteristics and variables describing land-use, site conditions, and mitigation activities. Data were collected on 45 NOWs and 51 MWs. Overall species richness was high (365 plant taxa), but more than 50% of the species present on both NOWs and MWs were introduced. Only 14 species occurred on more than half the sites, and nine of them were invasive introduced species. The mean number of native species per site did not differ between land-use categories (ANOVA, F = 0.62 at 3 and 88 df, p = 0.6031); however, wetlands surrounded by agricultural and commercial/industrial/transportation corridor uses had more introduced species per site than wetlands surrounded by undeveloped land (Fishers Protected LSD at 88 df, p -< 0.05). Although overlapping in floristic composition, NOWs and MWs had significantly different (MRPP. p < 0.01301) species assemblages that were identified using TWlNSPAN. MRPP analyses for all sites showed that watershed, land-use, HGM class, percent cover of water, and MW age were significantly related to the floristic composition of thc study wetlands. Canonical correspondence analyses further revealed that the primary gradient tor species distribution in NOWs was related to moisture; the secondary gradient was related to land-use. The primary gradient also described a strong relationship between percent cover of water and HGM class. For MWs, the primary gradient was related to watershed location and surrounding landuse; the secondary gradient was related to percent cover of water and MW age. Most MWs (44 out or 51 sites) were depressions in various settings, so while HGM class separates NOWs from MWs, it does little to distinguish MW assemblages. Our results show that wetlands in the urbanizing study area are floristically degraded. Further, current wetland management practices are replacing natural marsh and wet meadow systems with ponds, resulting in changes in the composition of plant species assemblages.

Understanding hydrologic requirements of native and introduced species is critical to sustaining native plant communities in wetlands of disturbed landscapes. We examined plant assemblages, and 31 of the most common species comprising them, from emergent wetlands in an urbanizing area of the Pacific Northwest, USA, in relation to in situ, fine-scale hydrology. Percent cover by plant species was estimated in 2208 1-m 2 plots across 43 sites, with water depth at time of vegetation sampling measured in 432 plots. Three years of bi-weekly hydrologic data from each of the 43 sites were used to estimate mean surface water level and mean absolute difference (MAD) in surface water level for every plot. Nine assemblages of plant species that co-occur in the field were identified using TWINSPAN. The assemblage richest in native species occurred under intermediate hydrologic conditions and was bracketed by pasture grass dominated assemblages at drier conditions with low water level variability, and Phalaris arundinacea L. assemblages with higher mean water levels and variability. Results suggest minor changes in average water levels ($10 cm) or in variability (±2 cm in MAD) could promote a shift from assemblages dominated by natives to those dominated by invasive or alien taxa. Canonical correspondence analysis segregated the species into four groups related to hydrologic gradients. Each species response group was typified by taxa with similar optima for a given environmental variable, with each group related to a characteristic suite of hydrologic conditions. The most common species (P. arundinacea, Juncus effusus L., and Typha latifolia L.), each representing a different response group, exhibited unique responses in occurrence/abundance in relation to water level variability, but were abundant over a wide range of water depth. The realized niches of other species in each response group were more restricted, with peaks in cover confined to narrower ranges of water depth and variability.

No abstract

We evaluated the importance of alien species in existing vegetation along wadeable streams of a large, topographically diverse river basin in eastern Oregon, USA; sampling 165 plots (30 · 30 m) across 29 randomly selected 1-km stream reaches. Plots represented eight streamside community types associated with varying elevation, precipitation, and landform. Mantel comparisons, non-metric multidimensional scaling (NMS), and Spearman correlation identified relationships of alien species (n = 60) distribution and relative alien cover (RAC) to native species (n = 355) composition, 28 vegetation descriptors, 31 environmental variables, and 30 disturbance conditions. Alien species occurred in 93% of sample plots, in all community types, and along all sampled stream reaches; with RAC ranging from 0.1% to 47% and 1 to 24 alien species occurring along individual stream reaches. Alien richness and RAC were positively related to native diversity where invasion was limited (RAC < 5%), but negatively where invasion was more severe (RAC > 5%). RAC differed among community types: greatest in arid associations (shrubland/grassland), followed by associations with limited tree canopy cover (meadows, dry forest), and lowest in moist, closed forest associations; suggesting differences in invasion status or vulnerabilities to alien invasion among community types. Alien species, as a group had wider ecological amplitude than natives, and species composition among community types was less distinct when both alien and natives were considered compared to native species only. RAC was negatively related to elevation, precipitation, and tree cover; while positive relationships of RAC occurred with grazing pressure, upstream watershed size, stream order, overall level of exogenous disturbance, limited vegetated buffer, agriculture in the upstream watershed, floodplain or south facing slope locations, and proximity to roads. Taken together, these results suggest possible ecological consequences to streamside vegetation related to alien species and identify indicators of conditions, where invasion may be greatest, offering potential for informing decisions for monitoring and managing alien species.Nomenclature follows the PLANTS Database (USDA and NRCS 2004).