Search

Ecosystem structure and processes of coastal temperate rainforests of the Pacific Northwest are thought to be strongly influenced by herbivory primarily of Roosevelt elk (Cervus elaphus roosevelti) and secondarily of Columbian black‐tailed deer (Odocoileus hemionus columbianus). Two large (0.5‐ha) exclosures were built in old‐growth coniferous rainforest communities in Olympic National Park, Washington, during 1979 to study these effects. Cover of shrubs, ferns, herbs, and graminoids and numbers of tree seedlings were described over 36 yr. Results show a sequence following ungulate exclusion of early release of shrubs, ferns, and herbs followed by eventual dominance of shrubs as other vegetation layers become shaded. Short‐term responses of individual species reflected functional traits related to ability to avoid or tolerate herbivory. Over the longer term, effects reflected changing competitive relationships among vegetation layers and other ecosystem dynamics such as the provision of fallen trees in the appropriate decay class to serve as establishment substrate for tree seedlings. In aggregate, vegetation composition shifted after 36 yr from a system dominated by herbaceous cover with a major graminoid component to one dominated by shrubs (5‐ to 6‐fold absolute increase) and ferns (5–7% increase in absolute cover), less absolute herb cover (15–20% loss), and almost no graminoids (<1.5% cover remaining in any plot) after 36 yr. These changes represented a substantial loss in plant community diversity with a loss of 46 of 74 species. Elk abundance outside of the exclosures began to decline in the 1990s leading to parallel changes in plant community trajectories outside of exclosures to those initially seen inside. While this suggests plant community responses inside the exclosures were also driven by elk exclusion, the strength of this response depends on elk abundance. [ABSTRACT FROM AUTHOR]

Early seral vegetation was studied on a former lake bottom after the removal of the 64‐m‐tall Glines Canyon Dam on the Elwha River. In 2015, vegetation cover of all vascular plant species was determined in 63 plots located on sites that emerged in 2011–2012. The sites had been planted and/or seeded, or were permitted to revegetate spontaneously. The plots were further classified by substrate texture: coarse sediments on the valley bottom and fine ones on the valley slopes. Plots were located randomly along random transects perpendicular to the former lake shore that extended into coarse sediment terraces perched above the floodplain. Additionally, 32 plots were sampled in surrounding native forests near these transects. Data were analyzed by detrended correspondence analysis and by canonical correspondence analysis. Substrate texture, that is whether fine or coarse, appeared to explain most of the variability in vegetation. The distance to forest and successional age, that is time since the site had been drained, were also significant explanatory variables, while assisted restoration by planting and seeding appeared to be insignificant to date. Spontaneous succession on fine sediments led to a species composition approaching that of adjacent natural forests. Invasive species were much less abundant than expected. Spontaneous restoration of vegetation on fine sediments in drained lake bottoms can rapidly produce a desirable vegetation composition and structure. On coarse sediments, active restoration may be useful to accelerate the development of native vegetation communities. [ABSTRACT FROM AUTHOR]

Identifying the relative contributions of physical and ecological processes to channel evolution remains a substantial challenge in fluvial geomorphology. We use a 74-year aerial photographic record of the Hoh, Queets, Quinault, and Elwha Rivers, Olympic National Park, Washington, USA, to investigate whether physical or trophic-cascade-driven ecological factors - excessive elk impacts after wolves were extirpated a century ago - are the dominant drivers of channel planform in these gravel-bed rivers. We find that channel width and braiding show strong relationships with recent flood history. All four rivers widened significantly after having been relatively narrow in the 1970s, consistent with increased flood activity since then. Channel planform also reflects sediment-supply changes, evident from landslide response on the Elwha River. We surmise that the Hoh River, which shows a multi-decadal trend toward greater braiding, is adjusting to increased sediment supply associated with rapid glacial retreat. These rivers demonstrate transmission of climatic signals through relatively short sediment-routing systems that lack substantial buffering by sediment storage. Legacy effects of anthropogenic modification likely also affect the Quinault River planform. We infer no correspondence between channel evolution and elk abundance, suggesting that trophic-cascade effects in this setting are subsidiary to physical controls on channel morphology. Our findings differ from previous interpretations of Olympic National Park fluvial dynamics and contrast with the classic example of Yellowstone National Park, where legacy effects of elk overuse are apparent in channel morphology; we attribute these differences to hydrologic regime and large-wood availability. Published 2016. This article is a U.S. Government work and is in the public domain in the USA [ABSTRACT FROM AUTHOR]

The Mahalanobis distance statistic (D2) has emerged as an effective tool to identify suitable habitat from presence data alone, but there has been no mechanism to select among potential habitat covariates. We propose that the best combination of explanatory variables for a D2 model can be identified by ranking potential models based on the proportion of the entire study area that is classified as potentially suitable habitat given that a predetermined proportion of occupied locations are correctly classified. In effect, our approach seeks to minimize errors of commission, or maximize specificity, while holding the omission error rate constant. We used this approach to identify potentially suitable habitat for the Olympic marmot (Marmota olympus), a declining species endemic to Olympic National Park, Washington, USA. We compared models built with all combinations of 11 habitat variables. A 7-variable model identified 21,143 ha within the park as potentially suitable for marmots, correctly classifying 80% of occupied locations. Additional refinements to the 7-variable model (e.g., eliminating small patches) further reduced the predicted area to 18,579 ha with little reduction in predictive power. Although we sought a model that would allow field workers to find 80% of Olympic marmot locations, in fact, <3% of 376 occupied locations and <9% of abandoned locations were >100 m from habitat predicted by the final model, suggesting that >90% of occupied marmot habitat could be found by observant workers surveying predicted habitat. The model comparison procedure allowed us to identify the suite of covariates that maximized specificity of our model and, thus, limited the amount of less favorable habitat included in the final prediction area. We expect that by maximizing specificity of models built from presence-only data, our model comparison procedure will be useful to conservation practitioners planning reintroductions, searching for rare species, or identifying habitat for protection. [ABSTRACT FROM AUTHOR]

Habitat loss and fragmentation are the leading causes of species’ declines and extinctions. A key component of studying population response to habitat alteration is to understand how fragmentation affects population connectivity in disturbed landscapes. We used landscape genetic analyses to determine how habitat fragmentation due to timber harvest affects genetic population connectivity of the coastal tailed frog ( Ascaphus truei), a forest-dwelling, stream-breeding amphibian. We compared rates of gene flow across old-growth (Olympic National Park) and logged landscapes (Olympic National Forest) and used spatial autoregression to estimate the effect of landscape variables on genetic structure. We detected higher overall genetic connectivity across the managed forest, although this was likely a historical signature of continuous forest before timber harvest began. Gene flow also occurred terrestrially, as connectivity was high across unconnected river basins. Autoregressive models demonstrated that closed forest and low solar radiation were correlated with increased gene flow. In addition, there was evidence for a temporal lag in the correlation of decreased gene flow with harvest, suggesting that the full genetic impact may not appear for several generations. Furthermore, we detected genetic evidence of population bottlenecks across the Olympic National Forest, including at sites that were within old-growth forest but surrounded by harvested patches. Collectively, this research suggests that absence of forest (whether due to natural or anthropogenic changes) is a key restrictor of genetic connectivity and that intact forested patches in the surrounding environment are necessary for continued gene flow and population connectivity. [ABSTRACT FROM AUTHOR]

1. Faecal surveys are used widely to study variations in abundance and distribution of forest-dwelling mammals when direct enumeration is not feasible. The utility of faecal indices of abundance is limited, however, by observational bias and variation in faecal disappearance rates that obscure their relationship to population size. We developed methods to reduce variability in faecal surveys and improve reliability of faecal indices. 2. We used double-observer transect sampling to estimate observational bias of faecal surveys of Roosevelt elk Cervus elaphus roosevelti and Columbian black-tailed deer Odocoileus hemionus columbianus in Olympic National Park, Washington, USA. We also modelled differences in counts of faecal groups obtained from paired cleared and uncleared transect segments as a means to adjust standing crop faecal counts for a standard accumulation interval and to reduce bias resulting from variable decay rates. 3. Estimated detection probabilities of faecal groups ranged from < 0·2–1·0 depending upon the observer, whether the faecal group was from elk or deer, faecal group size, distance of the faecal group from the sampling transect, ground vegetation cover, and the interaction between faecal group size and distance from the transect. 4. Models of plot-clearing effects indicated that standing crop counts of deer faecal groups required 34% reduction on flat terrain and 53% reduction on sloping terrain to represent faeces accumulated over a standard 100-day interval, whereas counts of elk faecal groups required 0% and 46% reductions on flat and sloping terrain, respectively. 5. Synthesis and applications. Double-observer transect sampling provides a cost-effective means of reducing observational bias and variation in faecal decay rates that obscure the interpretation of faecal indices of large mammal abundance. Given the variation we observed in observational bias of faecal surveys and persistence of faeces, we emphasize the need for future researchers to account for these comparatively manageable sources of bias before comparing faecal indices spatially or temporally. Double-observer sampling methods are readily adaptable to study variations in faecal indices of large mammals at the scale of the large forest reserve, natural area, or other forested regions when direct estimation of populations is problematic. [ABSTRACT FROM AUTHOR]

A vegetation chronosequence spanning over 300 years was established in unconstrained reaches of the lower Queets River in Olympic National Park, Washington, USA, for an examination of riparian successional patterns. The Queets is an unconstrained, dynamic, mountain river located within a temperate rain forest environment. Ongoing channel movements create intricate patterns in the physical structure of the valley. Twenty-one plots containing a total of 4359 trees were mapped and measured for structural and crown characteristics. Snags, logs, and understory vegetation were also quantified. Recent alluvial deposits are colonized primarily by early-successional trees Salix sitchensis and Alnus rubra. Conifer seedlings, primarily Picea sitchensis, generally invade after the initial cohort of hardwood trees begins senescence: 20-30 years for Salix and 40-60 years for Alnus. Through accumulation of sediments from floods and channel downcutting, surfaces become perched above the reach of annual floods after 40-80 years and are then slowly colonized by late successional tree species Acer circinatum, Acer macrophyllum, and Tsuga heterophylla. Diverse, old-growth forests ultimately develop after 200-250 years, containing some of the largest known trees in the Pacific Northwest. However, canopy and stem densities remain lower than comparative Pseudotsuga menziesii forests from the nearby Cascade Mountains. Vast individual crowns can develop, with occasional Picea up to 25 m wide and 70 m deep. Individual stands may accumulate >200 000 m³/ha of canopy volume—among the highest recorded on earth. Mixed among the generalized successional sequence are variations created by uncommon channel movements. Avulsions followed by channel incision form cobblefields in abandoned channels or other surfaces which are isolated from subsequent inundation and sediment deposition. These cobblefields embark on a different successional trajectory, which often includes conifer seedlings present in the initial cohort. Ultimately, whatever the initial trajectory, soils become productive due to soil conditioning by Alnus and the decomposition of other plant material. These biophysical complexities, interconnected patterns, and system-scale resilience are summarized in a multiple-pathway successional model that may be applicable to floodplain riparian forests throughout much of the Pacific coastal ecoregion. [ABSTRACT FROM AUTHOR]