Quantifying the effectiveness of regional habitat quality index models for predicting densities of juvenile Atlantic salmon ( Salmo salar L.) (original) (raw)
Related papers
2012
Decisions on managed flow releases in regulated rivers should be informed by the best available science. To do this, resource managers require adequate information regarding the tradeoffs between alternative methodologies. In this study, we quantitatively compare two competing multivariate habitat models for juvenile Chinook salmon (Oncorhynchus tschawytscha), a highly valued fish species under serious decline in a large extent of its range. We conducted large-scale snorkel surveys in the American River, California, to obtain a common dataset for model parameterization. We built one habitat model using Akaike Information Criterion analysis and model averaging, 'model G', and a second model by using a standard method of aggregating univariate habitat models, 'model A'. We calculated Cohen's kappa, percent correctly classified, sensitivity, specificity and the area under a receiver operator characteristic to compare the ability of each model to predict juvenile salmon presence and absence. We compared the predicted useable habitat of each model at nine simulated river discharges where usable habitat is equal to the product of a spatial area and the probability of habitat occupancy at that location. Generally, model G maintained greater predictive accuracy with a difference within 10% across the diagnostic statistics. Two key distinctions between models were that model G predicted 17.2% less useable habitat across simulated flows and had 5% fewer false positive classifications than model A. In contrast, model A had a tendency to over predict habitat occupancy and under predict model uncertainty. The largest discrepancy between model predictions occurred at the lowest flows simulated and in the habitats most likely to be occupied by juvenile salmon. This study supports the utility and quantitative framework of Akaike Information Criterion analysis and model averaging in developing habitat models.
Evaluation of IUCN spatial distribution metrics for a migratory species, Fraser River Sockeye salmon
Biological Conservation, 2014
The International Union for Conservation of Nature (IUCN) has developed criteria to assess extinction risk; the Committee on the Status of Endangered Wildlife in Canada (COSEWIC) is an independent agency that assesses the vulnerability of Canadian species using IUCN criteria. Criterion B uses extent of occurrence, area of occupancy, and number of locations to evaluate risks associated with restricted spatial distribution in conjunction with several subcriteria. Concerns have been expressed about the use of these distribution metrics for migratory aquatic species as they often have naturally restricted distributions that change at different life stages. The conservation status of Sockeye salmon (Oncorhynchus nerka) of the Fraser River, British Columbia, Canada is currently being assessed and distribution metrics have been estimated. We compared metrics for Sockeye salmon to similar values for 57 assessments of vertebrate, invertebrate and plant species conducted by COSEWIC. Relationships among the metrics were generally similar between Fraser River Sockeye salmon and COSEWIC assessments, suggesting that despite concerns about the applicability of occurrence and abundance metrics and corresponding methodologies to highly migratory and aquatic taxa, Fraser River Sockeye salmon do not demonstrate large departures compared to other species assessed in Canada. The majority of Fraser River Sockeye salmon distribution metrics fell below thresholds for endangered or threatened statuses. However, we also observed that reported values for species assessed by COSEWIC had little relationship to the status that was ultimately assigned. Thus, based on results of this study, the ultimate role of the distribution metrics in COSEWIC status assessment is uncertain.
Implications of territory size for the measurement and prediction of salmonid abundance in streams
Canadian Journal of Fisheries and Aquatic Sciences, 1998
Information about territory size is useful for both the measurement and prediction of salmonid abundance. Percent habitat saturation (PHS), the percentage of the stream area occupied by the territories of salmonid fishes, is a better measure of abundance than population density because the former integrates the effects of (a) several age-classes or species in a stream, and (b) variation in growth rate or sampling date. "Effective density" or "effective PHS," calculated by weighting crude density (no.·m-2) or PHS by the number of organisms in the sampling unit, more accurately reflects density from the organism's point of view than does crude density or PHS. Effective density and PHS of Atlantic salmon (Salmo salar) in Catamaran Brook, New Brunswick, increased by 0.4 fish per m2 and 4%, respectively, for each order of magnitude decrease in the area of the sampling unit. Literature data suggested that territory size is inversely related to food abundance and ca...
Do habitat measurements in the vicinity of Atlantic salmon (Salmo salar) parr matter?
Fisheries Management and Ecology, 2017
Atlantic salmon (Salmo salar) parr habitat characterization is usually performed by in situ measures of key environmental variables, taken at the exact fish location, or conversely, in large sampling sections, often ignoring variability in the immediate vicinity around individuals. These data may have a critical importance in development and validation of habitat preference models. The influences of seven increasing distances of measurements, the number of considered measures, and two depth of velocity measurement were tested in the calculations of HSI (Habitat Suitability Index) from a multiple-experts fuzzy model. The radius of 50 cm around the fish, an average measure of 6 measurements in the neighbouring environment and a velocity measured at 60% of the depth gave the highest HSI values. These results show some potential for the use of an intermediate study scale, between micro-and mesohabitat, and questions how the fish habitat conditions are currently measured.
Ecological Applications, 2007
The geographic distribution of stream reaches with potential to support highquality habitat for salmonids has bearing on the actual status of habitats and populations over broad spatial extents. As part of the Coastal Landscape Analysis and Modeling Study (CLAMS), we examined how salmon-habitat potential was distributed relative to current and future (þ100 years) landscape characteristics in the Coastal Province of Oregon, USA. The intrinsic potential to provide high-quality rearing habitat was modeled for juvenile coho salmon (Oncorhynchus kisutch) and juvenile steelhead (O. mykiss) based on stream flow, valley constraint, and stream gradient. Land ownership, use, and cover were summarized for 100-m analysis buffers on either side of stream reaches with high intrinsic potential and in the overall area encompassing the buffers. Past management seems to have concentrated nonindustrial private ownership, agriculture, and developed uses adjacent to reaches with high intrinsic potential for coho salmon. Thus, of the area in coho salmon buffers, 45% is either nonforested or recently logged, but only 10% is in larger-diameter forests. For the area in steelhead buffers, 21% is either non-forested or recently logged while 20% is in larger-diameter forests. Older forests are most extensive on federal lands but are rare on private lands, highlighting the critical role for public lands in near-term salmon conservation. Agriculture and development are projected to remain focused near high-intrinsic-potential reaches for coho salmon, increasing the importance of effectively addressing nonpoint source pollution from these uses. Percentages of larger-diameter forests are expected to increase throughout the province, but the increase will be only half as much in coho salmon buffers as in steelhead buffers. Most of the increase is projected for public lands, where policies emphasize biodiversity protection. Results suggest that widespread recovery of coho salmon is unlikely unless habitat can be improved in high-intrinsic-potential reaches on private lands. Knowing where high-intrinsic-potential stream reaches occur relative to landscape characteristics can help in evaluating the current and future condition of freshwater habitat, explaining differences between species in population status and risk, and assessing the need for and feasibility of restoration.
Transferability of habitat suitability criteria of juvenile Atlantic salmon ( Salmo salar )
Canadian Journal of Fisheries and Aquatic Sciences, 2002
We constructed generalized habitat criteria for juvenile Atlantic salmon (Salmo salar) based on four riverspecific suitability indices for depth, water velocity, and substrate to assess whether habitat criteria for juvenile Atlantic salmon are transferable across rivers. We first tested whether salmon are more likely to occupy higher-quality habitats than generally available in a stream reach based on these composite criteria. We then repeated the same procedure using the river-specific criteria of this study and the generalized habitat suitability criteria of Heggenes. As expected, the river-specific criteria were generally the most effective ones in predicting fish habitat use. However, both of the two generalized criteria also transferred fairly well to the test sites. Viewed across salmon size-classes (<9 cm and >9 cm), the river-specific criteria passed the test in ten of eleven cases (91%), and the two generalized criteria passed in nine (82%) of eleven. Thus, it appears that with respect to summertime habitat criteria for juvenile Atlantic salmon, criteria transference is conceivable at least on a regional scale, and perhaps even on a more "universal" scale.
Quantifying the effects of stream habitat on populations of breeding Pacific salmon
Canadian Journal of Fisheries and Aquatic Sciences, 2015
Recognizing the mechanisms by which environmental conditions drive population dynamics can greatly benefit conservation and management. For example, reductions in densities of spawning Pacific salmon (Oncorhynchus spp.) have received considerable attention, but the role of habitat characteristics on population sizes of breeding salmon is not fully understood. We studied relationships between habitat characteristics and stream population densities of spawning chum (Oncorhynchus keta) and pink (Oncorhynchus gorbuscha) salmon in 44 streams in the Great Bear Rainforest of coastal British Columbia, Canada, with individual streams as the unit of comparison. Our results indicate that a small number of habitat characteristics are important in predicting population density of spawning chum and pink salmon in streams, namely pH for chum salmon and riparian slope and large wood volume for pink salmon. This is the largest multivariable comparison to examine habitat–population relationships in a...
Improving models of juvenile Atlantic salmon habitat use through high resolution remote sensing
Ecological Modelling, 2006
The effect of spatial variation in substrate size on juvenile Atlantic salmon density in the Sainte-Marguerite, a major salmon-spawning river in QC, Canada, was investigated using ground surveys and remote sensing. Densities of both fry and parr were determined by single-pass electro-fishing at up to 48 sites along the length of the river in August of each of the years from 1997 to 2004. Substrate size was determined by two techniques. Firstly, a ground survey of median substrate size (D50) was conducted concurrently with the electrofishing, providing information on the substrate actually experienced by the salmon at the point of capture (a parcel area of 5 m × 20 m for each site). Secondly, an airborne survey was conducted along the length of the river in August 2002 using a helicopter-mounted digital camera. Images from this survey were then processed to show the mean D50 in the area of each image (an area from bank to bank, 60 m in length along the river) encompassing each parcel. Relationships between juvenile salmon density and substrate size were determined using preference models where the model partitioning had been determined by regression tree analysis. For comparison, preference models were also produced where the model partitioning was determined arbitrarily. The shape of the relationships between juvenile salmon density and parcel D50 were similar to those between juvenile salmon density and mean image D50. However, the relationship was stronger with mean image D50, suggesting that the habitat surrounding the location in which juvenile salmon were found had a direct influence on their density. Additionally, preference models determined from regression tree analysis had greater explanatory power than those determined using an arbitrary partitioning approach. These results suggest that remote sensing, alongside a robust approach for determining preference models, may be an effective tool in modelling juvenile Atlantic salmon habitat use.
We tested the validity of the predictions made by a habitat probabilistic index (HPI) developed using a description of the physical conditions (depth, flow velocity, grain size) used and avoided by parrs during days of different cloudiness. Thirteen surveys were designed to estimate the number and the distribution of parrs actively foraging within a 300-m reach of a river. During these surveys, the number of parrs actively foraging ranged from 12 to 118, cloud cover ranged from 5% to 100%, and water temperature ranged from 16.5°C to 21.7°C. The number of parrs actively foraging was negatively related to cloud cover (r 2 = 0.44 to 0.88) but was independent of water temperature. HPI models developed under low (<33%) and intermediate (34-67%) cloud cover explained 82% to 98% of the local variations of fish density. The HPI model developed under high cloud cover (67-100%) was unable to predict fish distribution observed during cloudy days. Our results suggest that HPI models developed when cloudiness is >67% may have a limited predictive power.