USGS ScienceSearch

SEARCH · USGS Science

Results for “American Forests”

Search indexed USGS publications on groundwater, aquifers, geologic maps, mineral resources and earthquakes. Explore source records by subject and place.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 379 records · Page 21Linked to original sources

Recovery of sediment characteristics in moraine, headwater streams of northern Minnesota after forest harvest

We investigated the recovery of sediment characteristics in four moraine, headwater streams in north-central Minnesota after forest harvest. We examined changes in fine sediment levels from 1997 (preharvest) to 2007 (10 years postharvest) at study plots with upland clear felling and riparian thinning, using canopy cover, proportion of unstable banks, surficial fine substrates, residual pool depth, and streambed depth of refusal as response variables. Basin-scale year effects were significant ( p < 0.001) for all responses when evaluated by repeated-measures ANOVAs. Throughout the study area, unstable banks increased for several years postharvest, coinciding with an increase in windthrow and fine sediment. Increased unstable banks may have been caused by forest harvest equipment, increased windthrow and exposure of rootwads, or increased discharge and bank scour. Fine sediment in the channels did not recover by summer 2007, even though canopy cover and unstable banks had returned to 1997 levels. After several storm events in fall 2007, 10 years after the initial sediment input, fine sediment was flushed from the channels and returned to 1997 levels. Although our study design did not discern the source of the initial sediment inputs (e.g., forest harvest, road crossings, other natural causes), we have shown that moraine, headwater streams can require an extended period (up to 10 years) and enabling event (e.g., high storm flows) to recover from large inputs of fine sediment.

Minnesota

Comparison of settlement-era vegetation reconstructions for STEPPS and REVEALS pollen–vegetation models in the northeastern United States

Reconstructions of prehistoric vegetation composition help establish natural baselines, variability, and trajectories of forest dynamics before and during the emergence of intensive anthropogenic land use. Pollen–vegetation models (PVMs) enable such reconstructions from fossil pollen assemblages using process-based representations of taxon-specific pollen production and dispersal. However, several PVMs and variants now exist, and the sensitivity of vegetation inferences to PVM selection, variant, and calibration domain is poorly understood. Here, we compare the reconstructions, parameter estimates, and structure of a Bayesian hierarchical PVM, STEPPS, both to observations and to REVEALS, a widely used PVM, for the pre–Euro-American settlement-era vegetation in the northeastern United States (NEUS). We also compare NEUS-based STEPPS parameter estimates to those for the upper midwestern United States (UMW). Both PVMs predict the observed macroscale patterns of vegetation composition in the NEUS; however, reconstructions of minor taxa are less accurate and predictions for some taxa differ between PVMs. These differences can be attributed to intermodel differences in structure and parameter estimates. Estimates of pollen productivity from STEPPS broadly agree with estimates produced for use in REVEALS, while comparison between pollen dispersal parameter estimates shows no significant relationship. STEPPS parameter estimates are similar between the UMW and NEUS, suggesting that STEPPS parameter estimates are transferable between floristically similar regions and scales.

Connecticut, Maine, Massachusetts, New Hampshire,

Groundwater hydrology and chemistry of Jamestown Island, Virginia—Potential effects of tides, storm surges, and sea-level rise on archaeological, cultural, and ecological resources

As the site of the first permanent English settlement in North America in 1607, Jamestown Island, Colonial National Historical Park (COLO), Virginia, contains a rich archaeological record that extends from the Paleoindian period (15,000 to 8,000 years ago) through the 20th century. The island is located on the lower James River near the mouth of Chesapeake Bay. Jamestown Island vegetation is dominated by upland forests surrounded by tidal, freshwater-to-oligohaline marshes. Along the Virginia coast, relative sea-level rise was more than 2.5 times the global average during the 20th century. Consequently, the National Park Service (NPS) has identified COLO as one of the 25 national parks most threatened by climate change. Surface waters across the island are hydraulically connected to the laterally continuous Surficial aquifer. The land-surface altitude of the island is low, with two-thirds of the island less than 5 feet (ft) above the North American Vertical Datum of 1988 (NAVD 88). Consequently, sea-level rise, combined with tides and storm surges, threatens the island and its resources as surface-water and groundwater levels rise, saltwater enters the Surficial aquifer, and groundwater chemistry changes. The impact of sea-level rise on the island’s surface-water resources has been well studied, but groundwater effects have been largely ignored. Quantifying the effects of tides, storm surges, and sea-level rise on groundwater levels and chemistry is essential to developing an effective strategy for managing climate-induced changes. The first step in developing a response strategy includes a parkwide general risk assessment for archaeological sites on the island, so that sites can be prioritized for management actions. The U.S. Geological Survey and the NPS began a study in 2015 to develop a long-term groundwater-monitoring program to evaluate this risk and to develop an updated management strategy. The groundwater-monitoring program consists of 45 wells and piezometers in two individual clusters and three transects across the island in different hydrologic and chemical settings. Samples for water quality were collected from the wells and piezometers from October 2015 through September 2018 at variable time intervals. Results of the monitoring identified disparate hydrologic and chemical responses to saltwater intrusion across the island. Specific conductance (an indicator of salinity) of groundwater beneath several marshes responded differently to changes in James River salinity. Groundwater response to changes in James River specific conductance appeared to be controlled by land-surface altitude and slope, differences in lateral and vertical sediment characteristics, distance from surface waters, and the degree of surface water/groundwater connectivity between channels and the aquifer. Groundwater chemistry data from monitoring wells at Black Point, a low-altitude, upland setting, are in contrast with conditions observed in Island House observation wells, a high-altitude, upland setting. Specific conductance (less than 200 microsiemens per centimeter [μS/cm]) and pH (greater than 5.0) of groundwater beneath much of the uplands that characterize the Island House observation wells are typical of groundwater in noncarbonate sedimentary aquifers recharged by precipitation. At Black Point, specific conductance ranged from 2,490 to 15,200 μS/cm, and pH ranged from 3.1 to 6.6 standard units. At the Black Point observation wells, the most saline and dense water was at the water table rather than deeper in the aquifer, causing a density inversion that persisted throughout the study. The density inversion likely resulted from differences in permeability between the shallow clay and fine-grained sands and the deeper coarse-grained sand and gravel. Groundwater with the lowest pH was at the water table. As saline groundwater flows through organic sediment beneath the marshes, bacterial biodegradation of organic matter creates anoxic conditions. Continued biodegradation concomitantly reduces iron-oxide minerals in the sediment and sulfate in saline water. When oxygen is reintroduced into groundwater, iron and sulfur can reoxidize to form sulfuric acid, locally lowering the pH of the water. This report describes the groundwater monitoring network design, rationale for site selection, monitoring approach, and results of monitoring from October 2015 through September 2018. Maps of inundation at selected water-level altitudes are included to identify the risk to archaeological, cultural, and ecological resources. The monitoring results of the hydrology and chemistry data are interpreted, and the different hydrologic and chemical settings are described. The implications of the study results for management decisions are presented, and suggestions for improving the monitoring network are included.

Virginia

Agriculture

Agricultural production is a fundamental activity conducted on 45% of the U.S. land area, 55% of Mexico’s land area, and 7% of Canada’s land area (World Bank 2016). Because of this vast spatial extent and the strong role that land management plays in how agricultural ecosystems function, agricultural lands and activities represent a large portion of the North American carbon budget. Accordingly, improved quantification of the agricultural carbon cycle, new trends in agriculture, and added opportunities for emissions reductions provide a critical foundation for considering the relationships between agriculture and carbon cycling at local, regional, continental, and global scales. More than 145 countries have specifically included agriculture in their targets and actions for mitigating climate change (FAO 2016), and agriculture has featured particularly prominently in recent target and action commitments made by developing countries to reduce greenhouse gas (GHG) emissions (Richards et al., 2015). Conversion of vast native forest and prairie to agriculture across North America between 1860 and 1960 resulted in carbon dioxide (CO2) fluxes to the atmosphere from biota and soils that exceeded those from fossil fuel emissions over the same period (Houghton et al., 1983). Correspondingly, soil organic carbon (SOC) declined in many soils during the 50 years following conversion from native ecosystems to production agriculture (Huggins et al., 1998; Janzen et al., 1998; Slobodian et al., 2002). Crop yields and corresponding above- and belowground biomass have steadily increased since the 1930s due to genetic and management innovations, which provide more organic input from which to build SOC ( Johnson et al., 2006; Hatfield and Walthall 2015). This, coupled with improved input-use efficiencies may reduce GHG-emissions per unit yield (GHG intensity), with additional improvements possible through management optimization (Grassini and Cassman 2012; Pittelkow et al., 2015). Options include reducing tillage, integrating perennials onto the landscape, reducing or eliminating bare-fallow land (i.e., land without living plants), adding cover crops, and enrolling lands in conservation easement programs. These options, originally proposed to control erosion, have potential co-benefits in terms of increased soil health, plant productivity, and soil carbon stabilization (Lehman et al., 2015). Conversely, returning lands previously enrolled in conservation easements (e.g., the Conservation Reserve Program [CRP] and other land set-aside efforts) to row-crop production, tillage, or aggressive harvesting of crop residues all risk degrading soil quality and exacerbating SOC loss. Of note is that the net results of land use and land management practices in an agricultural setting vary according to many factors, such as crop or production system type, soil type, climate, and the collection of practices at any given site. For example, many traditional practices followed by Indigenous people on tribal lands are based on an integrated approach to natural resource management and response to environmental change that may provide agricultural options uniquely suited to varied environmental settings (see Ch. 7: Tribal Lands, p. 303). Agricultural land in the United States totaled 408.2 million hectares (ha) in 2014, of which 251 million ha were in permanent meadows and pastures, 152.2 million ha were in arable land, and 2.6 million ha were in permanent crops (FAOSTAT 2016). Compared with the distribution in 2007, these numbers reflect a 4.7 million ha decline in total agricultural lands, driven by declines in arable land and permanent crops but partially offset by a modest increase in permanent meadows and pastures. Although arable lands have been declining, the combined acreage of the four major crops (corn, wheat, soybeans, and cotton) has risen slightly, with increases in land planted in corn and soybeans and decreases in cotton and wheat (see Figure 5.1, p. 232). Despite the overall slight decline in agricultural land area, the value of U.S. agricultural production rose over the past decade as a result of increased production efficiency and higher prices (USDA 2017a; see also www.ers.usda.gov). Canada has about 65 million ha of agricultural land, of which about 46 million ha are arable, accounting for only about 7% of the country’s total land area (FAOSTAT 2017). Prominent crops on Canada’s arable lands include cereals (e.g., wheat, barley, and maize), oilseeds (e.g., canola and soybeans), and pulses (e.g., peas and lentils). Natural and seeded pastures available for grazing in Canada make up about 20 million ha (Legesse et al., 2016). Agricultural land in Mexico makes up 107 million ha, of which 23 million ha are arable land, 2.7 million ha are permanent crops, and 81 million ha are permanent meadows and pastures (FAOSTAT 2017). Mexico’s major crops are fruits, corn, grains, vegetables, and sugarcane.

Report

Macroinvertebrate assemblage recovery following a catastrophic flood and debris flows in an Appalachian mountain stream

In June 1995, heavy rains caused severe flooding and massive debris flows on the Staunton River, a 3rd-order stream in the Blue Ridge Mountains (Virginia, USA). Scouring caused the loss of the riparian zone and repositioned the stream channel of the lower 2.1 km of the stream. Between 1998 and 2001, we conducted seasonal macroinvertebrate surveys at sites on the Staunton River and on White Oak Canyon Run, a reference stream of similar size and geology that was relatively unaffected by the flood. Our study was designed to determine the extent to which flood-induced changes to the stream channel and riparian habitats caused long-term changes to macroinvertebrate community structure and composition. Sites within the impacted zone of the Staunton River supported diverse stable benthic macroinvertebrate assemblages 3 y after the flood despite dramatic and persistent changes in environmental factors known to be important controls on stream ecosystem function. However, significant differences in total macroinvertebrate density and trophic structure could be attributed to the flood. In autumn, densities of most feeding guilds, including shredders, were higher at impacted-zone sites than at all other sites, suggesting higher overall productivity in the impacted zone. Higher shredder density in the impacted zone was surprising in light of expected decreases in leaf-litter inputs because of removal of riparian forests. In contrast, in spring, we observed density differences in only one feeding guild, scrapers, which showed higher densities at impacted-zone sites than at all other sites. This result conformed to a priori expectations that reduced shading in the impacted zone would lead to increased light and higher instream primary production. We attribute the seasonal differences in trophic structure to the effects of increased temperatures on food quality and to the relationship between the timing of our sampling and the emergence patterns of important taxa. ?? 2006 by The North American Benthological Society.

Journal of the North American Benthological Societ

Deserts of China

Improving arid land quality requires an understanding of the original state of the land and its relationship to wind, water, and plant regimes, as well as understanding of interactions within the present ecosystem. Chinese scientists and local residents have made significant advances in improving arid environments in gobi and sandy deserts and in less arid sandy lands. Wind patterns are being changed by planting forest belts to protect oases and sandy lands, and on a smaller scale by planting grasses and shrubs or constructing straw grids. Research on reclamation of deserts is now focusing on how sand-fixing plants may be adapted to local environments, and how the resources of grazing land and water may be effectively exploited without being overused.

American Scientist

Introgression of coyote mitochondrial DNA into sympatric North American gray wolf populations

Mitochondrial DNA (mtDNA) genotypes of gray wolves and coyotes from localities throughout North America were determined using restriction fragment length polymorphisms. Of the 13 genotypes found among the wolves, 7 are clearly of coyote origin, indicating that genetic transfer of coyote mtDNA into wolf populations has occurred through hybridization. The transfer of mtDNA appears unidirectional from coyotes into wolves because no coyotes sampled have a wolf-derived mtDNA genotype. Wolves possessing coyote-derived genotypes are confined to a contiguous geographic region in Minnesota, Ontario, and Quebec, and the frequency of coyote-type mtDNA in these wolf populations is high (>50%). The ecological history of the hybrid zone suggests that hybridization is taking place in regions where coyotes have only recently become abundant following conversion of forests to farmlands. Dispersing male wolves unable to find conspecific mates may be pairing with female coyotes in deforested areas bordering wolf territories. Our results demonstrate that closely related species of mobile terrestrial vertebrates have the potential for extensive genetic exchange when ecological conditions change suddenly.

Evolution

Bats of Mesa Verde National Park, Colorado: Composition, reproduction, and roosting habits

We determined the bat fauna at Mesa Verde National Park (Mesa Verde) in 2006 and 2007, characterized bat elevational distribution and reproduction, and investigated roosting habits of selected species. We captured 1996 bats of 15 species in mist nets set over water during 120 nights of sampling and recorded echolocation calls of an additional species. The bat fauna at Mesa Verde included every species of bat known west of the Great Plains in Colorado, except the little brown bat ( Myotis lucifugus ). Some species showed skewed sex ratios, primarily due to a preponderance of males. Thirteen species of bats reproduced at Mesa Verde. Major differences in spring precipitation between the 2 years of our study were associated with differences in reproductive rates and, in some species, with numbers of juveniles captured. Reduced reproductive effort during spring drought will have a greater impact on bat populations with the forecasted increase in aridity in much of western North America by models of global climate change. We radiotracked 46 bats of 5 species to roosts and describe the first-known maternity colonies of spotted bats ( Euderma maculatum ) in Colorado. All 5 species that we tracked to diurnal roosts relied almost exclusively on rock crevices rather than trees or snags, despite the presence of mature forests at Mesa Verde and the use of trees for roosts in similar forests elsewhere by some of these species. Comparisons with past bat surveys at Mesa Verde and in surrounding areas suggest no dramatic evidence for effects of recent stand-replacing fires on the composition of the bat community.

Colorado

North American tree migration paced by climate in the West, lagging in the East

Tree fecundity and recruitment have not yet been quantified at scales needed to anticipate biogeographic shifts in response to climate change. By separating their responses, this study shows coherence across species and communities, offering the strongest support to date that migration is in progress with regional limitations on rates. The southeastern continent emerges as a fecundity hotspot, but it is situated south of population centers where high seed production could contribute to poleward population spread. By contrast, seedling success is highest in the West and North, serving to partially offset limited seed production near poleward frontiers. The evidence of fecundity and recruitment control on tree migration can inform conservation planning for the expected long-term disequilibrium between climate and forest distribution.

PNAS

North American ducks

Increased predation and habitat degradation and destruction coupled with drought, especially on breeding grounds, have caused the declines of some duck populations. More than 30 species of fucks breed in North America, in areas as diverse as the Arctic tundra and the subtropics of Florida and Mexico. For many of these species, however, the Prairie Pothole region of the north-central United States and south-central Canada is the most important breeding area, although migratory behavior and the life histories of different species lead them to use many wetland habitats. Numerous sources of information are available on the status of duck populations in North America. The two most comprehensive and reliable sources are the Breeding Population and Habitat Survey, conducted since 1955 and encompassing the Prairie Pothole region, boreal forests, and tundra habitats from South Dakota to Alaska (Caithammer et al. 1993), and the Midwinter Survey, encompassing the United States and portions of Canada and Mexico at regular intervals. Results from these surveys are the basis for this article.

Book chapter

Defining opportunities for collaboration across data life cycles

Monitoring natural resources - water, forests, and animal populations—is required to support effective management of natural resources. However, because monitoring activities are often specific to a discipline, issue, or agency, it is typically difficult to integrate data to answer questions that transcend geopolitical and jurisdictional boundaries. How do we reach the better data integration we need?

Eos, Transactions, American Geophysical Union

Using macroinvertebrates to identify biota-land cover optima at multiple scales in the Pacific Northwest, USA

Macroinvertebrate assemblages and environmental variables were evaluated at 45 stream sites throughout the Puget Sound Basin, Washington, USA. Environmental variables were measured at 3 spatial scales: reach, local, and whole watershed. Macroinvertebrate distributions were related to environmental variables using canonical correspondence analysis to determine which variables and spatial scales best explained the observed community composition and to identify biota-land cover optima. The calculation of a biota-land cover optimum was a 2-step process. First, an individual taxon's optimum was estimated for a particular land cover by weighting the mean value for that land cover by the abundance of that taxon at all sites. Second, the biota-land cover optimum was determined as the point at which the greatest numbers of taxa, at their calculated optima, appeared for a particular land cover. Sampling reaches were located on streams in watersheds with varying levels of forest, agriculture, and urban/suburban land cover that represented the full range of physical conditions typically found in Puget Sound streams. At the reach scale, taxa composition was correlated with conductivity and mean velocity. At the local and whole-watershed scales, taxa composition was correlated with % forest and agricultural land cover and % forest and bedrock land cover, respectively. For all of the scales, the dominant environmental variables represented an anthropogenic gradient. There was little difference in the amount of variability explained by each spatial scale. At the local-watershed scale, a biota-land cover optimum of ???80 to 90% forest land cover was identified. The total number of taxa at their optima declined rapidly as forest land cover within the local scale declined below 80 to 90%. At the whole-watershed scale, a biota-land cover optimum of 70 to 80% forest land cover was identified. The total number of taxa at their optima declined rapidly as forest land cover within the whole watershed declined below 70 to 80%. Our results suggest that macroinvertebrates can be used both as quantitative indicators of environmental conditions at multiple scales and indicators of land cover optima. Further examination of these optima could be used to establish priorities for conservation and restoration efforts.

Journal of the North American Benthological Societ

Application of a rule-based model to estimate mercury exchange for three background biomes in the continental United States

Ecosystems that have low mercury (Hg) concentrations (i.e., not enriched or impactedbygeologic or anthropogenic processes) cover most of the terrestrial surface area of the earth yet their role as a net source or sink for atmospheric Hg is uncertain. Here we use empirical data to develop a rule-based model implemented within a geographic information system framework to estimate the spatial and temporal patterns of Hg flux for semiarid deserts, grasslands, and deciduous forests representing 45% of the continental United States. This exercise provides an indication of whether these ecosystems are a net source or sink for atmospheric Hg as well as a basis for recommendation of data to collect in future field sampling campaigns. Results indicated that soil alone was a small net source of atmospheric Hg and that emitted Hg could be accounted for based on Hg input by wet deposition. When foliar assimilation and wet deposition are added to the area estimate of soil Hg flux these biomes are a sink for atmospheric Hg. ?? 2009 American Chemical Society.

Environmental Science & Technology

The effect of coachwhip presence on body size of North American racers suggests competition between these sympatric snakes

When sympatric species compete, character divergence may help maintain coexistence. Snakes are often found in species-rich assemblages while exploiting similar resources; because snake body size is a relatively plastic trait that determines the range of prey sizes an individual may consume, divergence in body size between sympatric species may arise as a result of interspecific interactions. The North American racer, Coluber constrictor, and the larger coachwhip, Coluber flagellum, have a close taxonomic relationship and similar foraging strategies. Therefore, we hypothesized that C. constrictor would be smaller where they co-occur with C. flagellum, as compared to where C. flagellum is absent, throughout the southeastern extent of their range. To evaluate this hypothesis, we obtained data on body size for 2321 adult C. constrictor and 526 adult C. flagellum, along with habitat data and other potentially important factors influencing body size. Coluber constrictor was smaller than elsewhere when in peninsular Florida, in pine forests, on hydric soils and in the presence of the larger and potentially competing C. flagellum. Body size of C. flagellum did not vary by any measured habitat variables. The trends we documented are consistent with the hypothesis that C. constrictor body size is influenced by several variables, including co-occurrence with C. flagellum.

Alabama;Florida;Georgia;Louisiana;Mississippi;Sout

Modeling anuran detection and site occupancy on North American Amphibian Monitoring Program (NAAMP) routes in Maryland

One of the most fundamental problems in monitoring animal populations is that of imperfect detection. Although imperfect detection can be modeled, studies examining patterns in occurrence often ignore detection and thus fail to properly partition variation in detection from that of occurrence. In this study, we used anuran calling survey data collected on North American Amphibian Monitoring Program routes in eastern Maryland to investigate factors that influence detection probability and site occupancy for 10 anuran species. In 2002, 17 calling survey routes in eastern Maryland were surveyed to collect environmental and species data nine or more times. To analyze these data, we developed models incorporating detection probability and site occupancy. The results suggest that, for more than half of the 10 species, detection probabilities vary most with season (i.e., day-of-year), air temperature, time, and moon illumination, whereas site occupancy may vary by the amount of palustrine forested wetland habitat. Our results suggest anuran calling surveys should document air temperature, time of night, moon illumination, observer skill, and habitat change over time, as these factors can be important to model-adjusted estimates of site occupancy. Our study represents the first formal modeling effort aimed at developing an analytic assessment framework for NAAMP calling survey data.

Maryland

Pliocene-climate history of the western United States derived from lacustrine ostracodes

Pliocene sediments from several sites in western North America contain ostracodes indicating deposition in lacustrine and wetland settings. The ostracodes offer a means of reconstructing the aquatic paleoenvironment. Because water temperature, chemistry, and lake volume are coupled to climate, reconstruction of these parameters provides a direct insight into Pliocene climate. The site ages were determined from tephrochronology, paleomagnetics, and associated mammals. The morphology of many ostracode species also provides direct information about the paleoenvironment in which they lived. During the Pliocene (about 3.5-2.5 Ma) some species have unusually ornate carapace morphology indicative of large geologically stable lakes, which must have required a stable climate to sustain them. North American Pliocene climate changed from a modern-like state 4.5-3.5 Ma to a period with greater precipitation and less evaporation than today, 3.5-2.5 Ma. This wetter period, inferred from the large geologically long-lived lakes, implies a stable atmospheric circulation pattern. The stable circulation pattern collapsed around 2.5 Ma and climate returned to a modern-like situation.

western United States

Streams in Catskill Mountains still susceptible to acid rain

Precipitation in North America has become less acidic over the past 2 decades because of reduced power plant emissions and compliance with the Clean Air Act Amendments [ Sirois , 19937rsqb;. The 1990 Clean Air Act Amendments were developed to reduce the acidity of sensitive surface waters, which are primarily in upland forested environments, where acidified waters and associated high aluminum concentrations are toxic to many species of aquatic flora and fauna [ Schindler et al. , 1989]. Our studies show that in spite of less acidic precipitation, the buffering capacity of streams in upland forests of the Catskill Mountains in southeastern New York has not increased in recent years. These data suggest that long-term leaching by acid rain has lowered exchangeable calcium ion concentrations in the soil in upland areas, where the underlying, slow-weathering bedrock provides an inadequate supply of cations to neutralize acidity.

New York

Landscape models of brook trout abundance and distribution in lotic habitat with field validation

Brook trout Salvelinus fontinalis are native fish in decline owing to environmental changes. Predictions of their potential distribution and a better understanding of their relationship to habitat conditions would enhance the management and conservation of this valuable species. We used over 7,800 brook trout observations throughout New York State and georeferenced, multiscale landscape condition data to develop four regionally specific artificial neural network models to predict brook trout abundance in rivers and streams. Land cover data provided a general signature of human activity, but other habitat variables were resistant to anthropogenic changes (i.e., changing on a geological time scale). The resulting models predict the potential for any stream to support brook trout. The models were validated by holding 20% of the data out as a test set and by comparison with additional field collections from a variety of habitat types. The models performed well, explaining more than 90% of data variability. Errors were often associated with small spatial displacements of predicted values. When compared with the additional field collections (39 sites), 92% of the predictions were off by only a single class from the field-observed abundances. Among “least-disturbed” field collection sites, all predictions were correct or off by a single abundance class, except for one where brown trout Salmo trutta were present. Other degrading factors were evident at most sites where brook trout were absent or less abundant than predicted. The most important habitat variables included landscape slope, stream and drainage network sizes, water temperature, and extent of forest cover. Predicted brook trout abundances were applied to all New York streams, providing a synoptic map of the distribution of brook trout habitat potential. These fish models set benchmarks of best potential for streams to support brook trout under broad-scale human influences and can assist with planning and identification of protection or rehabilitation sites.

New York