Spatial modeling of harvest constraints on wood supply versus wildlife habitat objectives

We studied the effects of spatial and temporal timber harvesting constraints on competing objectives of sustaining wildlife habitat supply and meeting timber harvest objectives in a boreal mixedwood forest. A hierarchical modeling approach was taken, where strategic and tactical level models were used to project blocking and scheduling of harvest blocks. Harvest block size and proximity, together with short- and long-term temporal constraints, were adjusted in a factorial manner to allow creation of response-surface models. A new measure of the habitat mosaic was defined to describe the emergent pattern of habitat across the landscape. These models, together with multiple linear regression, were used to provide insight on convergence or divergence between spatial objectives. For example, green-up delay (defined as time required before a harvest block adjacent to a previously logged block can be scheduled for harvest) had an adverse effect on the amount of annual harvest area that could be allocated and blocked spatially, and habitat supply responded in an opposite direction to that of wood supply, where caribou, moose wintering, and marten habitat supply increased when harvest blocks were further apart, maximum block size smaller, and both a green-up delay and mesoscale stratification were applied. Although there was no "solution space" free of conflicts, the analysis suggests that application of the mesoscale stratification, together with a diversity of harvest block sizes and a between-harvest block proximity of 250 m, will perform relatively well with respect to wood supply objectives, and at the same time create a less fragmented landscape that better reflects natural forest patterns.     

Simulating Spatial and Temporal Context of Forest Management Using Hypothetical Landscapes

/ Spatially explicit models that combine remote sensing with geographic information systems (GIS) offer great promise to land managers because they consider the arrangement of landscape elements in time and space. Their visual and geographic nature facilitate the comparison of alternative landscape designs. Among various activities associated with forest management, none cause greater concern than the impacts of timber harvesting on the composition, structure, and function of landscape ecosystems. A timber harvest allocation model (HARVEST) was used to simulate different intensities of timber harvest on 23,592-ha hypothetical landscapes with varying sizes of timber production areas and different initial stand age distributions. Our objectives were to: (1) determine the relative effects of the size of timber production areas, harvest intensity, method used to extract timber, and past timber harvest activity on the production of forest interior and edge; and (2) evaluate how past management (in the form of different initial stand age distributions) constrains future timber production options. Our simulations indicated that the total area of forest interior and the amount of forest edge were primarily influenced by the intensity of timber harvest and the size of openings created by harvest. The size of the largest block of interior forest was influenced most by the size of timber harvests, but the intensity of harvest was also significant, and the size of nontimber production areas was important when harvests were numerous and widely dispersed within timber management areas, as is often the case in managed forests. Stand age-class distributions produced by past harvest activity limited the amount of timber production primarily when group selection was used, but also limited clear-cutting when recent harvest levels were high.KEY WORDS: Simulation modeling; Timber harvest; Historical context; Spatial context; Landscape pattern; Forest interior; Forest edge     

IMPROVING AQUATIC HABITAT CONDITIONS OVER TIME WHILE PRODUCING WOOD PRODUCTS: AN EXAMINATION OF OPTIONS1

ABSTRACT: A land management activity scheduling model that can perform a multi-period, simultaneous evaluation of aquatic habitat quality and commodity production goals was used to identify alternatives which would allow the improvement of aquatic habitat conditions over time, while producing wood products. The scheduling model has the ability to use stream sediment index levels, stream temperature index levels, and equivalent clearcut acres (ECA) levels as primary goals. A secondary goal imbedded in the model is the achievement of an even-flow of timber harvest volume, and a tertiary goal is the achievement of maximum efficiency (maximum net present value). The scheduling model utilizes a heuristic programming technique (Tabu search) to guide the selection of timber harvests and road standards. A 14,643 acre case study watershed in eastern Oregon is used to illustrate several policy scenarios. Activities considered include: clearcutting and partial cutting; cable, skyline, ground-based, and helicopter logging; road obliteration; requiring lower truck tire pressures on forest roads; and tree planting in riparian areas. The scheduling model produced land management plans which were spatially and temporally feasible over ten ten-year time periods. Stream temperature was shown to be dramatically reduced if tree planting is performed in all riparian areas, regardless of whether harvesting activities occurred, and including meadows and forested areas where shade density is low. Timber harvest volume levels decreased 31 to 43 percent, and net present value levels decreased 36 to 46 percent, from an unconstrained case, when any of the following occurred: ECA was constrained to 15 percent, sediment index levels were required to decrease by 1 percent per decade, or temperature levels were constrained to “no harvest” levels. The use of a heuristic programming technique is a departure from traditional techniques that are commonly used in management plan development. Yet the heuristic technique allows the inclusion of complex management goals, many of which may be prohibited when using more traditional mathematical programming techniques. In addition, decision variables which require spatial information, requiring them to take on integer or non-linear representations, can be accounted for without realizing the limitations of the traditional techniques.     

EFFECTIVENESS OF TIMBER HARVEST PRACTICES FOR CONTROLLING SEDIMENT RELATED WATER QUALITY IMPACTS1

ABSTRACT: Timber harvest best management practices (BMPs) in Washington State were evaluated to determine their effectiveness at achieving water quality standards pertaining to sediment related effects. A weight‐of‐evidence approach was used to determine BMP effectiveness based on assessment of erosion with sediment delivery to streams, physical disturbance of stream channels, and aquatic habitat conditions during the first two years following harvest. Stream buffers were effective at preventing chronic sediment delivery to streams and physical disturbance of stream channels. Practices for ground‐based harvest and cable yarding in the vicinity of small streams without buffers were ineffective or only partially effective at preventing water quality impacts. The primary operational factors influencing BMP effectiveness were: the proximity of ground disturbing activities to streams; presence or absence of designated stream buffers; the use of special timber falling and yarding practices intended to minimize physical disturbance of stream channels; and timing of harvest to occur during snow cover or frozen ground conditions. Important site factors included the density of small streams at harvest sites and the steepness of inner stream valley slopes. Recommendations are given for practices that provide a high confidence of achieving water quality standards by preventing chronic sediment delivery and avoiding direct stream channel disturbance.     

Assessing Land-Use Impacts on Natural Resources

/ Much information is available on changes that occur in natural resources from both spatially-explicit data on environmental conditions and models of the interactions of these conditions and resources with human activities. The strategy for assessing land-use impacts on natural resources developed in this paper provides a framework for using relevant data and models to address questions of how management practices can promote both use and protection of resources. This assessment strategy integrates spatially explicit environmental data using geographic information systems (GIS) with computer models that simulate changes in land cover in response to land-use impacts. The computer models also simulate susceptibility of species to changes in habitat suitability and landscape patterns. The approach is applied to management of limestone barrens on the Oak Ridge Reservation in East Tennessee. Potential limestone barrens habitats are identified by overlaying appropriate soils, geology, slope, and land-use/land-cover conditions. Their validity is tested against known sites containing rare species that occur in these habitats. The location of habitats at risk in the aftermath of human activities is determined by using an available area model that identifies the size and proximity of sites that particular types of species can no longer use as habitat. The resulting risk map can be used in land management planning. The approach uses readily available in situ and remotely sensed data and is applicable to a wide range of locations and land-use scenarios. This approach can be refined based on needs identified by land managers and on the sensitivity of the results to the resolution of available resource information.KEY WORDS: Land management; Assessment; Habitat characterization; Limestone barrens; Ecological modeling; Geographic information systems     

Integrating forage,wildlife, water,and fish projections with timber projections at the regional level: A case study in southern United States

The impact of timber management and land-use change on forage production, turkey and deer abundance, red-cockaded woodpecker colonies, water yield, and trout abundance was projected as part of a policy study focusing on the southern United States. The multiresource modeling framework used in this study linked extant timber management and land-area policy models with newly developed models for forage, wildlife, fish, and water. Resource production was integrated through a commonly defined land base that could be geographically partitioned according to individual resource needs. Resources were responsive to changes in land use, particularly human-related, and timber management, particularly the harvest of older stands, and the conversion to planted pine.     

Testing the Use of a Land Cover Map for Habitat Ranking in Boreal Forests

Habitat loss and modification is one of the major threats to biodiversity and the preservation of conservation values. We use the term conservation value to mean the benefit of nature or habitats for species. The importance of identifying and preserving conservation values has increased with the decline in biodiversity and the adoption of more stringent environmental legislation. In this study, conservation values were considered in the context of land-use planning and the rapidly increasing demand for more accurate methods of predicting and identifying these values. We used a k-nearest neighbor interpreted satellite (Landsat TM) image classified in 61 classes to assess sites with potential conservation values at the regional and landscape planning scale. Classification was made at the National Land Survey of Finland for main tree species, timber volume, land-use type, and soil on the basis of spectral reflectance in satellite image together with broad numerical reference data. We used the number and rarity of vascular plant species observed in the field as indicators for potential conservation values. We assumed that significant differences in the species richness, rarity, or composition of flora among the classes interpreted in the satellite image would also mean a difference in conservation values among these classes. We found significant differences in species richness among the original satellite image classes. Many of the classes examined could be distinguished by the number of plant species. Species composition also differed correspondingly. Rare species were most abundant in old spruce forests (>200 m³/ha), raising the position of such forests in the ranking of categories according to conservation values. The original satellite image classification was correct for 70% of the sites studied. We concluded that interpreted satellite data can serve as a useful source for evaluating habitat categories on the basis of plant species richness and rarity. Recategorization of original satellite image classification into such new conservation value categories is challenging because of the variation in species composition among the new categories. However, it does not represent a major problem for the purposes of early-stage land-use planning. Benefits of interpreted satellite image recategorization as a rapid conservation value assessment tool for land-use planners would be great.     

TIMBER HARVEST IMPACTS ON SMALL HEADWATER STREAM CHANNELS IN THE COAST RANGES OF WASHINGTON1

ABSTRACT: We evaluated changes in channel habitat distributions, particle‐size distributions of bed material, and stream temperatures in a total of 15 first‐or second‐order streams within and nearby four planned commercial timber harvest units prior to and following timber harvest. Four of the 15 stream basins were not harvested, and these streams served as references. Three streams were cut with unthinned riparian buffers; one was cut with a partial buffer; one was cut with a buffer of non‐merchantable trees; and the remaining six basins were clearcut to the channel edge. In the clearcut streams, logging debris covered or buried 98 percent of the channel length to an average depth of 0.94 meters. The slash trapped fine sediment in the channel by inhibiting fluvial transport, and the average percentage of fines increased from 12 percent to 44 percent. The trees along buffered streams served as a fence to keep out logging debris during the first summer following timber harvest. Particle size distributions and habitat distributions in the buffered and reference streams were largely unchanged from the pre‐harvest to post‐harvest surveys. The debris that buried the clearcut streams effectively shaded most of these streams and protected them from temperature increases. These surveys have documented immediate channel changes due to timber harvest, but channel conditions will evolve over time as the slash decays and becomes redistributed and as new vegetation develops on the channel margins.     

A method to create simplified versions of existing habitat suitability index (HSI) models

The habitat evaluation procedures (HEP), developed by the US Fish and Wildlife Service, are widely used in the United States to determine the impacts of major construction projects on fish and wildlife habitats. HEP relies heavily on habitat suitability index (HSI) models that use measurements of important habitat characteristics to rate habitat quality for a species on a scale of 0 (unsuitable) to 1.0 (optimal). This report describes a method to simplify existing HSI models to reduce the time and expense involved in sampling habitat variables. Simplified models for three species produced HSI values within 0.2 of those predicted by the original models 90% of the time. Simplified models are particularly useful for rapid habitat inventories and evaluations, wildlife management, and impact assessments in extensive areas or with limited time and personnel.     

Changes in Upland Wildlife Habitat on Farmland in Illinois 1920–1987

/ An index of upland wildlife habitat was developed to investigate patterns and changes in habitat over time, using four years (1920, 1940, 1964, 1987) and the state of Illinois as an example. The index was composed of two subdivisions that described, at the county level, the quantity of wildlife habitat and a third subdivision that described farming disturbances that impacted the quality of the habitat. Data came from the US Census of Agriculture. The first subdivision that reflected quantity of habitat was called the wildlife habitat subdivision and was the sum of percentage woodland on farms, percentage farmland in nonrow crops, and percentage farmland in set-aside programs. The second subdivision that reflected the quantity of habitat was termed the soil-related features subdivision and was the sum of the percentage of farmland that was not highly erodible, the percentage of farmland in soil-protecting crops, and the percentage of farmland in conservation tillage. The third subdivision, reflecting the quality of the habitat, was the farming disturbance subdivision and was the sum of the percentage of grazing and the percentage of land on which fertilizers, pesticides, and herbicides were applied. Overall, major decreases occurred between 1920 and 1987 in the subdivisions reflecting the quantity of wildlife habitat and a major increase occurred in the subdivision associated with farming disturbance, reflecting the intensification of agriculture in the state. However, there was variability throughout the state, with some counties being more favorable to wildlife (as measured by the subdivisions) than others. Most of the changes within the state for the subdivisions reflecting quantity of upland wildlife habitat occurred during 1940 while changes in the farming disturbance subdivision (reflecting habitat quality) occurred in 1964. By 1987, the western and southern parts of Illinois were the most favorable for wildlife as reflected in all three subdivisions. Upland wildlife harvest indices were related to the subdivisions in 1964 and 1987, when harvest indices were available. Cottontail and northern bobwhite harvests were higher in counties with higher amounts of the wildlife habitat subdivision in both years. Cottontail harvest was also higher in counties with lower levels of the farming disturbance subdivision in 1964 and higher levels of soil-related features subdivision in 1987. Indices at the county level have the potential to be used in a multiscale analysis to investigate the impact of policy changes on large-scale areas of the Midwest and to develop regional perspectives of the impacts of agriculture on upland wildlife and their habitats.KEY WORDS: Upland wildlife; Habitat; Agriculture; Illinois     

Modeling the Influence of Dynamic Zoning of Forest Harvesting on Ecological Succession in a Northern Hardwoods Landscape

Dynamic zoning (systematic alteration in the spatial and temporal allocation of even-aged forest management practices) has been proposed as a means to change the spatial pattern of timber harvest across a landscape to maximize forest interior habitat while holding timber harvest levels constant. Simulation studies have established that dynamic zoning strategies produce larger tracts of interior, closed canopy forest, thus increasing the value of these landscapes for interior-dependent wildlife. We used the simulation model LANDIS to examine how the implementation of a dynamic zoning strategy would change trajectories of ecological succession in the Great Divide Ranger District of the Chequamegon–Nicolet National Forest in northern Wisconsin over 500 years. The components of dynamic zoning strategies (number of zones in a scenario and the length of the hiatus between successive entries into zones) and their interaction had highly significant impacts on patterns of forest succession. Dynamic zoning scenarios with more zones and shorter hiatus lengths increased the average amount of the forest dominated by early successional aspen (Populus sp.). Dynamic zoning scenarios with two zones produced more late successional mature northern hardwoods than scenarios with four zones. Dynamic zoning scenarios with very short (30 years) or very long (120 years) hiatus lengths resulted in more late successional mature northern hardwoods than scenarios with intermediate hiatus lengths (60 and 90 years). However, none of the dynamic scenarios produced as much late successional mature northern hardwoods as the static alternative. Furthermore, the amounts of all habitat types in all dynamic zoning scenarios fluctuated greatly in time and space relative to static alternatives, which could negatively impact wildlife species that require a stable amount of habitat above some minimum critical threshold. Indeed, implementing dynamic zoning scenarios of different designs would have both positive and negative effects on wildlife species and for other objectives of forest management.     

Influences on Wood Load in Mountain Streams of the Bighorn National Forest,Wyoming, USA

We documented valley and channel characteristics and wood loads in 19 reaches of forested headwater mountain streams in the Bighorn National Forest of northern Wyoming. Ten of these reaches were in the Upper Tongue River watershed, which has a history of management including timber harvest, tie floating, and road construction. Nine reaches were in the North Rock Creek watershed, which has little history of management activities. We used these data to test hypotheses that (i) valley geometry correlates with wood load, (ii) stream gradient correlates with wood load, and (iii) wood loads are significantly lower in managed watersheds than in otherwise similar unmanaged watersheds. Statistical analyses of the data support the first and third hypotheses. Stream reaches with steeper valley side slopes tend to have higher wood loads, and reaches in managed watersheds tend to have lower wood loads than reaches in unmanaged watersheds. Results do not support the second hypothesis. Shear stress correlated more strongly with wood load than did stream gradient, but statistical models with valley-scale variables had greater explanatory power than statistical models with channel-scale variables. Wood loads in stream reaches within managed watersheds in the Bighorn National Forest tend to be two to three times lower than wood loads in unmanaged watersheds.     

Opportunity costs of implementing forest plans

Intellectual concern with the National Forest Management Act of 1976 has followed a course emphasizing the planning aspects of the legislation associated with the development of forest plans. Once approved, however, forest plans must be implemented. Due to the complex nature of the ecological systems of interest, and the multiple and often conflicting desires of user clientele groups, the feasibility and costs of implementing forest plans require immediate investigation. For one timber sale on the Coconino National Forest in Arizona, forest plan constraints were applied and resulting resource outputs predicted using the terrestrial ecosystem analysis and modeling system (TEAMS), a computer-based decision support system developed at the School of Forestry, Northern Arizona University, With forest plan constraints for wildlife habitat, visual diversity, riparian area protection, and soil and slope harvesting restrictions, the maximum timber harvest obtainable was reduced 58% from the maximum obtainable without plan constraints.Former Graduate Student at Northern Arizona University.     

Cumulative silvicultural impacts on watersheds: A hydrologic and regulatory dilemma

Because of the nature of watersheds, the hydrologic and erosional impacts of logging and related road-building activities may move offsite, affecting areas downslope and downstream from the operation. The degree to which this occurs depends on the interaction of many variables, including soils, bedrock geology, vegetation, the timing and size of storm events, logging technology, and operator performance. In parts of northwestern California, these variables combine to produce significant water quality degradation, with resulting damage to anadromous fish habitat.Examination of recent aerial photographs, combined with a review of public records, shows that many timber harvest operations were concentrated in a single 83 km² watershed in the lower Klamath River Basin within the past decade. The resulting soil disturbance in this case seems likely to result in cumulative off-site water quality degradation in the lower portion of the Basin.In California, both state and federal laws require consideration of possible cumulative effects of multiple timber harvest operations. In spite of recent reforms that have given the state a larger role in regulating forest practices on private land, each timber harvest plan is still evaluated in isolation from other plans in the same watershed. A process of collaborative state-private watershed planning with increased input of geologic information offers the best long-term approach to the problem of assessing cumulative effects of multiple timber harvest operations. Such a reform could ultimately emerge from the ongoing water quality planning process under Section 208 of the amended Federal Water Pollution Control Act.     

Using Satellite Imagery to Assess Large-Scale Habitat Characteristics of Adirondack Park, New York, USA

Satellite imagery is a useful tool for large-scale habitat analysis; however, its limitations need to be tested. We tested these limitations by varying the methods of a habitat evaluation for white-tailed deer (Odocoileus virginianus) in the Adirondack Park, New York, USA, utilizing harvest data to create and validate the assessment models. We used two classified images, one with a large minimum mapping unit but high accuracy and one with no minimum mapping unit but slightly lower accuracy, to test the sensitivity of the evaluation to these differences. We tested the utility of two methods of assessment, habitat suitability index modeling, and pattern recognition modeling. We varied the scale at which the models were applied by using five separate sizes of analysis windows. Results showed that the presence of a large minimum mapping unit eliminates important details of the habitat. Window size is relatively unimportant if the data are averaged to a large resolution (i.e., township), but if the data are used at the smaller resolution, then the window size is an important consideration. In the Adirondacks, the proportion of hardwood and softwood in an area is most important to the spatial dynamics of deer populations. The low occurrence of open area in all parts of the park either limits the effect of this cover type on the population or limits our ability to detect the effect. The arrangement and interspersion of cover types were not significant to deer populations.     

Linking linear programming and spatial simulation models to predict landscape effects of forest management alternatives

Forest management planners require analytical tools to assess the effects of alternative strategies on the sometimes disparate benefits from forests such as timber production and wildlife habitat. We assessed the spatial patterns of alternative management strategies by linking two models that were developed for different purposes. We used a linear programming model (Spectrum) to optimize timber harvest schedules, then a simulation model (HARVEST) to project those schedules in a spatially explicit way and produce maps from which the spatial pattern of habitat could be calculated. We demonstrated the power of this approach by evaluating alternative plans developed for a national forest plan revision in Wisconsin, USA. The amount of forest interior habitat was inversely related to the amount of timber cut, and increased under the alternatives compared to the current plan. The amount of edge habitat was positively related to the amount of timber cut, and increased under all alternatives. The amount of mature northern hardwood interior and edge habitat increased for all alternatives, but mature pine habitat area varied. Mature age classes of all forest types increased, and young classes decreased under all alternatives. The average size of patches (defined by age class) generally decreased. These results are consistent with the design goals of each of the alternatives, but reveal that the spatial differences among the alternatives are modest. These complementary models are valuable for quantifying and comparing the spatial effects of alternative management strategies.     

A harvest failure approach to assess the threat from an invasive species

We present the idea of using potential infringements on annual allowable harvest targets as an approach to estimate threats from invasive species to the forest products sector. The approach uses present-day harvest levels as a reference level to estimate when and where the impact of a nonnative forest pest could become economically damaging. We use a generic model that simulates spread and damage by nonnative invasive species, basic harvest and forest growth through time. The concept is illustrated with a case study of a new nonnative invasive pest, Sirex noctilio Fabricius on pine resources in eastern Canada. Impacts of invasion on wood supply, in particular, the point at which present-day harvest levels are not attainable, were identified for 77 non-overlapping geographical regions that delimit the primary wood supply areas around large mills and wood processing facilities in eastern Canada. The results identify the minimum area of a pest outbreak that could trigger harvest shortages (approximately 12.5–14 M ha of pine forests in Ontario and Quebec). Beyond this level, the amount of host resource available for harvesting in any given year declines rapidly. The failure to sustain broad-scale harvest targets may be an attractive and intuitive indicator for policy makers and regulators interested in developing control and “slow-the-spread” programs for non-native forest pests.     

The Effects of Population Density on Juvenile Growth Rate in White-Tailed Deer

Animal body size is driven by habitat quality, food availability, and nutrition. Adult size can relate to birth weight, to length of the ontogenetic growth period, and/or to the rate of growth. Data requirements are high for studying these growth mechanisms, but large datasets exist for some game species. In North America, large harvest datasets exist for white-tailed deer (Odocoileus virginianus), but such data are collected under a variety of conditions and are generally dismissed for ecological research beyond local population and habitat management. We contend that such data are useful for studying the ecology of white-tailed deer growth and body size when analyzed at ordinal scale. In this paper, we test the response of growth rate to food availability by fitting a logarithmic equation that estimates growth rate only to harvest data from Fort Hood, Texas, and track changes in growth rate over time. Results of this ordinal scale model are compared to previously published models that include additional parameters, such as birth weight and adult weight. It is shown that body size responds to food availability by variation in growth rate. Models that estimate multiple parameters may not work with harvest data because they are prone to error, which renders estimates from complex models too variable to detect interannual changes in growth rate that this ordinal scale model captures. This model can be applied to harvest data, from which inferences about factors that influence animal growth and body size (e.g., habitat quality and nutritional availability) can be drawn.     

Resampling methods for evaluating classification accuracy of wildlife habitat models

Predictive models of wildlife-habitat relationships often have been developed without being tested The apparent classification accuracy of such models can be optimistically biased and misleading. Data resampling methods exist that yield a more realistic estimate of model classification accuracy These methods are simple and require no new sample data. We illustrate these methods (cross-validation, jackknife resampling, and bootstrap resampling) with computer simulation to demonstrate the increase in precision of the estimate. The bootstrap method is then applied to field data as a technique for model comparison We recommend that biologists use some resampling procedure to evaluate wildlife habitat models prior to field evaluation.