|Forest Investment Account (FIA) - Forest Science Program|
|FIA Project Y062074|
|A framework to support landscape analyses of habitat supply and effects on populations of forest-dwelling species|
|Contributing Authors: Sutherland, Glenn D.; O'Brien, Daniel T.; Fall, S. Andrew; Waterhouse, F. Louise; Harestad, Alton S.; Buchanan, Joseph B.|
|Imprint: [Victoria, B.C.] : British Columbia Ministry of Forests and Range, 2007|
|Subject: Forest Investment Account (FIA), Spotted Owl, Habitat, British Columbia, Endangered Species|
|Series: Forest Investment Account (FIA) - Forest Science Program|
|Planning tools and decision-making processes to support sustainable forestry are an integral part of practicing good forest stewardship in British Columbia. The challenges when applying stewardship principles are often at their greatest when resource extraction activities and habitats of forest-dependent species overlap. Tools to represent and integrate information about both ecological processes and predicted consequences of forest management activities, and approaches for comparing costs and benefits of both economic and environmental values, are evolving to meet this challenge. In this document we present a spatial modelling framework designed to assist those confronting these challenges to sustainable forestry. Users can use this framework as a tool to evaluate hypotheses about the ecological and economic consequences of management strategies. Of particular interest is the capability of the framework to assist in the search for acceptable trade-offs between social and ecological values¡Xa necessary but challenging requirement of meeting good stewardship objectives in natural resource management.|
We illustrate application of the framework using an endangered species in British Columbia, the Northern Spotted Owl (Strix occidentalis caurina; SPOW). Our approach was designed to help decision-makers understand the probable roles of currently hypothesized threats to the population in modelled experiments conducted within the framework. We developed indicators representing the condition of the landscape, volumes of merchantable timber harvested from the landscape, and several types of indicators representing population-level status of Spotted Owls. The main questions we examined during the evolution of the framework were:
- What is a reasonable recovery goal for the study species (Spotted Owl) expressed as the number of breeding pairs?
- Is habitat loss a continuing threat, and if so, how?
- Is habitat recovery possible, and if so, when and where?
- Can potential outcomes for both the case study species and socioeconomic values using a suite of potential management policies be demonstrated?
- Is some suitable habitat of better quality than others? Does the definition of suitable habitat need to account for spatial locations of current and potential populations, a concept related to the idea of 'critical habitat'?
- Where should we place our species-specific management areas to capitalize on habitat?
- Can we better understand the relationship between the recovery goal, the current population size, and current habitat amount and configuration?
- Could Barred Owls (Strix varia varia; BDOW) be a significant threat?
To help answer these questions, we developed models for spatial landscape projection, ecological classification, cross-scale habitat assessment, population dynamics, and reserve selection. The modelling framework used to represent these components is necessarily a simplistic representation of a very complex reality (Walters Æ̀986). Sufficient empirical data needed to define functional relationships were not always available. Estimates of parameters, even where data are available, required care in their use and interpretation. These, combined with informed expert judgements about many key hypotheses and relationships, formed the basis of model building and testing. The following chapters outline the data and assumptions used to model the Spotted Owl, the development of the suite of tools for the framework, and the findings on both the model framework and the Spotted Owl as synthesized through the framework.
Section 1 presents an overview of the modelling framework, and describes the six integrated, spatially explicit model components. These are:
1. a landscape dynamics model for projecting forest growth and stand-replacing natural disturbances that is capable of fully spatial timber supply analyses;
2. a habitat supply model that can be tailored for particular species;
3. a spatial model for calculating locations of potential territories for a territorial species;
4. a structural connectivity model for assessing spatial arrangement and proximity of habitat, territories, and management areas;
5. a spatial population model for projecting population dynamics of a particular species on projected landscapes; and
6. an evaluation post-processor that implements rules for identifying and ranking potential habitat reserves based on biological and other criteria measured at multiple scales.
Section 2 describes the ecological and management problem of recovery planning for the Spotted Owl that formed the case study we used to develop and test the framework. Evidence indicates that the Spotted Owl population in British Columbia is small and declining. Currently known and potential threats to this species in British Columbia include:
1. loss of nesting and/or foraging habitat,
2. fragmentation of nesting and/or foraging habitat,
3. negative effects from environmental and genetic factors related to small population sizes,
4. competition from Barred Owls,
5. climate change, and
We used the components of the framework to test a number of ecological hypotheses about the first four of these threats to learn how projected outcomes behave in relation to our assumptions about the causal factors influencing the status of this species.
Sections 3-7 describe the primary ecological modelling components of the framework for projecting future ecological states. The landscape dynamics component (Section 3) combines a spatially explicit forest state model with a stand-replacing natural disturbance model to estimate sustainable harvest flows and to project spatial time-series of forest-state indicators (e.g., stand age, height, structure, disturbances) for a particular 'landscape change' scenario. The ecological consequences for the case study species (Spotted Owl) of the projected landscape dynamics under each scenario are then assessed using the finest spatial scale (termed site-scale) habitat classification models for foraging, nesting, and movement (Section 4) based on biophysical variables representing the influences of climate, topography, vegetation structure, and composition. We then evaluate habitat at the coarser scale of potential territories (Section 5), searching for those areas where the spatial configuration of habitat meet criteria for supporting a breeding site and territory. At a still coarser scale, the spatial proximity and clustering of habitat across the landscape is evaluated using spatial graph techniques for measuring connectivity (Section 6). The results at this scale of ecological assessment provide data on the effects of loss of connectivity on individuals or the population, and can also be used to investigate the efficacy of such management options as potential habitat corridors or reserves. In Section 7, we explore the consequences of the changing landscape structure upon individuals and the population using an individual-based spatially explicit population model. This model permits systematic study of alternative hypotheses of habitat change, demographic factors (e.g., recruitment, survival), and dispersion of nest sites on potential population trends. It can also be extended to assess effects of other threats (e.g., competition from Barred Owls, climate change).
Sections 8-10 demonstrate the post-processing analyses of indicators produced by each model component to inform decisions on the types of questions involved in recovery planning. Section 8 describes a habitat quality assessment tool built using a Bayesian belief network that weights selected habitat attributes measured at the site, territory, and population scales. It thus obtains an integrated measure of biological habitat quality for each spatial location that is deemed to be 'suitable habitat.' This habitat quality evaluation can be used to facilitate selection of critical habitat locations for the study species. In Section 9, we advance this concept further by using a resource location model that selects candidate habitat reserve areas that meet biological and/or risk criteria for recovery goals at different times in the future. This approach is particularly useful for land-use planning problems involving species conservation because it facilitates efficient selection of habitat that meets both current and future biological goals for the amount and spatial configuration of habitat and other biological criteria while minimizing impacts on other values. In Section 10, we illustrate how to apply the outputs of the framework to evaluate different policy options for forest and species management, and compare their ecological and economic costs and benefits.
Finally, in Section 11 we: (1) summarize the strengths and weaknesses of the design and implementation of the framework for spatial projections of large-scale ecological and management problems such as those found in recovery planning; and (2) present key findings for the current population, future population, habitat management for recovery, and habitat requirements derived for the Spotted Owl case study. This case study species is of significant conservation concern in Canada and in British Columbia as well as elsewhere in the Pacific Northwest; thus our findings are of interest well beyond simply demonstrating analytical and modelling approaches. The findings of the research must be considered collectively, as they apply to the issues of recovery of this species in British Columbia.
We conclude by noting that several aspects of the resulting framework build upon and extend previously developed model approaches and concepts. Our design approach of separating the main ecological, management, and analysis components of the system into relatively autonomous components (e.g., timber supply analysis, landscape dynamics, habitat supply, territory analysis, connectivity analysis, and population dynamics) allowed us to efficiently and rigorously explore different hypotheses about the causes of declines in Spotted Owl populations. In turn, careful design of modelling experiments allowed us to elucidate the relative influences of different factors (habitat, management, demographics) on recovery options. Looking beyond the specific analyses undertaken in a particular study or the conclusions drawn from the results, we believe that a substantial benefit of this project was the process formulated to develop the framework, which promoted communication and learning among stakeholders about the intricacies of a complex and difficult resource management problem. We are (and must be) fairly conservative in our interpretation of the findings obtained with the framework in our case study. From the outset, we did not expect spatial modelling results alone to provide a complete solution for recovery of either the British Columbia Spotted Owl population or indeed any species, because of uncertainties in biological parameters, in inventory data, and in describing and projecting all possible threats to populations. We argue that the structure of the framework is very amenable to further informing (and being informed by) long-term monitoring programs for recovering species designed to assess management strategies established to promote the chances of recovering an endangered species or population.
compiled and edited by G.D. Sutherland, D.T. O'Brien, S.A. Fall, F.L. Waterhouse, A.S. Harestad, J.B. Buchanan.
|Related projects:  FSP_Y051074|
Executive Summary (17Kb)
Technical Report 38 - Northern Spotted Owl
To view PDF documents you need Adobe Acrobat Reader, available free from the Adobe Web Site.
Updated August 16, 2010
Please direct questions or comments regarding publications to For.Prodres@gov.bc.ca