Assessing Demographic Change for Conservation Purposes


The Use of Population Viability Analyses in Conservation Planning. Sjögren-Gulve, P., and T. Ebenhard , editors . 2000 . Ecological Bulletins 48. $35.00 (hardback) . Wallin & Dalholm , Lund, Sweden. ISBN 87–16–16382–6 .

Modeling has become an important tool for guiding management and conservation planning. In particular, the use of computer programs to evaluate population viability has increased markedly in recent years. Several recent volumes on population viability analysis (PVA) demonstrate the demand for comprehensive reviews of this topic. An example of this effort is the outstanding volume resulting from the symposium “The Use of Population Viability Analyses in Conservation Planning,” organized by the Swedish Environmental Protection Agency and the Swedish Biodiversity Centre in Stockholm in December 1997. This well-organized book reviews the structure, assumptions, and applications of the three main approaches to population viability analysis: spatially realistic occupancy models, age- or stage-structured models, and individual-based models.

In the first chapter, Akçakaya and Sjögren-Gulve compare PVA with alternative methods of conservation assessment. The authors present compelling evidence supporting PVA as a rigorous methodology that uses different types of data, incorporates uncertainties and natural variability, and delivers products or predictions relevant to conservation goals. They emphasize as major disadvantages the single-species focus of PVA and its requirements for large quantities of data that may not be available for many species. Their guidelines are particularly informative for choosing appropriate PVA models, which range from simple occupancy metapopulation models to data-intensive, individual-based population models. Akçakaya and Sjögren-Gulve conclude that model choice depends on the question asked but is often constrained by limited information and financial resources.

The three basic approaches to PVA are reviewed separately in the next chapters. Flexibility is the main advantage of structured and individual-based models over occupancy models. Akçakaya's review of demographically structured models clearly suggests that, because these models group individuals into distinct classes ( based on demographic characteristics, their location, or both), they can incorporate multiple biological factors and can represent spatial structure in various ways. Even more detailed information can be considered in individual-based models. Lacy argues that as the number of individuals decreases, the number of processes significantly affecting endangered species persistence increases and becomes more idiosyncratic. Accurate PVA of small populations may often require individual-based models to simulate specific threats and their synergistic interactions. By the same token, the main disadvantages of these two PVA approaches when compared with occupancy models are their larger data requirements.

Sjögren-Gulve and Hanski describe three patch-occupancy metapopulation models. These models have the lowest data requirements and are most applicable to dynamic systems in subdivided landscapes. They evaluate changes in the proportion of habitat patches inhabited by a focal species. The classic model of Levins assumes that all patches are identical, whereas the logistic-regression metapopulation model and the incidence-function model are spatially realistic. Presence and absence data are used to parameterize incidence- function models, but colonization and extinction records are required for logistic-regression models. Sjögren-Gulve and Hanski favor a complementary use of different PVA approaches.

Plant studies are better represented here than in previous reviews of PVA. Menges stresses the role of several life-history peculiarities in determining special approaches for plant PVAs. Dormant stages, periodic recruitment or reproduction, and clonal growth are difficult to demonstrate or describe. Data limitations and conceptual complexity make it challenging to incorporate these stages in models. He also warns that the short duration of most available plant PVAs may not be adequate for reliable predictions. Population viability analysis models are particularly well-suited to describing the effects of disturbance on plant populations. Lennartsson's PVA of Gentianella campestris, based on a management experiment involving three levels of grazing intensity and two levels of microsite moisture, provides guidance for optimal grassland management of this species.

Four other case studies are also presented. Kindvall uses data on the bush cricket (Metrioptera bicolor) to compare three different spatially realistic metapopulation models. Similar good predictions of turnover rates and temporal changes in regional occupancy from an incidence-function model, a logistic-regression model, and a demographic model indicate that sophisticated models may not be necessary for patch-level analysis of viability. Berglind simulates five management scenarios for populations of the sand lizard (Lacerta agilis) in the northern limit of its range in central Sweden. His demographic simulations indicate that only extreme recovery strategies, including captive raising, captive breeding, and habitat management, may reduce risks of decline and extinction. Ebenhard reviews the application of individual-based models to the wolf, otter, and Peregrine Falcon, which allowed analyses of the effects of inbreeding depression and loss of genetic variation. He mentions that although, in general, PVAs work positively to influence policies adopted and actions taken by agencies, problems occur due to the lack of perceived ownership of PVA results by government agencies and communication of the probabilistic nature of PVAs. Vos, Ter Braak, and Nieuwenhuizen present an incidence-function model used to quantify spatial habitat requirements for the protection and long-term survival of the frog Hyla arborea. The model was useful for conservation assessment when different restoration scenarios were compared, and it provided general guidelines for management of tree-frog populations.

Description and evaluation of spatial patterns of biotic diversity are central to making wise decisions about land use and conservation strategies. Fleishman, Jonsson and Sjögren-Gulve describe a procedure using focal species correlated with species richness as indicators to facilitate monitoring and management. They apply this procedure to cryptogams in boreal forests in northern Sweden and butterflies in mountain ranges in the western United States. Their results suggest that the value of focal species varies with taxonomic group and ecosystem. But the selection of study variables also may play an important role. Focal species provided more reliable estimates of species richness for boreal cryptogams than for butterflies.

An interesting remark is made in the concluding chapter. Notwithstanding the potential value of PVA as a tool for identifying threatened species, it has seldom been used in the new red list system from the World Conservation Union. This is a common limitation shared with other species catalogues for species protection. Gärdenfors claims that inadequate data and unfamiliarity with PVA partially explain its limited use. He also argues that biased selection of species favoring vertebrates and an emphasis on small populations threatened by factors affecting individuals in a way that can be described by pure stochastic models limit the application of PVAs. Population viability analysis rarely is performed for plants or invertebrates, species with large but decreasing populations, or species affected by deterministic negative changes in habitat or other resources.

This is a carefully edited volume that includes practical information. The structure of Vortex, an individual-based simulation program, is explained in detail in the last chapter. Links to software available via the Internet are included in several chapters. This volume is likely to be useful even to the layperson. Color pictures illustrate study sites and species. Colored boxes call attention to descriptions of modeling details.

The Use of Population Viability Analyses in Conservaton Planning is an enlightening contribution that will help to disseminate PVA methods among a broad spectrum of people interested in conservation.