Society for Conservation Biology 16th Annual Meeting July 14-July 19 2002
co-hosted by DICE and the British Ecological Society


Abstracts for Planning and Reserve Design
Session One

Monday 15th July, 10.15 - 12.15, Grimond Lecture Theatre 2

Chair: Sandy Andelman



Conference Home Page | Session timetable


(BLOCK CAPITALS indicate the presenting author)


10.15 - 10.30
RAHN, MATTHEW, Holly Doremus, and James Diffendorfer. San Diego State University, Department of Biology, 5500 Campanile Drive, San Diego, California 92182, USA, <mrahn@sciences.sdsu.edu> (MR, JD), University of California, Davis, School of Law, 400 Mrak Hall Dr., Davis, CA 95616, USA (HD).

CHOOSING SPECIES FOR COVERAGE IN HABITAT CONSERVATION PLANS: WHERE’S THE SCIENCE?

Modern conservation planning focuses on large-scale multi-species programs intended to provide ecosystem-level protection. Habitat Conservation Plans (HCPs) prepared under section 10(a) of the U.S. Endangered Species Act provide an example of the difficulties in effectively planning for ecosystem protection. While HCPs primarily focus on federally listed species, they are increasingly being designed to cover proposed, candidate, rare, or declining unlisted species. According to the U.S. Fish and Wildlife Service, this serves to increase the biological value of HCPs through comprehensive multi-species or ecosystem planning providing early, proactive consideration for unlisted species. The lack of criteria for developing covered species lists, however, thwarts this objective. Selection criteria for covered species should reflect a stringent science based approach to multi-species monitoring and ecosystem conservation. In particular, the current lack of coverage of insects in HCPs constitutes a significant gap in "ecosystem-level" information. Covered species lists typically include few insects. Those insects included are generally chosen for their charisma rather than their value as ecological indicators. We propose a framework for developing ecologically valid covered species lists, providing appropriate treatment for insects. This framework is not limited to the HCP context; it could be usefully applied to any large-scale conservation effort.


10.30 - 10.45
HESS, GEORGE R., Ray C. Bode, Terri J. King, Matthew J. Rubino. North Carolina State University, Forestry Department, Raleigh, NC 27695-8002 USA grhess@ncsu.edu (GRH, RCB), North Carolina GAP Program, Raleigh, NC 27695-7617 USA (TJK), and EcoScience, 1101 Haynes Street, Suite 101 Raleigh, NC 27604 USA (MJR).

REGIONAL PLANNING FOR WILDLIFE USING A FOCAL SPECIES APPROACH

We are using a focal species approach to develop a wildlife conservation plan for a rapidly suburbanizing region of North Carolina, USA, because of the need to act quickly with incomplete information. While working in a specific region, our larger goal is to prototype an approach that is broadly applicable, relatively rapid and inexpensive, and scientifically defensible. The process consists of identifying focal species for planning; mapping potential habitat for each species using GIS technology; verifying the habitat maps in the field; testing how well the focal species serve as conservation umbrellas; and combining the maps to depict a regional network of open spaces for wildlife. A panel of wildlife experts identified six landscape types and nine associated focal species. We evaluated habitat requirements for each species and matched them with readily available GIS data to create habitat maps. To date, we have tested the habitat models for two of the focal species. Using a simple presence-absence approach, we found that 65-80% of the habitat patches contained the focal species. The patches served as an umbrella for only a small portion of known rare species, because most rare species were found in patches too small for focal species persistence.




10.45 - 11.00
RODRIGUEZ, JON PAUL and Christopher J. Sharpe. Centro de Ecología, Instituto Venezolano de Investigaciones Científicas, Apartado 21827, Caracas 1020-A, Venezuela <jonpaul@ivic.ve> (JPR), Provita, Apartado 47552, Caracas 1041-A, Venezuela (CJS).

TOWARDS THE FIRST REGION-WIDE NETWORK OF SURVEY SITES FOR MONITORING NEOTROPICAL BIRDS

Large scale tropical biodiversity surveys, though essential for conservation planning, have been unfeasible for lack of trained local personnel and financial resources. Roadside surveys, one potential alternative, cover large geographical areas quickly, at relatively low cost. The North American Breeding Bird Survey (BBS) has generated abundant data on avian population trends since 1966, while building a strong constituency among amateur and professional ornithologists. Compared to the temperate zone, however, Neotropical bird communities are more diverse, occupy habitats of higher complexity, and lack breeding synchrony -- characteristics which may disqualify roadside surveys in the Neotropics. BBS routes consist of 50 3-min point counts spaced along 40 km of road. Is it necessary to perform longer point counts to adequately quantify Neotropical bird communities? By implementing a series of variations on the standard BBS protocol, we explore this question in three contrasting Venezuelan habitats, differing both in bird richness and detectability. Our findings indicate that increasing the time of point counts does not improve the estimate of species richness: the BBS protocol detects most species, most accurately. We conclude that roadside surveys are adequate for monitoring Neotropical birds. Our next objective is to design and implement a Neotropical BBS throughout the region.




11.00 - 11.15
SMITH, ROBERT J., Peter S. Goodman, David N. Johnson and Nigel Leader-Williams. Durrell Institute of Conservation and Ecology, University of Kent at Canterbury, Canterbury CT2 7NS, UK, <R.J.Smith@ukc.ac.uk> (RJS). KwaZulu-Natal Wildlife, P.O. Box 13053, Cascades, 3202, South Africa (PSG). KwaZulu-Natal Wildlife, P.O. Box 13053, Cascades, 3202, South Africa (DNJ). Durrell Institute of Conservation and Ecology, University of Kent at Canterbury, Canterbury CT2 7NS, UK (NLW).

PROTECTING THE OBVIOUS: REDUCING THE EFFECTS OF SAMPLING BIAS IN CONSERVATION PLANNING

Rarity-based algorithms are commonly used in conservation planning to identify efficient protected area systems based on species distribution records. Unfortunately, such records are often affected by sampling bias, so that some widely distributed species are only recorded in heavily sampled areas. As a consequence, these species appear to have restricted ranges, leading to the selection of those areas where sampling effort was greatest. Therefore, it is important to identify species that are less affected by sampling bias and only use their distributional data in subsequent planning exercises. This approach was tested using recorded and modelled bird species lists for the 17 quarter-degree grid squares in Maputaland, South Africa. These data were used to calculate how accurately the distribution of each species had been recorded and to test whether this was influenced by their physical appearance, diet and habitat preferences. The results showed that species with a distinctive appearance and/or song were recorded more accurately, and were less affected by sampling bias. Furthermore, using recorded distribution data from distinctive species, instead of all species, greatly affected the results of a planning exercise involving the 171 quarter-degree grid squares found in KwaZulu-Natal.




11.15 - 11.30
ANDELMAN, SANDY J. and Michael R. Willig. National Center for Ecological Analysis and Synthesis, University of California, Santa Barbara, 735 State Street, Suite 300, Santa Barbara, CA 93101, USA, <andelman@nceas.ucsb.edu> (SJA), Department of Biological Sciences, Texas Tech University, Lubbock, Texas 79409-3131, USA (MRW).

ALTERNATIVE CONSERVATION RESERVE CONFIGURATIONS FOR PARAGUAYAN BATS: CONSIDERATIONS OF SPATIAL SCALE

The application of systematic and quantitative approaches to conservation planning is increasing; however, the quality and quantity of data available to planners remains inadequate. We used two databases on bat species distributions at 25 sites in Paraguay to illustrate some of the consequences of the spatial scale of sampling and data quality on decisions about reserve-siting. We used a simulated annealing algorithm to identify alternative scenarios for comprehensive representation of the nation’s bat fauna within a system of reserves and to evaluate the contribution of existing protected areas in Paraguay to this conservation goal. The location, efficiency, and level of protection (i.e., the number of populations of each species protected) were affected by both spatial scale and source of data. Our results suggest that systematic and intensive biodiversity surveys are an important element of efficient conservation planning for biodiversity conservation.




11.30 - 11.45
SPECTOR, SACHA. Center for Biodiversity and Conservation, American Museum of Natural History, Central Park West at 79th St., New York, NY 10024, USA, <spector@amnh.org>.

INTEGRATING REMOTE SENSING AND FIELD DATA TO PREDICT DUNG BEETLE ASSEMBLAGE STRUCTURE AND COMPOSITION

Remote sensing technologies hold enormous promise for the fields of biodiversity assessment, prediction and conservation. When coupled with field data, remotely sensed information can provide insights into natural patterns at a scale and resolution unobtainable by other practical means and permit predictions about the ranges of species or assemblages. I investigated the utility of Landsat TM and synthetic aperture radar data to estimate the structure and composition of dung beetle assemblages (Coleoptera: Scarabaeidae) within a region in eastern Bolivia. Using multiple linear regression, I developed response models of dung beetle biomass and estimated species richness, and I developed assemblage—level models using canonical correspondence analysis (CCA). Landsat TM Band 3 single-handedly explained 52% of the total variance in log—biomass/trap (R2 = 0.518), while 53% of the variance in ICE—predicted species richness was explained with a small number of variables. The CCA analysis explained 35% of the variance in the species counts with four remotely sensed variables. The results of this study suggest that detailed information on the ecological requirements of species may not always be necessary to estimate assemblage composition or structure. Instead, remotely sensed data may be utilised to construct generalised models of these biological phenomena.




11.45 - 12.00
MANNE, LISA L. and Paul H. Williams. Natural History Museum, Cromwell Road, London SW7 5BD, UK, <l.manne@nhm.ac.uk>.

THE BEST INDICATOR GROUPS MAY NOT BE TAXON-BASED

Indicator groups of species that predict the distribution of overall biodiversity have been sought for conservation planning for many years. Given constraints of time and money, consistently-performing indicator relationships would be a boon to conservation. As networks of sites encompassing complementary biotas represent much more overall diversity than sites chosen based on richness alone, the search for indicators of species richness has shifted to a search for consistent indicators of complementarity. Most of this work has been taxon-based, using birds to predict butterflies, etc. However, such indicator relationships have proven problematic, with resultant conservation networks sometimes not performing significantly better than choosing sites randomly. Are indicator groups a holy grail? In this paper, we hypothesise that there are quantifiable traits about single species that make these species desirable components of an indicator group: small range size, low overlap with other species' distributions, small body size, and species that are habitat specialists or show high ecological complementarity. We assess the utility of indicator groups chosen based on these traits, toward selecting networks for overall biodiversity in Europe, comparing to randomly chosen networks of areas, and to networks chosen using the more traditional taxon-based indicator groups.




12.00 - 12.15
HOWELL, CHRISTINE A. International Center for Tropical Ecology & Department of Biology, University of Missouri - Saint Louis, 8001 Natural Bridge Rd, Saint Louis, MO 63121, USA, <chowell@jinx.umsl.edu>.

MODELLING RARITY: IMPLICATIONS FOR RESERVE DESIGN

Species distribution modelling allows one to convert point locality data into the hypothetical range of a taxa; thus this methodology is increasingly being incorporated into the conservation reserve network design process. However, conservation planners are often most concerned with modelling distributions of rare species; yet these are the species with the least amount of information and the fewest locality data points. Moreover, species may be considered "rare" for many reasons including geographic extent, local population size, habitat specialization, anthropogenic effects, metapopulation dynamics, etc. Rabinowitz used combinations of these first three factors (geographic extent, population size, and specialization) to create a rarity classification scheme consisting of seven rarity forms. I used Rabinowitz's rarity classification scheme to designate South African plants into the seven rarity categories and then I modelled each species using a general additive modelling algorithm. The results show how well species distribution models predict distributions of species associated with different rarity categories. Results vary across categories, but species considered "sparse" were particularly difficult to model. Additionally I model the impacts of global climate change on species representative of each rarity category to determine if species with similar "rarity" characteristics respond similarly to climate change.

The SCB2002 pages are maintained by Christine Eagle
email: C.M.Eagle@ukc.ac.uk
Conference email: scb2002@ukc.ac.uk
Last updated: 30.06.02