home

Identifying the Gaps, Locating the Reserves: Some Thoughts on Getting 
Gap Analysis into Conservation Practice

Introduction

GAP seeks two objectives in mapping and categorizing the stew-
ardship of the U.S.: 1) to map the conservation network as we de-
fine it, and 2) to report the representation of elements of biodiversity
(species and plant communities) or lack thereof in the conservation
network. The purpose is to characterize who the conservation stew-
ards are, where their lands are, and the character of those lands in
terms of size and intent to maintain biodiversity. The latter involves
categorizing tracts of land according to the well-known four cat-
egories of biodiversity maintenance (Crist 2000) to assess the rela-
tive protection of, or risk to, the biotic elements found in or pre-
dicted to inhabit them. The result is a report, for each of the ele-
ments mapped, on the amount and percentage of their distribution
found in each ownership type and protection category.

This information has proven very useful in assessing general pat-
terns of protection afforded this country’s biodiversity. For example,
when the occurrence of nature reserves is mapped against elevation
and soil productivity, the underrepresentation of low elevations and
productive soils is very apparent (Scott et al. in press). Similarly,
early results from GAP projects primarily in the West indicate that
50-85% of the mapped land cover types have less than 10% of their
area in nature reserves. While 10% is a widely accepted standard
of representation, it is an arbitrary standard. Protection of the full
range of biodiversity may require 50-75% of the land base to in-
clude evolutionary potential of species at one end of the hierarchy
to ecosystem processes at the other.

Without trivializing the monumental mapping and modeling effort
to create the standard GAP information, when compared to GAP’s
mission, providing this information is only the initial phase of con-
servation planning. In the pioneering GAP monograph (Scott et al.
1993), 15 steps of the process were listed, including “13. Identify
minimum areas required for . . . species and vegetation types, 14.
Identify landscape corridors connecting vegetation types and cen-
ters of species richness,” and “15. Design [the] reserve network.”
Clearly some of these tasks are now outside the responsibilities of
an information program, but the U.S. lacks a government program
to undertake such tasks, and the land management agencies typi-
cally do not feel mandated to do so. Thus it appears the need still
exists to identify the geography of conservation need, not just tabu-
lar results (see, for example, Grue et al. and Stoms et al. below).
From here I describe the context of GAP’s origins to understand
how we reached the current situation and where we may be or should
be headed.

Past

GAP began in the U.S. Fish and Wildlife Service (FWS) to gener-
ally serve its land management and planning information needs.
Specifically GAP sought to help avoid endangered species crises
by aiding the FWS in the identification of elements requiring height-
ened scrutiny and ultimately to identify potential reserve locations
for underrepresented elements of biodiversity. Work in the early
and mid ‘90s by a variety of researchers suggested that this goal
was no easy feat (for review see Csuti and Kiester 1996, Kiester et
al. 1996, Merrill et al. 1995, Wright et al. 1994, also Stoms et al.
below). GAP’s early recommendation for using species-rich
hotspots was found to be a limited surrogate for biodiversity in gen-
eral, and the computational requirements for identifying represen-
tative locations were enormous. Improvements in technology and
technique combined with new concepts of “complementarity” for
representing all elements (biotic and nonbiotic) emerged as more
satisfying and robust answers (Pressey et al. 1997), but GAP has
yet to incorporate these concepts into its standard analyses. The
reasons for delay in pursuing more robust analyses range from the
perennial lack of funding to support the work at either the state or
national program level to the lack of demand for such information
from management agencies that continue reactive management
under legal duress (D. Osborne, pers. comm.).

Present

By the end of 2000, GAP will have 37 completed state projects
with data sets describing and quantifying the predicted “gap” status
of nearly all terrestrial vertebrates and plant communities for a large
portion of the nation (74% of the area of the 48 conterminous states).
This is a tremendous technical, logistical, and scientific accomplish-
ment, but to assess our success in increasing conservation it is illus-
trative to look at the FWS’s application of GAP. Despite legisla-
tion in 1997 that charges the FWS with guiding the future of its
refuges towards the “conservation of ecosystems” (U.S. 1997) the
FWS, GAP’s original home and client, has not made full use of the
GAP data. However, a few successful examples exist. The results
from the early GAP effort in Hawaii (see Brackney and Crist, this
issue) were used to create two new national wildlife refuges, and
Indiana-GAP used the data to help design the Grand Kankakee
Marsh National Wildlife Refuge (NWR; Clark and Slusher 2000).
GAP information could significantly aid the nationwide FWS pro-
cesses for identifying potential acquisition projects or the process
for prioritizing them for funding. We also developed a pilot deci-
sion support system for refuge planning, “Refuge-GAP”
(Herdendorf and Crist 1998), that spurred interest but has been slow
to receive any FWS support for further research and development.

Perhaps the most significant block to using the results of gap analy-
ses is that few agencies feel compelled to react to “gaps” that are
not legally recognized by the Endangered Species Act (ESA) as
threatened or endangered. GAP’s analyses may be viewed with
limited interest because we do not currently incorporate perceived
short-term “real” risk in the sense of reduced abundance and clear
and imminent threats such as logging of Red-cockaded Woodpecker
habitat (see Vilella and Minnis, this issue). All of this reflects the
fact that GAP did not emerge from national legislation as, for ex-
ample, the Superfund Project did to address a widely perceived
national crisis, but has instead been built from the bottom up to
address the need as seen by biologists and researchers.

National GAP is rapidly developing GIS capability to both region-
alize the state data sets and undertake basic gap analysis at the re-
gional level and hopefully to engage other researchers in more ro-
bust analyses. At present, though, we need to be cognizant of how
well data are being received and used by other agencies. Even so-
phisticated GIS users have had substantial difficulty conceptualiz-
ing how to integrate GAP’s novel data into decision-making pro-
cesses that have accrued over decades, using either very crude bio-
logical information or none at all. Simply publishing the informa-
tion has not demonstrated that it will be used, and certainly not for
the purposes for which it was created. In 1996 Crist began collabo-
ration with the University of Wyoming GAP team on the develop-
ment of decision support systems (DSS), including Refuge-GAP
described above. The intent of these systems is to provide integra-
tion of GAP data in a format friendly to users’ decision-making
processes and not requiring either biological or GIS expertise. DSS
has come of age, at least conceptually and in recognition of need,
in the arena of natural resources technology but is not yet recog-
nized as a normal funding item in most agencies.

Future

In the immediate future, GAP intends to publish a great deal of
information and publicize it well. An important part of that process
is publication of results by GAP researchers in the peer-reviewed
literature. These are only first steps of putting the information to
use, but they are important to show progress for the investment made.
The next practical step will be development of other data products
such as regional and element rangewide analyses (for examples see
Stoms et al. [1998] and current work in Idaho to analyze gaps in the
land cover of the 11 western states). Later work will identify
complementarity-based reserve areas as well as development of more
sophisticated DSSs for more user groups.

All of these activities will simply attempt to put current informa-
tion to better use. However, it is also time to “negotiate with” as
well as educate our partners. Educating them involves aiding the
integration of GAP results into decision making by explaining the
currently little-recognized need to consider common species and
rangewide gap status. Negotiation will recognize that political re-
alities call for us to identify those more urgent conservation issues
that will continue to take precedence. Clearly our “one-size-fits-
all” gap analysis requires change. Our process assumes an equal
reaction of all elements to a stewardship condition, while we are
well aware that some species thrive in agriculture and urbanization,
and some species perish in anything less than total protection.
Schemes for addressing individual element sensitivity have been
suggested (Crist et al. 2000; Redford and Richter 1999; see also
Gon et al., this issue) but such an approach may push the limits of
our knowledge of species requirements and would require an en-
tirely new level of partnership with primary land stewards to popu-
late the necessary databases. We do believe the process is tractable,
however, and probably necessary to provide truly useful informa-
tion for the continuing “crisis planning” conducted in the U.S.

Conclusion

Recently GAP has funded socioeconomic seed grants to aid us in
1) identifying types of products and analyses needed by society,
and 2) incorporating socioeconomic factors in gap analyses to ad-
dress the “short-term” risk issues. GAP information has already
been used for innumerable applications across the nation, and those
applications will increase exponentially as we publish the gigabytes
of information pouring into the national program. Of this we should
all be proud but also contemplative about why our information has
yet to be used (much) for its primary purpose. Clearly institutional
and societal education and changes are required as they are for all
aspects of biodiversity conservation, but the greater (and more man-
ageable) burden lies with GAP to conduct the type of analyses and
create the products and tools that will better serve societal needs.