Forgot your password?
Document Actions
Do conservation managers use scientific evidence to support their decision-making?
BIOLOGICAL
CONSERVATION
Biological Conservation 119 (2004) 245–252
www.elsevier.com/locate/biocon
Do conservation managers use scientific evidence to support
their decision-making?
a,*
, Teri M. Knight b, David A. Stone c, Kevin Charman c
Andrew S. Pullin
a
School of Biosciences, The University of Birmingham, Edgbaston, Birmingham B15 2TT, UK
b
Solihull Primary Care Trust, Mell House, Solihull B91 3BU, UK
c
English Nature, Northminster House, Peterborough PE1 1UA, UK
Received 14 June 2003; received in revised form 19 November 2003; accepted 20 November 2003
Abstract
Conservation involves making decisions on appropriate action from a wide range of options. For conservation to be effective,
decision-makers need to know what actions do and do not work. Ideally, decisions should be based on effectiveness as demonstrated
by scientific experiment or systematic review of evidence. Can decision-makers get this kind of information? We undertook a formal
assessment of the extent to which scientific evidence is being used in conservation practice by conducting a survey of management
plans and their compilers from major conservation organizations within the UK. Data collected suggest that the majority of
conservation actions remain experience-based and rely heavily on traditional land management practices because, many manage-
ment interventions remain unevaluated and, although some evidence exists, much is not readily accessible in a suitable form. We
argue that nature conservation along with other fields of applied ecology, should exploit the concept of evidence-based practice
developed and used in medicine and public health that aims to provide the best available evidence to the decision-maker(s) on the
likely outcomes of alternative actions. Through critical evaluation, we present the challenges and benefits of adopting evidence based
practice from the decision-makers point of view and identify the process to be followed to make it work.
Ó 2003 Elsevier Ltd. All rights reserved.
Keywords: Conservation practice; Conservation management; Conservation policy; Biodiversity management; Decision support
sound decisions, but in cases involving more funda-
1. Introduction
mental questions influencing the quality of the habitat,
Conservation action is increasing globally as the scale viability of a population or the functioning of an
of the threat to biodiversity is more widely recognized. ecosystem, how can the decision-maker compare the
Many organizations, both governmental and non-gov- effectiveness of possible alternative actions? For con-
ernmental, are currently reviewing policy and formu- servation practice to be effective decision-makers need to
lating conservation management strategies catalyzed by know what actions do and do not work, or how effective
the Convention on Biodiversity. Conservation manage- a given action has been in achieving objectives (Pullin
ment involves day-to-day decision making by a wide and Knight, 2001). Ideally, decisions should be based on
range of individuals from office-based policy formers to effectiveness of actions in achieving the objectives as
field-based site managers. All face decisions regarding demonstrated by scientific experiment. Can decision-
what actions they should take to achieve objectives and makers get this kind of information when they need it?
most will involve a level of uncertainty of outcome. In The volume of information on conservation practice
some cases the uncertainty may be minor and individual has increased enormously over the last 10–20 years. New
knowledge and experience may be good enough to make scientific peer-reviewed journals have appeared and rate
of paper publication increased. More practically based
journals and magazines focusing on conservation man-
*
Corresponding author. Tel.: +44-121-414-7147; fax: +44-121-414-
agement issues have also appeared and many in-house
5925.
magazines are in circulation to keep decision-makers
E-mail address: a.s.pullin@bham.ac.uk (A.S. Pullin).
0006-3207/$ - see front matter Ó 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biocon.2003.11.007
246 A.S. Pullin et al. / Biological Conservation 119 (2004) 245–252
and practitioners within larger organizations abreast of 2.1. Analysis of management plans
new developments. Added to this, the Internet revolu-
tion has made possible the rapid exchange of informa- Management plans were obtained from six organi-
tion that may have an influence on management policy. sations representing a mixture of statutory and non-
Is this all that decision-makers require? governmental conservation bodies. Plans were selected
Pullin and Knight (2001, 2003) argue that although on the basis of availability and the most recent were
good evidence for some actions does exist, indeed an selected in preference to older plans. Each was reviewed
increasing number of papers are providing scientific and analysed using a checklist for key issues related to
evidence to develop appropriate actions, in general, the gathering and use of information to support deci-
conservation actions lack thorough evaluation and are sion-making and the monitoring and evaluation of ac-
still based on anecdote, personal experience and inter- tions undertaken.
pretations of traditional land management practices.
They further hypothesized that this was not because 2.2. Questionnaire survey of management plan compilers
conservation bodies do not want to use evidence when it
is available, but because decision-makers do not have Questionnaires were sent to management plan com-
the time to access it nor a supporting framework that pilers from seven medium to large organisations (in-
provides the best quality information in a form they can cluding the six from which management plans were
readily absorb and use. obtained) representing both statutory and non-govern-
In this paper, we report on a test of the above hy- mental bodies. Most of the organisations sampled had
pothesis analysing how management plan compilers conservation as a primary objective, but for some it was
within the UK conservation community approach de- a secondary function. Compilers were asked to complete
cision making. The results provide evidence of the lack a questionnaire summarising their overall experience of
of an appropriate support system for decision-makers management plan compilation. Questions were asked
that would make scientific information easily accessible concerning their length of experience, use of informa-
in a usable form. To address this problem we propose tion, access to information, and evaluation of their
the adoption of an evidence-based framework adapted decision-making. The questionnaires were either dis-
from the fields of medicine, and more latterly, public tributed directly to the appropriate person in each or-
health and social sciences. Medicine and public health ganisation or given to a central co-ordinator within the
have recently gone through an Õeffectiveness revolutionÕ organisation for distribution.
in which the outcomes of actions have been evaluated
by experiment and decisions on future actions made on 2.3. Interviews with management plan compilers
the basis of scientific evidence of effectiveness (Coch-
rane, 1972; Stevens and Milne, 1997). The medical evi- After the questionnaire returns had been received and
dence-based framework provides the best available analysed, a subset of 20 respondents was followed up by
evidence to the decision-maker on the likely outcomes telephone interview. Further structured questions were
of alternative actions and enables decisions to be made asked to increase the understanding of earlier responses.
on the basis of evidence that has been critically evalu- The subset was selected on the basis of the responses
ated and widely disseminated in a format that is ac- being typical of the full range of replies.
cessible to policy makers and practitioners alike
(Dawes, 2000). 2.4. Data analysis
Data from the management plans and question-
naire responses were collated on a spreadsheet using
2. Methods
Microsoft Excel and subsequently analysed using SPSS
The extent to which scientific evidence is currently Version 11.
used in decision-making was investigated by examining
the way in which a selection of conservation organisa-
tions formulates Nature Reserve Management Plans. 3. Results and discussion
Our aim was to gain an overall impression of the range
of practice rather than to compare the practices of dif- 3.1. Management plan analysis
ferent organisations. Sources of information used by
management plan compilers to support decision-making A total of 38 management plans, written between
and information arising from their decisions was inves- 1996 and 2002, were analysed. Caution was exercised in
tigated in three ways; analysis of management plans, drawing information simply from reading the plan as
questionnaire returns from compilers, and interviews decisions may have been made and processes under-
with compilers. taken in its compilation that were not apparent in the
A.S. Pullin et al. / Biological Conservation 119 (2004) 245–252 247
Table 1
Percentage of management plans in which proposed actions were
justified by reference to the listed information sources
Information source type (%)
Primary scientific literature 11
Secondary reviews of literature 16
Habitat management handbooks 29
Biodiversity action plans 29
Accounts of traditional management 71
plan itself. However, information subsequently obtained
from compilers suggested that plans were a full reflec-
tion of the compilation process.
In 79% of cases, justification (usually by inclusion of Fig. 1. Questionnaire respondentÕs level of experience of compiling
a ÔRationaleÕ section) was given for the choice of action management plans.
and in 13% of cases it was unclear. In 8% justification
was not given. Where justification was given it was by
cluded from the analysis the return rate is 54%, ranging
reference to one or more of the sources given in Table 1.
from 30% up to 100% for a small NGO.
Of all practical management actions, 58% were direct
Almost half of the group had compiled or contrib-
continuations of traditional management and a further
uted to the compilation of between 3 and 10 manage-
26% were modifications of traditional management.
ment plans (Fig. 1), a quarter less than 3 and 10% had
This highlights a reliance on tradition as an indicator
written more than 20. When asked who was involved in
and guide to future management. In 66% of plans al-
the compilation of management plans, 29% responded
ternative actions did not appear to have been considered
that they always compiled their plans alone and a fur-
and in only 16% of plans were alternative actions dis-
ther 32% said that this was usually the case. This means
cussed. In only 8% of the plans was any attempt to re-
that approaching two thirds of the group are making
view the literature apparent and in no plan was it
decisions apparently without active input from col-
evident that the review had been extensive. In no plan
leagues. It is not known to what extent this is personal
was the quality of evidence in support of actions criti-
choice or organisational practice. In contrast 23% of the
cally evaluated. It therefore appears that most man-
respondents had never compiled plans on their own. The
agement plans are being compiled using a limited
latter reflects policy of some contributing organisations
amount of the total information available to support
to provide Ôin-houseÕ support to plan production. For
decision-making. The reasons for this were explored
28% of respondents it was the norm to use a working
through the questionnaire analysis.
group within their organisation to compile management
In 45% of plans, action to fill knowledge gaps was
plans. Only 11% normally used a working group derived
advocated, but in only 13% was this described in specific
from more than one organisation whilst 57% never did
terms, and only 5% of plans contained specific time
this (Fig. 2).
targets for this action. Methods to monitor the outcome
When asked about the extent to which they had to
were outlined in 53% of plans. In only 16% of plans was
decide between possible alternative actions in order to
it clear that monitoring was sufficient to evaluate effec-
achieve the objectives set in the plan, 67% said they
tiveness and outcome. These results suggest that com-
always or usually had to make these sorts of decisions;
pilers are well aware of gaps in evidence and the need to
4% said they never had to do so. It is clear that most
monitor outcomes from actions. However, the process
compilers have to actively consider a number of alter-
of addressing this lack of evidence was not formally
native actions.
included within the plans.
3.2.1. Information sources used in decision-making
3.2. Questionnaire of management plan compilers
Compilers were asked about the sources of informa-
tion used to support their decision-making. The most
A total of 141 returns were received from the seven
frequently used sources were existing management plans
organisations. Contributions ranged from 1 from the
(60% – the proportion that responded ‘‘always used’’ or
smallest NGO, to 44 from a major conservation body.
‘‘usually used’’), expert opinion from outside the compi-
Estimating the return rate was complicated by the fact
lation group (49%), published reviews, books or hand-
that one organisation was unable to provide an estimate
books (47%), and documentation or personal accounts of
of how many of their staff had actually compiled a
traditional management practices (46%) (Fig 3). Least
management plan for a nature reserve and were there-
frequently used sources of information were electronic/
fore potential respondents. If this organisation is ex-
248 A.S. Pullin et al. / Biological Conservation 119 (2004) 245–252
quently use secondary literature (published reviews,
books or handbooks), when it is available. Primary sci-
entific literature is infrequently accessed and 12% said
they never did so.
When a subgroup was asked at follow-up interview
why they did not access primary literature to help them
in their decision-making, the most frequent response
(65%) was that this literature is too time consuming to
locate and access. The majority (60%) also said this lit-
erature is too time consuming to read. A significant
number (25%) said primary literature is too technical
and difficult to interpret in the context of their decision-
making. Importantly, 25% stated that they rely on
Ôin-houseÕ advisors or expert groups to interpret infor-
mation from primary literature for them. This reflects
the fact that some conservation organisations have re-
cognised and tried to address the problem of informa-
tion retrieval and interpretation, although often not in a
systematic way.
3.2.2. Locating information sources
To locate published information only 8% of respon-
Fig. 2. The frequency with which management plans have been com-
dents routinely hand search library resources (Fig. 4)
piled by individuals (white bars), a working group within their orga-
and only 3% search library databases electronically.
nisation (grey bars), or a working group derived from more than one
organisation (black bars). Percentages are even lower for unpublished material.
The majority (72%) have never undertaken an electronic
search of a library database in connection with man-
agement plan compilation. Less than 1% routinely used
a web-search for publications and 76% have never done
so. Most respondents rely on literature recommended by
a colleague (42%) or use of their own or a colleagueÕs
personal collection (56%) to locate published material.
Figures are similar for unpublished material. Consider-
ing the time constraints on the respondent group, it is
Fig. 3. The frequency with which different sources of information are
accessed to support decision-making. Grey bars; ÔalwaysÕ or ÔusuallyÕ
used: black bars; ÔneverÕ used. EMPs ¼ existing management plans;
Pub Sci ¼ published scientific papers; Pub Pop ¼ published popular
articles; Pub Rev ¼ published reviews/books; Unpub ¼ unpublished
papers/reports; Web; web-based material; Exp Op ¼ Expert opinion
from outside compilation group; Trad Man; documentation or per-
sonal accounts of traditional management practices.
Fig. 4. A comparison of the frequency of use of different methods of
locating information to support management plan compilation. Grey
web-based materials (4%), published popular articles
bars; ÔalwaysÕ or ÔusuallyÕ used: black bars; ÔneverÕ used. Hand Lib ¼ -
(13%) and published scientific papers (23%). It appears
hand search of literature from library; Elect Lib ¼ electronic searching
that, in terms of written material, compilers rely heavily of library databases; Web ¼ web-based searching of publications da-
on current or traditional practices to guide them, together tabases; CollÕs Rec ¼ Literature recommendations from colleagues;
with ÔexpertÕ opinion. Interestingly, compilers do fre- Pers Coll ¼ use of personal collection (own or colleagues).
A.S. Pullin et al. / Biological Conservation 119 (2004) 245–252 249
not surprising that literature is not being systematically
sought out or reviewed, but it is of interest that they are
using literature that is immediately available to them.
3.2.3. Access to information
Of the respondents, 58% have easy access to a library,
either at work or elsewhere; 4% have no access. The
majority (65%) have easy access to the Internet, over
90% have some access, but 4% said they had access but
were not trained to use it. This suggests that the majority
could access electronic information if it was delivered in
a suitable form.
3.2.4. General experience of compilation process
Respondents were asked for the general experience of
Fig. 6. CompilersÕ perception of the amount of published material
finding relevant information to support decision-mak-
available to support their decision-making on a scale from 1 (none) to
ing. There was a relatively even spread between those 6 (enough).
who found it relatively quick and easy (22%) and those
who found it difficult and time consuming (16%), with
the majority expressing no strong preference (62%)
were 28% and 23%. A subset (18) who felt there was
(Fig. 5). A subset (17) who found the process quick and
enough published information were asked why. The
easy were selected for a follow-up interview to explore
majority (89%) thought this was partly or wholly be-
reasons for this. Nearly half (47%) replied that this was
cause they had enough information to remove all rea-
because they confined their search to material that was
sonable doubt about the course of action required to
immediately available within their organisation (cf. Ôlo-
meet their objectives. A third (33%) also felt that there
cating information sourcesÕ above). Interestingly, 24%
was enough in the sense that they would not have time
replied that they were already aware of all the infor-
to digest any more.
mation and 24% felt they had enough information after
relatively little search effort. Three (18%) benefited from
3.2.5. Relative inputs of experience and evidence to
having information provided by others and another
decision-making
three said the process was quick and easy because they
Respondents were asked to scale the relative inputs of
relied on their personal experience and that of their
Ôexperience-basedÕ information (e.g. qualitative descrip-
colleagues.
tion, expert opinion) versus Ôevidence-basedÕ information
When asked if there was enough published material
(experimental analysis and quantitative measurement).
to support their decision-making, 37% thought there
The majority (75%) thought that the greater input
was enough, 9% thought there was none or next to none
was from experience-based information (Fig. 7). Analy-
(Fig. 6). Equivalent figures for unpublished material
sed in a little more detail, 49% thought that experience-
based information was more influential, whilst only
5% thought evidence-based information was more
influential.
3.2.6. Responses to information deficit
When asked if they were able to identify knowledge
gaps in the compilation process where further research
was required, 63% responded that they were ÔalwaysÕ or
ÔusuallyÕ able to do so; 4% thought they were not able to
do so. Those that answered yes were then asked if they
were able to go further and describe the research re-
quired and 57% thought they were ÔalwaysÕ or ÔusuallyÕ
able to do so. A subset (20) of those able to identify
knowledge gaps were asked at interview whether this
extended to a description of the experimental design
required and 95% responded that it did not. A minority
Fig. 5. CompilersÕ general experience of finding relevant information to
(15%) said that they routinely asked experts from their
support decision-making on a scale from 1 (quick and easy) to 6
own organisation to design appropriate experiments.
(difficult and time-consuming).
250 A.S. Pullin et al. / Biological Conservation 119 (2004) 245–252
Fig. 7. CompilersÕ perception of the relative inputs of experience-based and evidence-based information to their decision-making on a scale from 1 (all
experience-based) to 6 (all evidence-based).
Those identifying knowledge gaps were asked in the up, but 35% were referred to in the next management
questionnaire if they had subsequently approached plan in some form.
anybody to get the research done and 36% had ÔalwaysÕ
or ÔusuallyÕ done so; 11% had ÔneverÕ done so. Those who
had approached other bodies were asked if any research 4. Providing decision support through an evidence-based
had subsequently been initiated. Over 80% said it had, framework
9% said it had not and 3% didnÕt know. A subset (18) of
those who said that research had been initiated were Our results suggest that management plan compilers
asked by interview what had been done with the results. are not making full or systematic use of the information
Half (50%) responded that research was ongoing and available to support their decision-making. Nor are they
results were awaited; 22% said some results had been fully monitoring and evaluating the effectiveness of ac-
published; whilst 72% responded that results had been tions and disseminating it for use by others. Realisti-
written up in a report but either not disseminated or cally, when faced with the day-to-day pressures of
only locally disseminated within the organisation or executing the actions, conservation decision makers
local area. Some respondents (22%) reported cases of (particularly those directly involved in practical man-
research being undertaken but not written up. agement) do not have sufficient time to access the pri-
mary information they need to judge effectiveness of
3.2.7. Monitoring and evaluation of actions alternative actions, let alone evaluate it. In such cases
Compilers were asked if the actions proposed in the they frequently rely on the status quo of continuing with
management plan had been implemented. Only 3% said an established but unevaluated practice. The hypothesis
all had, whilst 73% said most had. Asked if monitoring forwarded by Pullin and Knight (2001) is therefore lar-
programmes had been put in place to measure the out- gely supported.
come of implemented actions, 22% responded that this If we expect decision-makers, without the time to do
was ÔalwaysÕ done, a further 48% said this was ÔusuallyÕ their own information search, to be aware of evidence
done and 5% said it was ÔneverÕ done. Asked if they had relevant to their responsibilities and to apply that evi-
been able to evaluate the effectiveness of actions in their dence in seeking solutions to conservation problems,
management plans, 16% said all had, 36% said most had they require a framework with associated infrastructure
been evaluated, 37% said some had and 6% said none to support their decision-making. Pullin and Knight
had. A subset (20) of those who said they had done some (2001) drew attention to an existing framework for
evaluation of the effectiveness of their plans was asked supporting decision-making provided in the field of
how effectiveness was measured. Nearly half (45%) said medicine and public health. These disciplines have much
that evaluation was only qualitative and often experi- in common with conservation in that they are crisis
ence-based. A smaller number (35%) said they used disciplines that were established on experience of prac-
annual counts of species and species trends and 20% had titioners. In medicine it was recognized that even for
put in place direct monitoring of progress toward out- some of the commonest procedures there was little evi-
comes. Each was subsequently asked what has been dence for their effectiveness; choice of which treatment
done with the information arising from the evaluation to pursue, or surgical operation to perform, depended
and 25% said the information was written up as a report largely on the experience of the individual clinician. For
but in no case was this widely disseminated. The re- some medical interventions, research on effectiveness
maining 75% said evaluations were not formally written had been carried out but the results had little impact on
A.S. Pullin et al. / Biological Conservation 119 (2004) 245–252 251
practice. The challenge to develop Ôevidence-based work to increase the quality of evidence available (Pullin
practiceÕ was twofold: and Knight, 2003).
1. to ensure that the results of research impacted upon We do not contend that it will be an easy task to
practice; develop evidence-based practice in conservation. The
2. to increase good quality research into the effective- revolution will certainly be longer than the one experi-
ness of interventions. enced by medicine and it might possibly be a more
The concept of ‘‘evidence based medicine’’ has been vigorously fought one as conservation can be practiced
rapidly accepted and an industry supporting it has de- by anyone, anywhere and control over standards is more
veloped (Dawes, 2000). Fundamental to evidence-based lax. Some may view the apparent differences in the
practice is the systematic review in which research pa- professions as evidence that the same revolution cannot
pers selected on the basis of their relevance to the occur in conservation. Medicine is after all much less
question, are subjected to Ôcritical appraisalÕ using a complex than other social or ecological systems in only
standard protocol (Dawes, 2000; NHSCRD, 2001). This dealing with the human body. Although the impact of
covers the whole research process from the hypothesis to the evidence-based framework was initially experienced
be tested, study design, selection of subjects, data col- in medicine, the approach has quickly spread to public
lection and analysis. Studies that do not meet the re- health and the social sciences (Stevens et al., 2001). In
quired quality standard are either rejected or are further more complex systems such as these, evidence will al-
evaluated with their limitations in mind. Results from ways have to be interpreted and integrated within the
those remaining are summarized to enable common context of the systemÕs dynamics. This will certainly be
themes and messages to be drawn out. A specialized true of conservation as well, but we argue that the
form of this process is meta-analysis, increasingly com- paradigm shift in decision-making achieved through the
mon in the ecological literature (e.g. Bender et al., 1998; evidence-based approach in medicine and other fields is
Hartley and Hunter, 1998; Gates, 2002), the result of a template for significant improvement in conservation
which is a more powerful analysis than was possible practice (Pullin and Knight, 2003).
from the individual studies.
Systematic reviews are therefore not simply research
reviews of a chosen subject area, as published in many
Acknowledgements
ecological journals, but are reviews that result from sys-
tematic and explicit searches for evidence in the literature
The authors wish to acknowledge the financial sup-
that has a bearing on a specific question (Petticrew, 2001;
port of English Nature. The authors are grateful to the
Gates, 2002). In conservation, the amount of evidence
following organizations and their staff for their collab-
available is likely to be small and the quality of evidence
oration in this study: Butterfly Conservation, English
relatively low (although not always) compared to that in
Nature, UK Ministry of Defence, The National Trust,
medicine. In some cases, relatively few high quality ex-
Plantlife, Royal Society for the Protection of Birds, The
periments may be complemented by a larger number of
Wildlife Trusts.
lower quality data sets that provide valuable additional
evidence on the wider application of an action. In cases of
conflicting outcomes, the quality of the evidence is crucial
in the interpretation and conclusion on the effectiveness References
of the action (Stevens and Milne, 1997).
Two important points need to be made to avoid Bender, D.J., Contreras, T.A., Fahrig, L., 1998. Habitat loss and
population decline: a meta-analysis of the patch size effect. Ecology
misunderstanding of the evidence-based approach: 1.
79, 517–533.
The evidence-based model is a strategy to produce more
Cochrane, A., 1972. Effectiveness and Efficiency. Random Reflections
good quality evidence on which to base decisions and on the Health Service, Nuffield Provincial Hospitals Trust, Lon-
does not imply that decisions should not be taken if don.
good quality evidence is not available. 2. The fact that Dawes, M., 2000. Evidence Based Practice. Health Service Journal
Monographs No. 1. Emap Public Sector Management Publica-
we need a structured evidence-base to conservation
tions, London.
should not be used to undermine the credibility of efforts
Gates, S., 2002. Review of methodology of quantitative reviews using
currently being made to conserve species and habitats. It meta-analysis in ecology. Journal of Animal Ecology 71, 547–557.
does not mean that current actions are wrong. Hartley, M., Hunter Jr., M.L., 1998. A meta-analysis of forest cover,
Despite the anticipated difficulties in rising to a edge effects and predation rates of artificial nests. Conservation
Biology 12, 465–469.
standard of evidence-based conservation practice, the
National Health Service Centre for Reviews and Dissemination, 2001.
principle is established for care of our own species. We
Undertaking Systematic Reviews of Research on Effectiveness.
argue that effectiveness in conservation can improve by CRD Report 4, second ed. York Publishing Services, York.
working to that principle, both with the evidence cur- Petticrew, M., 2001. Systematic reviews from astronomy to zoology:
rently available to us and by putting in place a frame- myths and misconceptions. British Medical Journal 322, 98–101.
252 A.S. Pullin et al. / Biological Conservation 119 (2004) 245–252
Pullin, A.S., Knight, T.M., 2001. Effectiveness in conservation Stevens, A., Abrams, K., Brazier, J., Fitzpatrick, R., Lilford, R. (Eds.),
practice: pointers from medicine and public health. Conservation 2001. The Advanced Handbook of Methods in Evidence-Based
Biology 15, 50–54. Healthcare. Sage Press, London.
Pullin, A.S., Knight, T.M., 2003. Support for decision making in Stevens, A., Milne, R., 1997. The effectiveness revolution and public
conservation practice: an evidence-based approach. Journal for health. In: Scally, G. (Ed.), Progress in Public Health. Royal
Nature Conservation 11, 83–90. Society of Medicine Press, London, pp. 197–225.
CONSERVATION
Biological Conservation 119 (2004) 245–252
www.elsevier.com/locate/biocon
Do conservation managers use scientific evidence to support
their decision-making?
a,*
, Teri M. Knight b, David A. Stone c, Kevin Charman c
Andrew S. Pullin
a
School of Biosciences, The University of Birmingham, Edgbaston, Birmingham B15 2TT, UK
b
Solihull Primary Care Trust, Mell House, Solihull B91 3BU, UK
c
English Nature, Northminster House, Peterborough PE1 1UA, UK
Received 14 June 2003; received in revised form 19 November 2003; accepted 20 November 2003
Abstract
Conservation involves making decisions on appropriate action from a wide range of options. For conservation to be effective,
decision-makers need to know what actions do and do not work. Ideally, decisions should be based on effectiveness as demonstrated
by scientific experiment or systematic review of evidence. Can decision-makers get this kind of information? We undertook a formal
assessment of the extent to which scientific evidence is being used in conservation practice by conducting a survey of management
plans and their compilers from major conservation organizations within the UK. Data collected suggest that the majority of
conservation actions remain experience-based and rely heavily on traditional land management practices because, many manage-
ment interventions remain unevaluated and, although some evidence exists, much is not readily accessible in a suitable form. We
argue that nature conservation along with other fields of applied ecology, should exploit the concept of evidence-based practice
developed and used in medicine and public health that aims to provide the best available evidence to the decision-maker(s) on the
likely outcomes of alternative actions. Through critical evaluation, we present the challenges and benefits of adopting evidence based
practice from the decision-makers point of view and identify the process to be followed to make it work.
Ó 2003 Elsevier Ltd. All rights reserved.
Keywords: Conservation practice; Conservation management; Conservation policy; Biodiversity management; Decision support
sound decisions, but in cases involving more funda-
1. Introduction
mental questions influencing the quality of the habitat,
Conservation action is increasing globally as the scale viability of a population or the functioning of an
of the threat to biodiversity is more widely recognized. ecosystem, how can the decision-maker compare the
Many organizations, both governmental and non-gov- effectiveness of possible alternative actions? For con-
ernmental, are currently reviewing policy and formu- servation practice to be effective decision-makers need to
lating conservation management strategies catalyzed by know what actions do and do not work, or how effective
the Convention on Biodiversity. Conservation manage- a given action has been in achieving objectives (Pullin
ment involves day-to-day decision making by a wide and Knight, 2001). Ideally, decisions should be based on
range of individuals from office-based policy formers to effectiveness of actions in achieving the objectives as
field-based site managers. All face decisions regarding demonstrated by scientific experiment. Can decision-
what actions they should take to achieve objectives and makers get this kind of information when they need it?
most will involve a level of uncertainty of outcome. In The volume of information on conservation practice
some cases the uncertainty may be minor and individual has increased enormously over the last 10–20 years. New
knowledge and experience may be good enough to make scientific peer-reviewed journals have appeared and rate
of paper publication increased. More practically based
journals and magazines focusing on conservation man-
*
Corresponding author. Tel.: +44-121-414-7147; fax: +44-121-414-
agement issues have also appeared and many in-house
5925.
magazines are in circulation to keep decision-makers
E-mail address: a.s.pullin@bham.ac.uk (A.S. Pullin).
0006-3207/$ - see front matter Ó 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biocon.2003.11.007
246 A.S. Pullin et al. / Biological Conservation 119 (2004) 245–252
and practitioners within larger organizations abreast of 2.1. Analysis of management plans
new developments. Added to this, the Internet revolu-
tion has made possible the rapid exchange of informa- Management plans were obtained from six organi-
tion that may have an influence on management policy. sations representing a mixture of statutory and non-
Is this all that decision-makers require? governmental conservation bodies. Plans were selected
Pullin and Knight (2001, 2003) argue that although on the basis of availability and the most recent were
good evidence for some actions does exist, indeed an selected in preference to older plans. Each was reviewed
increasing number of papers are providing scientific and analysed using a checklist for key issues related to
evidence to develop appropriate actions, in general, the gathering and use of information to support deci-
conservation actions lack thorough evaluation and are sion-making and the monitoring and evaluation of ac-
still based on anecdote, personal experience and inter- tions undertaken.
pretations of traditional land management practices.
They further hypothesized that this was not because 2.2. Questionnaire survey of management plan compilers
conservation bodies do not want to use evidence when it
is available, but because decision-makers do not have Questionnaires were sent to management plan com-
the time to access it nor a supporting framework that pilers from seven medium to large organisations (in-
provides the best quality information in a form they can cluding the six from which management plans were
readily absorb and use. obtained) representing both statutory and non-govern-
In this paper, we report on a test of the above hy- mental bodies. Most of the organisations sampled had
pothesis analysing how management plan compilers conservation as a primary objective, but for some it was
within the UK conservation community approach de- a secondary function. Compilers were asked to complete
cision making. The results provide evidence of the lack a questionnaire summarising their overall experience of
of an appropriate support system for decision-makers management plan compilation. Questions were asked
that would make scientific information easily accessible concerning their length of experience, use of informa-
in a usable form. To address this problem we propose tion, access to information, and evaluation of their
the adoption of an evidence-based framework adapted decision-making. The questionnaires were either dis-
from the fields of medicine, and more latterly, public tributed directly to the appropriate person in each or-
health and social sciences. Medicine and public health ganisation or given to a central co-ordinator within the
have recently gone through an Õeffectiveness revolutionÕ organisation for distribution.
in which the outcomes of actions have been evaluated
by experiment and decisions on future actions made on 2.3. Interviews with management plan compilers
the basis of scientific evidence of effectiveness (Coch-
rane, 1972; Stevens and Milne, 1997). The medical evi- After the questionnaire returns had been received and
dence-based framework provides the best available analysed, a subset of 20 respondents was followed up by
evidence to the decision-maker on the likely outcomes telephone interview. Further structured questions were
of alternative actions and enables decisions to be made asked to increase the understanding of earlier responses.
on the basis of evidence that has been critically evalu- The subset was selected on the basis of the responses
ated and widely disseminated in a format that is ac- being typical of the full range of replies.
cessible to policy makers and practitioners alike
(Dawes, 2000). 2.4. Data analysis
Data from the management plans and question-
naire responses were collated on a spreadsheet using
2. Methods
Microsoft Excel and subsequently analysed using SPSS
The extent to which scientific evidence is currently Version 11.
used in decision-making was investigated by examining
the way in which a selection of conservation organisa-
tions formulates Nature Reserve Management Plans. 3. Results and discussion
Our aim was to gain an overall impression of the range
of practice rather than to compare the practices of dif- 3.1. Management plan analysis
ferent organisations. Sources of information used by
management plan compilers to support decision-making A total of 38 management plans, written between
and information arising from their decisions was inves- 1996 and 2002, were analysed. Caution was exercised in
tigated in three ways; analysis of management plans, drawing information simply from reading the plan as
questionnaire returns from compilers, and interviews decisions may have been made and processes under-
with compilers. taken in its compilation that were not apparent in the
A.S. Pullin et al. / Biological Conservation 119 (2004) 245–252 247
Table 1
Percentage of management plans in which proposed actions were
justified by reference to the listed information sources
Information source type (%)
Primary scientific literature 11
Secondary reviews of literature 16
Habitat management handbooks 29
Biodiversity action plans 29
Accounts of traditional management 71
plan itself. However, information subsequently obtained
from compilers suggested that plans were a full reflec-
tion of the compilation process.
In 79% of cases, justification (usually by inclusion of Fig. 1. Questionnaire respondentÕs level of experience of compiling
a ÔRationaleÕ section) was given for the choice of action management plans.
and in 13% of cases it was unclear. In 8% justification
was not given. Where justification was given it was by
cluded from the analysis the return rate is 54%, ranging
reference to one or more of the sources given in Table 1.
from 30% up to 100% for a small NGO.
Of all practical management actions, 58% were direct
Almost half of the group had compiled or contrib-
continuations of traditional management and a further
uted to the compilation of between 3 and 10 manage-
26% were modifications of traditional management.
ment plans (Fig. 1), a quarter less than 3 and 10% had
This highlights a reliance on tradition as an indicator
written more than 20. When asked who was involved in
and guide to future management. In 66% of plans al-
the compilation of management plans, 29% responded
ternative actions did not appear to have been considered
that they always compiled their plans alone and a fur-
and in only 16% of plans were alternative actions dis-
ther 32% said that this was usually the case. This means
cussed. In only 8% of the plans was any attempt to re-
that approaching two thirds of the group are making
view the literature apparent and in no plan was it
decisions apparently without active input from col-
evident that the review had been extensive. In no plan
leagues. It is not known to what extent this is personal
was the quality of evidence in support of actions criti-
choice or organisational practice. In contrast 23% of the
cally evaluated. It therefore appears that most man-
respondents had never compiled plans on their own. The
agement plans are being compiled using a limited
latter reflects policy of some contributing organisations
amount of the total information available to support
to provide Ôin-houseÕ support to plan production. For
decision-making. The reasons for this were explored
28% of respondents it was the norm to use a working
through the questionnaire analysis.
group within their organisation to compile management
In 45% of plans, action to fill knowledge gaps was
plans. Only 11% normally used a working group derived
advocated, but in only 13% was this described in specific
from more than one organisation whilst 57% never did
terms, and only 5% of plans contained specific time
this (Fig. 2).
targets for this action. Methods to monitor the outcome
When asked about the extent to which they had to
were outlined in 53% of plans. In only 16% of plans was
decide between possible alternative actions in order to
it clear that monitoring was sufficient to evaluate effec-
achieve the objectives set in the plan, 67% said they
tiveness and outcome. These results suggest that com-
always or usually had to make these sorts of decisions;
pilers are well aware of gaps in evidence and the need to
4% said they never had to do so. It is clear that most
monitor outcomes from actions. However, the process
compilers have to actively consider a number of alter-
of addressing this lack of evidence was not formally
native actions.
included within the plans.
3.2.1. Information sources used in decision-making
3.2. Questionnaire of management plan compilers
Compilers were asked about the sources of informa-
tion used to support their decision-making. The most
A total of 141 returns were received from the seven
frequently used sources were existing management plans
organisations. Contributions ranged from 1 from the
(60% – the proportion that responded ‘‘always used’’ or
smallest NGO, to 44 from a major conservation body.
‘‘usually used’’), expert opinion from outside the compi-
Estimating the return rate was complicated by the fact
lation group (49%), published reviews, books or hand-
that one organisation was unable to provide an estimate
books (47%), and documentation or personal accounts of
of how many of their staff had actually compiled a
traditional management practices (46%) (Fig 3). Least
management plan for a nature reserve and were there-
frequently used sources of information were electronic/
fore potential respondents. If this organisation is ex-
248 A.S. Pullin et al. / Biological Conservation 119 (2004) 245–252
quently use secondary literature (published reviews,
books or handbooks), when it is available. Primary sci-
entific literature is infrequently accessed and 12% said
they never did so.
When a subgroup was asked at follow-up interview
why they did not access primary literature to help them
in their decision-making, the most frequent response
(65%) was that this literature is too time consuming to
locate and access. The majority (60%) also said this lit-
erature is too time consuming to read. A significant
number (25%) said primary literature is too technical
and difficult to interpret in the context of their decision-
making. Importantly, 25% stated that they rely on
Ôin-houseÕ advisors or expert groups to interpret infor-
mation from primary literature for them. This reflects
the fact that some conservation organisations have re-
cognised and tried to address the problem of informa-
tion retrieval and interpretation, although often not in a
systematic way.
3.2.2. Locating information sources
To locate published information only 8% of respon-
Fig. 2. The frequency with which management plans have been com-
dents routinely hand search library resources (Fig. 4)
piled by individuals (white bars), a working group within their orga-
and only 3% search library databases electronically.
nisation (grey bars), or a working group derived from more than one
organisation (black bars). Percentages are even lower for unpublished material.
The majority (72%) have never undertaken an electronic
search of a library database in connection with man-
agement plan compilation. Less than 1% routinely used
a web-search for publications and 76% have never done
so. Most respondents rely on literature recommended by
a colleague (42%) or use of their own or a colleagueÕs
personal collection (56%) to locate published material.
Figures are similar for unpublished material. Consider-
ing the time constraints on the respondent group, it is
Fig. 3. The frequency with which different sources of information are
accessed to support decision-making. Grey bars; ÔalwaysÕ or ÔusuallyÕ
used: black bars; ÔneverÕ used. EMPs ¼ existing management plans;
Pub Sci ¼ published scientific papers; Pub Pop ¼ published popular
articles; Pub Rev ¼ published reviews/books; Unpub ¼ unpublished
papers/reports; Web; web-based material; Exp Op ¼ Expert opinion
from outside compilation group; Trad Man; documentation or per-
sonal accounts of traditional management practices.
Fig. 4. A comparison of the frequency of use of different methods of
locating information to support management plan compilation. Grey
web-based materials (4%), published popular articles
bars; ÔalwaysÕ or ÔusuallyÕ used: black bars; ÔneverÕ used. Hand Lib ¼ -
(13%) and published scientific papers (23%). It appears
hand search of literature from library; Elect Lib ¼ electronic searching
that, in terms of written material, compilers rely heavily of library databases; Web ¼ web-based searching of publications da-
on current or traditional practices to guide them, together tabases; CollÕs Rec ¼ Literature recommendations from colleagues;
with ÔexpertÕ opinion. Interestingly, compilers do fre- Pers Coll ¼ use of personal collection (own or colleagues).
A.S. Pullin et al. / Biological Conservation 119 (2004) 245–252 249
not surprising that literature is not being systematically
sought out or reviewed, but it is of interest that they are
using literature that is immediately available to them.
3.2.3. Access to information
Of the respondents, 58% have easy access to a library,
either at work or elsewhere; 4% have no access. The
majority (65%) have easy access to the Internet, over
90% have some access, but 4% said they had access but
were not trained to use it. This suggests that the majority
could access electronic information if it was delivered in
a suitable form.
3.2.4. General experience of compilation process
Respondents were asked for the general experience of
Fig. 6. CompilersÕ perception of the amount of published material
finding relevant information to support decision-mak-
available to support their decision-making on a scale from 1 (none) to
ing. There was a relatively even spread between those 6 (enough).
who found it relatively quick and easy (22%) and those
who found it difficult and time consuming (16%), with
the majority expressing no strong preference (62%)
were 28% and 23%. A subset (18) who felt there was
(Fig. 5). A subset (17) who found the process quick and
enough published information were asked why. The
easy were selected for a follow-up interview to explore
majority (89%) thought this was partly or wholly be-
reasons for this. Nearly half (47%) replied that this was
cause they had enough information to remove all rea-
because they confined their search to material that was
sonable doubt about the course of action required to
immediately available within their organisation (cf. Ôlo-
meet their objectives. A third (33%) also felt that there
cating information sourcesÕ above). Interestingly, 24%
was enough in the sense that they would not have time
replied that they were already aware of all the infor-
to digest any more.
mation and 24% felt they had enough information after
relatively little search effort. Three (18%) benefited from
3.2.5. Relative inputs of experience and evidence to
having information provided by others and another
decision-making
three said the process was quick and easy because they
Respondents were asked to scale the relative inputs of
relied on their personal experience and that of their
Ôexperience-basedÕ information (e.g. qualitative descrip-
colleagues.
tion, expert opinion) versus Ôevidence-basedÕ information
When asked if there was enough published material
(experimental analysis and quantitative measurement).
to support their decision-making, 37% thought there
The majority (75%) thought that the greater input
was enough, 9% thought there was none or next to none
was from experience-based information (Fig. 7). Analy-
(Fig. 6). Equivalent figures for unpublished material
sed in a little more detail, 49% thought that experience-
based information was more influential, whilst only
5% thought evidence-based information was more
influential.
3.2.6. Responses to information deficit
When asked if they were able to identify knowledge
gaps in the compilation process where further research
was required, 63% responded that they were ÔalwaysÕ or
ÔusuallyÕ able to do so; 4% thought they were not able to
do so. Those that answered yes were then asked if they
were able to go further and describe the research re-
quired and 57% thought they were ÔalwaysÕ or ÔusuallyÕ
able to do so. A subset (20) of those able to identify
knowledge gaps were asked at interview whether this
extended to a description of the experimental design
required and 95% responded that it did not. A minority
Fig. 5. CompilersÕ general experience of finding relevant information to
(15%) said that they routinely asked experts from their
support decision-making on a scale from 1 (quick and easy) to 6
own organisation to design appropriate experiments.
(difficult and time-consuming).
250 A.S. Pullin et al. / Biological Conservation 119 (2004) 245–252
Fig. 7. CompilersÕ perception of the relative inputs of experience-based and evidence-based information to their decision-making on a scale from 1 (all
experience-based) to 6 (all evidence-based).
Those identifying knowledge gaps were asked in the up, but 35% were referred to in the next management
questionnaire if they had subsequently approached plan in some form.
anybody to get the research done and 36% had ÔalwaysÕ
or ÔusuallyÕ done so; 11% had ÔneverÕ done so. Those who
had approached other bodies were asked if any research 4. Providing decision support through an evidence-based
had subsequently been initiated. Over 80% said it had, framework
9% said it had not and 3% didnÕt know. A subset (18) of
those who said that research had been initiated were Our results suggest that management plan compilers
asked by interview what had been done with the results. are not making full or systematic use of the information
Half (50%) responded that research was ongoing and available to support their decision-making. Nor are they
results were awaited; 22% said some results had been fully monitoring and evaluating the effectiveness of ac-
published; whilst 72% responded that results had been tions and disseminating it for use by others. Realisti-
written up in a report but either not disseminated or cally, when faced with the day-to-day pressures of
only locally disseminated within the organisation or executing the actions, conservation decision makers
local area. Some respondents (22%) reported cases of (particularly those directly involved in practical man-
research being undertaken but not written up. agement) do not have sufficient time to access the pri-
mary information they need to judge effectiveness of
3.2.7. Monitoring and evaluation of actions alternative actions, let alone evaluate it. In such cases
Compilers were asked if the actions proposed in the they frequently rely on the status quo of continuing with
management plan had been implemented. Only 3% said an established but unevaluated practice. The hypothesis
all had, whilst 73% said most had. Asked if monitoring forwarded by Pullin and Knight (2001) is therefore lar-
programmes had been put in place to measure the out- gely supported.
come of implemented actions, 22% responded that this If we expect decision-makers, without the time to do
was ÔalwaysÕ done, a further 48% said this was ÔusuallyÕ their own information search, to be aware of evidence
done and 5% said it was ÔneverÕ done. Asked if they had relevant to their responsibilities and to apply that evi-
been able to evaluate the effectiveness of actions in their dence in seeking solutions to conservation problems,
management plans, 16% said all had, 36% said most had they require a framework with associated infrastructure
been evaluated, 37% said some had and 6% said none to support their decision-making. Pullin and Knight
had. A subset (20) of those who said they had done some (2001) drew attention to an existing framework for
evaluation of the effectiveness of their plans was asked supporting decision-making provided in the field of
how effectiveness was measured. Nearly half (45%) said medicine and public health. These disciplines have much
that evaluation was only qualitative and often experi- in common with conservation in that they are crisis
ence-based. A smaller number (35%) said they used disciplines that were established on experience of prac-
annual counts of species and species trends and 20% had titioners. In medicine it was recognized that even for
put in place direct monitoring of progress toward out- some of the commonest procedures there was little evi-
comes. Each was subsequently asked what has been dence for their effectiveness; choice of which treatment
done with the information arising from the evaluation to pursue, or surgical operation to perform, depended
and 25% said the information was written up as a report largely on the experience of the individual clinician. For
but in no case was this widely disseminated. The re- some medical interventions, research on effectiveness
maining 75% said evaluations were not formally written had been carried out but the results had little impact on
A.S. Pullin et al. / Biological Conservation 119 (2004) 245–252 251
practice. The challenge to develop Ôevidence-based work to increase the quality of evidence available (Pullin
practiceÕ was twofold: and Knight, 2003).
1. to ensure that the results of research impacted upon We do not contend that it will be an easy task to
practice; develop evidence-based practice in conservation. The
2. to increase good quality research into the effective- revolution will certainly be longer than the one experi-
ness of interventions. enced by medicine and it might possibly be a more
The concept of ‘‘evidence based medicine’’ has been vigorously fought one as conservation can be practiced
rapidly accepted and an industry supporting it has de- by anyone, anywhere and control over standards is more
veloped (Dawes, 2000). Fundamental to evidence-based lax. Some may view the apparent differences in the
practice is the systematic review in which research pa- professions as evidence that the same revolution cannot
pers selected on the basis of their relevance to the occur in conservation. Medicine is after all much less
question, are subjected to Ôcritical appraisalÕ using a complex than other social or ecological systems in only
standard protocol (Dawes, 2000; NHSCRD, 2001). This dealing with the human body. Although the impact of
covers the whole research process from the hypothesis to the evidence-based framework was initially experienced
be tested, study design, selection of subjects, data col- in medicine, the approach has quickly spread to public
lection and analysis. Studies that do not meet the re- health and the social sciences (Stevens et al., 2001). In
quired quality standard are either rejected or are further more complex systems such as these, evidence will al-
evaluated with their limitations in mind. Results from ways have to be interpreted and integrated within the
those remaining are summarized to enable common context of the systemÕs dynamics. This will certainly be
themes and messages to be drawn out. A specialized true of conservation as well, but we argue that the
form of this process is meta-analysis, increasingly com- paradigm shift in decision-making achieved through the
mon in the ecological literature (e.g. Bender et al., 1998; evidence-based approach in medicine and other fields is
Hartley and Hunter, 1998; Gates, 2002), the result of a template for significant improvement in conservation
which is a more powerful analysis than was possible practice (Pullin and Knight, 2003).
from the individual studies.
Systematic reviews are therefore not simply research
reviews of a chosen subject area, as published in many
Acknowledgements
ecological journals, but are reviews that result from sys-
tematic and explicit searches for evidence in the literature
The authors wish to acknowledge the financial sup-
that has a bearing on a specific question (Petticrew, 2001;
port of English Nature. The authors are grateful to the
Gates, 2002). In conservation, the amount of evidence
following organizations and their staff for their collab-
available is likely to be small and the quality of evidence
oration in this study: Butterfly Conservation, English
relatively low (although not always) compared to that in
Nature, UK Ministry of Defence, The National Trust,
medicine. In some cases, relatively few high quality ex-
Plantlife, Royal Society for the Protection of Birds, The
periments may be complemented by a larger number of
Wildlife Trusts.
lower quality data sets that provide valuable additional
evidence on the wider application of an action. In cases of
conflicting outcomes, the quality of the evidence is crucial
in the interpretation and conclusion on the effectiveness References
of the action (Stevens and Milne, 1997).
Two important points need to be made to avoid Bender, D.J., Contreras, T.A., Fahrig, L., 1998. Habitat loss and
population decline: a meta-analysis of the patch size effect. Ecology
misunderstanding of the evidence-based approach: 1.
79, 517–533.
The evidence-based model is a strategy to produce more
Cochrane, A., 1972. Effectiveness and Efficiency. Random Reflections
good quality evidence on which to base decisions and on the Health Service, Nuffield Provincial Hospitals Trust, Lon-
does not imply that decisions should not be taken if don.
good quality evidence is not available. 2. The fact that Dawes, M., 2000. Evidence Based Practice. Health Service Journal
Monographs No. 1. Emap Public Sector Management Publica-
we need a structured evidence-base to conservation
tions, London.
should not be used to undermine the credibility of efforts
Gates, S., 2002. Review of methodology of quantitative reviews using
currently being made to conserve species and habitats. It meta-analysis in ecology. Journal of Animal Ecology 71, 547–557.
does not mean that current actions are wrong. Hartley, M., Hunter Jr., M.L., 1998. A meta-analysis of forest cover,
Despite the anticipated difficulties in rising to a edge effects and predation rates of artificial nests. Conservation
Biology 12, 465–469.
standard of evidence-based conservation practice, the
National Health Service Centre for Reviews and Dissemination, 2001.
principle is established for care of our own species. We
Undertaking Systematic Reviews of Research on Effectiveness.
argue that effectiveness in conservation can improve by CRD Report 4, second ed. York Publishing Services, York.
working to that principle, both with the evidence cur- Petticrew, M., 2001. Systematic reviews from astronomy to zoology:
rently available to us and by putting in place a frame- myths and misconceptions. British Medical Journal 322, 98–101.
252 A.S. Pullin et al. / Biological Conservation 119 (2004) 245–252
Pullin, A.S., Knight, T.M., 2001. Effectiveness in conservation Stevens, A., Abrams, K., Brazier, J., Fitzpatrick, R., Lilford, R. (Eds.),
practice: pointers from medicine and public health. Conservation 2001. The Advanced Handbook of Methods in Evidence-Based
Biology 15, 50–54. Healthcare. Sage Press, London.
Pullin, A.S., Knight, T.M., 2003. Support for decision making in Stevens, A., Milne, R., 1997. The effectiveness revolution and public
conservation practice: an evidence-based approach. Journal for health. In: Scally, G. (Ed.), Progress in Public Health. Royal
Nature Conservation 11, 83–90. Society of Medicine Press, London, pp. 197–225.