Conservation tools – back to basics

A rough guide to fauna surveys, distribution models, and PVAs

By Brendan Wintle (This article was first published in the February 2013 issue of Decision Point, The Monthly Magazine of the Environmental Decisions Group)

Imagine you’re a forest manager responsible for balancing (among other things) timber production and threatened species management in a large forest estate. You’ve been given the task of figuring out the potential impacts of a timber harvesting plan on a nationally listed threatened bandicoot. Or perhaps you’re more familiar with the problem of having to evaluate the impacts of a proposed urban expansion on nationally listed threatened species. Or if you’re really lucky, maybe you’re trying to figure out where and how you should restore native vegetation to achieve the best conservation outcomes to improve threatened species habitat and species’ viability.

Maybe these aren’t your issues at this moment but, as you read this, someone somewhere in Australia is grappling with problems just like these. They are big challenges, but they are not intractable. They all involve determining which species occur in areas likely to be impacted or improved by management, and working out what is likely to happen to these species over time. Solving these challenges involves the use of expert knowledge, the application of existing field survey data (supplemented with new field surveys as needed) usually combined with other environmental information within some sort of species distribution model.

Determining how strongly a species will be affected by a proposed management action might involve an analysis of the number of known occurrences or area of predicted suitable habitat that would be lost, gained or altered under a given action. In some instances it is appropriate to extend these analyses to predict the long-term population-level consequences of losing individuals and habitats using some form of population viability analysis model. The activities of (1) field survey data collection and analysis, (2) distribution and habitat modelling, and (3) population viability analysis are among the most basic and important tools in the conservation biologist’s tool kit.

Over the past 30 years, conservation scientists have invested a great deal of theoretical and empirical research effort developing, testing and refining these tools. Reflecting on the ancient history of conservation research, EDG researchers such as David Lindenmayer, Richard Hobbs, Jane Elith, Hugh Possingham, Mark Burgman, and Mick McCarthy have been world leaders in the development of tools, methods, and ways of thinking about the questions posed in the box (see box 1 below). Their strong track record of research into these applied questions is evident in many of our current NERP-ED and ARC-CEED research projects.

But there’s a BIG problem. Hardly any of the managers and practitioners, who we imagine to be routinely using these tools to answer fundamental biodiversity conservation questions, are actually using them. While lots of field surveys are conducted in the course of impact assessment and monitoring, the design of those surveys often falls well short of what might be recommended by texts such as Lindenmayer and Likens (2010). For example, when’s the last time you saw a replicated, before-after-control-impact monitoring of urban development impacts that took into account the estimated detection rates for a species of interest (see the stories by Georgia Garrard and Mick McCarthy in the Februrary 2013 Decision Point and on this blog for pointers on what needs to be considered)?

“When’s the last time you saw a replicated, before-after-control-impact monitoring of urban development impacts that took into account the estimated detection rates for a species of interest?”

The same goes for modelling habitats and species’ viability. When was the last time you saw a population viability analysis (PVA) model used to predict the long-term impacts of a coal mine on the viability of a nationally listed species? It almost never happens, despite the prevalence of published papers on the topic and freely available software for implementation.

A recent EDG workshop on the use of species distribution models (SDMs) in conservation decision making discovered that, despite the MASSIVE academic literature about species distribution modelling and the pervasive view among scientists that SDMs are one of the most important conservation science tools, surprisingly few people are actually using them in decision making. There’s an upcoming paper on this by ARC-CEED researcher Antoine Guisan and colleagues.

So, why are these tools not being used as often or as well as we would like? Limited time? Limited budgets? Limited number of trained professionals…? Maybe. Or maybe our research and our tools are completely irrelevant to real world management? This last possibility, however, is at odds with my own experience which is that if you are prepared to get in and help with the survey design, monitoring design, distribution modelling or viability analysis for an agency or an NGO, they’re very happy to get help, utilize the products, and pay for the service. But that’s not really the role of a researcher and therefore not really a solution to the problem in the long term.

Box 1 Five basic questions

Questions that arise when trying to evaluate the potential impacts or conservation benefits of a proposed action on threatened species revolve around the following:

  1. Which species are most likely to be affected by a proposed action?
  2. Of the species you expect to be affected, are there known occurrences of individuals or core habitats in the areas likely to be impacted or restored?
  3. What is the expected loss of threatened species’ individuals or habitat that would arise from an action going ahead?
  4. On the flip side, what gains in habitat quality and extent are expected to arise from a restoration investment?
  5. Given the expected losses and gains of individuals and/ or suitable habitat, what is the net effect on the long-term viability of local populations or species as a whole?

Answers to each of these questions involves understanding which species occur in areas likely to be impacted or improved by management, and how they are likely to respond to proposed management actions. Field surveys, distribution modelling and population viability analysis are among of the most basic and important tools involved in obtaining this understanding.

Optimistically, I like to think that if we could provide advice or guidance to managers and conservation practitioners on the use of key conservation tools in a digestible and intuitive way (i.e., not in textbooks or published academic papers) that uptake and use of tools such as species distribution modelling and population viability analysis might increase. My hope is that as managers become more knowledgeable about what’s involved in utilizing these tools properly, they may be more likely to utilize these tools themselves or more confident to employ someone to do so, confident that they know what they’re buying into.

Towards this end I have prepared three ‘rough guides’ on the basic tools of conservation decision making and we’ll be bringing them to you in the first three issues of Decision Point this year. The first is on how to set up a fauna survey, the second is on species distribution modelling, and the third is on population viability analysis. Each is in the form of a checklist of good practice.

These checklists could be anywhere between 150 and 450 pages shorter than your average textbook on these topics, so there will, of course, be big gaps. I apologise right here to everybody I insult by leaving out the element they consider most important or the topic of their most recent, excellent academic publication. However, I will maintain these checklists on my website and will add or subtract things over time (if I can be convinced that the changes are worthwhile). So, please write to me if you believe I have left out or stuffed up something fundamental that is key to good practice.

The first checklist deals with some basic aspects of fauna survey design. I have separated fauna and flora survey issues because despite there being many similarities in terms of underlying theory, there are a sufficient number of differences that I felt they warranted separate checklists. I have also chosen to focus specifically on the design and implementation of fauna surveys for understanding the distribution and/or abundance of fauna, NOT for monitoring changes in those properties. Again, monitoring is a large topic that deserves its own checklist.

So, I hope you find the checklists useful and I’d be interested in any feedback. Don’t expect them to be ‘laugh-out-loud’ entertaining, indeed it’s more the opposite, but keep in mind the stakes involved. If we can’t get the basics properly applied then nobody should have confidence that we can protect our threatened biodiversity.

This entry was posted in Conservation, Decision Point, Papers and tagged , , , . Bookmark the permalink.

4 Responses to Conservation tools – back to basics

  1. Pingback: Principles of population viability analysis (PVA): a checklist of the basics | Quantitative & Applied Ecology Group

  2. Pingback: Developing and interpreting species distribution models: a checklist of the basics | Quantitative & Applied Ecology Group

  3. Thanks Ian, this is a very important point you raise and you’re right, these are not easy questions to answer. I suspect that most managers (and many researchers) couldn’t answer these questions without digging into a text book, some papers and/or getting some help from a specialist. However, the questions included in the checklist reflect little more than the assumptions that underlie the use of fauna survey methods in impact assessment (and other management and scientific activities). I would argue that I’m not asking anybody to walk an extra mile, just to report on basic information required to assess the quality or sufficiency of a survey. By asking these questions, I’m asking the managers/consultants to be explicit about what they’re assuming when they report the results of a biological survey. To put a little more bluntly, if the these sorts of questions cannot be answered, then there is little reason to trust the results of an impact assessment survey: “how likely is it that I would detect the species during my survey if it’s present?”, “if I don’t detect the species, does that mean it’s probably not there?”, “how many surveys should I do to be 90% sure of detecting the species if it’s present?”….
    So, am I asking too much? My logic is that by laying out these issues in a short document, I’m at least providing a starting point for managers who want to assess the quality of the survey reports that cross their desk, but may not know where to start. I would even stick my neck out and say that if consultants convincingly answered each of the checklist questions when reporting the results of biological surveys, then a manager may have reason to feel confident about the results, and can then use them as a valid input to decision making (with at least some sense of the relevant uncertainty inherent in the data). So in summary, you are right that the checklists don’t bridge the science-management divide by providing the answers. However, by bounding the range of questions that need to be answered in order to feel confident about survey results, I hope that they make some contribution.

    Thanks heaps for reading and commenting on the checklist (and sorry I took so long to respond!)

    Cheers, Brendan.

  4. Ian Lunt says:

    Hi Brendan, thanks for a great blog and a great initiative. I don’t want to wear a ‘black hat’ but I’m worried about the gap between the skills that many (most?) managers possess and the amount of analytical knowledge that your questions assumes managers possess, or can easily acquire. For instance, the fauna survey info page asks many questions which I would imagine many managers have little potential to properly consider, let alone answer, given the degree of analytical knowledge they require (see sample of questions below). Do you have any thoughts on how this chasm can be bridged? Best wishes Ian

    Does the sampling design factor in the requirements and assumptions of the analytical methods to be employed? Is the survey effort sufficient to be confident in the findings of the survey? Is the choice of analytical method justified with respect to the data being analysed (and based on references to advice in the published literature)?  Have species detection probabilities been obtained from the literature or estimated from the data? Have the assumptions about the detectability of the species that are being used when designing the sampling strategy been made explicit? Has the probability that the species would have been detected in a single unit of survey effort if present been explicitly stated? Has the a priori probability of occupancy, based on knowledge about the quality of the habitat and the proximity to other known populations, been explicitly stated?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s