Amateur Ecological Monitors: the benefits and challenges are of using volunteers. Reprinted from the
Bulletin of the British Ecological Society.
Dr. Christina D. Buesching & Dr. Chris Newman
Wildlife Conservation Research Unit, Oxford University
"People will only conserve and protect things that they care about. And they will only care about the things they understand".
In April 2000 WildCRU implemented a portfolio "The Mammal Monitoring Project" to bring together its research programmes at Oxford University's SSSI "Wytham Woods" research site under a single umbrella, for example long-term work on the ecology of badgers, deer, mice and voles.
But the Project's long-term objectives extend beyond Wytham's 1000 acres. Equally important is the development of easy-to-use, yet accurate, mammal monitoring techniques, which could potentially be used to monitor mammals nationally.
In 1998 WildCRU produced a scoping report "Proposals for the future monitoring of British mammals" for the central government, highlighting the need to establish robust methods for quantifying the distribution and abundance of mammals in the UK. Many of this report's suggestions were implemented, via the government's Joint Nature Conservation Committee, in April 2004, with the launch of the national "Tracking Mammals Partnership", of which WildCRU is a member.
A key observation of the report was that it would not be possible to monitor mammals across the entire country using only professional ecologists; it would also be necessary to tap into the enthusiasm of the nation's many willing volunteers.
Drawing together these themes, WildCRU has been collaborating with the Earthwatch Institute, who recruit volunteers to assist scientists, and has fielded some 40 teams over the past 5 years, comprised by over 500 volunteers from all walks of life, representing a huge variety of occupations and nationalities.
Whereas many surveys are carried out by observing and identifying species directly, mammal surveys usually rely on the correct identification of the species field-signs. Surveying for field-signs generally requires more skill and training, and is also more easily prone to misidentification; therefore, most mammal surveys require some form of basic training or written instruction making it important to identify methods that are both ecologically useful and simultaneously interesting and rewarding for the volunteers.
The Volunteer Data Validation Study (VDVP)
Teams of 8 to 12 volunteers, supplied by Earthwatch Institute, assist the Project for six to 10 days, eight to 10 weeks each year. Whereas some of our volunteers had considerable background knowledge and conservation experience, others were embarking on a new interest.
For a subset of 155 volunteers, we set out to "monitor the monitors" and devised a system to evaluate volunteer performance on a 0-5 subjective ability scale (assuming a professional biologist would score 5/5) for a variety of tasks:
- assembly/ setting/ checking of Longworth traps;
- handling of small mammals;
- experimental deer dropping survey/ identification;
- badger censusing at setts, badger field-sign survey; and
- winter transect surveys.
One male and one female supervising scientist (CN & CDB) assessed volunteers and scores were averaged for each task. Scores also included the ability to understand the principles of the task, to execute the task correctly and efficiently, and to work reliably without supervision. Additionally, a subjective score was given for attention to information, fitness, and mental aptitude (defined as enthusiasm for the work). Though all volunteers were informed beforehand of the general principles of the assessment study, they were not told any details of the scoring criteria until the conclusion of the team visit. Quantitative measures were also scored where possible, such as ability to survey for deer droppings in artificially stocked 10m2 quadrats.
All "per task" scores were then combined to create overall means, which indicated the researchers' perception of the volunteers' usefulness. A second measure, defined as the volunteer ability score, included the assessment of each volunteer's physical and mental aptitude.
Training was provided primarily ‘hands-on' in the field, building on introductory lectures.
The Findings
In the UK, most species monitoring programmes on local, regional and national scales depend heavily on the involvement of amateurs. Yet many scientists still doubt the accuracy and reliability of field data collected by volunteers and detailed scientific studies evaluating volunteer data are rare (see Newman, Buesching & Macdonald, 2003).
We used volunteers to monitor three very different types of mammal: badgers, small rodents and deer. In these particular circumstances and with the particular variety of volunteers we tested, our analyses showed an average ability score for volunteers of 2.94 out of 5.0, s.d = 0.75), amounting to an efficiency of 59.6% compared to a professional researcher. In addition, the wide variability in volunteer-performance (min. = 23.2%, max. = 84%, average = 58.8%) was significantly moderated by gender, fitness and previous experience.
Influence of training
As most volunteer-based studies utilise people from a variety of social and educational backgrounds and with varying physical ability, it is important to understand what effects these factors have on volunteer performance, both from the standpoint of data accuracy and effective volunteer training and deployment. The necessity for basic field-training is recognised by many volunteer-based monitoring programmes, and wildlife societies often offer species specific training days to their members (e.g. Mammal Training Workshops of the Mammal Society, Training Days of the Wildlife Trusts). In our study, practical field-training and demonstrations proved to be essential for all the monitoring techniques used. For example, volunteers provided with only written instructions were unable to find/identify badger latrines. Generally, we found that after a half-day of basic training per focal species, including background theory, practical demonstration and initial close supervision, volunteers produced results which were within the range recorded by people using the same methods professionally.
Influence of volunteer characteristics
Team management and training had a significant effect on volunteer motivation and influenced data quality. Importantly, we found that varying attitudes towards different tasks appeared to be a crucial factor influencing performance. In this particular sample of volunteers, women were often more hesitant than men, and sometimes had to be ‘persuaded' to perform certain tasks (e.g. handle mice), thus, overall, men appeared to make more confident volunteers. However, with encouragement and additional tuition this hesitancy was usually overcome. By contrast, people from less privileged backgrounds, as exemplified in our sample by those from a drug rehabilitation programme, showed far less hesitation to undertake tasks; indeed, they were generally amongst our highest scoring volunteers.
Our volunteers varied widely in their physical abilities. On the whole, people at the limit of their physical fitness did not perform as well as volunteers who found the work physically easy. Once controlled for fitness, age had no significant influence on performance.
Previous experience, i.e., having prior involvement with practical conservation work, was not always a positive advantage. While experienced volunteers proved more able to trap and handle small mammals (both in terms of per capita efficiency and handling ability), they demonstrated a lower success rate at finding deer droppings in the experimental quadrats, not least due to being significantly quicker. Thus we hypothesise that their experience resulted in over-confidence as well as a lower threshold of boredom, with the result that more droppings were missed. Background knowledge of the natural history did, however, appear to give the more experienced volunteers a keener interest in instructive and explanatory sessions.
Benefits of using volunteers
While it is essential to calibrate the accuracy of volunteers this does not diminish their value. They allow numerous sites to be surveyed simultaneously and they can save time e.g. deer dropping survey and small mammal trapping. In this context, our training regime (alongside the training days of other wildlife societies) represents an economy of scale. Many volunteers can be trained simultaneously, representing a time- and cost-effective method for ultimately increasing the number competent ecological monitors in society. This is especially significant if the trained volunteers can make a long-term commitment to a project.
A pragmatic reason for using volunteers is that they are increasingly willing to pay to be involved in environmental science, unlocking new revenue for conservation. Additionally, the involvement of volunteers from the wider public in conservation-based research also increases the public's environmental awareness and understanding of scientific issues.
References
Macdonald, D.W., Mace, G. & Rushton, S. (Eds.) (1998). Proposals for future monitoring of British mammals. HMSO, London.
Newman, C. Buesching, C.D. & Macdonald, D.W. (2003) Validating mammal monitoring methods and assessing the performance of volunteers in wildlife conservation - "Sed quis custodet ipsos custoides". Biological Conservation, 113: 189-197
For further information on this project and national mammal monitoring initiatives please see:
Mammals of Nova Scotia
Wildlife Conservation Research Unit