Monitoring changes in species distribution and abundance is central to the sustainable management of natural resources and the definition of conservat
Lead Research Organisation:
University of Oxford
Department Name: Zoology
Abstract
Monitoring changes in species distribution and abundance is central to the sustainable management of natural resources
and the definition of conservation goals. However for many rare, widespread and cryptic species, monitoring over large
spatiotemporal scales remains a serious challenge. A growing and increasingly complex set of analytical tools exists that
can be used to estimate the abundance, density or occupancy of species across landscapes. The technical literature on
these methods emphasises the absolute necessity of controlling statistically for variation in detectability. Without this, real
trends in abundance cannot be separated from confounding trends in detectability, potentially leading to mistaken
conclusions. Robustly distinguishing variations in abundance from detectability requires that specific assumptions are
satisfied, which in turn requires rigorous design and application of field methods. In practice, this usually means that
surveys are designed and executed by professional ecologists, and have very substantial labour costs, particularly when
executed over large scales.
In response to this cost constraint, another strand of the monitoring literature emphasises practicality. While recognising the
desirability of robust, unbiased monitoring methods, these studies point out such approaches are pointless if they are too
expensive to be applied over sufficiently large spatial and temporal scales to provide the information needed for
management. This is particularly true in developing countries, where funding is usually limited and intermittent, and
management priorities often demand information over large areas of remote, inaccessible terrain.
Tension thus exists between those advocating monitoring methods that are, above all, accurate, and those who rather emphasise long-term cost effectiveness. While monitoring methods can be both accurate and affordable in some
circumstances, the tradeoff between cost and accuracy is nonetheless a pervasive and unresolved problem in biodiversity
monitoring. One approach to more sustainable monitoring has been involving local people. The benefits of this can be twofold:
reducing labour costs, and facilitating the translation of monitoring results into effective management action. However,
it remains important to structure information gathering to maximise accuracy. While there are some good examples of the
integration of local participation into ecological monitoring, it remains underdeveloped, and many questions remain about
its effective application.
This study will develop a new method for monitoring widespread animal populations in developing countries; interviewbased
occupancy analysis. We aim to: test the approach with a range of species at different levels of density and spatial
distribution; assess the sources of variation in detectability, both biophysical and observer-based; quantify trade-offs
between cost, precision and accuracy for different species and observer types (e.g. rangers vs pastoralist vs NTFP
collectors); make recommendations for future implementation of the approach.
The study will be carried out in partnership with a ZSL project in Benin, which aims to strengthen conservation of the WArly-
Pendjari conservation complex, globally recognised as a conservation priority. That project aims to support Protected
Area management and protection, improve understanding of local community use of natural resources and explore options
for integrating community centred resource management into land use planning. Specifically this PhD research will help
the project meet key objectives regarding wildlife population monitoring and engaging local communities in management.
More broadly this research will have widescale applicability by advancing methodologies appropriate for efficient monitoring
of wildlife populations at a large spatial scale under limited time and financial resources.
and the definition of conservation goals. However for many rare, widespread and cryptic species, monitoring over large
spatiotemporal scales remains a serious challenge. A growing and increasingly complex set of analytical tools exists that
can be used to estimate the abundance, density or occupancy of species across landscapes. The technical literature on
these methods emphasises the absolute necessity of controlling statistically for variation in detectability. Without this, real
trends in abundance cannot be separated from confounding trends in detectability, potentially leading to mistaken
conclusions. Robustly distinguishing variations in abundance from detectability requires that specific assumptions are
satisfied, which in turn requires rigorous design and application of field methods. In practice, this usually means that
surveys are designed and executed by professional ecologists, and have very substantial labour costs, particularly when
executed over large scales.
In response to this cost constraint, another strand of the monitoring literature emphasises practicality. While recognising the
desirability of robust, unbiased monitoring methods, these studies point out such approaches are pointless if they are too
expensive to be applied over sufficiently large spatial and temporal scales to provide the information needed for
management. This is particularly true in developing countries, where funding is usually limited and intermittent, and
management priorities often demand information over large areas of remote, inaccessible terrain.
Tension thus exists between those advocating monitoring methods that are, above all, accurate, and those who rather emphasise long-term cost effectiveness. While monitoring methods can be both accurate and affordable in some
circumstances, the tradeoff between cost and accuracy is nonetheless a pervasive and unresolved problem in biodiversity
monitoring. One approach to more sustainable monitoring has been involving local people. The benefits of this can be twofold:
reducing labour costs, and facilitating the translation of monitoring results into effective management action. However,
it remains important to structure information gathering to maximise accuracy. While there are some good examples of the
integration of local participation into ecological monitoring, it remains underdeveloped, and many questions remain about
its effective application.
This study will develop a new method for monitoring widespread animal populations in developing countries; interviewbased
occupancy analysis. We aim to: test the approach with a range of species at different levels of density and spatial
distribution; assess the sources of variation in detectability, both biophysical and observer-based; quantify trade-offs
between cost, precision and accuracy for different species and observer types (e.g. rangers vs pastoralist vs NTFP
collectors); make recommendations for future implementation of the approach.
The study will be carried out in partnership with a ZSL project in Benin, which aims to strengthen conservation of the WArly-
Pendjari conservation complex, globally recognised as a conservation priority. That project aims to support Protected
Area management and protection, improve understanding of local community use of natural resources and explore options
for integrating community centred resource management into land use planning. Specifically this PhD research will help
the project meet key objectives regarding wildlife population monitoring and engaging local communities in management.
More broadly this research will have widescale applicability by advancing methodologies appropriate for efficient monitoring
of wildlife populations at a large spatial scale under limited time and financial resources.
Publications
Brittain S
(2022)
Comparing interview methods with camera trap data to inform occupancy models of hunted mammals in forest habitats
in Conservation Science and Practice
Brittain S
(2020)
Ethical considerations when conservation research involves people.
in Conservation biology : the journal of the Society for Conservation Biology
Brittain S
(2022)
Power to the people: Analysis of occupancy models informed by local knowledge
in Conservation Science and Practice
Brittain S
(2022)
Using scenarios-based interviews to predict changes in wild meat hunting and consumption in response to different alternatives projects
in Conservation Science and Practice
Ibbett H
(2020)
Conservation publications and their provisions to protect research participants.
in Conservation biology : the journal of the Society for Conservation Biology
Whytock R
(2021)
Robust ecological analysis of camera trap data labelled by a machine learning model
in Methods in Ecology and Evolution
| Description | 1. We assessed the effects of hunter and village level variables on the spatial and temporal trends in bushmeat hunting at two contrasting study sites within the same ecosystem. We find that socioecological system (SES) frameworks are a useful tool to piece together evidence to help us understand hunting sustainability, given how challenging monitoring sustainability can be in forest environments. Hunters in both villages use hunting for food and livelihoods. However, both study sites present different sustainability challenges due to differences in the way that hunter, village and external pressures interact. The use of an SES framework allowed me to investigate the influence of variables at different scales. Future studies on bushmeat hunting will also benefit from using such an approach, which allows for a more holistic view of sustainability than biological measures alone, and allows for recommendations on how the system in question could be brought back in line with sustainability if required. 2. Experts are commonly called upon to make judgements and estimations in conservation based on their prior experience (Burgman et al 2011; Martin et al 2012). This is especially useful in the face of rapidly changing systems, remote locations and often restricted budgets which make collecting robust and up to date empirical data for conservation purposes at times highly challenging. This study provides the first known application of modern structured expert knowledge elicitation techniques outside of a conventional expert context. An important conclusion from this research is that pooled opinion, while not always capturing the 'truth', did often perform better than the top experts, who tended to overestimate more than the pooled estimate. This finding supports the 'wisdom of the crowd' phenomenon (Suroeiki 2004) and provides additional evidence that gathering judgements from groups results in more consistent estimates than relying on estimates from a handful of self and peer identified experts (Hemming et al 2018; van der Hoeven 2004; Armstrong, 2001). 3. Despite the growing popularity of combining interview data with occupancy analysis, there is a surprising lack of studies that assess local knowledge combined with occupancy analysis in forest environments or for monitoring species commonly hunted for wild meat. The motivation for this study therefore stems from this lack of focus in the literature, and the need for comparison studies that focus on uncertainty. Here, I evaluated the efficacy and robustness of occupancy models informed by observational data from local people, for a range of rare and commonly hunted mammal species in two contrasting community forests in Cameroon. Using a mixed-method approach, I triangulate estimates of detectability and occupancy from three different methods to assess their precision and accuracy at different scales and provide guidance on future use of these methods for monitoring both threatened and hunted mammals in forest habitats. I found that, where camera trap estimates are available, estimates of occupancy informed by both hunter diaries and interview data are broadly comparable to estimates derived from camera traps at the village scale, although with species level differences. Further, while estimates were often comparable at the village level, evidence for comparable estimates at the site level was less certain. The results revealed species level differences in the effectiveness of different monitoring methods as well as differences between methods in the significant variables that account for occupancy and detection. The inclusion of observer-based and environmental variables in the top occupancy models highlights the importance of accounting for variation in both occupancy and detection As such, I find that occupancy analysis should be used to account for biases in detection and occupancy in all methods (Van Strien et al. 2013).While this small-scale study permitted me to ensure spatial and temporal matching, a lack of robust estimates from the camera traps limited my ability to provide more extensive species comparisons. A greater camera trapping effort in terms of both the number of sites and the amount of time allocated may be required in future comparative research, to increase the detection of endangered and cryptic species that may avoid community forest areas. A larger survey area may also help to capture a wider range of environmental covariates and improve the site-level predictions of occupancy. 4. Finally, I wanted to explore the power of occupancy models to detect change under different survey designs and budgetary scenarios, informed by three monitoring approaches. I assessed the power of camera trap data, seasonal interviews and daily diaries collected from local people to detect change in occupancy for 14 different mammal species, either commonly hunted for wild meat or of conservation interest. I found that age, gender, the number of trips to the forest and the number of nights spent sleeping in the forest were all important detection variables included in the top models for the diary data, while the number of nights slept in the forest was insignificant in the top models for the interview data. Under the monitoring plan used in this study, all methods had sufficient power to detect 50% change for between 58-78% of the species captured by that monitoring method. Yet, where occupancy or detection was already very low, 80% power to detect change was rarely achieved, despite increasing survey effort. I found that a similar survey effort is needed to detect 80% and 50% changes in populations, but smaller levels of change are only detectable above a certain occupancy threshold, which again differs with method and species. Conservation practitioners need to weigh up the costs and benefits of attempting to detect small proportional changes for their target species, given the high cost and effort which may be required (Southwell et al 2018). The results identified species-level differences in the power to detect change across the different methods. It's interesting to note that, for primates and pangolins, diary data had the greatest power to detect change. Primates and pangolins are heavily hunted in this region. Since diary respondents were active gun or snare hunters, these results may reflect more reliable knowledge of the animals that participants of the hunter diaries are actively targeting (Martinez- Marti et al 2016). It may also be that hunters completing the diaries are actively visiting sites where giant pangolins are more likely to be found, therefore increasing their chances of detecting them. I also found significant differences in the financial investment required for monitoring with each method, regardless of the scenario implemented. While camera traps performed well for abundant ungulates and rodents, the cost to implement effective monitoring to detect 50% growths or declines in occupancy was prohibitive for all species except blue duiker and porcupine. These results reflect the findings of several studies which have now found monitoring that incorporates local knowledge to be a highly cost-effective option (Danielsen et al. 2010; Turvey et al. 2014; Parry & Peres, 2015), especially useful where data is lacking, or in challenging habitats such as forest environments (Turvey et al 2015; Martinez-Marti et al. 2016). |
| Exploitation Route | 1. For other researchers looking to assess the sustainability of busuhmeat hunting, I hope that this research will contribute to the body of work that shows how researchers must consider ecological and social data, as well as biological, in order to truely understand the drivers of change and assess the resilience of a system in the face of rapid change. This work will be written as a paper, and the findings will be presented at conferences over the next few years. 2. This research is the first to apply modern structured elicitation methods to help respondents to engage with uncertainty and provide the most robust estimates possible. I hope that the approach presented in this work, once written into a paper, will be used by researchers engaging with local people for conservation, to improve the robustness of data obtained from local experts. |
| Sectors | Environment,Other |
| Description | Paper on ethics has resulted in an improvement in the way some high-impact journals assess the ethical impact that conservation research can have on participants and researchers |
| First Year Of Impact | 2019 |
| Sector | Environment,Government, Democracy and Justice |
| Impact Types | Cultural |
| Description | Lessons learnt from my research will be integrated into a new wild meat monitoring training programme run by ZSL Cameroon. |
| Geographic Reach | Africa |
| Policy Influence Type | Influenced training of practitioners or researchers |
| Description | Jana Robeyst Small Grant |
| Amount | € 1,500 (EUR) |
| Organisation | Jana Robeyst Trust Fund |
| Sector | Charity/Non Profit |
| Country | Belgium |
| Start | 05/2017 |
| End | 05/2018 |
| Description | Rufford small grant |
| Amount | £5,000 (GBP) |
| Funding ID | 23269-1 |
| Organisation | Rufford Small Grants Foundation |
| Sector | Charity/Non Profit |
| Country | United Kingdom |
| Start | 09/2017 |
| End | 09/2018 |
| Title | Interview Based Occupancy Analysis |
| Description | Combining interview data with occupancy analysis is a method growing in popularity for the assessment of wildlife status, distribution and relative abundance. |
| Type Of Material | Improvements to research infrastructure |
| Year Produced | 2015 |
| Provided To Others? | No |
| Impact | Combining interview data with occupancy analysis is a method growing in popularity for the assessment of wildlife status, distribution and relative abundance. However, there are no studies that attempt to assess the robustness of this method and the ways in which the method can be best implemented for reliable results. My current research is contributing to this knowledge. As soon as my research is completed, the results will be published to help guide other researchers using this method to achieve the most reliable results possible. |
| Title | Monitoring the Illegal Killing of Elephants |
| Description | The findings of my published paper on forest elephant status, distribution and relative abundance in Cameroon is being incorporated in the the MIKE IUCN database. |
| Type Of Material | Database/Collection of data |
| Year Produced | 2016 |
| Provided To Others? | Yes |
| Impact | People accessing data on the IUCN MIKE database will have access to my data on forest elephant distribution, threats and status from 2008-2013 across a large forested area in Cameroon. This data can be used to guide targeted forest elephant conservation action, and better understand the threats posed to forest elephants in the region |
| Description | Supervision and CASE industry partner |
| Organisation | Zoological Society of London |
| Department | Institute of Zoology |
| Country | United Kingdom |
| Sector | Charity/Non Profit |
| PI Contribution | I have taken an active role in life at IoZ, traveling to london on a monthly basis to work with my colleagues in London and form good working relationships. I have given 2 presentations at the IoZ on my work and engaged with teaching groups of MSc students. |
| Collaborator Contribution | My second supervisor Dr Marcus Rowcliffe is based at IoZ in London. He has been actively involved in guiding me with the development of methods and assisted with analysis of data. |
| Impact | Paper coauthored by myself, Dr Marcus Rowcliffe at IoZ and my other supervisors. |
| Start Year | 2015 |
| Description | University partner in Cameroon |
| Organisation | Higher Institute of Environmental Sciences |
| Country | Cameroon |
| Sector | Public |
| PI Contribution | I have visited the university (HIES) to give training to the undergraduate and postgraduate students at the university, as well as to present the key initial findings of my research to the students in the hope of inspiring them to follow similar work. I have employed two students from the university as part of my research team training them in research methods and analysis. |
| Collaborator Contribution | The university have helped me to identify suitable and well qualified students to have on my research team for this research. |
| Impact | Two students from the university are currently employed as my research assistants, providing training and capacity building to them. I am working to build stronger ties with HIES to further promote the quality of their students to other researchers coming to Cameroon, and I'm trying to get a visa for Fabrice to come to the UK for a month internship. |
| Start Year | 2016 |
| Description | ZSL Cameroon |
| Organisation | Zoological Society of London |
| Country | United Kingdom |
| Sector | Charity/Non Profit |
| PI Contribution | I work closely with other ZSL Cameroon staff members when I am in the office in Cameroon. I have given training in occupancy analysis to the team, taken part in training days for safety in the field and write reports on my key findings and field experiences that are useful for other team members when they then engage in the same communities I've been working in. |
| Collaborator Contribution | ZSL Cameroon have greatly helped with my field logistics, including allowing me access to a car, driver and desk space when in Cameroon. |
| Impact | Written a paper with ZSL staff member, blogs for the ZSL Africa website and planning a further two papers with ZSL staff members on the monitoring of African Golden Cats in the Dja Faunal Reserve. |
| Start Year | 2015 |
| Description | Bushmeat specialist working group |
| Form Of Engagement Activity | A formal working group, expert panel or dialogue |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Postgraduate students |
| Results and Impact | Attended the bushmeat specialist working group meeting at ZSL in 2017, where I talked about my current research and how my work could feed into the working groups activities and research. |
| Year(s) Of Engagement Activity | 2017 |
| Description | Conservation Optimism |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Other audiences |
| Results and Impact | On the organizing committee for the Conservation Optimism conference, an interdisciplinary conference of talks, films, presentations, posters, art and performance that aimed to reach previously unengaged groups into a conversation round conservation, from a more positive perspective. Constant negative messaging can leave people feeling hopeless, and feel that there is nothing they can do on an individual level to contribute towards conservation. This conference aimed to share the need for conservation in a more upbeat way, learning from successes and failures to progress together. A survey was shared after the conference which showed that people left the conference with a more positive view of conservation and importantly, what they can do to help. |
| Year(s) Of Engagement Activity | 2017 |
| Description | ICN workshop, Oxford |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Postgraduate students |
| Results and Impact | I co-led and facilitated a session on ethics at the second Interdisciplinary Conservation Network workshop (ICN). The workshop lasted 2 days, during which our group discussed the key issues facing early career researchers carrying out interdisciplinary conservation fieldwork.The workshop results in a paper in a high-impact journal |
| Year(s) Of Engagement Activity | 2018 |
| Description | Mammal monitoring methods in the Congo Basin, Utrecht University |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Professional Practitioners |
| Results and Impact | Attended a 3 day workshop to identify the best-practice methods for monitoring mammals in tropical forests such as the Congo basin. The workshop will (eventually) result in a paper. |
| Year(s) Of Engagement Activity | 2018 |
| Description | Mentoring at ZSL |
| Form Of Engagement Activity | A formal working group, expert panel or dialogue |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Undergraduate students |
| Results and Impact | Gave a presentation about my work and mentored a group of students over the course of one week to help with a systematic review of literature around the use of local knowledge for population monitoring. Students were exposed to new literature and methods that they were previously unfamiliar with, and two students have since engaged in projects involving social methods for conservation research. |
| Year(s) Of Engagement Activity | 2016 |
| Description | Pangolin monitoring workshop, Cambridge |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Professional Practitioners |
| Results and Impact | 34 day workshop to identify the best current available methods for monitoring different pangolin species, and to identify knowledge gaps that can be filled by the Pangolin specialist group. The workshop will result in a collaborative paper, currently in prep. |
| Year(s) Of Engagement Activity | 2018 |
| Description | Presentation at University in Cameroon |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Undergraduate students |
| Results and Impact | Given two presentation at the Higher Institute of Environmental Sciences on the aim and objectives of my research and its importance (2016) and again presenting the key initial findings (2018). Students asked many questions at the end of the talk about the research and I have been since contacted my see of the pupils to find out more about my research. |
| Year(s) Of Engagement Activity | 2016,2018 |
| Description | Presentation at the ECCB, Finland |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Professional Practitioners |
| Results and Impact | approximately 120 participants attended the talk which sparked a good Q&A session and resulted in helpful and interesting conversations with leaders in my field. |
| Year(s) Of Engagement Activity | 2018 |
| Description | Teaching at Imperial College London |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Undergraduate students |
| Results and Impact | Taught a two hour session on my research and on issues of detectability when monitoring species diversity. I was asked to come back again this year to teach again and the students have since contacted me asking for volunteer work and to find out more about the method I am developing to use for their projects. |
| Year(s) Of Engagement Activity | 2017,2018 |