Module 6: Distribution Surveys
An on-line resource for practitioners
About this course
To estimate how many snow leopards there are, we need to know where they are. Assessment of the status and distribution of rare and elusive species such as the snow leopard is challenging. Recent surveys in some parts of the snow leopard range have indicated that our understanding of the species’ distributions might not be as accurate as previously thought. To minimize subjectivity and maximize replicability and reliability, it is important to address imperfect detection probability in estimating species distribution.
Module 6 aims to equip participants with the knowledge and tools to plan and carry out snow leopard distribution survey across large areas using interviews of key informants, sign surveys, camera traps or genetic surveys. We will discuss means and workflows to collect, process and organize data in a way that it can be used for occupancy analyses. Additionally, we will be sharing the latest Macro level methods and tools recommended by the PAWS GSLEP Programme. The Macro Level design tools provide recommendation for assessing snow leopard distribution across large area (5,000 sq km) and how to select sites for more intensive surveying. Distribution estimates of occupancy can be obtained by conducting interview surveys, camera trapping or sign based occupancy methods. The module will cover key concepts underlying occupancy models and taking into account detection probability. Module 6 builds on Module 1 and you may hear us repeat familiar themes.
Occupancy surveys can be implemented with a variety of sampling methods, spatial extents, and effort levels. Assessing the distribution of a species at local and regional scales may also help track changes over time and gauge the effects of potential threats by comparing local extinctions and colonizations over time. Probability of site use or occupancy, as a function of habitat can also help define strata for which specific intensive sampling strategies using spatial capture recapture methods can be developed.
This course is Module 6 of the Snow Leopard Network’s training initiative. This Module is offered thanks to the support of GSLEP, University of St Andrews and the Snow Leopard Trust. The course was conducted live through on-line sessions with Snow Leopard Network participants in December 2020. The training took place over 4 sessions (each 2.5 hours). The recordings from this live training are now available below. Do follow the outlined structure of the course as each session builds on each other. In total the course consists of 10 hours of video presentation and discussion. If you have any questions about the course or accessing the reading material please contact us.
Meet the Trainers
Dr. Koustubh Sharma: Koustubh is the International Coordinator of the Global Snow Leopard and Ecosystem Protection Program and a Senior Regional Ecologist with the Snow Leopard Trust. He, along with Justine help coordinate Population Assessment of the World’s Snow Leopards (PAWS) as a GSLEP initiative. He holds a PhD in Wildlife Zoology from Mumbai University, and a Masters degree in Physics. He has undergone training on spatial capture recapture methods at the Centre for Research in Ecological and Environmental Research (CREEM), University of St. Andrews, and on advanced applications of ArcGIS by ESRI. He has been involved with colleagues and partners in developing training tool-kits and delivering training workshops for a suite of ecological methods relevant for snow leopard research and conservation.
Dr. Justine Shanti Alexander: Justine is the Executive Director of the Snow Leopard Network. She provides technical support to GSLEP for the population assessment of the worlds snow leopards (PAWS) and other efforts related to snow leopard conservation. Justine also acts as the Regional Ecologist for the Snow Leopard Trust and supports research and conservation work across the snow leopard range. She holds a PhD in snow leopard population assessments from Beijing Forestry University and a MSc in Conservation Science from Imperial College London.
Dr. Ian Durbach: Ian is part statistician, part operations researcher. He is a research fellow at the Centre for Research into Ecological and Environmental Modelling at the University of St Andrews, where he currently works on two projects: modelling behavioural responses of whales to sonar, and designing the camera trap component of the first range-wide survey of snow leopards. He is also adjunct associate professor in the Department of Statistical Sciences at the University of Cape Town, where he is part of the Centre for Statistics in Ecology, the Environment, and Conservation (SEEC). Ian is also interested in multi-criteria decision modelling (MCDM) for supporting decisions between options whose outcomes are uncertain, and applying machine learning to ecological classification tasks involving images, audio, and video.
We also thank Dr. Abishek Ghoshal and Muzaffar Ahmad for joining us as guest speakers.
I analyze data for conservation projects. Although I have become proficient enough in
diverse methods to obtain animal distribution and abundance, I am self-taught. Going over those methods in a more formal framework with specialists has provided me with an invaluable theoretical background, that I sometimes lose sight of.
The online toolkit modules are designed for practitioners in the field and were very hands-on and informative. Even for those who have previous experience in field data collection and population modelling, they will learn something new. The balanced way these sessions were conducted was formal enough for people to learn but informal enough to keep the participants engaged and interested in what otherwise is quite an academically and rigorous subject. The presenters had an excellent grip on the subject and I would certainly join another another module. Great work SLN and keep it up!
Session 1: Introduction to Distribution Surveys
As preparation for this session please refresh yourself with the principles of sampling and watch this presentation from Module 6. We also recommend the following publications that review presence only methods as reference:
- Maxent is Poisson’s Regression, but in an unreliable way (Biometrics 2013 link)
- Likelihood analysis of species occurrence probability from presence‐only data for (MEE 2012 link)
- Presence-only modeling using MAXENT: when can we trust the inferences? (MEE 2013 link)
- MaxEnt versus MaxLike: empirical comparisons with ant species distributions (Ecosphere 2013 link)
- A comparison of Maxlike and Maxent for modelling species distributions (MEE 2014 link)
- Fitting and Interpreting Occupancy Models (Plos One 2013 link)
- Ignoring Imperfect Detection in Biological Surveys Is Dangerous: A Response to ‘Fitting and Interpreting Occupancy Models’ (Plos One 2014 link)
- Imperfect detection impacts species distribution models (Global Ecology and Biogeography 2013 link)
- Is my species distribution model fit for purpose? (Global Ecology and Biogeography 2015 link)
- Opportunities for improved distribution modelling via a strict maximum likelihood interpretation of MaxEnt (Ecography 2014 link)
- Maxent is not a presence–absence method: a comment on Thibaud et al. (MEE 2014 link)
- MaxEnt’s parameter configuration and small samples: are we paying attention to recommendations? A systematic review (PeerJ 2017 link)
Part 1: Introduction to PAWS & Detection Probability
Part 2: Species Distributions
Part 3: Introduction to Occupancy
Session 2: Occupancy Data Collection
As preparation for this session please refer to the following links:
- Darryl debunks the myth of grid cell size and home range size: Link
- MacKenzie, D. I., Nichols, J. D., Lachman, G. B., Droege, S., Royle, J. A., & Langtimm, C. A. (2002). Estimating site occupancy rates when detection probabilities are less than one. Ecology, 83(8):2248–2255.
Part 1: Sites & Surveys
Part 2: Occupancy Exercise
Part 3: Site & Survey Covariates
Part 4: Interviews for occupancy
Session 3: Occupancy Analysis
As preparation for this session please refer to:
& the following snow leopard occupancy publications:
- Taubmann, J., Sharma, K., Uulu, K., Hines, J., & Mishra, C. (2016). Status assessment of the Endangered snow leopard Panthera uncia and other large mammals in the Kyrgyz Alay, using community knowledge corrected for imperfect detection. Oryx, 50(2), 220-230. doi:10.1017/S0030605315000502,
- Ghoshal, A., Bhatnagar, Y., Pandav, B., Sharma, K., Mishra, C., Raghunath, R., & Suryawanshi, K. (2019). Assessing changes in distribution of the Endangered snow leopard Panthera uncia and its wild prey over 2 decades in the Indian Himalaya through interview-based occupancy surveys. Oryx, 53(4), 620-632. doi:10.1017/S0030605317001107
- Alexander JS, Gopalaswamy AM, Shi K, Hughes J, Riordan P (2016) Patterns of Snow Leopard Site Use in an Increasingly Human-Dominated Landscape. PLoS ONE 11(5): e0155309. https://doi.org/10.1371/journal.pone.0155309
Part 1: Occupancy Assumptions
Part 2: Occupancy Analysis 1
Part 3: Occupancy Analysis 2
Session 4: Occupancy Applied
As preparation for this session please refer to the following publications that look at wider applications of occupancy:
- Bailey, L.L., MacKenzie, D.I. and Nichols, J.D. (2014), Advances and applications of occupancy models. Methods Ecol Evol, 5: 1269-1279.
- Goswami VR, Medhi K, Nichols JD, Oli MK. Mechanistic understanding of human-wildlife conflict through a novel application of dynamic occupancy models. Conserv Biol. 2015 Aug;29(4):1100-10. doi: 10.1111/cobi.12475. Epub 2015 Mar 10. PMID: 25757801.
- Lachish, S., Gopalaswamy, A.M., Knowles, S.C.L. and Sheldon, B.C. (2012), Site‐occupancy modelling as a novel framework for assessing test sensitivity and estimating wildlife disease prevalence from imperfect diagnostic tests. Methods in Ecology and Evolution, 3: 339-348.
- Sharma, K., Wright, B., Joseph, T. and Desai, N. (2014). Tiger poaching and trafficking in India: Estimating rates of occurrence and detection over four decades. Biological Conservation, 179: 33-39
- Vasudev, D., Goswami, V.R., & Oli, M.K. (2021) Detecting dispersal: A spatial dynamic occupancy model to reliably quantify connectivity across heterogenous conservation landscapes. Biological Conservation, 253:108874
Part 1: From Field Surveys to Analysis
- Please refer to the example Questionnaire for survey on Past and Present distributions of large predators and their prey species and People’s perception- Download PDF