Distance sampling when animals are missed at zero distance
Conventional distance sampling methods assume that all animals at zero distance are detected with certainty. However, in some situations this assumption is violated, and some animals are missed, causing a negative bias in the density and abundance estimates. This occurs on some shipboard surveys of marine mammals, for example, where animals may be missed on the trackline because they are underwater while the survey vessel passes, or because the sea conditions are rough so that even some animals right on the line are missed. The former (being underwater and so not available for detection) is often called availability bias and the latter perception bias.
In this two-day workshop, we cover survey protocols and analytical methods for dealing with both perception bias and availability bias. The workshop is intended for those already familiar with the basics of distance sampling.
24th and 25th August 2020
17:00-19:30 British Summer Time // 12:00-14:30 Eastern Daylight Time (EDT) // 09:00-11:30 Pacific Daylight Time (PDT)
The number of participants is limited to 15, on a first-come, first-served basis. Registration is confirmed on receipt of a completed a registration form and payment in full. Registration and payment details are below.
Format and delivery method:
The workshop will be a mix of live online interactive sessions, delivered via videoconference software (Microsoft Teams), and computer exercises. There will be two sessions in total, on successive days, with a total class time of 4 hours (including breaks) and an additional computer exercise to complete after the first day and before the second. Exercises can be undertaken using Distance for Windows or R (via the mrds package).
- Day 1, 24th Mark recapture distance sampling. The session will begin with an interactive lecture. We introduce the general problem area of missing animals at distance zero and its implications for abundance estimation, before moving on to solutions based on mark-recapture distance sampling (MRDS, also called double-platform distance sampling). MRDS is designed to deal with perception bias and, in some circumstances, availability bias. After introducing the concepts, we review the available software for estimation. We then take a break from the videoconference for you to undertake a computer exercise (with online text-based support if required). We reconvene on video to discuss the exercise, before moving to a second interactive lecture where we deal with further topics including a review of assumptions, methods for duplicate identification and point transect methods. We close with a discussion session.
- Day 2, 25th Dealing explicitly with availability bias. In some contexts, availability bias may be the main source of bias, and in such situations you may need to deal with it explicitly rather than relying on MRDS methods. An extreme example is aerial video surveys of marine mammals, where it might be reasonable to assume that with sufficient care in processing video footage, perception bias in the narrow field of view of the camera(s) is negligible and that the only reason animals are missed is because they are underwater while within the camera’s view. We will describe a variety of methods of dealing explicitly with availability bias, in an interactive lecture covering methods ranging from applying simple availability “correction factors”, to methods that simultaneously model the availability and detection processes. The methods will be illustrated using applications to real datasets, although at present no user-friendly software is available for participants to experiment with on their own. We will also briefly cover newly-developed methods designed for drone surveys in which duplicates are not identified.
You should already know about distance sampling, for example by having taken an introductory workshop or online (either interactively or via our free pre-recorded lectures). You should also have either experience using Distance for Windows, or of using R, and have the relevant software downloaded and installed in advance of the workshop. We will offer pre-workshop group meetings to check the video-conferencing software, and that you have the software correctly installed.
Registration and Payment
The Registration Fee is 150.00 GBP
Payment should be made using the online store at the University of St Andrews. Please follow this link to make payment. You will be asked to register if it is your first visit to the store.
Please feel free to email Len Thomas for more information.
Terms and Conditions:
Information about the University of St Andrews Online Payment Services is available here.