Impact case study database

.

The distance sampling methods developed at the University of St Andrews are widely used as an abundance estimation tool by natural resource managers for monitoring the status of populations of conservation concern (example impact: actions triggered by trends in abundance estimates under the US Marine Mammal Protection Act), for managing the take of harvested populations (example impact: quotas for Atlantic bluefin tuna set based on aerial survey estimates of biomass), and for routine, often statutory, monitoring of biodiversity trends (example impact: Defra policy is influenced by biodiversity trends as measured by the Wild Bird Indicator, one of 15 headline UK Quality of Life Indicators).

Providing the global standard software for estimating animal abundance from sightings surveys

The Distance team at St Andrews has disseminated its work through a multi-pronged, sustained strategy: 1. Publish methodological developments in top journals. 2. Publish introductory and advanced texts. 3. Provide user-friendly software Distance. 4. Back this up with responsive support via email and an online forum. 5. Provide annual training workshops, and a free online course. 6. Publish papers that promote best practice in ecology and taxon-specific journals, targeting wildlife managers. 7. Provide a consultancy service to key user groups. The Royal Statistical Society recognised the success of this strategy by awarding Buckland with the Guy Medal in Gold in 2019. Part of the citation for this award read: “ Stephen has opened up advanced statistical methods to the user community by leading the development of freely-available software (notably Distance now with over 50,000 [people] users from 130 countries) and associated core textbooks.” [S1].

Our Distance Windows software is the global industry standard and has been downloaded from 22,150 distinct email addresses from 181 countries between 1st August 2013 and 31st July 2020. (The figures in the above quote correspond to an overlapping but mostly earlier time period.) The official reference for Distance is R6, and as of 26 August 2020, Google Scholar showed 2,075 citations for this paper. The software Unmarked is the only serious competitor; its official reference is Fiske and Chandler (J Stat Software, 2011), and had 1,409 citations in Google Scholar on the same date. As Unmarked is for use on a wider set of methods, many of those citations do not relate to distance sampling.

Although research papers establish the credibility and utility of methods, it is crucial to publish textbooks and tutorial-style papers to make the methods more accessible to the user community. Buckland is senior author of all four books that are dedicated to distance sampling: an initial book in 1993, an introductory book in 2001, an advanced book in 2004, which have led to the current a book on methods and applications, “ Distance Sampling: Methods and Applications” published in 2015. We also offer a series of annual training workshops, which include online workshops. Software development plays a key part in making methods accessible. Underpinning research (Section 2) is implemented in the software on an ongoing basis, and since August 2013, the following features have been available or implemented: a link to the free statistical software R, providing a major expansion in analytical capabilities; a density surface modelling (DSM) analysis engine, written as a library in R; a simulation engine and a multi-analysis engine; point transect analyses for the MRDS (mark-recapture distance sampling) and DSM engines; a Shape Import Wizard. In parallel with these developments, the methods of Distance for Windows have progressively been made available through R packages (Miller et al., J Stat Software, 2019; http://distancesampling.org/).

Supporting MCS compliance of the eastern tropical Pacific tuna fishing industry

The eastern tropical Pacific (ETP) is a massive area that extends from Central America to Hawaii, and from California to Peru. ETP tuna purse seine fleets use fishing methods that result in some dolphin mortality. There is a need to monitor trends in the affected species of dolphins, to assess whether the industry is having a negative impact on populations. In the absence of such a study, tuna exports from Mexico to the US are under threat. (Mexican exports of tuna are worth over USD150,000,000 per year, with the US accounting for well over half.) The US used to conduct surveys at a cost of around USD10,000,000 per year, but have now stopped, due to budget cuts. Our multiple-covariate distance sampling methods were used to analyse the sightings data ( https://www.int-res.com/abstracts/meps/v291/p1-21/). Their final survey was in 2006. On the basis of the US stock assessments that use our methods and other evidence, the Marine Stewardship Council (MSC) for sustainable fishing awarded the Pacific Alliance for Sustainable Tuna (PAST) MSC certification in 2017. One of the conditions of the award required: “ significant financial investment in an international research program to assess dolphin populations in the Eastern Tropical Pacific Ocean”. A research programme is now underway to satisfy this condition. Oedekoven (PI) and Buckland (CI) are working closely with the Inter-American Tropical Tuna Commission and PAST to implement a new survey, and they coordinated a trial shipboard and aerial drone survey off Mexico in November 2019. The methods of analysis for the data are based on our model-based distance sampling research (R2), using both multiple-covariate and mark-recapture distance sampling. The subsequent main survey has been delayed by the Covid-19 lockdown. The Director of the Inter-American Tropical Tuna Commission, notes that our methods are “ key to determining whether the purse-seine fleet that catches tunas found in association with dolphins can continue to fish for tuna with the confidence that the fishery is sustainable” [S2].

Quantifying the effect of climate change on marine mammals in the Arctic for monitoring purposes

We have also worked with scientists in Denmark and Norway to design and analyse surveys of populations of beluga whales (survey conducted in 2012, estimates published in 2017), bowhead whales (survey conducted in 2006, revised estimates published in September 2013) and polar bears (survey conducted in 2015, estimates published in 2017). These surveys are important as they allow population trajectories to be estimated, and the impact of ice loss in the Arctic to be assessed. Our methods (R6) allow estimation of abundance from the data recorded on aerial and shipboard surveys. The Head of the Norwegian Polar Institute’s Polar Bear Programme, which advises Norwegian authorities on matters concerning polar environmental management, comments: “ Your methods have played a key role in establishing the status of the Barents Sea polar bear population, so that we can monitor the impact on this keystone predator as summer sea ice cover reduces …” [S3], while a Professor of the Greenland Institute of Natural Resources comments: “ Your [Buckland’s] methods have played a key role in understanding the effects of sea-ice loss on beluga whale distribution and abundance, and on the population trajectories of bowhead whales and narwhals off Greenland.” [S4]. By quantifying size of populations and time trends in the size, our methods allow the impact of climate change on Arctic marine mammals to be monitored.

Changing professional practice for monitoring biodiversity of terrestrial vertebrates

In 2017, we developed camera-trap distance sampling (CTDS) to allow estimation of population sizes (R5). The method is being applied to monitor vertebrate diversity in a range of terrestrial environments, from the Himalayas to desert. Its greatest impact has been in quantifying biodiversity of rain forests, for which wildlife surveys are particularly difficult to conduct and biodiversity trends problematic to estimate, given the dense vegetation. A Senior Scientist and Project Manager of the Max Planck Institute for Evolutionary Anthropology notes that the following organisations have switched to using CTDS for monitoring in rain forests: World Wildlife Fund, Wildlife Conservation Society, Wild Chimpanzee Foundation, the government of Guinea, and the Office Ivoirien des Parcs et Réserves [S5]. The first two organisations monitor rain forests across Africa, Asia and the Americas.

Providing a scientific basis for two proposed Scottish Marine Protected Areas

To safeguard marine biodiversity around Scotland, Scottish Natural Heritage (renamed NatureScot in 2020) proposed the Southern Trench Marine Protected Area (MPA) in 2014 and the North-east Lewis MPA in 2019, whose boundaries were chosen based on SNH commissioned reports 573 and 594, published in 2014, which used our distance sampling and density surface modelling methods (R3). These reports were commissioned because at that time, no quantitative evidence was available on which to propose MPAs and their boundaries [S6, S7].

Contributing to wildlife management on a global scale with diverse applications of Distance software

Our Distance software and distance sampling methods are used for a wide range of surveys. Between 1 Aug 2013 and 26 August 2020, Google Scholar lists 2,075 publications that cite R6, and most of these are studies that have used our software to address wildlife management issues. We list a small sample of the studies published in the first 5 months of 2019 that do so to illustrate the diversity of applications and wildlife issues addressed: management of declining populations of Arabian gazelles; assessing impact of poaching on ungulate populations in Iran; estimating the distribution and abundance of guanacos (a species of conservation concern) in Patagonia; managing livestock and populations of large wild mammals in the Yaeda Valley in Tanzania; estimating chimpanzee densities in the Ivory Coast so that population trends can be monitored; assessing viability of Madagascar lemur populations; managing rabbit populations on the Iberian Peninsula; estimating feral cat densities in an urban environment; improving ringed seal density estimates, so that harvests can be managed more effectively; modelling the spatial abundance of fin whales in the Gulf of St Lawrence; estimating blue whale density off California so that their recovery may be assessed; effect of windfarms on upland bird densities; evaluating methods for assessing abundance of a forest songbird in the US; monitoring forest bird populations in New Zealand; estimating the sizes of raptor populations in Uganda’s conservation areas; response of reptiles to fire in NW Africa [S8].