About This Project
Explanation
Our project assesses how conservation organizations are using new data technologies in an era when data and analytical capabilities are both proliferating and questioned. Big data and algorithmic analysis may equip organizations with the ability to track species, monitor ecosystem services, and target ecological restoration with unprecedented precision at large geographic scales. Data can be communicated and shared in highly visualized ways with government agencies and the general public, offering support for a variety of natural resource-related decisions. This era of ‘digital conservation’ is seen as transformative, with the potential to dramatically improve the efficiency, efficacy, and timeliness of environmental management.
Is this potential being realized? If so, under what conditions? If not, why?
The research on digital conservation suggests there are many opportunities offered by different data technologies:
- More real-time, spatially-precise, and, ultimately, objective, insights: "Artificial intelligence has revolutionized the capacity to classify data into formats useful for environmental decision-making at higher spatial and temporal resolutions." (Scoville et al.)
- A democratization of conservation initiatives: "Digital support systems and e-governance also have a potential for democratisation and social empowerment, particularly with regards to under-represented communities and rural people" (Arts et al.)
- New data sources can "provide a cost-efficient information source for addressing the grand challenges of biodiversity conservation in the Anthropocene epoch." (Toivonen et al.)
- Data technology and infrastructure may help conservation organizations focus on what matters and scale up: "By potentially saving thousands of hours of manual data analysis, deep learning could allow conservationists to focus on more pressing tasks" (Lamba et al.)
The research also suggests some challenges as well as practical and ethical concerns:
- Conservation still needs people and fieldwork: "Without involvement of people on the ground it is unlikely that conservation solutions identified by algorithm will work." (Adams)
- Shifting power dynamics in the sector: "AI’s increasing centrality to conservation will distribute power to actors in the private sector who control access to the platforms that increasingly mediate environmental decision-making." (Scoville et al.)
- Financial and technical capacity: "Big data challenges the capacity of underfunded conservation organizations." (Rissman et. al)
- Digital conservation has an environmental footprint: "Many electronic devices are built with planned obsolescence, and resulting e-waste is largely exported to developing countries where it can create environmental problems (Maffey et al. 2015)." (Arts et al.)
The above perspectives are largely from academics or media. Do conservationists and conservation organizations share them?
Methods
Our project consists of several parts:
- First, we’re reviewing online media to understand the conversation around digital conservation – what tools are (not) often talked about, alongside what other conservation topics (be it approaches like working lands conservation or specific ecosystem types like forests), and in what ways? To do this, we are accessing public tweets mentioning key terms such as "machine learning" and "nature conservation". We aren’t analyzing the tweets themselves – instead, we follow whatever page(s) they link to, be it a conservation organization’s blog post or a report from National Geographic. In short, we’re using Twitter as a kind of search engine for finding conservation-relevant discussions. We’ve even built web crawling and scraping software specifically for this (though its applications go beyond the project!): https://github.com/ericnost/observatory https://github.com/ericnost/digital_conservation
- Next, we are surveying conservation organizations across the US and Canada – from local land trusts to household names – to gauge their use of data technologies as well as their goals in using these, the barriers to use they face and what they see as solutions, as well as their future plans.
- Finally, we’ll be surveying individual conservationists to understand their own perspectives on the promise of digital conservation.
The rest of this brief write-up focuses on our website review.
Findings
Trends
Discussions of data technologies are growing, judging from the conservation-relevant webpages we gathered. This is true both in terms of how many times we see them mentioned (average number of times per 100 pages) and how often they are discussed on each page they are mentioned (frequency). Here’s one example: mentions of machine learning on pages linked to by Twitter users tweeting about digital conservation, by year:
Organizations
While the above trends take into account all sorts of webpages we found through our search of Twitter, we also wanted to know what conservation organizations specifically are saying when it comes to digital conservation. So we took a look at the websites of eight organizations (a fuller sample is forthcoming): the rare Charitable Trust, the David Suzuki Foundation, the Grand River Conservation Authority, WWF-Canada, Ecotrust, and the Nature Conservancy of Canada (NCC). Since the use of artificial intelligence techniques has raised ethical concerns related to privacy and surveillance in fields such as health care, we wanted to see if something similar is being addressed within conservation. Only four of these organizations (Suzuki Foundation, WWF-Canada, Ecotrust, and NCC) mention "ai", "artificial intelligence", or "machine learning" on their websites. This may not be all that surprising given that they are the four largest of the bunch. What the figure below shows us is that of them, only WWF-Canada also mentions privacy and surveillance when discussing AI.
Meaning
Obviously, counting terms on websites is not a great indication of conservation organizations’ use of data technologies, much less what they think of as the opportunities in and barriers to these tools. As next steps, we will be more closely reading related webpages where key terms are mentioned together to develop a fuller understanding. We’re also surveying a wide set of conservation organizations in US and Canada and we will reach out to hear from individual conservationists about their own perspectives too.
GEOG*6030 Conservation Tools and Technology
I teach this course at the University of Guelph as part of the Master of Conservation Leadership program in the Department of Geography, Environment and Geomatics. The program trains mid-career professionals in the sector to become leaders amidst the changing dynamics of the field. I’ll be reporting back on the results of our surveys to the class and I hope to take what we find in terms of training or expertise barriers and address them in this classroom setting.
Next steps
- Finish survey of conservation organizations regarding goals, challenges, and planned future uses of data technologies.
- Launch "Q" survey targeted towards individual conservationists
- Find and analyze where conservation organizations are talking about key terms together, to better understand how they are portrayed.