Research Spotlight: November 2020

Mon, 2020-11-02 10:56 -- ablackwell
 
Comparing the use of species occurrence data and information systems by academics and government professionals
 
Contributed by: Elizabeth Martín, Shari Ellis, Larry Page
 
Over the past 20 years with increased use of the web, we have seen the proliferation of web-based information systems that provide access to biodiversity data, primarily species occurrence data. Species occurrence data are records on the occurrence (e.g. presence or absence) of an organism at a particular geographic location and time. For this study, we defined web-based species occurrence information system as any databases and their data retrieval systems (including applications software and application programming interfaces) provided via websites and the Internet to discover, access, visualize, download or summarize species occurrence data. 
 
Even though web-based species occurrence information systems are currently very prevalent, research on users of those systems has remained limited. As with any technology, understanding users’ needs and perspectives is a precursor to identifying future improvements and enhancements to systems that may aid the user, and to eventually evaluate the success of data delivery mechanisms. 

Academic researchers and government professionals in natural resource fields are often considered primary audiences for biodiversity data providers and data aggregators’ activities. As such, many of the existing web-based species occurrence information systems are targeted to these key user communities. Our recent paper (https://doi.org/10.1371/journal.pone.0236556) surveyed academics and government professionals in the United States about their reuse of species occurrence data and use of information systems that provide those data.
 
We analyzed responses from 941 participants. Of these, 35% were from academia and 65% were from the government sector. The groups were similar in terms of gender (over 60% were males), average age, year of last degree, and region of residence (primarily the South, West and Midwest regions of the U.S.) of participants. The groups differed on highest level of education, with academics holding mostly doctorate degrees while government professionals were evenly split between those holding master’s degrees and doctorate degrees. Academics worked primarily in scientific research (86%) with most working in the biological sciences (81%), while government professionals worked primarily in natural resource-type professions (79%), with over half of them (55%) working on species management and conservation. They also differed in the geographic scope of the work they did, with academics working mostly on activities of national and international levels, while government professionals in our survey worked mostly on state level types of activities.
 
When asked about their preferences in reuse of species occurrence data, we found that academics preferred original / untransformed (60%) data, while government professionals had preference for two types of data: original / untransformed (35%) and summarized / synthesized (29%) data. Both academics and government professionals reused mostly observational data (87% academics, 98% government professionals) and species ranges & distributions (86% academics, 91% government professionals), followed by instrument data (67% for each sector). Specimen data (63% academics, 46% government professionals) and citizen science data (44% academics, 62% government professionals) were the data types reused least by survey participants. Figure 1 provides frequencies of reuse for these types of data.
 
Figure 1. Frequency of reuse of species occurrence data types by academic and government participants
 
Academics used mostly the web (89%), publications (81%), and colleagues (80%) as sources of data, while government professionals used colleagues (92%), reports (92%), and the web (90%) as their main sources of species occurrence data. When asked how they learned about the existence of web-based information systems that provide species occurrence data, they overwhelmingly selected colleagues and only distantly did they select web search engines and other sources (Figure 2).
 
Figure 2. Source of information where participants learned about web-based species occurrence information systems.
 
Academics and government professionals used web-based species occurrence information systems on average between 3-4 hours per week. Academics used a greater percentage of data from these systems than government professionals, with 40% of academics obtaining more than three quarters of their data from these systems compared to 30% of government professionals.
 
When asked about ease or difficulty they experienced with ‘Finding Systems’, ‘Accessing Systems’, ‘Using Systems’, ‘Identifying Data in Systems’, ‘Understanding Data Provided by Systems’ (context of data), ‘Evaluating Quality of Data Provided by Systems’ (data quality and trustworthiness), and ‘Retrieving Data in Needed Format’, both academics and government professionals indicated that their greatest difficulty was with aspects related to the data (content) provided by the systems rather than with use of the systems themselves.
 
Participants were also asked about their awareness and use of a selected few national and international species occurrence information systems. We found that awareness and use of these systems varied by sector of work, and the systems used by academics differed from the systems used by government professionals. The most used systems by academics were the Global Biodiversity Information Facility (GBIF) (51%) and Encyclopedia of Life (EOL) (47%), while for government professionals were NatureServe Explorer (52%) and the PLANTS Database (43%). 
 
Our results indicate that although views and perceptions held by academics and government professionals about use of web-based species occurrence information systems tend to be similar, these groups differ on how they want species occurrence data to be presented to them and on the types of systems they prefer to use. By acknowledging these differences and taking them into consideration when developing and implementing web-based species occurrence information systems, we could move closer to the goal of meeting the data needs of diverse user communities.