This study was conceived by a group of conservation practitioners as part of a global effort to promote the professionalisation of protected area personnel, including rangers. We had contributed to studies highlighting the poor working conditions and employment terms endured by many rangers, and witnessed first-hand the impacts of staff shortages on effective management.
We were aware that the forthcoming 15th Conference of Parties of the Convention on Biological Diversity (COP 15) was likely to agree a to target to conserve at least 30% of the planet by 2030 (’30 by 30’). But it was clear to us that achieving 30% is not just a scientific, political and financial challenge, it is also a workforce issue, and that this was not being addressed. In the absence of up-to-date information on personnel numbers, we decided to collaborate on a survey of rangers and other protected area personnel. The resulting information would be vital for guiding protected area management agencies on personnel needs, for rational and more efficient resource planning and allocation, for calculating training and other needs, and for advocating for better recognition of and support for a workforce that we suspected was far too small.
Assembling the data
We assumed that getting the information would take around a year; in fact, it took more than four! We were astonished by how many countries did not appear to collect protected area workforce data in a systematic way, or even at all. Assembling the information was often like detective work, searching for clues, following false leads, thinking laterally and assembling facts from different sources. Particularly challenging were federal countries such as the United States, Australia and Germany, where many protected areas are managed at the state level, and countries with highly decentralized protected area systems (e.g. the UK). In several countries, officials were reluctant to share directly ‘sensitive’ information on government employee numbers, although in many cases we were able eventually to find the required information in obscure official publications in the local language. Almost half the results eventually came through personal communications with officials and other contacts (See Figure 1). At least 150 people provided or pointed us to information; those who secured information from multiple countries in their region of the world were added as co-authors.
The COVID-19 pandemic has had profound negative impacts on protected areas1, but had it not been for the pandemic and the suspension of personal travel and field work schedules, it is unlikely that we would have had the time to secure the quantity or quality of information we did; nor would we have met the deadline of publication before COP15, which was postponed until December 2022.
The dataset we obtained was large, but inevitably ‘messy’. It covered different proportions of protected areas in different countries, at various levels of detail, from a single undifferentiated number to comprehensive staff rosters. For some countries the necessary information proved unobtainable. To submit our work to a journal (rather than publish a report), we needed to derive some reliable global and regional estimates from this dataset and we therefore invited two researchers from academia with expertise in statistics to join the team.
The statistical task of ‘filling the holes in the data’ proved easier said than done. Numerous methods exist to infer missing data, but each makes different assumptions that are often impossible to assess with real world data. It also became apparent that not all ‘holes’ were the same. Inferring missing numbers of personnel for partially surveyed countries turned out to be very different from inferring missing numbers of personnel in countries for which we had no data. Similarly, it was not immediately clear how to handle data from countries that differentiated rangers and non-rangers alongside that which comprised a single undifferentiated number. Finally, some of the best available predictive methods do not estimate the uncertainty of the predictions they produce, which was important for us to document.
Consequently, the task soon became quite complex. The final analysis involved close to 100 programming functions representing around 5,000 lines of computer code! We even had to rename the R package we created to wrap all of this together because the name we initially choose – rangeR – turned out to conflict with another package we were using called ranger for a different reason (“RANdom forest GEneRator", a random forest being a collection of decision trees and not a place where real rangers ever venture). Eventually we selected two methods to predict the numbers that could not be further apart in their inner workings. They even ended up relying on different predictors correlated to personnel numbers. Despite these differences, they produced very similar global and regional estimates, which gave us strong confidence in our results.
We expected the numbers to be low, but were surprised at how low they are (550,000 personnel, including 285,000 rangers), far fewer than our ‘guesstimates’ of a million rangers. As well as our published comparisons with agriculture and forestry, we made other comparisons (not included in the paper) with sectors where information is just a few clicks away. For example, the global ranger workforce is outnumbered by employees working at golf courses and country clubs in the United States (380,000)2 and by employees in the UK hair and beauty industry (288,000)3. Worldwide the private security industry employs around 20 million people4, 36 times more than work in protected areas. These comparisons are far from frivolous; they indicate first how easy labour market information can be obtained for other sectors, and second how little is comparatively invested in sustaining the natural systems that are increasingly recognized as being vital for human wellbeing.
Our discussion of the results stimulated numerous questions and potential analyses, but to expedite publication we restricted our interpretation of the results to broad sectoral and economic factors. We anticipate that the paper will stimulate and enrich further research and analysis on protected area management, and support global and national advocacy for the needs of those directly engaged in area-based management (our raw data for each country surveyed are provided in the Supplementary Data for the paper). The evidence provided by this study will make a major contribution to the goals of the Universal Ranger Support Alliance (URSA; www.ursa4rangers.org), a collaboration of international NGOs that includes the organisations of many of the co-authors. URSA works globally to promote ranger professionalisation by developing common standards for capacity, employment, equality and conduct in order to build a diverse, professional, accountable and competent ranger workforce, whose contributions are formally recognised and respected.
To ensure widespread awareness of our findings we are planning press events, distribution of an infographic, a webinar hosted by the IUCN, and a side event at COP 15 in Montreal in December 2022.
- Hockings, M. et al. COVID-19 and protected and conserved areas. Parks. 26. 7-24 (2020)
Mike Appleton is lead author of the paper. He is Director of Protected Area Management with the NGO Re:wild and Vice-Chair for Capacity with IUCN World Commission on Protected Areas. firstname.lastname@example.org
Alexandre Courtiol led the statistical analysis. He is a researcher at the Leibniz Institute for Zoo and Wildlife Research in Berlin, Germany email@example.com