GIS/GPS capacity development and support for Anti-Poaching efforts in the Dzanga Sangha Special Reserve

From June 15th until July 29th 2018, William Ouellette, Senior GIS Engineer at TomTom BV in Gent, Belgium, was deployed to an on site mission in the Central African Republic (CAR). The request for a volunteer came from WWF (World Wildlife Fund for Nature). The Dzanga Sangha Special Reserve in the south west of the country is the only area in which they operate, and WWF has fostered a long standing relationship with the Ministry of Water and Forest in the country. The reserve is in fact co-managed by the Ministry (mandated with governing the park) and the WWF (who plays an advisory role to the ministry).

William provided support to the Anti-Poaching activities of Dzanga Sangha in various ways, namely the creation of a new topographic map to support planning and operations, GIS/GPS and SMART capacity development, database management and restructuring, as well as reporting in line with donor requirements. The main objective of the mission was to train the GIS/SMART focal point (the appointed staff to handle GIS and SMART data) to effectively support patrol planning efforts by leveraging observation data collected by the eco-guards (a fancy name they have given to rangers). The SMART software is a tailor-made tool for conservation data management, developed by a consortium led by WCS (Wildlife Conservation Society), and although it has drawbacks, it has been developed by conservationists for conservationists, and addresses the needs of anti-poaching in Africa.

Data has been collected using SMART since 2013, meaning that almost 5 years of geo-located data have been stored in the database without being utilized operationally for patrolling route optimization. Data usage was limited to generating summary reports for donors. Although this allowed the securing of funds for anti-poaching activities, it did not tackle the problem at its root, which is to reduce the number of poached animals by understanding where, when and how poachers operate. Therefore, extracting observation information from the database and enabling its spatial visualisation was William’s main focus when conducting GIS training. The challenge was to make the appointed staff understand what they were looking at, and how to interpret what they were seeing to best plan and optimize their patrolling efforts.

Disclaimer: All opinions expressed in this report are that of the volunteer William Ouellette, and in no way represent the position of WWF or of the Ministry of Water And Forest of the Central African Republic.

Arrival

Flying on the only direct twice-a-week flight from continental Europe to the Central African Republic (will be known as CAR for simplicity from now on), I was surprised to see so many empty seats. The flight incidentally happens to lay over in Bangui, dropping off a mix of autochthones, NGO workers and UN soldiers, and continuing onwards to Yaoundé in Cameroon, after looping the loop back to Paris on the same day. I was told by the pilot on my way out of the flight that they fly to Bangui daytime first, so that they can both land and take off in daylight. This might have been for security reasons during the civil war, but it turns out the main reason is simply because the Bangui M’Poko airport is not equipped to handle night landings and take-offs.

I was relatively impressed by how fast we had gotten to the luggage pick-up area. However, the efficiency of the immigration was severely undermined by that of the baggage handlers. Maybe they were under-staffed? The heat made it difficult to work, but it still took more than an hour for the luggage of roughly half the plane’s passengers to make it from the plane (parked 50 m away from the terminal) to the conveyor belt. I finally got my bag, went through the last custom check for the bag content, but they only seemed to open suitcases which are easy to check the content of, and let rucksacks go. I smoothly got out, met by the WWF country coordinator Jean Bernard Yarissem, and we were finally off down the normally bustling Avenue des Martyrs (it was Eid-al-Fitr that day, which is a bank holiday in CAR), an unusually broad and sturdy street unlike many other roads you can find in Bangui or elsewhere in Africa. Turns out it was the landing strip of the previous airport…

I was kindly dropped off in the city center at what would be my hotel for the following 2 nights, “Hotel du Centre”, the former Novotel which shut down at the onset of the civil war. Spartan, but does what it says on the tin. Some intermittent Wi-Fi, fried fish from the Oubangui river one hundred meters away, an actual shower and even air conditioning. And to top it off, Canal+ Sport to show the World Cup! The Portugal-Spain game kept me in for the night. The discovery of Bangui would have to wait until tomorrow…

Bangui didn’t have much under the hood unfortunately, but I was met by Rod Cassidy, a notorious character in conservation in CAR. He owns the Sangha lodge in the Dzanga Sangha Special Reserve (which we will call APDS from now on, the french acronym for the area). He had to run some errands before flying back on the same flight as me on Sunday morning. We ended up spending the day together, going left and right buying last minute things, having a beer in between whenever seemed appropriate, and chatting the day away about various conservation topics relative to the project I was about to get into.

The following day, we flew to Bayanga, the main village in the APDS. From there I was met by Guillaume Duboscq, the technical advisor for anti-poaching activities for WWF, who would be my “boss” during my stay. We spent the day like a typical sunday, spreading lunch over the afternoon, lazing around and briefly discussing some of the topics upfront, although that would be mostly kept for the following day.

On-site team and Accommodation

The first Monday of work was mostly spent getting acquainted with the people I would be working with for the following 6 weeks. Here goes the team:

• Guillaume Duboscq – CTC (WWF) – Technical adviser for anti-poaching activities and my boss
• Christian Ndadet – Park Head Warden (Ministry) – Main recipient of the GIS capacity development
• Jonathan Apila – SMART/GIS Focal Point (Ministry) – Main recipient of the GIS capacity development
• Faustin Motako – Anti-Poaching Counselor & Captain (Ministry) – Main recipient of the GIS capacity development
• Gervais Sana – Assistant Warden (Ministry) – Main recipient of the GIS capacity development
• Luis Arranz – CTP (WWF) – Overall project adviser on WWF side
• Martial Betoulet – Responsable activités Ndima Kali  – Participated in the GIS training to support his community mapping activities as part of a local NGO called Ndima Kali.

The Sangha-Mbaere district of CAR, tucked away in the south-west of the country on the border with Cameroon and Congo-Brazza, is one of the few stable regions in the country, amidst a generally lawless state. This was emphasized from the very beginning in order to reassure GISCorps that it was perfectly safe to send volunteers here. In fact, the only turmoil the region has known was in 2013 and very brief, when the Seleka rebels occupied Bayanga for a few weeks. Israeli and French mercenaries were hired to negotiate their departure, and they were gone as quick as they came, incurring very little plundering and destruction. The project facilities were composed of several buildings scattered across a clearing in the forest.

The main base included the anti-poaching command center (where I worked on most days), the WWF office, the Ministry office, the old infirmary (now rehabilitated into more offices), the Pacebco building, built by a Cameroonian bank of the same name administrating donor funds for the Tri-National conservation efforts, known as the Tri-National Sangha (TNS), across Lobéké National Park (Cameroon), Sangha-Ndoki (CAR, a subset of the APDS) and Nouabalé-Ndoki (Congo-Brazza). The staff was housed in cute little houses all ranging between 100 m to 500 m distance from the main base, but tucked away in the forest, which gave a feeling of remoteness nonetheless. The monkeys were swinging from tree to tree around the house, and tropical fruit trees (papaya, jack-fruit, mango, bread fruit…) were growing all around.

Core Activities

I have spent most of my time teaching Jonathan Apila, the GIS/SMART focal point hired by Guillaume Duboscq just over a year ago. He is young and motivated, and shows great motivation to learn. He dedicated a lot of his work and non-work hours to practice GIS and GPS exercises I would give him. Based on his attitude and the results of his work, I am confident that he will be able to continue carrying out the GIS-related activities of the anti-poaching unit. He was systematically able to reproduce all exercises, and did not hesitate to ask whenever there were things he did not understand.

All the capacity development was carried out using open-source tools and software, namely QGIS and SMART. Although most functionalities of SMART were mastered by Jonathan, we had to start with QGIS from scratch.

Here is a list of the achievements we made together:

• Assistance and optimization of day-to-day tasks. Some of the daily practices were painfully inefficient, and we reviewed some general computer literacy issues which helped him do his job more efficiently such as consistency in folder structuring, saving files with logical naming conventions, cleaning up folders, etc. 

• Load GPX, SHP and CSV files into QGIS and displaying the data into a map. This was the necessary basis to build upon. We have done many exercises from loading data, rendering it with custom styles, and displaying it in a conform map with scale, orientation, legend, source and title. Once this was mastered, the rest followed. This has enabled Jonathan to help Luiz Arranz (CTP for WWF) to display elephant collar positions in near-real time on a map based on CSV files received daily.

• Creation of briefing maps to support the preparation of patrols heading to the field. This map is visual support to explain to the eco-guards what the objectives of the mission were and show them ahead of time the itinerary they needed to walk.

• Creating debriefing maps displaying the GPX tracks and observations collected by the eco-guards during their patrols. This is particularly important to verify that the eco-guards respected their objectives, and to question them when they deviated from them (e.g. adopted a different itinerary than planned), or when the observations collected seem odd or unlikely.

• Generation of summary reports using SMART. Those reports are particularly important as they are the tools which guarantee donor satisfaction. It summarizes the patrol efforts over the span of a month, trimester or year, and provides donors with an overview of the teams’ performances. Unfortunately, no report was monitoring team performance between the years, mostly because there was no significant improvement in anti-poaching practices, but now that data is used to improve patrol intervention, it would be interesting in the future to implement such a report.

• Loading/unloading GPX data to/from Garmin GPS receivers. The loading of GPX itineraries is important, as it provides the eco-guards with the itinerary to follow on the field. Therefore, upon return, it is easy to see whether the tracks captured diverged from the planned itinerary. Custom Maps in .kmz formats were created and loaded on all GPS devices showing topography, patrolling sectors and other administrative units, outposts and camps, as well as main roads and forest tracks. This is good for patrol’s positional awareness in the field. The unloading of the data captured by patrols from the GPS receiver to SMART was already mastered by Jonathan, but some additional work was done to ensure that he verifies every single track that comes out of the devices, in order to ensure that the device wasn’t turned on/off between two different locations.A long straight line will appear between the two locations on the gpx track if so, which pollutes the data and the statistics generated in the reports regarding distances driven and walked, as well as patrolling efforts per unit area. If this is the case, those vertices are to be deleted, because in this particular case, it is better to have no data than wrong data.

We spent every morning with Jonathan working on the above topics because he is the one responsible for carrying these activities in the future. However, in order to ensure that this knowledge is distributed across various people in case of sickness or leave, GIS trainings were carried out with other participants every afternoon between 15h and 17h. The scope was more general as certain participants were not affiliated to anti-poaching activities.

Other sessions included a SMART-only session for Gervais Sana (Assistant Warden) and Faustin Motako (anti-poaching counsellor) between 14h and 15h, who needed to be able to work with the basic functionalities of SMART, including patrol data encoding, database query running and exporting as well as report generation.

Moreover, I also sat down with Christian Ndadet (Head Warden) between 07h30 and 09h30 to teach him sensibly the same things as I would teach Jonathan every morning and the others in the afternoon session. I accepted to do this because he needed specific attention to boost his confidence as head of the anti-poaching unit, and some of the information we were dealing with in our exercises regarding intelligence information and patrol planning was too sensitive to be shared in the classroom.

Guillaume, being on leave back to France for most of the duration of my stay, was updated on the progress by emails and I contacted him on any inquiries I had related to the objectives of the mission. Although he wasn’t there to learn along with the others, I sat down with him a few mornings before his departure to establish the objectives of the mission, as well as to teach him a few useful functionalities of QGIS. Jonathan is the lead recipient for GIS knowledge, but it is useful for Guillaume to know some of the ins and outs of the software to be able to create maps on demand for donors or any other custom request he may have.

Besides the training of the staff directly involved in anti-poaching activities, I also offered to train other people who had keen interest in applying GIS/GPS knowledge to their own projects. One of them was Gnamey, a Central African intern writing her Bachelor Thesis from Sanghor University in Alexandria, Egypt. She was keen on learning on to make maps of the areas she visited during her internship and include it in her manuscript. Moreover, Martial Betoulet was interested to learn how to display geo-localized data captured by GPS receivers and smartphone apps as part of the community participatory mapping project he is involved with at Ndima Kali.

Ndima Kali is a local NGO focused on teaching the autochthonous populations (Pygmies, or Ba’akas in the local language) their disappearing arts and crafts. Indeed, Ba’aka elders are not passing on the indigenous knowledge to the youth, and this is making them more and more unfit for their environment, namely the equatorial forest, as well as threatening their livelihoods and existence altogether. Ndima Kali is helping them in regaining pride and reconnecting with their roots.

On top of capacity development, I also contributed to an array of other activities.

Firstly, I took a look at the data model in SMART, and restructured it to simplify the encoding forms. Jonathan is highly trained for encoding observation data from returning patrols, but only because he has acquired a lot of practice. It proved to be extremely difficult for newbies like Gervais and Faustin to encode the data using the former data model. I simplified it as much as I could without changing the intrinsic structure of the model, in order to ensure the continuity of the dataset. In fact, the data acquired between 2013 and 2017 was not encoded with the same data model than from 2017 onward, making the data of previous very hard to exploit jointly with the recent data. I made sure that with the new structure, data collected would remain inter-operable with the 2017-2018 data.

The second reason why the data model had to be simplified was to prepare for the implementation of Cybertracker, a data collection app on android smartphones enabling the filling out of the observation form in the field directly, thus suppressing the need for paper forms and manual encoding upon return. Although Cybertracker is not yet in use, the new data model is ready for use with the device and app. The eco-guards will require training to use it, but this should simplify the encoding task in the future, and reduce the number of potential error sources in the encoding phase (e.g. encoding errors by Jonathan, eco-guards reporting incomplete, or even forgetting to report, observations in the field) by making some reporting fields mandatory in the form itself.

Secondly, I mapped the elephant collar positions which had been installed while I was there. Surprisingly enough, they darted the elephants to install the GPS collars, but didn’t plan for capacity to track their positions in near-real time. I offered to do the initial mapping of the positions for the first monthly report to donors. I also showed Jonathan how to do it so he could reproduce for the following month, until they find a permanent monitoring solution.

Thirdly, I made a new topographic map for the control centre. It is 220 * 150 cm in size. The map contains all the information we jointly agreed to include with Guillaume and Christian. The goal is to use this map on a magnetic board, so magnets can be moved to the real time location of the patrols. Anyone wanting an overview of the dispatched patrols at any given time can then walk in and take a glance. The information is currently available through the Garmin InReach online service, but it is inconvenient to consult. This Garmin InReach service is useful for the radio operator, which can then update those positions based on what the web service tells him, and based on the Delorme messages he receives. These GPS frequency text messages are sent by the patrols, thus not relying on the traditional telecom network which is pretty much non-existent in the area. The Garmin InReach base layer was also complemented with additional information layers such as park boundaries, patrolling sectors, logging concessions and roads, much like what is shown on the topographic map.

Fun and Free Time

If I am boring you with the serious matters above, rejoice with the fun stuff I got to do while I was there. I got the chance of the lifetime to be there, and even though I was there to help out, Guillaume, Christian and all other WWF/Ministry staff made sure I would be able to discover all the natural wonders of Dzanga Sangha such as Dzanga Bai, a visit to the habituated gorilla families, camping in the forest with the Ba’akas, drinking Mulengue and dancing around the fire, enjoying a beer on the patio of the house on a warm evening, and all the fun evenings spent with people around the project.

Conclusion

All in all, I feel like I have given as much as I have taken from this experience. I am critical in my analysis of the current modus operandi, but I am grateful I was able to provide my humble building block to the construction that is the APDS project.

The former part of the report is positive and full of achievements in terms of GIS capacity development. Key members of staff involved in the anti-poaching activities have received precious knowledge which I am confident they will put to practice and continue fostering over time, if it requires an additional intervention from a GIS expert, be it me or someone else. On top of the capacity development, some structural changes to the SMART data model were made, as well as the creation of additional summary reports for donors. Finally, the cherry on top was the creation of a 220 * 150 cm topographic map to support operations.

The latter part deals with the intrinsic issues of practicing conservation in an area like Dzanga Sangha. Education, nutrition, and management of personal finances, as I presented them, are all interlinked, and present a few hurdles to successful conservation outcomes.

As an ending comment, I want to finish with the acknowledgement. I thank everyone who made sure I had an amazing time while there, and more specifically Guillaume Duboscq who offered me the chance to take a deep dive in the conservation world and lend my expertise to the project. I hope my small and humble contribution will go a long and have a lasting impact on the anti-poaching activities.

I also want to thank Shoreh Elhami and Leslie Zolman for seeing me as the adequate candidate for this mission. I hope I didn’t make them regret their decision upon reception of Guillaume’s feedback and through reading this report (and my blog, for those interested: https://remotemakesense.com). Moreover, I want to thank Luiz Arranz for allowing me to get the full experience by organizing all activities for me. I also want to thank Jonathan Apila, Martial Betoulet, Christian Ndadet, Gervais Sana, Gnamey, Franck and Faustin Motako for having the patience to bear with my chronic impatience. In the end, I had a lot of fun teaching the lot, and I hope it was reciprocal and that they got something out of it. I also want to thank Tiffany Handford, Bas Voorhout, Liz Hall, Prisca, Julie, Ana, Tobias and Therese, the other volunteers/researchers for the fun time we spent together sharing thoughts, food and good stories. Another thank you to Terence Fuh for having me over for the Belgium World Cup Games! Finally, a big thank you to all I forgot to mention, but will identify themselves nonetheless. I made a lot of new friendships, and I hope to stay in touch and maybe one pay another visit, who knows?