Amphibious Drone Field Report, Belize. In partnership with the Turneffe Atoll Sustainability Association, Zoological Society of London, the Marine Management Organisation and the Bertarelli Foundation

Abstract

A multi-criteria trial aiming to explore the use of a novel fixed-wing Unmanned Aerial Vehicle (UAV) as an enforcement and monitoring tool in the Turneffe Marine Atoll (Belize), was conducted in partnership between the Turneffe Atoll Sustainability Association (TASA), the Zoological Society of London (ZSL) and the Marine Management Organisation (MMO).

Full text

Report prepared by Melissa Schiele and Dr Tom Letessier
Zoological Society of London
Amphibious Drone
Field Report, Belize
In partnership with the Turneffe Atoll Sustainability Association, the
Marine Management Organisation and the Bertarelli Foundation
Melissa.schiele@ioz.ac.uk

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
1
Acknowledgments
We thank the Bertarelli Foundation and the Marine Management Organisation (MMO)
for funding supporting salary, equipment and field-work. We thank the Turneffe Atoll
Sustainability Association’s Director, Valdemar, and the Education and Outreach
programme coordinator, Eliceo, for their continued support and vision throughout the
expedition. We would like to thank Jayron and Maurice, two dedicated Belizean
Conservation Officers, who made sure our trials were a success in the field, whilst
eagerly embracing us, new technology and the project. Without the engineering
capabilities and comradery from Mauricio and Julian at Aeromao, our UAV would not
be the success it is, today.
How to cite this report
Schiele MA, Letessier TB. 2019. Amphibious Drone Field Report, Belize. Scientific and
field report to the Bertarelli Foundation, Marine Management Organisation and
Turneffe Atoll Sustainability Association.

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
2
Executive Summary
A trial aiming to explore the use of a novel fixed-wing Unmanned Aerial Vehicle (UAV)
as an enforcement and monitoring tool in the Turneffe Marine Atoll (Belize), was
conducted in partnership between the Turneffe Atoll Sustainability Association
(TASA), the Zoological Society of London (ZSL) and the Marine Management
Organisation (MMO). The Turneffe marine protected area (MPA) in Belize, was
delineated in 2012, but is difficult to manage, in part due to illegal fishing, its
remoteness, and high running costs.
The trial was conducted from February 18th-March 15th and was field-led by UAV
technician and ZSL researcher, Melissa Schiele, under project PI Dr Tom Letessier.
Twenty-four flights were completed, average length of 10.9km and with total transect
lengths of 263km. The UAVs maintained full waterproofing which enabled multiple
flights per day. The live front camera link uses a large parabolic antenna which
provides a clear image, which can be recorded on either the camera or its viewing
screen, although the antenna needed to be manually directed at the drone at all times.
We were able to detect megafauna (turtles, manatee, eagle rays), vessels, lobster
fishing gear and multiple habitats. Transects were flown at heights of 75m, 85m, 95m
and 110m to test the detection capabilities of the camera at different altitudes, both for
ecology and surveillance. Deployment from a small vessel was successful and the
UAV reached a furthest point Beyond Visual Line of Sight (BVLOS)
1
of 11.3km. Two
Conservation Officers responsible for enforcement patrols were trained in deployment,
flying, and transect planning. Reports from the Conservation Officers highlighted
benefits for using it as part of the reserve enforcement strategy, such as reducing the
number of skiff patrols needed, which would save fuel. The weather was suitable for
flying most days (n=13). Winds were deemed too strong (>15 knots) on 7 days. Sea
state ranged between Beaufort 2 and 3, occasionally 4, for the duration of the trip.
Future development of the UAV now prioritises on improving the following: 1)
telemetry link between the UAV and the computer, in order to maintain its pre-
programmed trajectory and avoid the UAV entering ‘fail-safe’ and return-to-home; 2)
developing a system where the antenna for the live video link does not need to be
manually directed at the UAV, most likely by installing a omnidirectional antenna into
the system; and 3) establishing features enabling use of a Dynamic Home Point, for
deployments from moving vessels.
Key Achievements
1. First deployment of our water-landing fixed-wing UAV in the Caribbean.
2. First time a live-link fixed-wing is used in the marine reserve.
3. First two Belizean Conservation Officers trained to use our UAV system.
1
BVLOS; Beyond visual line of sight. Our UAV is BVLOS after around 1.8km.

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
3
4. Reconnaissance images were gathered successfully.
5. First ever deployment of the UAV from a small, moving skiff.
6. Detection of eagle rays, manatee, birds, cushion star, bone fish and turtles.
7. Twenty-four transects flown, with a total flight length of 263km.
Expedition roster
Melissa Schiele
Zoological Society of London
Drone technician, pilot and
postgraduate research assistant
Dr Tom
Letessier
Zoological Society of London
Principle Investigator and post-
doctoral researcher
Sarah Keynes
Marine Management
Organisation
Marine Technology Senior Innovation
Lead, Blue Belt Programme
Julian Villada
Aeromao Ltd, Canada
Engineer
Patricia
Rybarczyk
Three Wise Monkeys
Productions
Film crew2
Sacha Bennett
Three Wise Monkeys
Productions
Film crew
Theirry Volant
Three Wise Monkeys
Productions
Film crew
Russell
Pierpoint
Evolved Media
Director and digital asset
management expert
Sophia Ellis
Essex University/Zoological
Society of London
MSc Student
2
Film crew, Three Wise Monkeys Productions, followed the project for one week of the trials in Belize
at their own cost. It is hoped that material produced may be used in future productions to raise
awareness of the of project.
The Gulf of Mexico
Turneffe Atoll
An overview map of the Caribbean, showing
Belize in orange, with inset of the Turneffe Atoll

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
4
Stakeholder roster
Valdemar
Andrade
TASA
Executive Director
Eliceo Cobb
TASA
Education and outreach programme
coordinator
Gil Williams
TASA
Finance manager
Jayron Young
TASA
Chief Conservation officer
Maurice Westby
TASA
2nd in command, Conservation officer
Dr. Leandra
Cho Ricketts
University of Belize
Administrative Director & Science
Director, Marine

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
5
Contents
1. Introduction ................................................................................................................................. 6
2. The Amphibious Fixed-Wing UAV ......................................................................................... 7
2.1 Background................................................................................................................................. 7
2.2 Operation .................................................................................................................................... 8
3. TASA ............................................................................................................................................ 10
3.1 Previous UAV use in the atoll ................................................................................................ 10
3.2 Criteria ....................................................................................................................................... 10
3.3 Results ...................................................................................................................................... 11
3.3.1 Land use change .............................................................................................................. 11
3.3.2. Boat routes ....................................................................................................................... 11
3.3.3. Habitat mapping capabilities.......................................................................................... 12
3.3.4. Vessel detection .............................................................................................................. 12
3.3.5. Lobster pot/shade detection .......................................................................................... 13
3.3.6. To detect hiding spots .................................................................................................... 13
3.3.7. Team performance .......................................................................................................... 14
3.4 Feedback .................................................................................................................................. 14
3.5 Drone patrol cost vs boats ..................................................................................................... 14
3.6 Future suggestions .................................................................................................................. 15
4. MMO ............................................................................................................................................. 16
4.1 Criteria and results .................................................................................................................. 16
5. ZSL ............................................................................................................................................... 17
5.1 Criteria ....................................................................................................................................... 17
5.2 Flights and methods ................................................................................................................ 18
5.3 Observations ............................................................................................................................ 18
6. UAV based infra-red camera use at night .......................................................................... 19
7. Digital Asset Management ..................................................................................................... 21
8. References ................................................................................................................................. 22
Appendix 1. Flight Log ................................................................................................................... 23
Appendix 2. Figures ........................................................................................................................ 25

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
6
1. Introduction
The Turneffe Atoll is situated off the coast of Belize within the 274km of coastline and
is a uniquely biodiverse atoll with around 25% of its reefs covered with mangroves.
Different mangroves types are represented here (red, black, white and buttonwood)
as well as large sea grass beds, coral reefs, sandy lagoons and deep water >200m
(Healthy Reefs, 2018). The atoll measures around 52km long by 16km at its widest
point. It is part of the Mesoamerican Barrier Reef System, which in part, stretches the
length of the Belizean coast, and is the largest barrier reef in the Northern Hemisphere.
The atoll has around 150 islands (Fairweather and Myvett, 2014) and the area is
distinctly multi-strategy in management (App2Fig. 1.), which is a system often used to
manage fisheries as it aims to placate all stakeholders. Fully protected species of
interest include manatees, cetaceans, salt water crocodiles, turtles and corals
(extraction is prohibited). Species which require permits to extract and/or have
restricted fishing seasons include lobster, conch and shark. There is currently a full
ban on extraction and exportation of sea cucumbers, also.
Belize’s population is around 380,000 people, and 2000 of these are registered
fisherfolk
3
(Fujita et al., 2019). In the Turneffe Atoll, the main fisheries are for
Caribbean Spiny Lobster (Panulirus argus) and Queen Conch (Strombus gigas). The
lobster season is closed from 14th February to 13th June and conch season is closed
from 1 July to 30 September (Huitric, 2005), allowing for breeding and recovery of
these species based on research done by various research groups (Wildtracks, 2012).
The fishers in Turneffe are mainly from Belize City and use unbaited lobster traps
made from palmetto (a small palm) wood slats. It takes two people to install these
traps, which lobsters are attracted to, for shelter. However, fisher folk from the North
tend to come in sail boats and target finfish using hand lines (correspondence with
Jayron).
A permit to fish in specific zones within the marine reserve are available for a B$25
fee. Shark permits are available but are expensive in comparison to lobster and conch
and few are sold (B$100). Fishing in the Turneffe Atoll has happened historically, and
fishers have between themselves, delineated their ‘turfs’ within the zones over the
years. As new legal fishers come to the area, increased pressure is put on the stability
of the target species, and as a result, some fishers try to outcompete each other and
maximise their profit by participating in illegal activities. Illegal fishing activity includes,
but is not exclusive to, taking target species outside of their fishing seasons, illegal
fishing methods and landing undersized species.
Belize to date has protected 4.5% of its waters soon to increase to 11.6% in the form
of more preservation zones which are fully ‘no-take’ (BBN Staff, 2019). These will be
put in place within the existing marine reserves but also in the EEZ, to protect important
reef based biological functions which underpin the fisheries industry in Belize (BBN
Staff, 2019). Belize has aspiration to protect its waters, by reassessing its fisheries
management strategies and increasing the size of its preservation zones. However,
3
Fisher-folk is a term which encompasses all people who fish

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
7
even with seemingly robust plans and marine reserves in place, enforcement of the
rules and laws which underpin the strategy decisions, are essential to ensure the
success of the marines protected areas.
The development of methodologies using UAVs is being explored by many research
groups, but none use a fixed-wing which water lands (Hodgson, Peel and Kelly, 2017;
Schofield et al., 2017; Colefax et al., 2019). The ZSL, in collaboration with a Canadian
drone producer, has developed an amphibious UAV. The UAV has the ability to
conduct wildlife surveys whilst providing live vessel surveillance (Schiele, 2018). The
UAV is hand-launched and can water-land, making it especially useful in island
locations, where deployment from large catapults, and landing in nets or on large
swathes of land, is not possible. The trials reported here determine the success of the
UAV as a tool for enforcement and ecological monitoring in the Turneffe Atoll.
2. The Amphibious Fixed-Wing UAV
2.1 Background
The Turneffe Atoll Sustainability Association (TASA) visited the UK in September 2018
and attended the Bertarelli Symposium of Marine Science at the Royal Geographical
Society. They met Dr Tom Letessier and Melissa Schiele from the Zoological Society
of London (ZSL) UAV team, as well as representatives from the Marine Management
Organisation (MMO). TASA invited the UAV team to come and trial the newest
iteration of the Amphibious UAV at the Turneffe Atoll Marine Reserve (TAMR); an area
which contains delineated zones and is managed using a primarily, multi-use strategy
(Coastal Zone Management Authority & Institute, 2013).
The original prototype of the UAV was trialled in partnership with the MMO in the British
Indian Ocean Territories (BIOT) from May-July 2018 by Melissa Schiele and Dr Tom
Letessier, with support from the BIOT Senior Fisheries Patrol Officer. The BIOT marine
protected area is 640,000km2 and consists of around 60 small islands. The initial
criteria for trialling the UAV was two pronged. Firstly, the MMO’s interest in using the
UAV as a surveillance tool for detection of illegal, unreported and unregulated (IUU)
fishing in BIOT, to gather intelligence and evidence of illegal activity (such as
photographs of fishers, and the ID of the vessel with a geotag
4
). The secondary criteria
pertained to the ZSL’s interest in using the UAV as a tool for ecological surveying and
was specifically assigned to detect megafauna (large marine animals) in areas where
traditionally, catch-data would be used to determine animal populations (Schiele,
2018).
A technical report was produced after the trial was completed, which clearly outlined
improvements needed. The main issues with the UAV were related to waterproofing,
camera quality and live camera link signal. The new iteration of the UAV has been fully
4
Geographical coordinates associated with an image, taken by a camera.

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
8
waterproofed, and both the live camera and nadir camera have been upgraded. The
live link signal to the viewing screen has been vastly improved allowing for excellent
quality of video, which is recordable both on the camera itself and by the viewing
screen.
The UAVs are engineered by a Canadian company, Aeromao, using specifications
and technical lead from the ZSL-run field trials from BIOT, and now, Belize, with input
from MMO leading on requirements for the drone as a compliance and enforcement
tool.
2.2 Operation
The following are the UAV specifications from the Aeromao website:
• Construction: EPO foam (durable, easy to repair, carbon fiber reinforced).
• Wingspan: 2 m.
• Weight:3600 grams.
• Endurance:2 Hours.
• Launch: Very easy to hand-launch.
• Take off: Fully automatic.
• Landing: Automatic belly landing on water or land. Parachute landing on land
available.
• Range: +30kms. Tested in real BVLOS flight operations (Beyond Visual Line of
Sight).
• Max. Altitude: About 4500 masl.
• Cruise speed: 62Km/h.
• Camera: 20Mp stills or up to 4K video, nadir orientation.
• Flight modes: Manual, Stabilize, Return to Home, Fly by Wire (automatic
holding altitude and airspeed), Auto.
• Wind Tolerance: up to 45 km/h. for flight and autonomous belly
landing. 25km/h for parachute landing.
• Maximum speed: +85 Km/h – Great stability and minimal roll & pitch oscillations
in gusty conditions = no gimbal required for cameras & sensors.
• Failsafe: Automatically returns to home & loiter upon loss of RC link. Other fail-
safe routines available.
• Telemetry: Battery status, alt, ground speed, compass, altitude, distance
travelled, time on air (speech enabled), and more than +200 parameters more.
• Moving map display/telemetry: UAV position and heading, commands on map,
fly to point on map, altitude changes, remote spot camera trigger, mission route,
and more.
• Weather: All-weather performance. -20oC to +40oC

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
9
The UAV’s flight transects are planned using Mission Planner software, and are also
controlled using an RC controller, on a 2.5Ghz frequency. The live camera link has
been improved, and a large parabolic aerial allows for a clearer signal and good quality
footage. When the UAV lands on water, it is retrieved by two people on a small vessel.
Twenty-four flights were successfully completed (Fig. 1).
Operation of the drone and flight planning required consideration and understanding
of wind patterns, strength and flight theory, in order to plan trajectories safely. Take-
off and landing occurred against the prevailing wind, to a) generate lift on take-off and
b) slow the UAV down on landing. This is considered good practise in UAV flying. In
Turneffe, the prevailing wind was easterly (north or south) and the transect designs
reflect these environmental variables.
Fig. 1. Twenty-four transects were flown in the marine reserve. Only two of the
protected zones were unreachable, due to an airstrip prohibiting flight and lack of a
suitable deployment spot.

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
10
3. TASA
The Turneffe Atoll Sustainability Association is a dedicated association for the
management and monitoring of the Turneffe Atoll, under the Belizean government,
and is co-managed in collaboration with the University of Belize, the Coastguard and
the Fisheries Department. TASA has three field stations within the atoll; Mauger Caye
in the North, Calabash Caye on the Eastern central point and Caye Bokel in the South
and a headquarters based in Belize City. There will be 3-5 Conservation Officers
based at each field station at any one time, and they conduct pre-planned patrols
around the atoll and produce reports outlining observations and interventions. They
have the power to confiscate vessels and arrest individuals who have broken the atoll
laws pertaining to fishing. Laws pertaining to the killing of (for example) crocodiles and
land-use change fall under the jurisdiction of the environment department and wildlife,
so they can only report observations of infractions to these relevant teams.
3.1 Previous UAV use in the atoll
The fisheries department in conjunction with the Wildlife Conservation Society (WCS)
has previously used a fixed-wing UAV. This UAV was smaller than the Aeromao talon
(by around 25%) and had its propeller on the front
5
.
Recently, a ‘moving tower’ strategy has been employed, using a DJ Mavic 2
quadcopter, as part of the regular skiff patrols: the UAV is flown to a selected altitude
and rotates on the spot, like a lighthouse beacon, surveying an area. It is also used to
inspect areas where suspect activity and stealth are required, such as subtle scoping,
or to locate (for example) the source of a fire. There are currently no regular planned
enforcement surveys or surveillance being done with the quadcopter, only these
occasional opportunistic flights (App2Fig. 2).
3.2 Criteria
TASA have specific areas of governance that may be improved, using UAV
technology. The criteria below were derived following discussions between the ZSL
and the TASA director.
1. To monitor land-use change of resorts and other developments within the
TAMR
5
A brief video of this project is available here https://news.nationalgeographic.com/2016/12/UAVs-
fig.ht-pirate-fishing-belize-conservation/.

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
11
2. To assess the most commonly used boat routes in order to ascertain best
areas for mariculture (seaweed farming)
3. To habitat map the atoll (seagrass and mangrove preference) for carbon
stock measuring
4. To survey for illegal fishers’ boats
5. To see if it’s possible to detect lobster catching gear from UAVs
6. To detect hiding spots of illegal fishers in the mangroves
7. To train TASA staff to fly the UAV and assess its suitability as a tool for
enforcement
3.3 Results
3.3.1 Land use change
TASA have concerns that some of those who have permits to build within the atoll are
disobeying the restrictions within their permit and wanted to ascertain if the UAV could
fly, undetected and gather reconnaissance at selected sites. We opted to conduct a
reconnaissance flight over a local resort being built, due to reports of mangroves
clearings for an unlicensed airstrip.
The UAV was able to detect that an airstrip had been built recently. Considering that
the flight plan was a general survey of the site, it was noted that the UAV images were
able to detect a lot of the airstrip, enough to estimate its length, as it ran the length of
a pre-known, manmade channel. The UAV was also able to identify potentially
unlicensed mangrove clearance, (App2Fig. 3 and App2Fig. 4). Ordinarily, the
Conservation Officers may use a quadcopter to gather information on land-use change
in places which are difficult to land a skiff on.
3.3.2. Boat routes
The commonly used navigation routes used by the fisherfolk are known to the TASA
Conservation Officers and fisheries departments and are typically used as the basis
of for their skiff patrols. These routes connect to the 25 or so fishing camps which are
scattered around the atoll. Boats will leave ‘scars’ through the seagrass if the boat is
in such shallow water that the propellers churn up the seagrass. These ‘scars’ were
detected by the UAV and highlighted previously unknown areas of elevated activity at
on the eastern edge of Zone V, which is a fully no-take preservation zone. The shape
of the ‘scars’ at the mouth of the channel (App2Fig. 5) indicate a vessel has had
difficulties navigating the shallow waters, suggesting an inexperienced captain, new
to the area.

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
12
3.3.3. Habitat mapping capabilities
The UAV has a Sony RX0 nadir camera with 20mp and was able to capture high
definition images of habitats. The UAV’s nadir camera has a 10-degree angle of
inclination, which was a feature recommended by the remote sensing team at the
Royal Botanical Gardens, Kew, to help with habitat mapping of terrestrial areas. The
images were geo-referenced retrospectively, using the image time-stamp and the UAV
telemetry log. We were able to capture a variety of habitats, and to identify
environmental issues, such as accumulation of plastic debris, and of sargassum
seaweed which can cause deoxygenation of the water and block out sunlight
(App2Fig. 6).
3.3.4. Vessel detection
A key requirement for the UAV is to be able to see and detect vessels, using both the
live front camera and the nadir survey camera. We explored flying at different altitudes,
in order to determine any effect on the photographs of the vessels. However, a primary
concern regarding clarity of photos, was the white glare of the fiberglass in the sun,
which made it difficult to identify any details such as ID number. Figure 7 A, B, and C
(Appendix 2) were all taken using the nadir camera. The UAV is as manoeuvrable as
it was in BIOT and can change flight plan to loiter over an area of choice, informed by
the live camera link. It is this dynamic feature that increases the operational
applications of the UAV, making it suitable for inspection of a situation and not just
reconnaissance of an area on a planned route. The improved live-link front camera
enabled the ground team to have a clear view of the UAV’s route (Fig. 2).

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
13
Fig. 2. Video still taken from the live front video camera showing a small skiff and a
launch (small speed boat used for patrols by the coast guard). The UAV live cam can
see 2-3km in front of it. This image is from 60m altitude.
3.3.5. Lobster pot/shade detection
Lobster shades are large zinc sheets cut into rectangles which are placed in the
lagoons during the lobster season. The lobster traps are set beneath them, and the
lobsters are attracted by the shade from the sun. We were able to detect the shades
(App2Fig. 8, A-D). When a shade is removed at the end of the lobster season, it leaves
a rectangle of bare sand where grass has died over the course of the season. We
were able to detect these as well.
3.3.6. To detect hiding spots
Some fishers in the area are suspected of making hiding spots within in the mangroves
themselves, where they are cleaning fish and potentially storing them, in ice boxes,
outside of the designated fishery season, especially for conch and lobster, as has been
previously observed by the Conservation Officers. Such an area was potentially
located on the edge of a fully no-take zone, where an unnatural looking patch of

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
14
mangrove, and clusters of garbage, suggest interference and clearing (App2Fig. 9).
The unnatural nature of the area was concluded by both TASA staff and a visiting
Fisheries department worker. These are the types of locations which the patrol team
will dedicate time to survey for with a UAV, and when found, can feed into patrol plans
for the area, using skiffs.
3.3.7. Team performance
Two participants from TASA were involved in the specific drone deployments and
participated on all flights, becoming trained and integral members of the flight team
(App2Fig. 10 (A)), (Jayron and Maurice). Using their experiences working at sea and
therefore being able to understand weather patterns and wind, they were able to
quickly grasp aviation concepts and plan flight routes safely, in the Mission Planner
software. Both Maurice and Jayron were also able to deploy (App2Fig. 10 (B)) and
recover the UAV safely. They were able to offer constructive insight into the problems
faced by TASA and were able to show us where they thought our UAV could be useful
in detecting something of interest. The good relationship that the TASA team maintain
with the local fishing camps was evident, as we were able to deploy from four different
ones and borrow kit from another and using one hut as a make-shift hangar (App2Fig.
10 (C)).
3.4 Feedback
The enforcement team at TASA, who were involved in the UAV flights, have given
positive feedback in terms of the usage and the potential applications of the UAV. The
TASA officers saw immediate value with the live camera function also. They viewed
this tool as a way to extend their vision of the atoll and were particularly interested to
understand how the telemetry system could be improved, allowing for other stations
to watch the live link camera even if a different field station base were operating it.
Feedback from Valdemar and Eliceo continues to be positive and is underpinning the
start of a mutually beneficial partnership between TASA and the ZSL. We are confident
that the UAV can become a valuable tool for the enforcement team, and we are
working to raise the necessary funds to realise this vision.
3.5 Drone patrol cost vs boats
The average cost of fuel per gallon varies in Belize. Each patrol is a 49km average
round trip and the cost of fuel is B$11 per gallon and each trip is 10 gallons of fuel,
meaning each trip is costing them c. B$110. Multiplied across three bases, this gives
a conservative daily fuel cost of B$330. If TASA are patrolling 5 days a week, that’s
B$1650 per week, or B$6,600 per month. In times of season closure or opening,

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
15
holidays and other leads, these patrols may increase in length and frequency,
increasing expenditure to around B$10,000 per month (from correspondence with
TASA conservation officers).
The UAV is electricity powered, supplied by two 4500 mAh batteries for the engine,
and one for the live camera (5200 mAh), and one for the RC controller (2000 mAh). It
takes 2 hours to fully charge the batteries and overnight for the RC controller. As the
base runs on solar energy, the cost to charge the batteries for the UAV are negligible.
It was estimated by Jayron, that around a third of the patrols currently done could
immediately be switched to a drone transect and flown from each of the three bases.
That alone, would save TASA B$2200, per month, or a yearly saving of B$26,400, or
$13,200 US, almost enough to buy a new fixed-wing amphibious drone, or several
Mavic Pro Duos.
3.6 Future suggestions
TASA are looking to improve their effectiveness as an organisation, in terms of marine
protected area management. This challenge is not unique to the Turneffe Atoll and
one where innovation is needed. The challenges that were highlighted to us included
locating illegal vessels, monitoring land-use changes and ensuring that reefs are not
being over used.
Ultimately, we envision the UAVs to be integrated into the everyday enforcement
strategy and flown by the TASA Conservation Officers. Currently, TASA have been
documenting their relatively recently introduced patrols, which have the explicit
purpose of policing the atoll. Not only are they logging their patrols, but also wildlife
sightings and where they are carrying out fisheries interventions, such as giving
warnings, fines, searches or other activities.
Using the data, it is possible to produce a heatmap to highlight where most of the
enforcement activities were occurring in 2018. The areas around Calabash Caye,
Caye Bokel and Cockroach Caye are likely hotspots for enforcement intervention. We
were able to fly multiple transects at these sights easily (Appendix 2) suggesting that
the use of the UAVs can be implemented and integrated simply and have a higher
chance to detect fishers in those areas.
A technical report for the second iteration of the UAV was produced after the Belize
trial and will be used by the development team as a reference point for future UAV
improvements.

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
16
4. MMO
The MMO is a UK Government department responsible for licensing, regulating and
planning marine activities in the seas around England, so that they are carried out in
a sustainable way.
The MMO is working in partnership with ZSL on this drone trial project, as part of the
Blue Belt Programme. The Blue Belt Programme supports delivery of the UK
Government’s manifesto commitment to provide long term protection of over four
million square kilometres of marine environment across the UK Overseas Territories
(UKOTs). The Programme is being delivered in partnership between the MMO and
Centre for Environment, Fisheries and Aquaculture Science (CEFAS), with the aims
of:
• Improving scientific understanding of the marine environment around the
UKOTs;
• Developing and implementing evidence-based, tailored marine management
strategies including surveillance and enforcement; and
• Ensuring management is sustainable and long term (MMO and CEFAS, 2017).
The Programme is funded through the Conflict, Stability and Security Fund (CSSF)
and will support the UKOTs to develop, implement and enforce marine protection
strategies.
The Programme is initially focused on Territories which have designated, or committed
to designating, large scale areas (>50% of their maritime zone, or of recognised
international importance) for marine protection and ensuring these areas are
effectively managed and enforced. These UKOTs are: the British Indian Ocean
Territory; South Georgia & the South Sandwich Islands; the British Antarctic Territory;
Pitcairn; St. Helena; Ascension Island and Tristan da Cunha.
4.1 Criteria and results
As part of the Blue Belt Programme, the MMO are investigating how new innovative
technology can support compliance and enforcement of marine protected areas
around the UKOTs, particularly in relation to surveillance and prevention of IUU
fishing. As such, they are primarily interested in the UAV as a tool for IUU vessels
detection and many of their criteria align with those outlined for TASA.
MMO are specifically interested in understanding the viability of using the UAV as a
tool within the UKOTs. They are keen to explore, for example: robustness and
durability; the range of the UAV; whether it is possible to deploy it from a boat; whether
it can circle in on vessels it detects as ‘targets’ – to identify vessels and collect
evidence of illegal activity; the level of control of the pilot; and the level of training

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
17
required for operation. The main proportion of their criteria has been addressed in the
MMO ‘Essential Requirements’ document, which has not been included in this report,
as it is a government working document. The remote locations of the UKOTs that the
MMO is working with make it challenging to trial the drone in these locations, so Belize
provided an opportunity to test against essential requirements in a relatively more
accessible location. It is hoped the final iterations of the drone trials will take place in
one of the UKOTs, in conjunction with enforcement tasking.
During the Belize trial, we were able to do several successful launches of the UAV
from a small vessel, which has never been done before. Although the UAV was
launchable from the skiff, it is still not possible to power-on and calibrate the UAV on
the vessel, as the movement is too much for the gyros and GPS components. A
solution for this is being currently devised, which will enable the UAV control station to
become mobile. This feature would enable increased manoeuvrability for patrol design
and would also mean that technically difficult flights would be possible as the take-off
point can be moved into the wind.
The UAV reached a maximum distance of 11.3km (round trip of 25km) which is
approximately a quarter of the entire length of the atoll. This distance is adequate for
the needs of the TASA team, but the UAV will need to fly further to meet MMO needs
for operation around the UKOTs. The UAV has a system which is capable of 30km+
telemetry and live link radio. Limitations to the system arose when the signal was
disrupted by high mangrove trees and potentially, humidity. We plan to address this
by elevating the telemetry receiver and the live link aerial to ensure that a better signal
strength between the UAV and the control station. This may be achieved using a
telescopic pole or more novel solutions such as tethering from a kite. It is however,
also important to ensure security of these channels, to avoid fisherfolk potentially
tuning in and gathering information from the Conservation Officers which could inform
them of planned patrols or strategies. The MMO are also keen to test the UAV in more
open ocean conditions and in a scenario where it is fully operated from a vessel out at
sea, in future trials.
5. ZSL
5.1 Criteria
The objectives were chosen to reflect the ongoing work which began in BIOT. The
ZSL team are interested to test the new onboard camera systems, and the MSc
student’s work will contribute to understanding of the types of ecological data that can
be derived from the image sets.

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
18
The main objectives of the expedition were:
1. To conclude whether the new UAV iteration was fit for use out at sea.
2. To ascertain whether MPA officers could be trained to use the UAV, within a
short space of time.
3. To explore if an ecological systematic transect design detected different
numbers of megafauna, in comparison to the IUU patrol style transect, which
are used to target illegal fishers, only.
4. MSc student’s project: To test whether varying altitudes of height influence
habitat classification from UAV images.
5.2 Flights and methods
Melissa trained the team of 3 to set up and assist in flying the UAV, conducting 24
successful flights around the atoll over three weeks (Appendix 1). The first week, UAV
engineer Julian Villada was also present and helped teach Melissa about the
modifications that been installed, including a homing beacon system, which was tested
on foot and to the live link camera system and the fuselage.
All the technical issues from the last version of the drone which was tested in BIOT
(Schiele 2018) had been fixed, which meant that the new drones were fully
waterproofed and functional.
Each evening, a plan for the next day, based on the overall criteria of the expedition
was created by Melissa and the two conservation officers. It was ascertained where
the most at-risk zones were at the time, and where flights should be conducted. The
TASA team would plan the transect for the surveying flight in Mission Planner, also
known as the IUU transect and Melissa would create a corresponding ecological
systematic transect for the same area. Melissa also planned transect routes for
Sophia’s MSc project which required 3 flights at the altitudes of 75m, 85m and 110m
over the reefs, at three sites. All marine zones within the TAMR were flown in, except
for zone IV, which is too close to a landing strip and airfield, disqualifying us from flying
there as per our permit.
5.3 Observations
The ZSL’s primary ecological interest were in being able to detect if observation
opportunities are lost if only IUU transects are flown and if ecological systematic
transects can detect more megafauna. The images are yet to be analysed, but
preliminary observations show that the UAVs new camera can photograph a good
selection of megafauna.

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
19
On flights 20 and 21 at Long Bough (Appendix 1) an IUU transect and an ecological
style transect were flown and within the ecological transect, a manatee was detected,
which was absent in the IUU survey (Fig. 3).
Figure 3. Megafauna observations from the image sets, thus far, with altitudes. A;
bony fish at 20m. B; two long barracuda-like fish, consistent with those seen during
ground-truthing, at 85m. C; manatee at 85m. D; eagle rays at 85m. E; green turtle at
85m.
6. UAV based infra-red camera use at night
A self-funded partner of ZSL, Russel Pierpoint, who is the Director of Evolved Media
6
,
joined the expedition for a week, bringing with him two quadcopter drones, where one
was the new DJI Mavic Pro Duo FLIR, which has the capability to fly and record in
RGB and infra-red (Fig. 4).
All stakeholders had interest in seeing the capability of this new UAV, as currently, no
other UAV on the market has FLIR cameras and is within an affordable price bracket.
This UAV costs around £3000 (US $3,900). Ten flights were conducted over the
duration of one week, either at dusk or at night, at the Calabash Caye Field station or
Caye Bokel (private property). Russell controlled the UAV using a hand-held controller
6
Evolved Media are a digital asset management firm, developing and creating workflows for vast
amounts of imagery http://www.evolvedmediasolutions.co.uk/
A
B
C
D
E

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
20
and was able to fly up to 4km away. The endurance of this UAV is around 23 minutes.
The UAV also has very bright spotlights which can be used to illuminate suspect
vessels, and this allows the drone to gather information about the vessel, such as type,
ID and number of people on board (if batteries allow, a closer inspection of the suspect
vessel will allow for high enough resolution images to warrant facial ID). The UAV has
a unique feature where a voice message can be recorded by the pilot, and then played
out from a speaker on the drone itself. This could act as a huge deterrent and to give
instructions to illegal vessel captains (i.e. “move away from this area, you are
committing an infraction and will be cautioned”).
It was concluded by Russel, Melissa and Jayron, that a dual-system, where the
Amphibious UAV is flown at day and the Mavic Pro Duo at night, would make for a
robust 24-hour surveillance system which would be easy and workable for the TASA
enforcement team. The power of UAVs as deterrents for illegal fishing activity is
posited but would require more qualitative research to conclude.
Figure. 4. A) A vessel is detected on the infra-red FLIR camera on the DJI Mavic
quadcopter (21:00hrs). B) An example of the Mavic capability during the day (RGB
camera) and the same image in IR. C) Vessels at a jetty (20:00hrs). The fiberglass
hulls retain heat for many hours after sunset
A
B
C

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
21
7. Digital Asset Management
As drones become more affordable, and camera technology better, (Colefax, Butcher
and Kelaher, 2017), we face the challenge of ever-increasing sizes of image sets from
expeditions. From BIOT, the total size of all the flight images was around 60GB. Our
image set from Belize is over 440GB. Having multiple external hard drives for the
storage and moving of data from the field to the office, is unstable and risky, and many
ecologists and science teams rely on this method (personal correspondence). There
is scope to move into using hard drives which work by copying images directly into a
cloud, when back at basecamp, as the hard drive creates its own hotspot connection
(as commonly used by the film and photography industry). This allows for large
amounts of images to be accessible to anyone in the world, within hours of the initial
image collection process, either via a cloud or dedicated database system, though to
use this feature a basic internet connection is required.
Russell’s team specialise in building databases and creating online storage facilities
for huge amounts of images (millions). The databases are known as digital asset
management (DAM) systems, and Russell’s team specialise in the ELVIS DAM
system, which was originally developed by Woodwing
7
. The vision shared by Russell
and the ZSL team, is to create a dynamic and AI driven database, which allows fast
and smart interrogation of the images, through easy navigation and customisable
components. For example, locating all the images which have ‘land’ in them could be
done by selecting and filtering, using the metadata tag of ‘land’, or you can ask the
system to search for all the images which contain ‘manatees’. These are very basic
examples of the things this system can do. It can also be linked with google and other
AI platforms, which will help our database identify things in the images, such as ‘birds’,
as we take advantage of the online training databases available. It will also be easy to
share access to this system; we would be fully able to select which parts of the
database would be accessible to whom and for what duration of time. This
interdisciplinary solution to our ever-increasing amounts of images will make the entire
process manageable and will free up valuable time for the scientists. We may see this
DAM system become a benchmark for other UAV users, also.
7
https://www.woodwing.com/en/digital-asset-management-system

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
22
8. References
BBN Staff (2019) Cabinet approves expansion of Fisheries Replenishment (No-Take)
Zones, Breaking Belize News. Available at:
https://www.breakingbelizenews.com/2019/04/03/cabinet-approves-expansion-of-
fisheries-replenishment-no-take-zones/ (Accessed: 9 April 2019).
MMO and CEFAS (2017) ‘Introducing the Blue Belt Programme’.
Coastal Zone Management Authority & Institute (2013) ‘Turneffe Atoll Coastal Zone
Management Guidelines’.
Colefax, A. P. et al. (2019) ‘Reliability of marine faunal detections in drone-based
monitoring’, Ocean & Coastal Management. Elsevier, 174(October 2018), pp. 108–
115. doi: 10.1016/j.ocecoaman.2019.03.008.
Colefax, A. P., Butcher, P. A. and Kelaher, B. P. (2017) ‘The potential for unmanned
aerial vehicles (UAVs) to conduct marine fauna surveys in place of manned aircraft’,
ICES Journal of Marine Science, 75(February), pp. 1–8. doi: 10.1093/icesjms/fsx100.
Fairweather and Myvett, G. (2014) ‘Lobster Trapping Tenureship Survey’, (February),
p. 20 + Appendix.
Fujita, R. et al. (2019) ‘Viability and Sustainability of Small-Scale Fisheries in Latin
America and The Caribbean’, 19, pp. 177–195. doi: 10.1007/978-3-319-76078-0.
Healthy Reefs (2018) ‘Mesoamerican Reef Report Card’, p. 17. Available at:
www.healthyreefs.org.
Hodgson, A., Peel, D. and Kelly, N. (2017) ‘Unmanned aerial vehicles for surveying
marine fauna: Assessing detection probability’, Ecological Applications, 27(4), pp.
1253–1267. doi: 10.1002/eap.1519.
Huitric, M. (2005) ‘Lobster and Conch fisheries of Belize: A history of sequential
exploitation’, Ecology and Society, 10(1). doi: 10.5751/ES-01319-100121.
Schiele, M. (2018) ‘Development and assessment of an amphibious fixed-wing UAV
for marine megafauna surveys and enforcement’, (December), pp. 1–46.
Schofield, G. et al. (2017) ‘Detecting elusive aspects of wildlife ecology using drones:
New insights on the mating dynamics and operational sex ratios of sea turtles’,
Functional Ecology, 31(12), pp. 2310–2319. doi: 10.1111/1365-2435.12930.
Wildtracks (2012) ‘Turneffe Atoll Marine Reserve Management Plan’, Belize Fisheries
Department, p. 293.

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
23
Appendix 1. Flight Log

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
24

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
25
Appendix 2. Figures
Fig. 1. The delineated zones are designated either as conservation zones,
preservation zones (100% no-take), special management areas or general use zones.

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
26
Fig. 2. A Conservation Office retrieving a quadcopter after a short reconnaissance
flight to check for land-use change.

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
27
Fig. 3. Potentially unlicensed airstrip within the TAMR.
Fig. 4. Previously unknown clearing of mangrove wood.

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
28
Fig. 5. Propellers from vessel engines leave ‘scars’ through the seagrass beds.

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
29
Fig. 6. Examples of different types of terrestrial and marine habitats within Turneffe.
.
A. Diving vessel at 75m
B. Patrol vessel, detected
opportunistically on the UAV camera,
95m
C. Vessels at the jetty,
from 65m altitude.

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
30
Fig. 7. A, B and C. Images of boats, taken using the nadir 20mp camera.
Fig. 8. A) lobster shades and areas where shades were, 110m altitude, B) 60m
altitude, large starfish and tyres, in shallow water near a fishing camp, C) seaweed
frames at 75m altitude and D) Fragments of unknown origin, 110m altitude
C
D
A
B

Amphibious Drone Field Report, Belize. Schiele & Letessier, 2019.
31
Fig. 9. A patch of dead/cleared mangrove is detected. Circled is waste and plastic.
Location is on the eastern edge of Zone V, which is within a fully no-take preservation
zone.
Fig. 10. A) Julian (Aeromao) teaching the TASA team about the drone, B) Maurice
(TASA) deploying the drone for the first time, C) Using a fishermen’s hut as a make-
shift hangar.
A
B
C