article / 8 May 2017

Getting SMART in Cambodia

In 2016, Cambodia reached a landmark for marine conservation when a 405 km2 Marine Fisheries Management Area was declared around the islands of Koh Rong and Koh Rong Sanloem, creating the country’s first large-scale marine protection. This area was also the first marine site in Southeast Asia to implement fully-operational SMART methods.  In this talk, Kate West, Manager of the Marine Conservation Group, shares her key lessons learnt from implementing SMART in a marine context and dives deep into the limitations and considerations that need to be taken into account when implementing the SMART approach. This is a valuable talk for any team considering deploying SMART in a marine context. 

In 2016, Cambodia reached a landmark for marine conservation when the Minister of Ministry of Agriculture, Forestry and Fisheries signed a proclamation declaring a 405 km2 Marine Fisheries Management Area (MFMA) around the islands of Koh Rong and Koh Rong Sanloem, creating the country’s first large-scale marine protection.

The Fisheries Administration and conservation organisations have been working tirelessly for more than five years within the archipelago to consult with local stakeholders and communities and gather baseline data about the area’s biodiversity to support the designation of the site. Fauna & Flora International and other organisations including Song Saa Foundation and Save Cambodian Marine Life have also played an important part in protecting the site and supporting the designation of the MFMA.

The results of these efforts have been impressive; community fishery patrols have received support and innovative techniques have been employed. These include the use of drones for habitat monitoring and the introduction of the Spatial Monitoring And Reporting Tool [SMART). The Koh Rong Archipelago has been the first marine site in Southeast Asia to implement fully-operational SMART methods, and it is changing the face of marine conservation research and management in the country.

Using the tool, the Community Fisheries have been able to collect and record patrol data, and analyse these to identify hotspots of illegal activity and patrol activity patterns. In turn, this information is being used by the Fisheries Administration to make more effective management decisions.

The Koh Rong Archipelago’s implementation of SMART is a good example of meaningful conservation management. It is an effective technique for locally-led fisheries patrols and management and, because it was in place before the declaration of the MFMA, it can be used as a basis for comparison, allowing us to evaluate the effectiveness of conservation work in this protected area.

Kate West, the Project Manager of FFI Cambodia’s Coastal & Marine Conservation Project and the WILDLABS.NET Marine Conservation Group Manager, gave a talk about their experiences implementing SMART in this marine project, with the goal of sharing honestly the lessons learnt along the way. Crucially, she outlines the key considerations teams looking to implement SMART should take into account in the planning stage of their projects, and identified the key limitations teams should be aware of when considering SMART for their project needs. 

If you have questions about using SMART for marine conservation, Kate is keen to share her experiences and hear about your work in our Marine Conservation Group. Join the group and start a discussion to connect with other members using SMART. 

Presentation Transcript

Site Information

Fauna & Flora International are using SMART in the Koh Rong Archipelago to support management of a Marine Fisheries Management Area (MFMA). In the map (below), the boundary of the MFMA is demarcated by the dark blue line. Within it, there are three community fisheries: Prek Svay Community Fishery in the north, the Daem Thkov Community Fishery in the middle, and finally in the south of the archipelago there is the Koh Rong Sanloem Community Fishery. These community fisheries were the only form of local management before the MFMA was declared. Within their management structures there are no-take zones (called conservation areas in this context) in the red and fish refugia which are the yellow. When the MFMA was declared, it brought all of the buffer zone (darker blue area) into the MFMA boundaries. Other areas demarcated in the archipeligo map include a protected area in green, where only family scale fishing is allowed, and a recreational area that only allows scuba diving and snorkelling, with no fishing allowed to take place in that site.

Cambodia MFMA Map

Project Timeline

FFI began working with marine conservation in Cambodia back in 2010. However, the work with the MFMA properly kicked off in April, 2012 with a Darwin Initative Grant. This was really focused on providing technical support and capacity building to the government and local partners to gather baseline data that would support the creation of a MFMA. This area is, for all intents and purposes, a multiple-use MPA. That’s really when our engagement began in full with the community fisheries, so revisitng their government structures, looking at the existing committees and doing re-elections, and then using these as a basis for bottom-up management at the site.

Project Lifecycle MFMA

Patrols began in a fairly non-structured way in 2014. These patrols were simply the community going out with a GPS, and the GPS would record a track of where they went, and sometimes they might record if they encountered someone. It was a very basic data collection which didn’t provide us, as managers, with any information about whether or not the community were actually going out. You could just take a GPS on any boat, so it wasn’t really giving you any information about what was actually happening when the patrols were taking place.

In October 2015, we started collecting data using the SMART logbooks. This began formalising the process, as it meant patrols began taking down information about their encounters with illegal fishing boats, for example. In January 2016, Phallen and I attended a regional workshop in Phnom Penh on SMART. This was really the point where the data model was designed, and when we really introduced SMART in full. In June 2016, the MFMA was officially proclaimed. Any of the SMART patrolling that was done before this was simply within the community fisheries (the three large zones on the map). Once the MFMA was declared that enabled patrolling for the whole site.

We’ve just had some follow up training provided by WCS Cambodia and also WCS Bangladesh, where they are using SMART in a terrestrial-mangrove riverine system, so they are using it in an aquatic environment. This has made us think a bit more carefully about our data model. Your data model frames what you collect on a patrol and then it structures the analysis of your data. For example, if you are going out on a patrol what was found, and from the analysis side, you can pull up things like how many patrols per month or how many instances of illegal activity.

How does SMART work? 

Just to simply provide an overview of how SMART works: 

You go out and collect your data, hopefully by boat in this place. In our case the community then writes the data down into logbooks, and then this data is inputted into the computer by one of our counterparts from the fisheries administration. Sophie Benbow (Marine Project Manager, FFI Eurasia and WILDLABS Member) is using CyberTracker which prevents this slightly clunky, paper-form system. Obviously there can be issues transferring data from paper into digital records, but for now this is the system we are using.

When you’ve put your data in, you can run an analysis based on what your data model allows you to collect. The program then allows you to create reports. In our case, these reports are submitted to both provincial and national government, we share them internally, and we also feed them back to the CFR – the community fishery.

This analysis will allow you to ultimately plan future patrols based on the feedback you are getting, so essentially you will be using it in an adaptive management way. You can also use the analysis for other purposes, such as monitoring and evaluation or in winding-up a project.

Information Captured

You can get information on the patrol itself. This include information about locations: where people went, quite importantly where are people in relation to the zoning, whether they are inside or outside a specific zone. If you encounter a fishing boat or a type of gear, then you can find out what type of interaction took place, did you inform them of the rules or fine them, the people that were on board, the types of fishing gear was used, and then the action that was taken


The first graph is from 2016, showing the number of patrols over the years. You can see as we got to the end of the year there was a decrease. This was in large part due to one of the community fisheries having an issue with their boat. But it was also due to some internal issues, which we are trying to deal with at the moment but there are provincial elections taking place, so we can’t re-elect the community fisheries committee until the provincial elections have taken place. This is a bit of a blockage at the moment, so we are trying to see if we get other communities to fill that gap.

The next chart shows how many boats were moved from a protected zone. This is simply if boats were in a conservation area and shouldn’t be there, they are told to move out.    

The final chart shows the number of verbal vs written warnings over the course of the year. There are quite a large number of written warnings being issued. The first time you get a verbal warning and then there is a written warning, and then they can have a more formal fines taking place later.


Terrestrial Tool

SMART was ultimately designed with terrestrial patrolling in mind. As a result, a lot of the data models that are available originate from a terrestrial setting. This is one of the challenges we came against when designing our data model, in that we were being informed by terrestrial protected area managers who didn’t have that much insight into the challenges you face in the aquatic environment.

  • Data models designed with forests in mind

Detection Power

Your detection power with SMART can be quite limited. With any kind of patrolling your detectability is going to be influenced by your frequency of patrolling. It’s also something worth considering, not just for SMART but for any kind of enforcement tool, is thinking about the time of day or the areas that illegal fishing might be taking place, and your detectability is going to vary. How you interpret your results about the effectiveness of your enforcement is going to be influenced by detection. For example, if you are doing all of your patrols in the day but illegal fishing is happening at night, then your results might show that there isn’t illegal fishing happening but actually that isn’t accurate.

  • Consider: Time of Day, Location, Frequency of Patrols

Night Patrols

Also, night patrols can be a limitation. Again this is not just in the context of SMART but with patrolling and enforcement in general: you need more resources to go out at night in order to ensure that your crew or patrol agency is safe. So this is something worth considering if a lot of illegal fishing or illegal activity is happening at night.

  • Need more resources (lights, more safety equiment)

Community vs. Government Lead Patrols

In our case, we have the communities conducting the patrols. Sometimes they are accompanied by government staff. If they are accompanied by government staff then they have the power to issue fines and written warnings. If they are not accompanied, then they can only inform people of the rules and move out, they don’t have the legal authority to move people from the area.

  • Communities feel they have less power/authority
  • Cost and manpower to consider

Not Real Time

Another limitation is that SMART is not in real time, which I will come on to address in a moment.

  • Identity, sex, boat, etc. 

Limited International and Regional pllciation in Marine Protected Areas

At the moment, or until really quite recently, there is really limited international and regional application of SMART in Marine Protected Areas. This means it is quite hard to find learning from other sites. So we tried to contact WCS in Belize but didn’t really get anywhere in terms of sharing data models or any kind of lessons learnt. Hopefully now that is changing a bit, but when we started out over a year ago there really wasn’t much out there.

  • Less learning and exchange
  • More trial and error


Considerations For Implementing SMART in Other Areas

Moving on from those limitations there are quite a few considerations for those looking to use SMART in other areas.

How fast do you need data vs. how fast can you respond?

Firstly, thinking about how fast you need data. SMART is not a real time data tool so you need to look at how fast you need data versus how fast you can respond. Vessel monitoring systems (there are a number of different options out there) will provide you with real time data. But one of the main uses of those will be to provide rapid information that you can respond rapidly to. In our case, we don’t have the manpower or resources to respond rapidly, so SMART is more appropriate.

  • Vessel Monitoring Systems = realtime data. Useful for rapid/immediate response to information

It might also be worth thinking about what can be done from the land. In Belize, I know there has been some work looking at more strategic position of lookouts and using those as the way to identify where illegal fishing is happening, rather than just going out on patrol in the hope that you might encounter an illegal fishing boat. Also intelligence, if you have a good intelligence network, responding to that could be more effective in some ways and getting your patrol team to respond to intelligence and being more targeted in your activity.

  • What can be done from land? (e.g. Look outs (Belize), intel)

SMART in the context of wider Monitoring and Evaluation

Where we are at the moment as well in this review we are doing internally about SMART and how we use it, and moving forward, is thinking about SMART in terms of wider monitoring and evaluation for projects. For us, we want our SMART data to help us to show spatial and temporal information about incursions into no take zones and to try and compare that to the biological indicators we get from these sites so that we can say whether or not we are having an impact.

  • Spatial and temporal information on incursions into NTZ - compare with biological indicators

Another thing to consider, and this is something we need to consider, is what does SMART tell you about the actual fishery

SMART is really quite limited in that capacity because you’re monitoring your patrols, you’re not monitoring your fishing boats. So vessel monitoring systems or very simple boat tracking systems might be more suitable if you wanted to look at what’s happening with the fishery, where fishing pressures are occurring and the kinds of landings you’re getting as a result of that fishing pressure. This is being done in Myanmar at the moment where they have trackers on individual boats and then they are measuring the landings from individual boats. You can’t do that with SMART. If you want your enforcement tool to have other purposes, then that is another thing to consider. But fishermen generally are not that keen to have vessel monitoring systems, or it is a conversation to be had.

  • Limited capacity to tell you about fishery landings - VMS or boat tracking may be more suitable

Data Model Design

The next thing to consider is your data model design. Paula and Phallin are working very diligently at the moment on thinking about whether we really need to totally restructure our data model. This basically will affect how we look back on our previously collected data because there are certain changes in the data model that will mean we can’t analyse the data in the same way if we change it at this point.

It’s worth thinking about who is collecting the data: is it a community member, or is it a ranger, or is it someone who has a relatively high level of education? That may affect or influence the system that you use. CyberTracker, for example, can be made to be very simple and pictorial, so that people can just click on icons rather than necessarily needing to read and write. Then the next level up, thinking about who is going to be processing and managing that data. Can they use a computer? If they can’t use a computer very well then do you really want them to be inputting lots of data?

  • Who are your data collectors and data processors/managers?

I would really recommend seeking a review for your data model from other marine sites. This is where we are at the moment; we’re seeking reviews from the wider SMART community, just to see if there are things we have overlooked.

  • Seek review from other marine sites using SMART

How does SMART fit in the wider context of local marine/fisheries management? 

Another consideration is thinking about how SMART fits within the wider context of local marine or fisheries management. For example, in Cambodia vessel licensing is very poor across both the small scale and commercial fisheries. In many fisheries, you would hope that if you were to encounter a boat, you would be able to say ‘Okay this is boat XXX’, and then if you saw boat XXX again you would have a record of that number. In our case there isn’t vessel licensing in a particularly formal way so you have use the identifying features to note repeat offenders. So this would be perhaps the Captain’s ID, or their name. But this is harder because you are tracking people, and people move around. When you are tracking a boat you know it’s the same boat, whereas the captain might change or the crew might change. So thinking about how SMART fits within your wider governance of marine resources is worth considering.

  • e.g. information on repeat offenders: vessel licensing
  • Data capture

Who is your enforcing agency?

Looking at who is your enforcing agency, again that relates back to capacity, resources, and skills.

Can the data or information be captured in another way?

SMART can be used for collecting lots of other things in addition to patrol data, such as biological information records. This might be very useful in your site, or it may be overcomplicating things.

Join the SMART Marine Taskforce

Finally, I encourage people to join the SMART Marine Taskforce. As FFI is not a partner of SMART, the SMART taskforce offers us and other marine users a way to be involved and share our experiences. I encourage all users to join this taskforce and help improve how we're able to use the SMART tool in marine contexts. 

The SMART Marine Taskforce is planning to hold a symposium at the 4th International Marine Protected Areas's Congress which is being hosted this year in Chile from the 4th to the 8th of September 2017. Fauna & Flora International will share experience of using SMART from Gokova Bay in Turkey and in Cambodia. Get in touch with Drew Cronin, SMART Program Manager, to find out more. We look forward to seeing some of you there!

About the Author

Kate West is the Project Manager of FFI Cambodia’s Coastal & Marine Conservation Project. After graduating from the University of Oxford with a degree in Biological Sciences, Kate worked on a small island nature reserve in the Seychelles before returning to study for a Conservation Science Masters at Imperial College London. Her Masters research took her to the coastlines of Senegal and Guinea in search of the West African seahorse, which had never been studied or photographed before in the wild. This experience in West Africa helped her to secure an exciting short-term role with an organisation leading the campaign against illegal fishing before she joined FFI in early 2013. Kate has worked for FFI as both an independent consultant and a Marine Project Officer in Cambridge.

Add the first post in this thread.

Want to share your own conservation tech experiences and expertise with our growing global community? Login or register to start posting!