Thank you Christine!
Remember this post?
It did generate a spirited debate where I once again reiterated my personal distaste for the all-pervasive duplication and lack of coordination in Shark research, and the resulting frustration of both the donors and the divers alike. Once again, the gist of my post is not that I believe that Ryan's initiative is a bad thing in itself - what I bemoan is that lacking the necessary coordination with other data bases, it further fragments the global effort which ultimately leads to worse, not better science.
Now Christine has attempted to chime in, but her comment was truncated due to the space limitations of the comments section of Blogger.
Here it is in its entirety.
A few comments from my experience…
- If people are taking their (limited) time to submit data, the goal should be to get as much useful information efficiently.
In 2004, I started collecting a lot of information, but quickly found that there were too many missing records for many of the variables that were being collected, such as sex, behaviour, habitat, depth, etc. Generally, if a question will generate >5% missing data then it is removed from the analysis and therefore a waste of time to collect in the first place. Data for data sake is not my goal. People are busy and donating their time, so I only aim to collect data that I can use.
- To reiterate Mike’s point, I think we can safely move beyond collecting data on single species groups, such as sharks.
I started with just sharks and rays as proof that recreational divers could provide useful data that the scientific community and governments would accept and use to inform conservation and management strategies. And they have! I was told that the results of the eManta_eShark survey (paper was in review at the time) were used in back room conversations at CITES COP 2013 and were influential for at least 2 countries in voting to list mantas on Appendix II. Similarly, the Bahamas protected their sharks following a paper that was published in 2010 that used the REEF.org data – recreational diver generated dataset - which showed the uniqueness of the Bahamas shark populations.
Note: The REEF.org dataset is the most valuable dataset that I use and I could not do my work without it. It is the template that I used to create eShark and allows me to fill the gap that Johann M identifies, which is the accuracy of reports. However…
- I think we – those coming from the scientific side of things - need to be careful not to suggest that scientific divers are perfect, or that recreational divers are not good at identification.
Way back when, when I started this work, many scientists said I couldn’t use recreational diver data because they couldn’t identify/count them properly …something about them “worrying about keeping their masks on their face”, which is so ridiculous in my opinion. So, I tested the shark scientists, fish biologists, recreational divers, and the general public on their shark identification abilities – covering people from around the world. Let's just say that the only person that got 100% on the test had never dived in their life and was only half way through their undergraduate degree - she just loved sharks.
In fact, throughout my interviews, I have been amazed at the accuracy of not only the identifications (photo verification), but of people’s memories. Many people have photos that correspond with their observations. I’ve used these dated photos to verify memory accuracy. Especially for the rare observations, some people could remember the exact month when an observation took place more than 5 years earlier. However, I always err on the side of my memory ability, which is really not very good. These inaccuracies, whether by recreational divers, dive masters/instructors or scientists, can be managed with good survey design, reasonable assumptions, awareness of limitations and responsible analysis.
- When (the late) Ransom Myers and I developed the idea of using divers to census shark and ray populations, an important first step was to ask the dive community about potential concerns for data.
I specifically asked this question when I was at DEMA and ADEX with Project Aware, which enabled me to cover a wide variety of perspectives and regions of the world. The most important concern raised was that I was creating a dataset that included the location of species that were vulnerable/endangered and still targeted. So, although I think scientists should be open to collaboration, and I certainly am, I think it is irresponsible to make these data freely available or open source and I think we need to be careful who we share it with.
To get around this, I provide data summaries to different people or groups that request it - e.g., I give (almost) monthly summaries to Shark Guardian on their progress in Thailand - and have contributed data to OBIS, which is open source, but the data are shared according to OBIS recommendations for sensitive species (10x10 degree cells - you can't even tell what country they are in!) so that the lat/lon and seasonality cannot be used by people that aim to do harm to these populations.
- I agree with you Mike – more collaboration would be valuable.
I am always on the lookout for other groups that share the my mission and are aligned with eShark. There are a few, including REEF.org, eBird.org, Marine Debris Tracker, and Redmap –Australia that are great and run by very good scientists that know how to use the data. Three of these I featured on eOceans.org for an Oceans Day event in 2014. I didn’t have time to organize this event again this year…
- I am encouraged by the number of different citizen science surveys now available – we have come a long way since those naysayers of 2004, some of which have actually gone on to create their own diver surveys now!
But I do appreciate and share Mike’s comments and concerns about diluting and frustrating the people that want to contribute in the best way they can. Of course people cannot contribute to all the existing surveys – I can barely keep track of them all at this stage.
The advice I have for potential contributors is to look at the goals and results of the organization running the survey and ask:
- i) Do they have publications to support their motivation to collect the data?
- ii) Do they have the scientific and statistical expertise to use the data?
- iii) Do they specifically state who will use the data, or do they just generally say ‘collecting for science’?
- iv) Does the organization intend to use the data responsibly and keep it protected from people that may wish to exploit it, or do they have it open source/freely available?
I personally value locally collected data that feeds into a larger picture, which is why Shark Guardian and the Great Fiji Shark Count are so effective and both received awards from Project Aware for their efforts.
Anyway, this is my two cents.
Thanks Mike for prompting me to comment.