[Coral-List] Assessing reef health through fish ID

Peter Sale sale at uwindsor.ca
Tue Mar 17 15:55:53 EDT 2015

Hi Gus, and list,
A good question which rattled my pure ecology nerves, so here goes.

If you want to monitor fish to assess reef condition, using volunteers, 
you need to limit the task to something tractable.  Best way to do this is 
to have a defined list of species to count, and to use a standard method 
of sampling so that your results are comparable with those of others.  The 
two main classes of sampling method are use of belt transects and use of 
point sampling -- I much prefer the former except for cases when a small 
number of highly trained people will be able to manage the challenges of 
sampling in an unbiased way, while counting fish within an imaginary 7.5m 
radius cylinder surrounding them.  I also much prefer narrow transects (1m 
or 2m) because wider ones, with width estimated, fluctuate in width as the 
interesting fishes come into view or swim away again.   (Other fish 
ecologists will probably dispute this; I just do not think we are that 
capable.)  Multiple 30 m long transects are more useful than fewer 100 m 
long ones.  (All of these comments are aspects of design that help to 
produce results with some reasonable precision.) 

The list of species should include ones that are closely associated with 
coral, rather than being broadly distributed, and ones that can be 
reliably identified in the field by volunteers with limited training.  The 
butterfly fishes are an excellent family from these perspectives.  So are 
the wrasses, surgeonfishes, and damselfishes (although some of the latter 
can be difficult to discriminate in the field, and some others persist in 
occurring in large schools).  Parrotfishes are more problematic except for 
the mature males.  Larger species may often be preferred because they are 
more 'useful' fish, but if relatively rare, you will sample lots of 
transects without counting any.  There are a number of papers on how to 
sample reef fish which would be worth reading, and several widely used 
monitoring methodologies.  You'd be wise to check out what methods are 
being used in East Africa, and favor those over others.  Tim McClanahan, 
Wildlife Conservation Society, Kenya would be worth contacting in this 

You initially asked how to sample fish to monitor reef health.  Realize 
that 'reef health' is a shorthand for 'reefs that look like what we think 
they should look like'.  Monitoring fish of particular species will tell 
you whether the sites sampled are ones that support large numbers, small 
numbers or none, and large individuals or small individuals, of those 
species of fish.  Fish are not instruments for measuring the health of a 
coral reef, any more than frogs measure the health of a puddle.  If you 
want to know about corals, and how well coral species are coping on your 
reefs, measure the corals rather than the fish.  Same for crabs, starfish, 
or anything else.  There are some totally overfished reefs in the world, 
with coral almost absent because of storm damage, bleaching or other 
events, and with little of the complex architecture of a reef with 
abundant living corals.  For the abundant macroalgae, and the fish species 
that associate with macroalgae, they are luxuriantly alive, and, so far as 
I know, quite healthy places.  The fish that live there may even be happy. 
 But now I am digressing.

Peter Sale

sale at uwindsor.ca                 @PeterSale3
www.uwindsor.ca/sale           www.petersalebooks.com

More information about the Coral-List mailing list