Quantitative Description of the Diets

1
Quantitative Description of the Diets

• S. H. Bowen
• Chapter 17

2
Why do we care about fish diets? knowledge of
food types and quantities required by important
or competing species. knowledge of food webs
and interactions. Evidence of changes in the
food web or quality of food or habitat over
time. Changes in size structure of prey?
3
Things to consider - Sample gears should
minimize stress/regurgitation and target species
and sizes desired. - Cessation of digestion
following capture by flash freezing, icing, or
chemical preservatives. Sampling. Diel
effects Seasonal effects Fish size effects
territoriality. Differential digestion rates
(on diff. Prey in stomach)
4
IDENTIFICATION OF DIET COMPONENTS Identification
of partly digested prey - look for
characteristic part of organism that is resistant
to digestion (e.g. vertebral counts, head
capsules, hooks, etc) Level of Identification -
some go to species, but only go as detailed as
needed for study objectives and as practical. -
for some studies only relative size of prey may
be important (usually do weigh or measure
individual prey in stomachs as well as count).
5
Ways of Describing the Diet. 1. Frequency of
Occurrence. - looks at how often individual fish
feed on each prey type. E.g. if 96 out of 200
bluegills had eaten at least 1 dragonfly larvae,
the FO would be 96/200 0.48 or 48. 2.
Percent Composition by Number. - Is just the
percent of the total of individual prey items
that one category makes up. E.g. if the
bluegills ate 2000 individual prey, and 200 were
dragonfly larvae, then 200/2000 0.10 or
10. 3. Percent Composition by Weight. (Wet or
Dry ????) - Percent of the total weight of
stomach contents made up of an individual prey.
E.g. if DFL weighed 14.0 g and total stomach
weights were 247 g, then DFL 14/247 0.057 or
5.7
6
Selectivity Indices. Do fish feed at random or
are they selective? - originally presented by
Ivlev (1961), now modified by Chesson (1983) and
Strauss (1979). The Strauss Index L ri -
pi where ri is the relative abundance of prey
type i in the diet (proportion of total number),
and pi is the relative abundance of prey type i
in the environment. Possible values range from
1 to -1. Strauss provides means of estimating
CIs and for statistical tests.
7
Diet Overlap Indices. One of the 1st was
Schoener (1971) Cxy 1 - 0.5 ( S pxi - pyi
) where Cxy is the index value, pxy is the
proportion of food type i used by species x, pyi
is the proportion of food type i used by species
y. Can also use Spearmans Correlation
Coefficient to test overlaps statistically.
8
Other Hybrid Diet Indices. - Due to
confounding effects of numerical domination and
biomass domination in diets, many sought a way
to reconcile the two and provide a
better interpretation of diet composition
data. - Pinkas et al. (1971) proposed the Index
of Relative Importance (IRI) IRI ( by number
by volume) X ( frequency of occurrence)
This is still in use today as is the George
Hadley (1979), Relative Importance Index (Aia)
see pp 526-27