waders and their estuarine food supplies - Vlaams Instituut voor de ...

Discussion
Response of Knot to spatial variation in their
harvestable food supply
Knot most often visited plots with abundant Macoma
(Fig. 7B). presumably because their rate of food intake
was highest in these areas. The relation between prey
density and the rate at which probing Knot encounter
Macoma (Fig. 11 A) can be used to estimate intake
rates over the wide range of prey densities found in the
field. The predicted intake rate is derived from three
variables: encounter rate with prey, handling time and
amount of flesh in Macoma of each si/e class (Zwarts
& Blomert 1992). The predicted intake rate rises
sharply with prey density (Fig. I IB). The two measurements
of intake rales actually made correspond
o
•7 60
>.
a.
| 40
searching ti
ro
o o
-®9
•®
•*-** 1*2
to
•
6)
•§-0.8
rate
intake
O
\ ••V8783
'.-^•o'
/r*lMS
•7/8/81
1 1 1 1 1 I 1 1 I 1 1 1
2 4 6 8 10 12
biomass (g M balthicam' 2 ) harvestable for knol
Fig. 11. Calidris canutus. A. Searching lime per prey (s) and B. intake
rate ling AFDW s'l as a function ol ihe harvestable biomass
(g AFDW nr : of Macoma 6 to 16 mm long living in the upper 2 cm).
The biomass dala correspond lo Ihe 1981 data shown in Fig. 7. The
two field measurements on feeding Knot i»i were laken in August
I981 (Zwarts unpubl.) and Augusl 1983 (Zwarts & Blomen 1992
The line in A is ihe inverse of the encounter rate, as derived from the
random (ouch model (Figs. 10 to 12 in Zwarts & Blomen 1992).
The line in B is based on predictions derived from a model with
three variables: handling lime, encounter rate and flesh content per
Size Class fjHg- 14 in Zwarts & Blomen 19921
VARIATION IN FOOD SUPPLY OF KNOT
298
well wilh the predicted values (Fig. 11 A), as did ihe iniake
tales of Knot feeding on a variable density of Macoma
in semi-captive conditions (Piersma pens,
comm.). Piersma found that intake rate reached a
plateau of 2 mg AFDW s ' when Macoma occurred at
the extremely high biomass of more than 40 g AFDW
m*-. Since the harvestable biomass of Macoma normally
vanes from 0 to 10 g AFDW m : , it would always
be worthwhile for Knot to search for sites with
the densest food supply. From this, we conclude lhat
choice of site must be one of the most crucial feeding
decisions made by Knot, because it has such a direct
and large effect on intake rate.
In late summer 1981. when Macoma were abundant.
Knot avoided sites where the biomass of harvestable
Macoma was less than 6 to 7 g AFDW m :
(Fig. 7). This means lhai sites with an iniake rate of
less than 0.7 mg AFDW s ' were under-used and lhat
the majority of Knot fed on sites where they could
maintain an intake rate of more than I mg AFDW s' 1 .
The following section discusses whether Knot used the
same decision rule to choose between feeding areas in
years other than 1981. when Macoma were less abundant.
Response of Knot to annual variation in their
harvestable food supply
Knot hardly fed in the study area in two years when the
biomass of harvestable Macoma was less than 4 g
AFDW m : . but did so extensively in the other four
years, when prey were abundant (Fig. 12). However,
the numbers of Knot at the roost did not vary, apart
from the exceptionally low number counted in 1981
(Fig. 12). Therefore. Knot did not leave this part of the
Wadden Sea when the harvestable food supply in the
study area was very low. but instead must have found
other feeding sues in the immediate vicinity. Two
questions are therefore discussed below: (1) What decision
rule is used by Knot when ihey exploit Macoma!
(2) Are there years in which Knot arriving in
the Wadden Sea find insufficient food?
When plotted on an annual scale. Knot scarcely fed
on Macoma when the harvestable food supply was below
about 4 g AFDW m - (Fig. 12). much lower than
the apparent acceptance threshold in 1981 (Fig. 7). It
seems, therefore, that the acceptance threshold is not
fixed, but goes down when there is not much food