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

since the bird was also observed during 3 preceding
low water periods (Table 4). The feeding rate on ihe
experimental day was 19'? above the average in the
control situation, which was achieved by a 25*/r increase
of N|Vvk. Hardly any worm broke during the experiment,
which also suggests that Curlew F was able
io seize Nereis easily from the surface.
Discussion
Prey availability
The distinction between N ^ (prey detected ai ihe surface)
and Npn))x. (prey taken from its burrow alter a
track reveals its presence in the mud) is not an absolute
one. On the one hand, a Curlew might detect a surfacefeeding
worm but be too late to catch it and subsequently
try to catch the prey in iis burrow. If dial were
to occur often, no large deviation would be found in
the frequency with which both prey types are taken
from dry and wet surfaces. As this difference, however,
is very large (Fig. 6B). it is unlikely lhat prey of the
Nprobc *yP*-- are of'*-' 11 delected as surface-feeding
worms. On the other hand, worms present close to the
entrance might be taken just in a single peck and be
classified as N|Hvk. It is reasonable to assume that this
kind of prey breaks more often than worms taken at the
surface. Indeed c. 59c of Np^ from a dry surface broke
as compared to c. 30% if this prey type is taken from a
substrate covered by water (Fig. 6A), indicating that
the latter category also includes at least some tillerfeeding
worms. We assume that this problem of inter­
VERSATILITY OF CURLEWS FEEDING ON NEREIS
pretation is more likelv in summer (when filleting is
common) man in winter (when filter feeding is rare),
since the handling time of N^ and the proportion of
broken N^.,., decreased from late summer to winter
(Fig. 12). ln other words, all worms of ihe Nprpbctype
are presumably captured when liliei feeding, while
N^ are surface-feeding and sometimes filler-feeding
worms.
ln this section we will estimate die density ol available
prey, if a Curlew selects solely worms of the Npcct
or Npniht. variety. Nereis occur w ith a lotal density of at
least 200 io 400 nr : in the study area during late summer
(Fig. 9A. Table 2), but the density relevant for
Curlews is reduced to e. 100 m 2 on average, since prey
< 7 cm are ignored.
Ihe density of worms available as Np,.ck depends
mainly on the surface activity of Nereis. Surface activity
was measured during 5 field days in September and
October I Esselink & Zwarts 1989). Not a single worm
was recorded during 3 low waler periods, while during
ihe other 2 low water periods, surface aciivilv look up
only 0.1 to (1.2' i of ihe time. This means thai a Curlew
can expect not more than 1 worm at die surface per 5
ot 10 III— nt any one time. Moreover, the densit) of surface-reeding
worms actuall*. accessible to feeding
Curlews is even lower since Nereis retreat quickly into
their burrows if they detect any visible stimulus or vibration
in the substrate (Linke 1939, Vader 1964. own
observations).
Because of this anti-predator reaction, the frequency
at which worms appear al the surface is probably
a more relevant measure of prev availability than
A 50m hose was used to bring 301 of a slurry of minced meat from the observation lower to the feeding termor, of a curlew The photo sin .us
the preparations al receding tide, before the arrival of ihe bird.
329