Geoscience Reference
In-Depth Information
Stomach contents of the 10most abundant fish species were analysed for two
of the stations (i.e. Tarfia and Bonanza) with a total of 96 samplings (12
months
4 samplings). Most fishes caught were small-sized, includ-
ing mainly postlarvae and juveniles, but also adults of small resident species.
Every individual mysid prey found in the stomach content of each of the 5725
small (total length (TL)
2stations
<
140 mm) fish was identified to species level.
3. Individual Connectivities
From the samplings of individual fish predators belonging to the 10 most
abundant fish species in all the temperature and salinity combinations studied
(n
5725 and the species list was composed by Aphia minuta, Argyrosomus
regius, Cyprinus carpio, Pomatoschistus spp., Dicentrarchus punctatus, Engrau-
lis encrasicolus, Liza ramada, Liza saliens, Pomadasys incisus,andSardina
pilchardus),
¼
three main prey within the Mysidae family were found
(m
18264 and the species list was composed by Neomysis integer, Rhopa-
lophthalmus tartessicus and Mesopodopsis slabberi). We can represent this data
in a individual diet binary matrix, A
m,n
¼
¼
1 if individual prey i
was found in the stomach of individual predator j and 0 otherwise. Because
one individual prey cannot be present in two individual predators, the size of A
is A
18264,5725
and each row contains a single non-zero entry (
Figure 4
). In this
matrix, individual connectivity is defined as the total number of prey indivi-
duals found in the stomach contents of each individual predator.
We decomposed this matrix for each of the nine salinity-temperature com-
binations, so each submatrix, A
m
s
;
n
T
, contained the number of sampled individ-
ual preys (m) and predators (n) in each salinity-temperature combination, with
S
[a
ij
], where a
ij
¼
[1(9-16), 2 (16-22), 3 (22-29)], respec-
tively. The size of these nine submatrices is as follows: A
56
1
;
250
1
, A
4
2
;
97
1
,
A
1622
3
;
1012
1
, A
815
1
;
717
2
, A
956
2
;
472
2
, A
2622
3
;
1025
2
, A
4931
1
;
987
3
, A
2629
2
;
443
3
,andA
4629
3
;
772
3
.
Because individual samplings represent the instantaneous rate or the last
predation events of each individual, equal per-individual predation rate can be
used as a neutral expectation. At steady state, we used the equal per-individual
predation rates given by Eq.
(26)-(28)
to assign all the observed mysid and fish
prey items to the individual fish predators sampled in each of these nine
temperature-salinity combinations. This sampling procedure was done using
the 10 most common fish and 3 most common mysid species. We then com-
pared the expected number of individual prey per predator with the observed
number by sorting each individual according to their connectivity and generat-
ing the distribution of the individual rank in connectivity for the expected and
observed data. 10
2
replicate simulations were completed in each temperature-
salinity combination and we have used the Kolmogorov-Smirnov test to study
whether the observed and expected individual rank in connectivity come from
the same distribution.
¼
[1 (0-6), 2 (6-12), 3 (12-36)] and T
¼
