Ecological Modelling 276 (2014) 64–79

Contents lists available at ScienceDirect

Ecological Modelling
journal homepage: www.elsevier.com/locate/ecolmodel

Effects of biomass changes in the supply–demand balance of energy

in aquatic food webs
Luis A. Salcido-Guevara ∗ , Francisco Arreguín-Sánchez
Laboratorio de Dinámica de Ecosistemas Acuáticos, Centro Interdisciplinario de Ciencias Marinas del IPN, Av. Instituto Politécnico Nacional s/n,
Col. Playa Palo de Santa Rita, Apdo. Postal 592, La Paz 23096, B.C.S., Mexico

a r t i c l e i n f o a b s t r a c t

Article history: Due to the complex interactions between the community and its environment, understanding the
Received 24 January 2012 behaviour of ecosystems is a difficult and laborious task. In this paper, we study several aspects of the
Received in revised form ecosystem characterised by food webs and how their energy balance is affected by changes in biomass and
22 November 2013
the availability of resources required for self-maintenance. Ecosystem behaviour, as expressed through
Accepted 7 January 2014
the energy flows between compartments, and the respiration flows and biomass of 124 trophic models
are analysed. The metabolism of food webs is characterised by measuring respiration flows and scaling
Keywords:
biomass to the 3/4 power. On the basis of this scaling, 16 food webs were selected to make a comparative
Aquatic ecosystem
Supply–demand balance
analysis of the system dynamics, assuming that metabolism, relative to the size and independence of the
Oversupply, Ecopath environment, is the main source of change in the supply–demand balance of energy. The results demon-
Metabolism strate the importance of variation in the biomass of primary producers for the availability of resources
Food web in the system, which is related to bottom-up control, whereas a decrease in the supply of resources by
top predators is associated with top-down control.
© 2014 Elsevier B.V. All rights reserved.

1. Introduction the only study in the literature describing relationships between

Studying ecosystems through an analysis of food webs allows
for the complementary use of holistic and reductionist approaches,
thus creating a synthesis between the general aspects and individ-
ual parts of a system. The aggregation of species into compartments
or functional groups in the food web allows us to map a network of
interactions that might initially hide key mechanisms in the func-
tioning of ecosystems. To reveal these networks, there are a variety
of ecological indicators that need to be considered in describing the
behaviour of the ecosystem.
Bendoricchio and Palmeri (2005) proposed an index of
supply–demand balance (SDB) to measure the energy state of an
ecosystem. Using Ecopath software (Christensen and Pauly, 1992),
they built trophic models assuming that food webs are systems
that transport resources with an efficiency affected by the form
and size of the network (Banavar et al., 1999). Bendoricchio and
Palmeri (2005) indicate that the SDB index calculates the balance
in the rates of supply and demand resources, thus reflecting the
distance between the ecosystem and its possible optimal state.
The SDB index was recently proposed and has not been suffi-
ciently explored. The work of Bendoricchio and Palmeri (2005) is
∗ Corresponding author. Tel.: +52 612 34658; fax: +52 612 25322.
E-mail addresses: lsalcidog@ipn.mx, salcidog@gmail.com (L.A. Salcido-Guevara).
the supply and demand of energy with population or ecosystem
attributes. The authors take the theoretical framework of SDB index
and assume that food webs with indices below 0.66 correspond
to trophic models characterised by a poorly representation of the
network due to the quality of data used or a mass imbalance of
the modelled system. An index greater than 0.75 corresponds to
food webs with a high proportion of supply/demand of resources, a
high recycling, greater resilience and a high cost of network main-
tenance.
The energy flow of an ecosystem may be governed by differ-
ent control mechanisms. Power (1992) provides a review about it,
indicating the importance of limited resources at different trophic
levels and their relationship to bottom-up and top-down forces.
Polis and Strong (1996) developed a conceptual model positing
the existence of a number of donor-controlled (sensu DeAngelis,
1980) resources and alternative pathways (allochthonous inputs)
controlling the system flow, thus indicating that the dynamics
expressed in a food web are primarily governed by the availability
of resources (bottom-up control sensu Huxel and McCann, 1998).
However, other studies have suggested that predation at higher
trophic levels plays an important role in the flow control of the
network from the top-down (Hairston et al., 1960; Oksanen, 1991
sensu Huxel and McCann, 1998). These ideas seem opposed to each
other, but both limit the widespread influence of natural variabil-
ity. As a single system, they offer either the possibility of bottom-up
control when there are significant energy subsidies (allochthonous

0304-3800/$ – see front matter © 2014 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.ecolmodel.2014.01.008
 L.A. Salcido-Guevara, F. Arreguín-Sánchez / Ecological Modelling 276 (2014) 64–79 65

inputs) that prevent a decline in the abundance or renewal rate

temperature (◦ C)
of prey by predators, or top-down control, when the subsidies are
reduced or removed (sensu Polis and Strong, 1996).
Considering that the ecosystems have control mechanisms on

Unknown
Unknown

Unknown

Unknown
Unknown
Unknown
the energy flow (Polis and Strong, 1996; Huxel and McCann, 1998),

Water

25.2b
27.7b

16.5c
27.5c
27.5
14.6
24.5

25.5

24.7
this leads us to ask if the flow control by certain functional groups

4.0
might be related to the energy balance and consequently to the
resources availability of the system. In this study, we test the sensi-

Latitude

24.655

−30.283

20.816

43.761

−22.192
27.250
3.260

−8.570

−26.024

5.630
−22.360

7.064

7.909
19.300
20.000

0.000
tivity of the SDB index (Bendoricchio and Palmeri, 2005) to changes
in the biomass of the compartments of 16 marine and freshwater
aquatic food webs. Our aim is to determine if the balance in the
supply and demand of energy is controlled by a particular group in
the ecosystem, and if a pattern of control mechanisms between

Longitude

−71.548
32.766
−90.393

1.319
38.443

80.982
−47.888
−83.250

114.760

−87.240

0.013
−52.840
112.500

−81.400

30.200
170.000
ecosystems could exist. The study ensures that the comparison
between different types of ecosystems is relative to the ecosystems’
metabolic state (explained below) and assumes an implicit and neg-

Mean temperature from the International Lake Environment Committee (http://wldb.ilec.or.jp/). Biomass, respiration and B/P are estimates of the static model Ecopath.
ligible influence from factors such as the taxonomic aggregation

36,290,000
Study area

Unknown
level of the functional groups and environmental variability.

933,000
170,000

Mean temperature obtained for the same area and period study through the Earth Science Information Partner Federation (Ocean ESIP: http://poet.jpl.nasa.gov).
76,668
(km2 )

3126

1100

250
28

25
30

60

90
2. Materials and methods

6
Food web models used to analyse marine and freshwater ecosystems. For their selection, biomass and heterotrophic respiration attributes were considered.
2.1. The food web model

Study period

1995–1998

1978–1989
1972–1985

1992–1995
1990–1998

1990–1991
2003–2004
Unknown

Unknown

Unknown
The food webs analysed in this paper are based on the approach

1970s

1970s
(year)

1972

1971
1990

1970
proposed by Polovina (1984) and improved by the Ecopath model
of Christensen and Pauly (1992). The data were obtained using the
Ecopath with Ecosim software (EwE version 5.1.218, Christensen

0.132
0.137

0.176

0.157
0.051
0.019
0.015
0.034

0.024
0.063

0.024
0.084
0.021

0.041
0.026
0.300
B/Pa
and Walters, 2004). The Ecopath model is based on a system of
linear equations that represent the mass-balance of species in

B/P is the biomass/production ratio and reflects the average size of organism in a system (sensu Christensen et al., 2005).
an ecosystem through compartments and energy flows that are
expressed quantitatively as,
(t/km2 year)
Respiration

5977.33
268.92

466.65

13494.73
3571.81
1281.32
1687.36
318.63
249.73
315.86
1619.64
856.37

59.42
1574.90

24904.01

7188.10
P  n 
 Q

Bi EEi − Bj DCji − Yi − Ei − BAi = 0 (1)
B i B j
j=1

where Bi and Bj are the biomass of the prey and the predator
Biomass

group, respectively; Pi , Qj , Yi , Ei and BAi are the rates of produc-

451.16
9.63

12.85
3876.92

183.34
51.81
76.17

7.21

72.97
41.45
76.49
1.25
59.90

2401.60

10.10

30.08
(t/km2 )

tion, consumption, fisheries capture, net migration and biomass

accumulation, respectively; DCji is the fraction of the prey i in
the average diet of the predator j. The sum describes the preda-
tor consumption; for its prey, it is equivalent to the mortality rate
Functional

of predation (M2i ); EEi is the ecotrophic efficiency and indicates

groups

the proportion of production used in the system. The model is

59
29
26
14
20
18
18
10
16
13
40
10
14
14
17
11
balanced under the following thermodynamic condition of con-
sumption (Qi ):
Mexican Caribbean (Alvarez-Hernández, 2003)

Paraná River (Angelini and Agostinho, 2005)

Broa Reservoir (Angelini and Petrere, 1996)

Qi = Pi + Ri · Di (2)
Sri Lankan Reservoir (Moreau et al., 2001)
Central Pacific Ocean (Cox et al., 2002)

Garonne River (Palomares et al., 1993)

Coast of Sarawak (Garces et al., 2003)

Celestun Lagoon (Chávez et al., 1993)

Such that,
Looe Keys (Venier and Pauly, 1997)

Maputo Bay (De Paula et al., 1993)

George Lake (Moreau et al., 1993)

Pi = Yi + M2i · Bi + Ei + BAi + M0i · Bi (3)

Bali Strait (Buchary et al., 2002)
West Florida (Okey et al., 2004)

Sakumo Lagoon (Pauly, 2002)

Awassa Lake (Fetahi, 2005)

where Ri , Di and M0i are the rates of respiration, waste and

Tongo Bay (Wolff, 1994)

mortality from different causes related to fishing and predation,

respectively. The energy units are expressed in terms of wet weight
biomass.
Food web

2.1.1. The chosen models

Out of 124 available Ecopath trophic models (Table A1), we
selected 16 to represent several types of aquatic ecosystems, such
as oceans, the continental shelf, reefs, bays, coastal lagoons, rivers,
Continental Shelf

lakes and reservoirs. These systems have several characteristics

Coastal lagoon

that might influence the comparison (Table 1) between their own

Ecosystem

Coral reef

implicit natural variability (space–time) and the criteria used for

Reservoir
Oceanic

the construction of each model. The food web used to experiment

Table 1

River

Lake
Bay

with biomass changes of the functional groups were selected based

a

on their biomass and respiration flow.

66 L.A. Salcido-Guevara, F. Arreguín-Sánchez / Ecological Modelling 276 (2014) 64–79
2.1.2. Transfer efficiency between trophic levels
Unlike EwE, which calculates the transfer efficiency of a discrete
trophic level (limitation of the model), the transfer efficiency (TEi )
is calculated for each heterotrophic compartment in order to have
an estimate related to a control mechanism. It excludes the pri-
mary producers and detritus groups that do not have an input flow
as consumption in the remaining groups and respiration (output
flow). According to equation 3 (Christensen and Walters, 2004), TEi
can be calculated as follows.
Yi + M2i · Bi + Ei + BAi
TEi =
(Q/B)i
TEi is the proportion of energy transferred by predation and
export (catch and migration). Other reasons for the loss of trans-
ferred energy from causes of mortality other than predation, such
as disease and senescence (M0i ), are added to the energy lost by
respiration and to the flow to detritus. In this sense, by subtrac-
ting M0i from the production of each group, TEi is obtained as the
ratio of outflows, which is also the available production (Pi − M0i ·Bi )
between inflows to a compartment and which serves as the rate of
consumption (Q/B)i . TEi is calculated to relate the biomass changes
of each group or specie with the SDB index. BAi is not calculated
because all estimates are made in static trophic models.
2.2. Energy supply–demand balance
Bendoricchio and Palmeri (2005) used the Ecopath food web
approach (Christensen and Pauly, 1992) to propose an ecological
index that measures the balance between supply and demand (SDB)
of the energy in an ecosystem. The SDB index is the exponent of the
relationship between the outflows (Eout ) as given by the sum of pro-
duction, respiration and detritus flow (Pi + Ri + Di ) and the biomass
(Bi ) of each compartment is expressed as
Eout ∝ BSDB
The ˛ (or SDB index) is the product of ˛ (1 + xs − xd ), where
˛ = D/(D + 1) with D = 3 for the three-dimensional Euclidean space
occupied by a food web; xs and xd are the supply and demand rates
of energy, respectively. In this sense, a network that is perfectly bal-
anced in its supply and demand rates (xs = xd ) can yield an SDB = 3/4.
If xs = 0, then xd = 1/D2 (=1/9) and SDB = 2/3; in this case (and when
xd = 0), the biomass can vary with a constant energy flow. If xs < xd ,
then 2/3 ≤ SDB < 3/4, which indicates that when supply is lower
than demand at a certain time, then the system can be rearranged
to increase supply or decrease demand for energy. If xs > xd , then
SDB > 3/4, indicating that there is more energy than required by the
system. This energy imbalance represents the cost of the mainte-
nance of the ecosystem, but it also allows for a greater stability to
act as a buffer that mitigates external pressures (Bendoricchio and
Palmeri, 2005).
2.2.1. Sensitivity analysis of the SDB index
For each compartment in a food web (static model), a routine
of 1000 runs was performed in a spread sheet with macros in MS
Office Excel. The SDB index was obtained by performing a random
change in biomass (Bc ) during each run. To calculate the change in
biomass, the following equation was used:
Bc = Bi + CBi (1 − 2X)
where Bi is the biomass of group i; C is the maximum ratio of
change allowed (between 0 and 1); and X is a random number
with a uniform distribution (0 ≤ X < 1) obtained with the function
RND. If X < 0.5, the biomass increases (Bi < Bc ≤ 2·Bi ); if X > 0.5, then
the biomass decreases (0 < Bc < Bi ); and if X = 0.5, the biomass does
not change. We analyse the sensitivity of SDB with the change of
biomass of each group without rebalancing the trophic model. With
(4)
Fig. 1. The relationship of potential scaling of respiration flows and the total biomass
of the heterotrophic groups of the 123 aquatic food web of marine and freshwater
ecosystems. Black dots indicate the models used to analyse the SDB index.
this we try to reduce the variability that could result in a new bal-
ance. In this regard, for each functional group or compartment in
the food web, ten categories of change in the biomass were con-
sidered in increments of 10% from C = 0.1 (10%) to C = 1 (100%). In
each class, the mean of the SDB index and its 95% confidence inter-
val were computed. This analysis demonstrated the effect of the
biomass of each functional group on the behaviour of the energy
balance of the food web as measured by the SDB index. Thus, if
an oversupply of resources in the system is related to a particular
trophic level, then we could define a type of control, be it top-
down, bottom-up or mixed, from the supply–demand level of the
resources of the food web.
(5)
3. Results
Considering that the data of the heterotrophic biomass and the
respiration of the 124 food webs had a normal distribution (K–S;
p > 0.2), we proceeded with a simple linear regression analysis of the
log–log relationship of both attributes. We obtained a slope of 0.804
(R2 = 0.764, n = 124) without significant differences from the expo-
nent 3/4 of the scaling rule of interspecific metabolism (b = 0.75,
Student’s T-test, 95% CI = 0.067, p = 0.093). Because the biomass and
respiration data from the Terminos Lagoon food web (model 92,
Table A1) had a very low value in the dispersion, the standardised
residuals were analysed by identifying five food webs as potential
outliers (models 19, 62, 92, 114, 120 in Table A1). However, only the
Terminos Lagoon food web was excluded from our analysis because
it was the furthest point of the global trend. In this case, a slope of
0.753 (R2 = 0.708, n = 123) was estimated with no significant differ-
ences of 0.75 (Student’s T-test, 95% CI = 0.073, p = 0.474). From this
potential scaling relation between respiration and biomass (Fig. 1),
those food webs closer to the central tendency were chosen to com-
pare their supply–demand balance of energy. We considered at
least two models for each type of ecosystem, resulting in a total
of 16 networks (Table 1).
Each of the 16 food webs showed an oversupply in the energy
(6)
according to the mean values of the index (SDB > 0.75). The changes
in the biomass (increases and decreases) of the compartments
varied significantly from the mean SDB index (Fig. 2). Regarding
decrements, the SDB index decreased due to changes in the biomass
of the consumer groups at the top of the food web. For exam-
ple, in the oceanic systems, the Central Pacific Ocean food web
had the largest decrease because of the top predators, large sharks
(TL = 4.66) and Blue Marlin (TL = 4.61), followed by the small blue
 L.A. Salcido-Guevara, F. Arreguín-Sánchez / Ecological Modelling 276 (2014) 64–79 67

Fig. 2. Variation in the supply–demand balance index of energy (SDB) with the stochastic changes of biomass of the functional groups of marine and freshwater aquatic
ecosystems. Each whisker represents the confidence interval of the SDB mean per class (95% CI, n = 1000).

shark (TL = 3.98) and other high trophic level groups. In the Bali
Strait food web, the main higher trophic levels groups were Res
dolphins (TL = 3.91), sea birds (TL = 3.99) and medium pelagics
(TL = 3.66; Table 2).
Almost all of the increases in the mean values of the SDB index
were mainly due to changes in the biomass of the primary pro-
ducer groups (TL = 1). In only two food webs was the increase due
to consumer groups: in the Tongo Bay network, the groups were
zooplankton, small benthos <1 cm and suspended scallops; and for
the Sakumo Lagoon network, the groups were oysters and adult
tilapia. Half of the networks analysed (Maputo Bay, Bali Strait, Coast
of Sarawak, Celestun Lagoon, Lake George, Awassa Lake, Sri Lankan
Reservoir and Broa Reservoirs) had an increase due to the primary
producers. Of these networks, the last four belong to freshwater
systems and showed consistency because the increase in SDB was
due only to benthic producers.
For most food webs, the mean values of the SDB index for each
class of biomass change (10–100%) of the compartments were
inversely related to the trophic level of the groups (Table 3; see
example of extreme correlation in Fig. A1). In most of these classes,
with the exception of the Looe Keys and the Mexican Caribbean
(both in class 10%), as well as the Tongo Bay (classes 30, 80, 90
and 100%) food webs, a negative slope was obtained. In the Central
Pacific Ocean network, as well as in the Bali Strait, West Florida,
Coast of Sarawak, Lake George and Mexican Caribbean networks
(Table 3), a significant negative relationship was observed in all
classes (Fig. A1). In more than half of its classes, the Broa Reservoir
network had a negative and significant relationship. In contrast,
the Tongo Bay network, as well as the Looe Keys, Sakumo Lagoon,
Garonne River, Parana River and Awassa Lake (Table 3), had no sig-
nificant relationship in any of the classes (Fig. A1). There were a few
significant and negative relationships in the Maputo Bay, Celestun
Lagoon and Sri Lanka Reservoir networks.
When analysing the relationship between the SDB index and
the transfer efficiency of the functional groups, we did not find a
significant relationship in most food webs (Table A2). There were
68 L.A. Salcido-Guevara, F. Arreguín-Sánchez / Ecological Modelling 276 (2014) 64–79

Table 2
The trophic level of groups with lower mean values of the SDB index of 16 marine and freshwater food webs. The highest trophic level in the system are in brackets. Only the
SDB mean is obtained from the simulations.

Ecosystem Food web Functional group Trophic level SDB mean

Continental shelf West Florida Seabirds 4.23 1.0493

(4.72) Manatees 2 1.0494
Ichthyoplankton 2.94 1.0500
Coast of Sarawak Serranids 3.99 1.2000
(4.4) Small crustaceans 2.42 1.2013
Tuna 4.37 1.2019
Oceanic Central Pacific Ocean Large sharks 4.66 1.1815
(4.66) Blue marlin 4.61 1.1816
Small blue shark 3.98 1.1816
Bali Strait Res dolphins 3.91 1.0755
(3.99) Seabirds 3.99 1.0755
Medium pelagics 3.66 1.0757
Coral reef Looe Keys Sea turtles 2.79 1.0791
(4.29) Sharks/Rays 3.93 1.0808
Decomp/Microf 2 1.0808
Mexican Caribbean Birds 4.02 0.9712
(4.02) Sea turtles 2.95 0.9728
Sharks and Rays 3.63 0.9730
Bay Tongo Bay Bacteria 2 1.1410
(4.31) Birds 3.28 1.1433
Zooplankton 3.98 1.1460
Maputo Bay Zooplankton 2.02 1.1790
(3.49) Scombridae 3.12 1.1808
Sharks 3.49 1.1821
Coastal lagoon Celestun Lagoon Penaeus 2.3 1.0317
(3.44) Phytoplankton 1 1.0323
Polychaeta 2.18 1.0334
Sakumo Lagoon Clibanarius 2.35 1.3263
(3.36) Penaeids 2.31 1.3283
Misc benthos 2.05 1.3294
River Paraná River Loricariichthys platymetopon 2.3 1.0581
(4.02) Schizodon altoparanae 2 1.0586
Steindacchnerina insculpta 2.1 1.0587
Garone River Carp 2.4 1.2495
(3.66) Bars 3.66 1.2508
Phytoplankton 1 1.2509
Lake Awassa Lake Herbivore zooplankton 2.05 0.8104
(3.4) Carnivore zooplankton 2.72 0.8129
Phytoplankton 1 0.8130
Lake George Birds 3.17 1.0170
(3.17) Zooplankton 2.05 1.0191
O. leucostictus 2.01 1.0208
Reservoir Sri Lankan Reservoir ZP littoral 2.01 0.8774
(3.21) ZP open waters 2.01 0.8775
Fish eat. Birds 3.21 0.8797
Broa Reservoir Zooplankton 2 1.0160
(3.7) Hoplias malabaricus 3.7 1.0171
Gymnotus carapo 2.7 1.0173

only two exceptions. In the Mexican Caribbean food web, there
were significant positive slopes and correlations greater than 0.5.
In the Celestun Lagoon, there were significant negative slopes and
correlations around 0.5.
4. Discussion
The study of an ecosystem assumes the taxonomic resolution
as part of the logistics used, and the prevailing environmental
variabilities, both spatial and temporal, are inherent factors that
influence and determine the system status. The food webs ana-
lysed in this work reflect such factors with a diversity of species
in their compartments. Beyond the influence that the variability of
trophic levels might have on the dynamic and structure of the sys-
tem (Ecosystem Exploitation Hypothesis; Polis and Strong, 1996),
the specific composition of the compartments reflect several orders
of magnitude that would affect the mean size of the organisms in
each system (e.g., B/P sensu Christensen et al., 2005; Table 1). In
nature, the mean size of the organisms in a community increases
with the substitution of r strategists for k strategists (Odum, 1985).
In relation to the undifferentiated increase in the production and
biomass rates of an ecosystem (Odum, 1985).
The differences in the mean size of organisms in each food
web suggest that there are implications for the metabolism of
each ecosystem. Brown et al. (2007) have reported that the body
size of individuals affects the structure and dynamics of aquatic
ecosystems, indicating that the biological processes of organisms
are influenced by their metabolic rates, which, in turn, are strongly
affected by both body size and temperature. Considering these find-
ings, the 16 trophic models chosen to investigate the sensitivity of
the SDB index were distributed along the trend described by the
metabolism of 123 food webs, which document a potential scaling
relationship between the respiration flows and the total biomass
of the heterotrophic groups of the system. For this study, it was
assumed that respiration is a collective property (Salt, 1979; Odum
and Barrett, 2005), which allowed us to define metabolism by the
amount of total respiration of the heterotrophic groups of each food
web, thus reflecting the overall cost of ecosystem maintenance.
Odum (1985) described community metabolism through respi-
ration and production attributes to indicate the types of autotrophic
 L.A. Salcido-Guevara, F. Arreguín-Sánchez / Ecological Modelling 276 (2014) 64–79 69

Table 3
The relationship between the mean values of the supply–demand balance index (SDB) and the trophic level of 16 marine and freshwater aquatic food webs.

Change of ba 95% CI rb p ba 95% CI rb p

biomass (%)

West Florida (nc = 55) Coast of Sarawak (nc = 28)

10 −6.1 × 10−6 3.7 × 10−6 0.356 0.0038 −4.1 × 10−5 1.9 × 10−5 0.573 0.0007
20 −3.1 × 10−5 1.1 × 10−5 0.540 <0.0001 −5.9 × 10−5 5.2 × 10−5 0.350 0.0337
30 −6.6 × 10−5 2.0 × 10−5 0.615 <0.0001 −0.0003 0.0001 0.649 <0.0001
40 −0.0001 4.5 × 10−5 0.467 0.0002 −0.0006 0.0002 0.689 <0.0001
50 −0.0002 7.5 × 10−5 0.476 0.0001 −0.0008 0.0003 0.613 0.0003
60 −0.0003 9.7 × 10−5 0.530 <0.0001 −0.0011 0.0005 0.576 0.0007
70 −0.0003 9.7 × 10−5 0.530 <0.0001 −0.0011 0.0005 0.576 0.0007
80 −0.0005 0.0002 0.507 <0.0001 −0.0026 0.0010 0.652 <0.0001
90 −0.0008 0.0003 0.559 <0.0001 −0.0035 0.0014 0.642 0.0001
100 −0.0011 0.0004 0.519 <0.0001 −0.0046 0.0020 0.606 0.0003

Central Pacific Ocean (nc , d = 25) Bali Strait (nc = 13)

10 −2.2 × 10−5 1.2 × 10−5 0.536 0.0029 −6.3 × 10−5 3.6 × 10−5 0.690 0.0045
20 −9.4 × 10−5 2.3 × 10−5 0.828 <0.0001 −0.0002 9.8 × 10−5 0.661 0.0070
30 −0.0003 6.1 × 10−5 0.838 <0.0001 −0.0004 0.0002 0.714 0.0031
40 −0.0004 0.0001 0.790 <0.0001 −0.0009 0.0004 0.754 0.0015
50 −0.0006 0.0002 0.797 <0.0001 −0.0012 0.0006 0.725 0.0025
60 −0.0008 0.0002 0.764 <0.0001 −0.0016 0.0009 0.691 0.0045
70 −0.0008 0.0002 0.764 <0.0001 −0.0016 0.0009 0.691 0.0045
80 −0.0020 0.0005 0.832 <0.0001 −0.0038 0.0020 0.724 0.0026
90 −0.0026 0.0006 0.822 <0.0001 −0.0047 0.0023 0.749 0.0016
100 −0.0040 0.0010 0.822 <0.0001 −0.0064 0.0035 0.702 0.0037

Looe Keys (n c = 19) Mexican Caribbean (nc = 17)

10 5.4 × 10−8 2.6 × 10−5 0.001 0.4986 4.8 × 10−6 1.9 × 10−5 0.112 0.3341
20 −1.7 × 10−5 9.3 × 10−5 0.079 0.3746 −0.0001 8.2 × 10−5 0.534 0.0136
30 −9.1 × 10−5 0.0002 0.213 0.1910 −0.0002 0.0001 0.610 0.0047
40 −7.9 × 10−5 0.0003 0.117 0.3163 −0.0002 0.0002 0.462 0.0309
50 −0.0002 0.0004 0.250 0.1511 −0.0006 0.0004 0.632 0.0033
60 −0.0003 0.0006 0.206 0.1985 −0.0007 0.0006 0.499 0.0207
70 −0.0003 0.0006 0.206 0.1985 −0.0007 0.0006 0.499 0.0207
80 −0.0007 0.0013 0.207 0.1973 −0.0017 0.0011 0.577 0.0077
90 −0.0008 0.0016 0.195 0.2118 −0.0023 0.0014 0.576 0.0078
100 −0.0010 0.0023 0.188 0.2209 −0.0036 0.0022 0.583 0.0070

Tongo Bay (n c , d = 17) Maputo Bay (n c = 9)

10 −3.9 × 10−7 6.2 × 10−5 0.003 0.4956 −4.2 × 10−6 8.5 × 10−5 0.035 0.4645
20 −5.5 × 10−5 0.0001 0.240 0.1765 −0.0003 0.0004 0.476 0.0975
30 1.2 × 10−5 0.0004 0.014 0.4791 −0.0009 0.0009 0.576 0.0524
−5
40 −3.1 × 10 0.0008 0.017 0.4742 −0.0011 0.0012 0.541 0.0664
50 −2.5 × 10−5 0.0011 0.011 0.4840 −0.0020 0.0020 0.588 0.0480
60 −2.2 × 10−5 0.0017 0.006 0.4912 −0.0025 0.0036 0.445 0.1153
70 −2.2 × 10−5 0.0017 0.006 0.4912 −0.0025 0.0036 0.445 0.1153
80 2.5 × 10−5 0.0031 0.004 0.4945 −0.0052 0.0064 0.504 0.0835
90 3.8 × 10−5 0.0043 0.004 0.4940 −0.0072 0.0089 0.498 0.0861
100 0.0008 0.0057 0.066 0.4002 −0.0072 0.0115 0.411 0.1358

Celestun Lagoon (nc = 15) Sakumo Lagoon (nc = 12)

10 −9.7 × 10−7 3.5 × 10−5 0.014 0.4807 −5.3 × 10−5 0.0002 0.135 0.3378
20 −0.0001 0.0002 0.361 0.0933 −0.0002 0.0005 0.267 0.2005
30 −0.0005 0.0004 0.496 0.0302 −0.0004 0.0007 0.296 0.1751
40 −0.0007 0.0008 0.423 0.0583 −0.0009 0.0020 0.234 0.2317
50 −0.0010 0.0012 0.390 0.0752 −0.0018 0.0036 0.277 0.1918
60 −0.0018 0.0018 0.442 0.0493 −0.0024 0.0048 0.275 0.1931
70 −0.0018 0.0018 0.442 0.0493 −0.0024 0.0048 0.275 0.1931
80 −0.0034 0.0036 0.420 0.0594 −0.0039 0.0092 0.235 0.2315
90 −0.0058 0.0058 0.440 0.0506 −0.0051 0.0130 0.219 0.2469
100 −0.0076 0.0087 0.398 0.0710 −0.0034 0.0170 0.114 0.3627

Paraná River (nc = 39) Garone River (nc = 9)

10 −9.4 × 10−6 1.1 × 10−5 0.239 0.0712 −0.0001 0.0001 0.535 0.0691
20 −8.5 × 10−6 2.4 × 10−5 0.098 0.2757 −0.0003 0.0004 0.410 0.1367
30 −7.3 × 10−5 7.6 × 10−5 0.257 0.0571 −0.0006 0.0013 0.326 0.1957
40 −8.8 × 10−5 0.0001 0.215 0.0947 −0.0013 0.0022 0.396 0.1460
50 −0.0001 0.0002 0.205 0.1053 −0.0019 0.0027 0.450 0.1123
60 −0.0002 0.0003 0.219 0.0898 −0.0023 0.0039 0.383 0.1544
70 −0.0002 0.0003 0.219 0.0898 −0.0023 0.0039 0.383 0.1544
80 −0.0004 0.0006 0.173 0.1463 −0.0053 0.0092 0.382 0.1553
90 −0.0006 0.0008 0.195 0.1176 −0.0077 0.0116 0.431 0.1236
100 −0.0007 0.0012 0.168 0.1533 −0.0065 0.0143 0.309 0.2091

Awassa Lake (nc = 13) Lake George (nc = 13)

10 −6.2 × 10−7 2.7 × 10−5 0.012 0.4841 −0.0001 8.7 × 10−5 0.659 0.0071
20 −0.0001 0.0002 0.374 0.1039 −0.0002 0.0001 0.579 0.0190
30 −0.0005 0.0006 0.427 0.0730 −0.0003 0.0003 0.520 0.0342
40 −0.0004 0.0006 0.315 0.1473 −0.0009 0.0005 0.734 0.0021
50 −0.0011 0.0011 0.459 0.0571 −0.0014 0.0009 0.656 0.0074
70 L.A. Salcido-Guevara, F. Arreguín-Sánchez / Ecological Modelling 276 (2014) 64–79

Table 3 (Continued)

Change of ba 95% CI rb p ba 95% CI rb p

biomass (%)

60 −0.0017 0.0019 0.435 0.0689 −0.0018 0.0012 0.615 0.0126

70 −0.0017 0.0019 0.435 0.0689 −0.0018 0.0012 0.615 0.0126
80 −0.0030 0.0038 0.394 0.0912 −0.0050 0.0028 0.692 0.0044
90 −0.0039 0.0054 0.365 0.1099 −0.0070 0.0047 0.628 0.0108
100 −0.0060 0.0085 0.358 0.1152 −0.0066 0.0065 0.483 0.0473

Sri Lankan Reservoir (nc = 16) Broa Reservoir (nc = 10)

10 −7.8 × 10−5 5.9 × 10−5 0.530 0.0174 −1.6 × 10−5 2.2 × 10−5 0.444 0.0996
20 −0.0002 0.0002 0.370 0.0793 −0.0003 0.0002 0.597 0.0343
30 −0.0005 0.0005 0.425 0.0503 −0.0005 0.0006 0.524 0.0601
40 −0.0007 0.0008 0.360 0.0853 −0.0008 0.0008 0.578 0.0401
50 −0.0015 0.0014 0.453 0.0388 −0.0015 0.0014 0.564 0.0448
60 −0.0016 0.0020 0.343 0.0970 −0.0022 0.0021 0.584 0.0380
70 −0.0016 0.0020 0.343 0.0970 −0.0022 0.0021 0.584 0.0380
80 −0.0038 0.0038 0.426 0.0498 −0.0039 0.0034 0.595 0.0349
90 −0.0040 0.0055 0.320 0.1133 −0.0058 0.0063 0.519 0.0623
100 −0.0060 0.0073 0.359 0.0857 −0.0079 0.0090 0.497 0.0719
a
The value b is the slope of simple linear regression model with 95% confidence interval and the probability value p of Student’s t-test (p < 0.05 is significant)
b
r is the Pearson’s correlation coefficient.
c
n is the number of functional groups for each kind of biomass change.
d
See example of extreme correlation in Fig. A1.

and heterotrophic succession experienced by an ecosystem. In our
case, the ecosystem metabolism is represented by the respira-
tion flows of heterotrophic groups, as the groups of detritus and
the primary producers were not included. In this experiment, the
metabolism of the aquatic ecosystems were analysed according to
the relationship between the flows of respiration and the biomass
of the heterotrophic groups of 123 food webs. The results yielded
an exponent of 0.753, which obeys the 3/4 allometric scaling rule
described for species metabolism (Kleiber, 1932; Hemmingsen,
1960).
SDB has a foundation in the theorem of Banavar et al. (1999),
from which a general relationship between the size and flow rates is
derived from an arbitrary network that, like a resource transporta-
tion system, describes an efficiency. Considering an n-dimensional
Euclidean space in which they both operate, a physical network
(e.g., river system) is analogous to a biological network (e.g., vas-
cular system), and it is demonstrated that the system efficiency
depends on the network geometry and that a metabolic rate pro-
vides the resources by scaling the size M to a power of 2/3 for a
physical network (two-dimensional. D = 2) and 3/4 for a biological
network (D = 3) with the form B ∝ MD/(D+1) . In a later work, Banavar
et al. (2002) explain in detail the importance of the supply and
demand rates that serve a particular region of the transport net-
work and show how they affect the interspecific 3/4 metabolic
scaling originally described by Kleiber (1932).
Bendoricchio and Palmeri (2005) analysed 33 aquatic ecosys-
tems that generally displayed values of SDB > 0.75. SDB or ˛ is the
scaling power of the energy flows with the biomass of the compart-
ments in a food web. The 3/4 exponent (=0.75) is a scaling property
of the direct networks of optimal transport, as expressed by a per-
fect balance in the rates of supply and demand flows (Banavar et al.,
2002). In this sense, as well as in the development of an organism,
it could be implied that a waste of resources is characterised by a
supply rate that is greater than a demand rate (Weibel et al., 1998
sensu Banavar et al., 2002). The values of SDB > 0.75 suggest an over-
supply of resources, which is characteristic of aquatic ecosystems
with an unlimited energy source from the sun (without down-
playing the nutrient limitation) and is provided through primary
producers.
The existing description of the food webs analysed in this study
might suggest a control mechanism for similar ecosystems. For
example, temporal simulations of the biomass changes of the func-
tional groups of an ecosystem in the great extension of the Pacific
Ocean (Cox et al., 2002) demonstrate the decreasing effect of
predation on some groups due to the decrease of top predators, thus
suggesting some sensitivity in predation mortality rates, a type of
top-down control if we assume increases in the biomass of prey
resources (Cox et al., 2002). However, other oceanic systems of the
same type suggest a different control (wasp-waist Buchary et al.,
2002). Although robust evidence is required to assume such control
mechanisms, the search for a pattern in the same type of system is
still difficult because of factors that affect their comparison, such as
environmental variability (space–time) and the level of taxonomic
resolution in the studies.
The food webs were selected considering their metabolism (res-
piration) relative to its size (biomass), seeking to reduce the effect to
the environmental variability implicit in the simulated food webs.
We assume that the activity of an organism can be characterised
within the limits of its basal and maximum metabolism (sensu
Peters, 1983). It is then assumed that ecosystems represented by
food webs exhibit a greater and lesser activity above and below
the potential trend described by the respiration and biomass of the
heterotrophic groups. If the community respiration may represents
the maintenance cost of the ecosystem (Odum, 1985), then the res-
piration flows of the heterotrophic groups could be an attribute
of the ecosystem’s metabolism. Moreover, an analogy would exist
between the biomass and the ecological 3/4 rule (Kleiber, 1932) or
with the interspecific metabolic scaling described for the species
(Hemmingsen, 1960), where the respiration of organisms of dif-
ferent species’ potentially scaled body mass can be determined
through the use of an allometric exponent (3/4) that remains
unchanged by several orders of magnitude. In this sense, if the
selected food webs show a state of activity or metabolism rela-
tive to its size, then it can be concluded that the changes in the
energy balance are largely explained by the system’s own attributes
(those that implicitly express their physicochemical environment)
rather than by the influence of its taxonomic resolution, as in the
construction of models.
Many factors affect the abundance of community species such
as limiting nutrients, solar radiation and temperature influence
mainly on the production rates of autotrophic organisms. The phe-
nomenon of density dependence is reflected in the rates of natural
mortality and growth of populations in all ecosystem trophic lev-
els (Walters et al., 2000; Begon et al., 2006). Trophic relationships
are a key factor in the regulation of the community; strong con-
sumptions leads to alternations of high and low biomass between
successive levels (Polis and Strong, 1996). This can be modelled
according to predation vulnerability exhibited by preys, which
 L.A. Salcido-Guevara, F. Arreguín-Sánchez / Ecological Modelling 276 (2014) 64–79
allows simulations of temporal changes in abundance of func-
tional groups of an ecosystem (Walters et al., 2000). Some groups
may have changes in abundance close to 100% of their biomass,
for example, whales (Mysticetae; Ainsworth et al., 2002) and fish
species such as greenland halibut and polar cod (Blanchard et al.,
2002). Smaller species such as phytoplankton and zooplankton can
have large interannual fluctuations (Begon et al., 2006). In this
study, we obtained magnitudes within a range of 100% change
in the biomass of functional groups, also incorporating a factor
stochastic. The equation used for this purpose (Eq. (6)) is not a pre-
diction model, it does not incorporate any adjustment factor like
an, environmental forcing such as temperature or nutrient uptake.
The equation only allows the resampling of the original data of each
Ecopath trophic model (static) and to obtain random samples of the
biomass data of each group. This makes possible to analyse the sen-
sitivity of the index SDB to changes in biomass, and to have a rough
idea of the importance of each species or group in the system.
In all of the food webs, the experiments of random changes in the
biomass of the compartments generated increases and decreases
in the oversupply of the system resources. For example, in the
SDB index of the Celestun Lagoon network (Chávez et al., 1993),
the mean value obtained for each class of change of biomass was
higher than reported by Bendoricchio and Palmeri (2005) for the
same system as a steady state (SDB = 0.82). These authors found an
energy oversupply in most food webs, noting that this is due to an
intrinsic property of these systems having a feedback mechanism
(high detritus flow) that becomes a highly indirect network. The
SDB index is sensitive to seasonal succession as a measure of system
organisation, and it might reflect changes in structure and function
(Bendoricchio and Palmeri, 2005). In the analysed food webs, the
biomass, as part of the system structure, changed independently of
its magnitude and was translated into significant variations of SDB.
Chávez et al. (1993) demonstrated the importance of the detritus
group in respect to phytoplankton in terms of their positive trophic
impacts on other groups, suggesting the detritus group’s impor-
tance as an available source of energy for groups of lower trophic
levels. Groups such as polychaetes and penaeids demonstrate that
detritus can cause the greatest decline in the energy supply of the
lagoon system, but phytoplankton has the same effect, and it does
so with the same intensity, even when its trophic impacts are not
as relevant. As a holistic indicator of importance of the group in
terms of the availability of system resources that are independent
of biomass, this is an interesting feature of SDB. However, it was
observed in all food webs that the major declines in the SDB index
were related to groups with the lowest biomass, which shows its
importance for the system. These kinds of groups and their trophic
impacts are considered an important attribute in identifying key-
stone species in the system (Libralato et al., 2006).
According to Polis and Strong (1996) the diversity of consumers
and resources affects the dynamics of the network, allowing for the
propagation of the effects of consumption and productivity rather
than for a concentration on a particular trophic level. However, a
holistic indicator such as SDB allows for an evaluation of the effect of
each element of the system and relates this evaluation to a particu-
lar trophic level, allowing us to understand the importance of each
functional group through a measure with statistical significance.
All food webs showed significant variations in the oversupply of
energy and were higher in energy when the biomass increased. In
particular, the increase in the oversupply was related to the trophic
level of primary producers, with the exception of two food webs,
in which the oversupply was due to different groups with lower
trophic levels. The behaviour experienced by the system when the
biomass of primary trophic groups is affected suggests a bottom-
up control. However, most consumer groups, especially the higher
trophic levels, decreased the supply of energy. In these situations,
the behaviour suggests a top-down control in which predation can
71
significantly affect the availability of system resources. Consider-
ing that the same system has the capacity to experience different
types of control, this would suggest that efficient groups such as
the primary producers favoured an oversupply of energy in the sys-
tem because the high production rates (being independent of the
density-dependent factor) affected the heterotrophic (food-limited
sensu Hairston et al., 1960) groups; they also responded with a
greater availability of resources than secondary production due to
a high subsidy of (unlimited) energy supplied by solar radiation
(Odum, 1985; Begon et al., 2006).
Polis and Strong (1996) also indicate that the increase in the
number of consumers could reduce the resources for lower levels
if the donor control itself were absent. This is consistent with the
results obtained through the comparison between the magnitude of
the increase and decrease in the oversupply of energy, where con-
sumers exerted the greatest decrease and the primary producers
exerted the greatest increase. These variations in the oversupply
could be associated with a type of donor-controlled mechanism
(sensu DeAngelis, 1980) identified as a bottom-up control, when an
increase in the oversupply promotes the availability of resources
in the system in such a way that the density of predators does
not decrease the prey because there is a prey density with a low
dependence on predators. In this sense, the food webs with con-
sumer groups exercising a greater decrease in SDB could promote a
top-down control, as suggested by Hairston et al. (1960), in which
predators that control herbivore populations should therefore limit
their own resources with a food-limited group.
Not all consumer groups caused a decrease in the oversupply of
energy, just as not all primary producers caused an increase. For
example, for the West Florida food web (Okey et al., 2004), the
only consumer group that caused an increase in oversupply was
sessile epibenthos, which has the largest biomass on the system
after detritus. The benthic producers sea grass and macroalgae also
increased oversupply. These groups represent 80% of the biomass
of primary producers and have an important effect on the sys-
tem because, in addition to microphytobenthos (which increased
slightly the oversupply), they constitute 91% of the biomass of
primary producers, whose mortality causes a reduction in the
biomass of most groups (Table 2 in Okey et al., 2004). Adversely,
the phytoplankton group caused a decrease in oversupply. This
behaviour was also observed in the freshwater ecosystems, which
consistently increased the oversupply of resources with respect
to benthic producers and decreased the oversupply of resources
with pelagic producers (phytoplankton groups). In relation to these
groups with declining benthic production (the microphytobenthos,
sea grasses and macroalgae), Okey et al. (2004) discuss the impor-
tance of shading by phytoplankton blooms. This idea is supported
by empirical studies (e.g., Okey et al., 2004) and suggests an inverse
relationship between the behaviour of the SDB index of benthic
and pelagic producers, particularly for relatively closed systems.
While oversupply is increased by benthic producers, which could
favour the availability of resources in the system, an oversupply is
decreased due to the changes in phytoplankton biomass.
Hairston et al. (1960) conclude that the limitation of resources
for species at different trophic levels indicates the difficulty
for producers such as carnivores and decomposers in avoiding
competition at a particular level. Functional groups with major
omnivorous behaviour consume prey of different trophic levels
(sensu Pimm, 1982), and are mainly observed in the middle of
the trophic levels (Arreguín-Sánchez et al., 2002; Zetina-Rejón
et al., 2003; Salcido-Guevara and Arreguín-Sánchez, 2007). In
this sense, being omnivorous allowed the system to distribute
energy in a greater number of ways, which in terms of complex-
ity (see, e.g., MacArthur, 1955), provides greater stability. Unlike
Pimm (1979), who does not find any evidence in favour of the
MacArthur’s hypothesis (1955), we considered low variations in
72 L.A. Salcido-Guevara, F. Arreguín-Sánchez / Ecological Modelling 276 (2014) 64–79
the supply–demand balance of energy by most omnivorous groups
(where there are more ways to obtain resources) to be evidence
that they might contribute to the stability of the system, particu-
larly through biomass changes of higher trophic level groups, which
cause the largest decrease in energy supply. This phenomenon was
observed in marine and freshwater food webs, with the exception
of coastal lagoons, where the fall in the energy supply is related to
the groups of low trophic levels (phytoplankton, penaeid, hermit
crabs and polychaetes). This finding suggests that those systems
with a greater number of compartments, which have lower rates
of omnivores and a high energy demand (higher prey resource
dependence) or are resource-limited (Hairston et al., 1960), could
reflect less efficiency in the system’s ability to respond to envi-
ronmental and anthropogenic disturbances. This possibility was
also suggested by Ulanowicz (2009) when describing the auto-
organisation capacity of ecosystems.
We explored a linear relationship between the SDB mean index
and trophic level of each of the 16 selected food webs, and almost
half of the networks there were significant cases. We also analysed
the relationship between SDB and transfer efficiency and only two
cases were significant. Christensen and Pauly (1993) reported a
nonlinear relationship between trophic level and transfer efficiency
in several Ecopath trophic models, showing the highest efficiency
in the discrete third trophic level (TL = III). The trend described by
the authors could indicate that biomass changes in the third trophic
level groups could exert a greater effect on the mean transfer effi-
ciency of the system, and therefore in the availability of resources.
However, the experiments of biomass changes suggest that both
top predators at different trophic levels (TL = III and TL = IV) and
secondary consumers (TL = II) can exert an important influence on
the availability of resources, that caused the largest reductions in
energy supply (Table 2) and was more frequently by groups of
the second trophic level. Not observed a clear trend of transfer
efficiency in relation to changes in the SDB index, such attribute
may respond to the system dynamic, and describe a clear trend
to be explored by the dynamic version of Ecopath model (Ecosim;
Walters et al., 2000).
5. Conclusions
The scope of this work allows us to conclude that the weight of
the compartments in the food web interactions is important and
that this importance can be expressed in the general behaviour
of the energy balance of the system. Just as the flow behaviour
in the basal groups (primary producers) is shown to increase the
availability of resources in a system, the other components (con-
sumers) are shown to decrease. Consequently, a general pattern
representing the inverse relationship between trophic levels with
respect to the SDB index emerges, particularly with an oversupply
of energy that has the greatest decline with the functional groups
with lower abundance (biomass). It is important to recognise that
the results of this study are the product of an analysis performed
on static trophic models (1st limitation of the study) in which the
biomass of each compartment was modified by keeping the other
groups unchanged. Although they were not rebalanced the trophic
model with each change of biomass of the groups (2nd limita-
tion of the study), we assume that low biomass changes should
reflect the behaviour of SDB in balanced models. Therefore, a fur-
ther study that combined temporary simulations over the whole
structure with the functioning of food webs analysed in this paper
could provide useful information that would strengthen our con-
clusions.
Acknowledgements
We thank the National Polytechnic Institute for the support
granted through the PIFI programme and projects SPI-20113417
and 20110785. We also thank the National Council of Sci-
ence and Technology (CONACYT CVU No. 162337) and the
support granted through projects SEP-CONACyT 104974 and
ANR-CONACyT 111465. We also thank the Mexican Network of
Integrated Management of Marine Coastal areas. Finally, special
thanks go to Dr. Just T. Bayle Sempere for his invaluable com-
ments.
 L.A. Salcido-Guevara, F. Arreguín-Sánchez / Ecological Modelling 276 (2014) 64–79 73

Appendix A.

Fig. A1. The relationship between the mean values of the supply–demand balance index (SDB) and trophic level. The ocean food web shows significant changes in SDB for
all cases (Student’s t-test; p < 0.05; see Table 3). The dots represent each compartment with 1000 random changes around its original biomass (percentage values).
 74
Table A1
Ecopath trophic models used in the analysis of aquatic ecosystems around the world.

No Trophic model Groups Approximate Country Study area (km2 ) Period (year) Bi Ma Ca Bo Cr Mo Zo Ph Im Ba De Fi
location
(longitude,
latitude)

Continental Shelves
1 Alaska Gyre 25 −144.976 55.424 Canada – 1980 × × × × × × × × × × ×
2 Eastern Bering Sea 1950 25 −168.500 58.100 Canada 484,508 1950 × × × × × × × × × × ×
3 Eastern Bering Sea 1980 25 168.000 57.000 Canada 484,508 1980 × × × × × × × × × × ×
4 Brunei Darussalam 13 114.600 4.990 Philippines 7396 1989–1990 × × × × × × × × ×
5 Northern British Columbia 1750 53 −131.210 53.670 Canada 70,000 1750–1750 × × × × × × × × × × ×
6 Northern British Columbia 1900 53 −131.210 53.670 Canada 70,000 1900–1900 × × × × × × × × × × ×
7 Northern British Columbia 1950 53 −131.210 53.670 Canada 70,000 1950–1950 × × × × × × × × × × ×
8 Northern British Columbia 2000 53 −131.210 53.670 Canada 70,000 2000–2000 × × × × × × × × × × ×
9 Rivers Inlet 1950 32 −128.167 51.383 Canada 25 1950–1950 × × × × × × × × ×

L.A. Salcido-Guevara, F. Arreguín-Sánchez / Ecological Modelling 276 (2014) 64–79

10 Rivers Inlet 1990 32 −128.600 51.430 Canada 25 1990–1990 × × × × × × × × ×
11 Northern Gulf of St. Lawrence 32 −63.000 48.500 Canada – 1985–1987 × × × × × × × × × × ×
12 Central Chile 1992 21 −73.165 −35.666 Chile 50,042 1992–1992 × × × × × × × ×
13 Central Chile 1998 21 −73.165 −35.666 Chile 50,042 1998–1998 × × × × × × × ×
14 Upwelling Gulf of Salamanca 18 −74.419 11.069 Colombia – 1993–1998 × × × × × × × × ×
15 Gulf of Nicoya 21 −84.931 9.986 Costa Rica 1530 – × × × × × × × × × × ×
16 Golfo Dulce 21 −83.265 8.586 Costa Rica 750 1993–1994 × × × × × × × × × × ×
17 Weat Greenland 12 −53.040 64.260 Greenland 63,500 1991–1992 × × × × × × × ×
18 Iceland’s fisheries 21 −19.220 66.730 Iceland 758,000 1997–1997 × × × × × × × × ×
19 Iceland 1950 24 −19.220 66.730 Iceland 376,766 1950–1950 × × × × × × × × × ×
20 Terengga 13 103.164 5.345 Malaysia 9656 1984–1985 × × × × × × × × × ×
21 Central Gulf of California 27 −110.890 27.600 México 27,900 1978–1979 × × × × × × × × × × ×
22 Campeche Sound 19 −92.450 18.980 México 65,000 1988–1994 × × × × × × × × ×
23 Upper Gulf de California 29 −114.470 31.350 México 7200 – × × × × × × × × × ×
24 North Sea 29 4.000 56.000 – 570,000 1981–1981 × × × × × × × ×
25 South Orkneys/Georgia 30 −50.670 −60.940 British Territories – 1990–1990 × × × × × × × × ×
26 Cantabrian Sea 28 −6.000 44.000 Spain 16,000 1994–1994 × × × × × × × × ×
27 Gulf of Paria 24 −62.110 10.380 Trinidad 7600 1945–1945 × × × × × × × × × ×
28 English Channel 1973 45 −2.016 50.149 England 89,607 1973–1973 × × × × × × × × ×
29 English Channel 1995 50 −2.016 50.149 England 89,607 1995–1995 × × × × × × × × × ×
30 West Florida Shelf 59 −83.250 27.250 United States 170,000 – × × × × × × × × × × ×
31 Eastern Bering Sea 38 −164.880 56.240 United States 485,000 1980–1985 × × × × × × × × × × ×
32 Western Bering Sea 36 −178.600 64.400 United States 254,000 1980–1985 × × × × × × × × × × ×
33 Prince William Sound 48 −147.020 60.610 Alaska 9000 1994–1996 × × × × × × × × × × ×
34 Prince William Sound (old mod.) 19 −147.020 60.610 United States 8800 1980–1989 × × × × × × × × ×
35 Venezuela Shelf 16 −63.300 11.100 Venezuela 30,000 1988–1988 × × × × × × × ×
36 Campeche Sound 25 −92.450 18.980 México 65,000 1988–1994 × × × × × × × × × ×
37 Southern Brazil 13 −48.198 −28.848 Brazil 28,661 1950–1994 × × × × × × × ×
38 North Sea 1880 46 4.000 56.000 – 570,000 1980–1980 × × × × × × × × × × ×
39 Barents Sea 1990 41 40.210 75.130 – 1,400,000 1990–1990 × × × × × × × × × × ×
40 Barents Sea 1990 (includ. juv.) 16 40.210 75.130 – – 1990–1990 × × × × × × × × × ×
41 Barents Sea 1995 41 40.210 75.130 – 1,400,000 1995–1995 × × × × × × × × × × ×
42 Barents Sea 1995 (includ. juv.) 16 40.210 75.130 – – 1995–1995 × × × × × × × × × ×
43 West Scotland 37 −2.016 50.149 England 31,085 2000–2004 × × × × × × × × × × ×
44 Western English Chanel 52 −3.960 49.540 United Kingdom 56,452 1994–1994 × × × × × × × × × ×
45 Southwest Coast of India 11 75.630 10.600 India 75,390 1994–1996 × × × × × ×
46 North Coast of Central Java 27 109.000 −6.500 Indonesia – 1979–1979 × × × × × × × × ×
47 Peninsula Malaysia 15 100.400 4.200 Malaysia 37,781 1987–1991 × × × × × × × ×
48 Coast of Sarawak 29 112.500 3.260 Malaysia 76,668 1972–1972 × × × × × × × × ×
49 Coast of Sabah 29 116.120 6.320 Malaysia 20,209 1972–1972 × × × × × × × × ×
50 Strait of Georgia 27 −122.938 48.866 – 7000 1950–1950 × × × × × × × × × ×
51 South Brazil Bight 25 −46.639 −24.874 Brazil 97,000 1998–1999 × × × × × × × × × × ×
52 Jalisco and Colima Coast 38 −104.594 19.124 México 700 1995–1996 × × × × × × × × × × ×
Table A1 (Continued)

No Trophic model Groups Approximate Country Study area (km2 ) Period (year) Bi Ma Ca Bo Cr Mo Zo Ph Im Ba De Fi
location
(longitude,
latitude)

53 Perú 50 20 −79.514 −8.406 Perú – 1953–1959 × × × × × × × ×

54 Perú 60 20 −79.514 −8.406 Perú – 1960–1969 × × × × × × × ×
55 Gulf of Mexico 15 −96.700 25.957 – – – × × × × × × × ×
56 Southern Gulf of California 37 −106.580 23.330 México 6250 1994–1997 × × × × × × × × × ×

Lakes
57 Tanganyica Lake, Burundi 7 29.317 −3.883 Africa 800 1980–1983 × × × × ×
58 Ontario Lake 14 −77.661 43.630 Canada 19,500 1991–1991 × × × × × ×
59 Awassa Lake 14 38.443 7.064 Ethiopia 90 2003–2004 × × × × ×
60 Aydat Lake 11 2.986 45.664 France – 1990–1992 × × × × × ×
61 Veli Lake 14 76.888 8.510 India – – × × × × × × ×

L.A. Salcido-Guevara, F. Arreguín-Sánchez / Ecological Modelling 276 (2014) 64–79

62 Kinneret Lake 14 35.592 32.809 Israel 168 – × × × × × × ×
63 Turkana Lake 1973 8 36.060 3.500 Kenya 7560 1973–1987 × × × × ×
64 Turkana Lake 1987 8 36.060 3.500 Kenya 7560 1973–1987 × × × × ×
65 Victoria Lake 16 34.042 −0.288 Kenya 70,000 1971–1972 × × × × × × ×
66 Malawi Lake 9 34.420 −11.941 Malawi 30,800 1977–1981 × × × ×
67 Laguna de Bay 1820 30 121.250 14.250 Philippines 900 1820–1820 × × × × × × × × × ×
68 Laguna de Bay 1920 26 121.250 14.250 Philippines 900 1920–1920 × × × × × × × × ×
69 Laguna de Bay 1950 21 121.250 14.250 Philippines 900 1950–1950 × × × × × × × × ×
70 Laguna de Bay 1968 16 121.250 14.250 Philippines 900 1968–1968 × × × × × × ×
71 Laguna de Bay 1980 17 121.250 14.250 Philippines 900 1980–1980 × × × × × × ×
72 Laguna de Bay 1990 20 121.250 14.250 Philippines 900 1990–1990 × × × × × × × ×
73 George Lake 14 30.200 0.000 Uganda 250 1970–1970 × × × × × ×
74 Chad Lake 15 14.100 13.330 Chad 25,000 1969–1972 × × × × × × × ×
75 Kariba Lake 10 28.554 −16.563 Zimbabwe 5364 1980–1980 × × × × × ×
76 Malawi Lake 26 34.420 −11.941 Africa 28,800 1977–1996 × × × × × ×
77 Victoria Lake 18 34.042 −0.288 – – – × × × × × × ×
78 Tanga 75 7 29.317 −3.883 – – 1975–1975 × × × ×
79 Victoria Lake 85 16 34.042 −0.288 Kenya 70,000 1985–1986 × × × × × ×

Coastal lagoons
80 Caeté Estuary 19 −46.621 −0.932 Brazil 220 – × × × × × × × × × ×
81 Maspalomas Lagoon 7 −15.594 27.738 Spain – 1993–1994 × × ×
82 Etang de Thau 11 3.617 43.404 France 88 1980–1980 × × × × × ×
83 Sakumo Lagoon 13 0.013 5.630 Ghana 1 1971–1971 × × × × × × ×
84 Orbetello Lagoon 12 11.214 42.441 Italy 27 1995–1996 × × × × × ×
85 Huizache-Caimanero Lagoon 26 −106.070 22.960 México 175 1984–1986 × × × × × × × ×
86 Terminos Lagoon 20 −91.500 18.600 México 2500 – × × × × × × × ×
87 Mandinga Lagoon 20 −96.070 19.010 México – 1982–1983 × × × × × × ×
88 Tampamachoco Lagoon 23 −97.356 21.007 México 15 – × × × × × × ×
89 Celestun Lagoon 16 −90.393 20.816 México 28 – × × × × × × × ×
90 Mangrove in Celestun Lagoon 19 −90.393 20.816 México – – × × × × × × ×
91 Sandy Barrier Lagoon Chiku 13 120.074 23.137 Taiwan 10 – × × × × × × × ×
92 Terminos Lagoon (seag./mang.) 16 −91.500 18.600 México – – × × × × × ×

Bays
93 Tongo Bay 17 −71.548 −30.283 Chile 60 1978–1989 × × × × × × × × ×
94 Bohai Sea 13 120.000 39.000 China 77,000 1982–1983 × × × × × × ×
95 Bay of Somme 9 1.530 50.227 France 50 1998–1998 × × × × × × ×
96 La Paz Bay 22 −110.500 24.500 México – 1992–1998 × × × × × × ×
97 Maputo Bay 10 32.766 −26.024 Mozambique 1100 1972–1985 × × × × × × × × ×
98 Kuosheng Bay 17 121.667 25.217 Taiwan 8 1998–2001 × × × × × × × ×
99 Monterey Bay 16 −122.000 36.780 United States – – × × × × × × × × × × ×
100 San Miguel Bay 16 123.180 13.860 Philippines 1115 1992–1994 × × × × × × × × ×

75
 76
Table A1 (Continued)

No Trophic model Groups Approximate Country Study area (km2 ) Period (year) Bi Ma Ca Bo Cr Mo Zo Ph Im Ba De Fi
location
(longitude,
latitude)

101 Ascencion Bay 19 −87.580 19.650 México 260 – × × × × × × × × × ×

L.A. Salcido-Guevara, F. Arreguín-Sánchez / Ecological Modelling 276 (2014) 64–79

102 Bay of Bengal 15 91.000 21.000 Bangladesh 166,000 1984–1986 × × × × × × × ×
103 San Pedro Bay 16 125.110 11.160 Philippines 625 1994–1995 × × × × × × × × ×

Oceans
104 Baltic Sea 16 20.499 58.949 – – 1974–2000 × × × × × × ×
105 Newfoundland 1450 50 −53.250 52.520 Canada 495,000 1450–1450 × × × × × × × × × ×
106 Newfoundland 1900 50 −53.250 52.520 Canada 495,000 1900–1900 × × × × × × × × × ×
107 Newfoundland 1995 50 −53.250 52.520 Canada 495,000 1985–1985 × × × × × × × × × × ×
108 Central Pacific Ocean 26 170.000 20.000 – – 1990–1998 × × × × ×
109 Central Pacific (sharks) 22 170.000 37.500 – 30,000 1990–1998 × × ×
110 Bali Strait 14 114.760 −8.570 – – – × × × × × × × ×

Coral reefs
111 Central Great Barrier Reef 25 147.000 −18.000 Australia – – × × × × × × × ×
112 Northern Great Barrier Reef 25 143.447 −11.561 Australia 10,000 1993–1994 × × × × × × × × ×
113 Barrier Reef of Tiahura 43 −149.901 −17.486 French Polynesia – 1971–1989 × × × × × × ×
114 Fringing Reef of Tiahura 46 −149.901 −17.490 French Polynesia – 1971–1989 × × × × × × × ×
115 Bolinao Coral Reef 26 119.917 16.436 Philippines 240 – × × × × × ×
116 Looe Keys Coral Reef 20 −81.400 24.655 United States 30 – × × × × × × × ×
117 Coral Reef Mexican Caribbean 18 −87.240 19.300 México – 1990–1990 × × × × × × × × × ×
118 Virgin Islands 21 −73.000 20.000 – – – × × × × × × × × ×

Reservoirs
119 Broa Reservoir 11 −47.888 −22.192 Brazil 6 – × × × × ×
120 Sri Lankan Reservoir 17 80.982 7.909 Sri Lanka 25 1970–1970 × × × × × ×
121 Ria Formosa Lagoon 14 −7.804 37.033 Portugal – 1996–1997 × × × × × × ×

Rivers
122 Garonne River 10 1.319 43.761 France 5 1990–1991 × × × × × × ×
123 Paraná River Floodplain 40 −52.840 −22.360 Brazil – 1992–1995 × × × × ×

Island
124 Aleutians Islands 1963 40 −179.779 51.577 Alaska 56,936 1963–1963 × × × × × × × × × × ×

Bi, birds; Ma, mammals; Ca, cartilaginous fishes; Bo, bony fishes; Cr, crustaceans; Mo, molluscs; Zo, zooplankton; Ph, phytoplankton; Im, Infauna/meiofauna; Ba, bacteria; De, detritus; Fi, fishery. More detail on the models on
http://www.ecopath.org and http://www.incofish.com.
 L.A. Salcido-Guevara, F. Arreguín-Sánchez / Ecological Modelling 276 (2014) 64–79 77

Table A2
The relationship between the mean values of the supply–demand balance index (SDB) and the transfer efficiency of 16 marine and freshwater aquatic food webs.

Change of ba 95% CI rb p ba 95% CI rb p

biomass (%)

West Florida (nc = 51) Coast of Sarawak (nc = 26)

10 7.4 × 10−6 3.3 × 10−5 0.053 0.356 0.0001 0.0002 0.177 0.194
20 0.0001 0.0001 0.187 0.094 0.0001 0.0006 0.064 0.378
30 3.1 × 10−5 0.0002 0.035 0.404 −0.0005 0.0010 0.156 0.224
40 0.0002 0.0005 0.092 0.261 −1.3 × 10−5 0.0022 0.002 0.496
50 0.0004 0.0007 0.130 0.182 −0.0002 0.0033 0.025 0.453
60 0.0004 0.0010 0.086 0.274 −0.0006 0.0054 0.039 0.424
70 0.0004 0.0010 0.086 0.274 −0.0006 0.0054 0.039 0.424
80 0.0011 0.0021 0.120 0.201 −0.0012 0.0090 0.046 0.412
90 0.0012 0.0028 0.102 0.237 2.3 × 10−5 0.0144 0.001 0.499
100 0.0015 0.0043 0.082 0.284 −0.0044 0.0193 0.080 0.349

Central Pacific Ocean (nc = 24) Bali Strait (nc = 12)

10 0.0001 0.0003 0.076 0.361 −2.8 × 10−6 0.0002 0.007 0.492
20 4.1 × 10−5 0.0008 0.019 0.465 0.0005 0.0010 0.277 0.191
30 −1.4 × 10−5 0.0021 0.003 0.495 0.0013 0.0020 0.347 0.134
40 −0.0003 0.0034 0.034 0.438 0.0023 0.0036 0.336 0.143
50 −0.0002 0.0051 0.014 0.474 0.0025 0.0057 0.241 0.225
60 0.0007 0.0074 0.036 0.433 0.0026 0.0078 0.190 0.277
70 0.0007 0.0074 0.036 0.433 0.0026 0.0078 0.190 0.277
80 −0.0008 0.0159 0.019 0.465 0.0044 0.0153 0.162 0.308
90 0.0031 0.0219 0.052 0.404 0.0106 0.0197 0.296 0.175
100 −0.0001 0.0334 0.001 0.498 0.0078 0.0279 0.157 0.313

Looe Keys (nc = 17) Mexican Caribbean (nc = 15)

10 0.0001 0.0002 0.192 0.230 0.0002 0.0002 0.342 0.106
20 0.0003 0.0005 0.245 0.171 0.0013 0.0010 0.549 0.017
30 0.0001 0.0011 0.050 0.424 0.0025 0.0013 0.678 0.003
40 0.0004 0.0017 0.103 0.347 0.0031 0.0022 0.569 0.013
50 0.0013 0.0022 0.255 0.162 0.0073 0.0041 0.660 0.004
60 0.0010 0.0039 0.119 0.325 0.0113 0.0054 0.718 0.001
70 0.0010 0.0039 0.119 0.325 0.0113 0.0054 0.718 0.001
80 0.0029 0.0079 0.165 0.264 0.0211 0.0126 0.635 0.005
90 0.0031 0.0095 0.147 0.287 0.0298 0.0149 0.701 0.002
100 0.0055 0.0131 0.185 0.238 0.0430 0.0249 0.647 0.005

Tongo Bay (nc = 15) Maputo Bay (nc = 7)

10 0.0001 0.0007 0.090 0.374 −0.0002 0.0016 0.105 0.411
20 −0.0003 0.0012 0.119 0.336 0.0001 0.0049 0.024 0.480
30 0.0020 0.0047 0.203 0.234 0.0049 0.0143 0.293 0.262
40 0.0051 0.0096 0.254 0.181 0.0119 0.0210 0.455 0.153
50 0.0084 0.0121 0.324 0.119 0.0088 0.0243 0.309 0.250
60 0.0113 0.0201 0.266 0.169 0.0105 0.0552 0.169 0.359
70 0.0113 0.0201 0.266 0.169 0.0105 0.0552 0.169 0.359
80 0.0098 0.0371 0.128 0.324 0.0314 0.0932 0.290 0.264
90 0.0289 0.0490 0.278 0.158 0.0351 0.1242 0.247 0.297
100 0.0443 0.0613 0.334 0.112 0.0271 0.1554 0.155 0.370

Celestun Lagoon (nc = 13) Sakumo Lagoon (nc = 11)

10 −0.0003 0.0006 0.299 0.161 −0.0011 0.0047 0.142 0.339
20 −0.0018 0.0015 0.531 0.031 −0.0017 0.0110 0.095 0.390
30 −0.0035 0.0050 0.361 0.113 −0.0041 0.0158 0.157 0.322
40 −0.0082 0.0066 0.561 0.023 −0.0056 0.0442 0.077 0.411
50 −0.0105 0.0096 0.510 0.038 −0.0103 0.0792 0.079 0.409
60 −0.0172 0.0171 0.478 0.049 −0.0086 0.1069 0.049 0.443
70 −0.0172 0.0171 0.478 0.049 −0.0086 0.1069 0.049 0.443
80 −0.0398 0.0329 0.547 0.026 −0.0315 0.1996 0.096 0.389
90 −0.0455 0.0465 0.468 0.054 −0.0633 0.2802 0.137 0.344
100 −0.0773 0.0548 0.607 0.014 −0.1365 0.3515 0.231 0.247

Paraná River (nc = 36) Garone River (nc = 7)

10 −1.6 × 10−5 0.0001 0.066 0.351 0.0004 0.0008 0.394 0.191
20 0.0001 0.0002 0.079 0.323 0.0020 0.0036 0.440 0.162
30 −0.0001 0.0007 0.064 0.356 0.0065 0.0094 0.525 0.113
40 −0.0001 0.0010 0.018 0.458 0.0099 0.0156 0.497 0.128
50 −0.0002 0.0017 0.039 0.412 0.0119 0.0183 0.505 0.124
60 −0.0004 0.0025 0.044 0.400 0.0158 0.0254 0.489 0.133
70 −0.0004 0.0025 0.044 0.400 0.0158 0.0254 0.489 0.133
80 −0.0003 0.0054 0.015 0.465 0.0434 0.0654 0.513 0.120
90 −0.0006 0.0074 0.025 0.442 0.0524 0.0795 0.511 0.121
100 0.0014 0.0109 0.038 0.413 0.0492 0.0786 0.491 0.131

Awassa Lake (nc = 11) Lake George (nc = 11)

10 0.0002 0.0002 0.446 0.085 0.0011 0.0014 0.436 0.090
20 −0.0002 0.0008 0.164 0.314 −0.0008 0.0020 0.233 0.245
30 −0.0006 0.0021 0.165 0.314 −0.0011 0.0033 0.202 0.275
40 −0.0005 0.0033 0.083 0.404 0.0007 0.0065 0.067 0.422
50 0.0010 0.0043 0.134 0.347 0.0058 0.0134 0.257 0.223
78 L.A. Salcido-Guevara, F. Arreguín-Sánchez / Ecological Modelling 276 (2014) 64–79

Table A2 (Continued)

Change of ba 95% CI rb p ba 95% CI rb p

biomass (%)

60 0.0017 0.0061 0.171 0.307 0.0016 0.0183 0.055 0.437

70 0.0017 0.0061 0.171 0.307 0.0016 0.0183 0.055 0.437
80 −0.0008 0.0126 0.038 0.456 0.0185 0.0415 0.263 0.217
90 −0.0082 0.0216 0.227 0.251 0.0546 0.0692 0.434 0.091
100 −0.0121 0.0318 0.226 0.252 0.0434 0.0884 0.287 0.196

Sri Lankan Reservoir (nc = 12) Broa Reservoir (nc = 7)

10 0.0003 0.0010 0.176 0.292 4.0 × 10−5 0.0004 0.087 0.426
20 0.0014 0.0022 0.336 0.143 −0.0006 0.0015 0.339 0.229
30 0.0033 0.0051 0.353 0.130 0.0001 0.0033 0.027 0.477
40 0.0038 0.0110 0.194 0.272 −0.0008 0.0036 0.185 0.345
50 0.0151 0.0165 0.464 0.064 −0.0014 0.0093 0.134 0.388
60 0.0224 0.0241 0.470 0.061 −0.0044 0.0079 0.452 0.154
70 0.0224 0.0241 0.470 0.061 −0.0044 0.0079 0.452 0.154
80 0.0416 0.0438 0.478 0.058 −0.0046 0.0190 0.215 0.322
90 0.0458 0.0708 0.347 0.134 −0.0029 0.0288 0.091 0.423
100 0.0531 0.0796 0.357 0.127 0.0021 0.0350 0.053 0.455
a
The value b is the slope of simple linear regression model with 95% confidence interval and the probability value p of Student’s test.
b
r is the Pearson’s correlation coefficient.
c
n is the number of functional groups for each kind of biomass change.

References
Ainsworth, C., Heymans, J.J., Pitcher, T.J., Vasconcellos, M., 2002. Ecosystem models
of Northern British Columbia for the time periods 2000, 1950, 1900 and 1750.
FCRR 10 (4), 41.
Alvarez-Hernández, J.H., 2003. Trophic model of a fringing coral reef in the southern
mexican caribbean (Modelo trófico para un arrecife de coral de tipo borde-
barrera en el sur del caribe mexicano). FCRR 11 (6), 227–235.
Angelini, R., Agostinho, A.A., 2005. Food web model of the Upper Paraná River Flood-
plain: description and aggregation effects. Ecol. Modell. 181, 109–121.
Angelini, R., Petrere Jr., M., 1996. The ecosystem of Broa reservoir, Sao Paolo state,
Brazil, as described using ecopath. Naga ICLARM Q. 19 (2), 36–41.
Arreguín-Sánchez, F., Arcos, E., Chávez, E., 2002. Flows of biomass and structure in
an exploited benthic ecosystem in the Gulf of California, Mexico. Ecol. Modell.
156, 167–183.
Banavar, J.R., Maritan, A., Rinaldo, A., 1999. Size and form in efficient transportation
networks. Nature 399, 130–132.
Banavar, J.R., Damuth, J., Maritan, A., Rinaldo, A., 2002. Supply–demand balance and
metabolic scaling. Proc. Natl. Acad. Sci. U.S.A. 99 (16), 10506–10509.
Begon, M., Townsend, C.R., Harper, J.L., 2006. Ecology from Individuals to Ecosystems,
4th ed. Blackwell Publishing, United Kingdom.
Bendoricchio, G., Palmeri, L., 2005. Quo vadis ecosystem. Ecol. Modell. 184, 5–17.
Blanchard, J.L., Pinnegar, J.K., Mackinson, S., 2002. Exploring marine mammal-fishery
interactions using ‘Ecopath with Ecosim’: modelling the Barents Sea ecosystem.
In: CEFAS Sci. Ser. Tech. Rep. No. 117, Lowestoft, p. 52.
Brown, J.H., Allen, A.P., Gillooly, J.F., 2007. The metabolic theory of ecology and the
role of body size in marine and freshwater ecosystems. In: Hildrew, A.G., Raf-
faelli, D.G., Edmonds-Brown, R. (Eds.), Body Size. The Structure and Function of
Aquatic Ecosystems. Cambridge University Press, New York, pp. 1–10.
Buchary, E.A., Alder, J., Nurhakim, S., Wagey, T., 2002. The use of ecosystem-based
modelling to investigate multi-species management strategies for capture fish-
eries in the Bali Strait, Indonesia. In: Pitcher, T., Cochrane, K. (Eds.), The Use
of Ecosystem Models to Investigate Multispecies Management Strategies for
Capture Fisheries. FCRR, Vancouver, pp. 24–32.
Chávez, E.A., Garduño, M., Arreguín-Sánchez, F., 1993. Trophic dynamic structure of
Celestun lagoon, Southern Gulf of Mexico. In: Christensen, V., Pauly, D. (Eds.),
Trophic Models of Aquatic Ecosystems. ICLARM Conf. Proc., Manila, pp. 186–192.
Christensen, V., Pauly, D., 1992a. ECOPATH II – a software for balancing steady-
state ecosystem models and calculating network characteristics. Ecol. Modell.
61, 169–185.
Christensen, V., Pauly, D., 1993. Flow characteristics of aquatic ecosystems. In: Chris-
tensen, V., Pauly, D. (Eds.), Trophic Models of Aquatic Ecosystems. ICLARM Conf.
Proc., Manila, pp. 338–352.
Christensen, V., Walters, C.J., 2004. Ecopath with Ecosim. Methods, capabilities and
limitations. Ecol. Modell. 172, 109–139.
Christensen, V., Walters, C.J., Pauly, D., 2005. Ecopath with Ecosim, a User’s Guide,
November ed. Fisheries Centre, University of British Columbia, Vancouver.
Cox, S.P., Essington, T.E., Kitchell, J.F., Martell, S.J.D., Walters, C.J., Boggs, C., et al., 2002.
Reconstructing ecosystem dynamics in the central Pacific Ocean, 1952–1998. II.
A preliminary assessment of the trophic impacts of fishing and effects on tuna
dynamics. Can. J. Fish. Aquat. Sci. 59, 1736–1747.
DeAngelis, D.L., 1980. Energy flow, nutrient cycling, and ecosystem resilience. Ecol-
ogy 6 (4), 764–771.
De Paula, E., Silva, R., Sousa, M.I., Caramelo, A.M., 1993. The Maputo bay ecosystem,
Mozambique. In: Christensen, V., Pauly, D. (Eds.), Trophic Models of Aquatic
Ecosystems. ICLARM Conf. Proc., Manila, pp. 214–223.
Fetahi, T., (master thesis) 2005. Trophic Analysis of Lake Awassa Using Mass-balance
Ecopath Model. Addis Ababa University.
Garces, L.R., Alias, M., Abu Talib, A., Mohamad-Norizam, M., Silvestre, G.T., 2003. A
trophic model of the coastal fisheries ecosystem off the west coast of Sabah
and Sarawak, Malaysia. In: Silvestre, G., Garces, L., Stobutzki, I., Ahmed, M.,
Valmonte-Santos, R.A., Luna, C., et al. (Eds.), Assessment, Management and
Future Directions of Coastal Fisheries in Asian Countries. WorldFish Center Conf.
Proc., pp. 333–352.
Hairston, N.G., Smith, F.E., Slobodkin, B., 1960. Community structure, population
control, and competition. Am. Nat. 94 (879), 421–425.
Hemmingsen, A.M., 1960. Energy metabolism as related to body size and respiratory
surfaces, and its evolution. Rep. Steno Mem. Hosp. Copenhagen 9, 1–110.
Huxel, G.R., McCann, K., 1998. Food web stability, the influence of trophic flows
across habitats. Am. Nat. 152 (3), 460–469.
Kleiber, M., 1932. Body size and metabolism. Hilgardia 6, 315–353.
Libralato, S., Christensen, V., Pauly, D., 2006. A method for identifying keystone
species in food web models. Ecol. Modell. 195, 153–171.
MacArthur, R.H., 1955. Fluctuations on animal populations and a measure of com-
munity stability. Ecology 36 (3), 533–536.
Moreau, J., Christensen, V., Pauly, D., 1993. A trophic ecosystem model of lake George,
Uganda. In: Christensen, V., Pauly, D. (Eds.), Trophic Models of Aquatic Ecosys-
tems. ICLARM Conf. Proc., Manila, pp. 124–129.
Moreau, J., Villanueva, M.C., Amarasinghe, U.S., Schiemer, F., 2001. Trophic rela-
tionships and possible evolution of the production under various fisheries
management strategies in a Sri Lankan reservoir. In: De Silva, S.S. (Ed.), Reser-
voir and Culture-based Fisheries, Biology and Management. Australian Centre
for International Agricultural Research, Canberra, pp. 201–214.
Odum, E.P., 1985. Fundamentos de ecología. Interamericana, México.
Odum, E.P., Barrett, G.W., 2005. Fundamentals of Ecology, 5th ed. Thomson Brooks,
Belmont.
Okey, T.A., Vargo, G.A., Mackinson, S., Vasconcellos, M., Mahmoudie, B., Meyer, C.A.,
2004. Simulating community effects of sea floor shading by plankton blooms
over the West Florida Shelf. Ecol. Modell. 172, 339–359.
Oksanen, L., 1991. Trophic levels and trophic dynamics: a consensus emerging?
Trends Ecol. Syst. 6 (2), 58–60.
Palomares, M.L., Yulianto, B., Puy, L., Bengen, D., Belaud, A., 1993. A preliminary
model of the Garonne river, Toulouse, France ecosystem in the spring. In: Chris-
tensen, V., Pauly, D. (Eds.), Trophic Models of Aquatic Ecosystems. ICLARM Conf.
Proc., Manila, pp. 172–179.
Pauly, D., 2002. Spatial modelling of trophic interactions and fisheries impacts in
coastal ecosystems. A case study of Sakumo Lagoon, Ghana. In: McGlade, J.M.,
Cury, P., Koranteng, K.A., Hardman-Mountford, N.J. (Eds.), The Gulf of Guinea
Large Marine Ecosystem- Environment Forcing and Sustainable Development of
Marine Resources. Elsevier Science, Amsterdam, The Netherlands, pp. 289–297.
Peters, R.H., 1983. The Ecological Implications of Body Size. Cambridge University
Press, USA.
Pimm, S.L., 1979. Complexity and stability. Another look at MacArthur’s original
hypothesis. Oikos 33, 351–357.
Pimm, S.L., 1982. Food Webs Population and Community Biology Series. Chapman
and Hall, London.
Polis, G.A., Strong, D.R., 1996. Food web complexity and community dynamics. Am.
Nat. 147 (5), 813–846.
Polovina, J.J., 1984. Model of a coral reef ecosystem. I. The ECOPATH model and its
application to French Frigate Shoals. Coral Reefs 3, 1–11.
Power, M.E., 1992. Top-down and bottom-up forces in food webs: do plants have
primacy. Ecology 73 (3), 733–746.
 L.A. Salcido-Guevara, F. Arreguín-Sánchez / Ecological Modelling 276 (2014) 64–79
Salcido-Guevara, L.A., Arreguín-Sánchez, F., 2007. A benthic ecosystem model of the
Sinaloa continental shelf, Mexico. In: Le Quesne, W.J.F., Arreguín-Sánchez, F.,
Heymans, S.J.J. (Eds.), INCOFISH Ecosystem Models: Transiting from Ecopath to
Ecospace. FCRR, Vancouver, pp. 170–188.
Salt, G.W., 1979. A comment on the use of the term emergent properties. Am. Nat.
113 (1), 145–148.
Ulanowicz, R.E., 2009. The dual nature of ecosystem dynamics. Ecol. Modell. 220
(16), 1886–1892.
Venier, J.M., Pauly, D., 1997. Trophic dynamics of a Florida Keys coral reef ecosystem.
Proc. Int. Coral Reef Symp. 1, 915–920.
Walters, C., Pauly, D., Christensen, V., Kitchell, J.F., 2000. Representing density
dependent consequences of life history strategies in aquatic ecosystems: EcoSim
II. Ecosystems 3, 70–83.
Weibel, E.R., Taylor, C.R., Bolis, L., 1998. Principles of Animal Design: The Optimiza-
tion and Symmorphosis Debate. Cambridge University Press, Cambridge.
Wolff, M., 1994. A trophic model for Tongoy Bay – a system exposed to suspended
scallop culture (Northern Chile). J. Exp. Mar. Biol. Ecol. 182, 149–168.
Zetina-Rejón, M.J., Arreguín-Sánchez, F., Chávez, E.A., 2003. Trophic structure and
flows of energy in the Huizache-Caimanero lagoon complex on the Pacific coast
of Mexico. Estuar. Coast. Shelf Sci. 57, 803–815.
79
Web references
Christensen, V., Pauly, D., 1992b. Ecopath II – a software for balancing steady-state
ecosystem models and calculating network characteristics. Ecol. Modell. 61,
169–185, http://www.ecopath.org/models (accessed May, 2008).
Dr. Rainer Froese. Incofish Co-ordinator, FB Marine Ökologie, Fischereibi-
ologie, Leibniz Institut für Meereswissenschaften, Kiel, Germany. Dr.
Francisco Arreguín-Sánchez. Ecosystem modelling Leader (Workpackages
4). http://www.incofish.org/Workpackages/WP4/WP4.php (accessed May,
2008).
Information on temperature data maintained by the NASA JPL Physical Oceanogra-
phy DAAC, Pasadena, CA. http://poet.jpl.nasa.gov (accessed April, 2011).
This is a test version of ILEC’s (International Lake Environment Committee) new
“World Lake Database” system which will in due course replace the current
version, called “World Lakes Database” accessible at http://www.ilec.or.jp/
database/asi/asi-45.html (accessed April, 2011). http://wldb.ilec.or.jp/.
Temperature data for Lake George in http://www.ilec.or.jp/database/afr/afr-10.html.
Temperature data for Sri Lankan Reservoir in http://www.ilec.or.jp/database/
asi/dasi45.html. Address: 1091 Oroshimo-cho Kusatsu-shi, Shiga 525-0001
Japan.