Progress in Oceanography: Marc H. Taylor, Matthias Wolff, Jaime Mendo, Carmen Yamashiro

View metadata, citation and similar papers at core.ac.
uk brought to you by CORE

provided by Electronic Publication Information Center
Progress in Oceanography 79 (2008) 336–351
Contents lists available at ScienceDirect
Progress in Oceanography
journal homepage: www.elsevier.com/locate/pocean
Changes in trophic flow structure of Independence Bay (Peru) over an ENSO cycle
Marc H. Taylor a,*, Matthias Wolff a, Jaime Mendo b, Carmen Yamashiro c
a
Center for Tropical Marine Ecology, Fahrenheitstrasse 6, 28359 Bremen, Germany
b
Facultad de Pesqueria, Universidad Agraria La Molina, Lima, Peru
c
Instituto del Mar del Perú, Esq. Gamarra y Gral. Valle s/n, La Punta, Callao, Peru
a r t i c l e i n f o a b s t r a c t
Article history: During the strong warm El Niño (EN) that occurred in 1997/98, Independence Bay (14°S, Peru) showed a
Accepted 14 October 2008 ca. 10 °C increase in surface temperatures, higher oxygen concentrations, and clearer water due to
Available online 21 October 2008 decreased phytoplankton concentrations. Under these quasi-tropical conditions, many benthic species
suffered (e.g. macroalgae, portunid crabs, and polychaetes) while others benefited (e.g. scallop, sea stars,
Keywords: and sea urchins). The most obvious change was the strong recruitment success and subsequent prolifer-
El Nino phenomena ation of the scallop Argopecten purpuratus, whose biomass increased fiftyfold. To understand these
Scallop fisheries
changes, steady-state models of the bay ecosystem trophic structure were constructed and compared
Steady state
Trophic relationships
for a normal upwelling year (1996) and during an EN (1998), and longer-term dynamics (1996–2003)
Humboldt current were explored based on time series of catch and biomass using Ecopath with Ecosim (EwE) software.
Peru Model inputs were based on surveys and landings data collected by the Instituto del Mar del Perú (IMA-
Independence Bay RPE). Results indicate that while ecosystem size (total throughput) is reduced by 18% during EN, mainly
as a result of decreased total primary production, benthic biomass remains largely unchanged despite
considerable shifts in the dominant benthic taxa (e.g. scallops replace polychaetes as secondary consum-
ers). Under normal upwelling conditions, predation by snails and crabs utilize the production of their
prey almost completely, resulting in more efficient energy flow to higher trophic levels than occurs dur-
ing EN. However during EN, the proliferation of the scallop A. purpuratus combined with decreased phy-
toplankton increased the proportion of directly utilized primary production, while exports and flows to
detritus are reduced. The simulations suggest that the main cause for the scallop outburst and for the
reduction in crab and macroalgae biomass was a direct temperature effect, whereas other changes are
partially explained by trophic interactions. The simulations suggest that bottom-up effects largely control
the system.
Ó 2008 Elsevier Ltd. All rights reserved.
1. Introduction nearshore species are of high commercial value and the fishery
supports thousands of fishers and their families (Wolff et al., 2003).
The Humboldt Current System (HCS), located in the south east Under ‘normal’ upwelling periods, near-seafloor oxygen con-
Pacific along the coasts of Chile and Peru, is one of the most pro- centrations <0.5 ml l1 are typical on the continental shelves of
ductive marine systems in the world. This high productivity is Peru (<100 m; Zuta et al., 1983). This is due to the oxygen mini-
the result of ‘coastal upwelling’ – a phenomenon driven by south- mum zone which occurs below the shallow, uplifted Peruvian ther-
erly trade winds that brings cold, nutrient-rich water from 40 to mocline (OMZ; 50–600 m) and the sinking of decomposing organic
80 m up into the euphotic zone where it supports phytoplankton matter from the overlying euphotic zone (Arntz et al., 2006). Bac-
growth (Barber et al., 1985; Arntz et al., 1991; Pennington et al., teria such as the filamentous ‘spaghetti’ bacteria (genus Thioplaca),
2006). As a result, the system supports a large biomass of small are commonly found in association with the OMZ (Arntz et al.,
planktivorous pelagic fish – comprising the bulk of catches by a 1991). At shallower depths, oxygen concentrations increase and
large purse seining fleet. An important fishery also exists down are able to support a higher benthic biomass.
to 15–30 m and in the intertidal areas (Arntz and Valdivia, These coastal phenomena propagate into Peru’s bays, where
1985a; Arntz et al., 1988). Despite a relatively low annual harvest much artisanal fishing occurs. This is seen in Independence Bay
(ca. 200,000 t yr1) compared to the pelagic system, the exploited (14°S) where the deeper regions of the bay (>30 m) are of low
biomass while the bay’s shallower perimeter is targeted by the
artisanal fishery. These shallow areas contain valuable molluscan
* Corresponding author. Tel.: +49 4212380056; fax: +49 4212380030. and crustacean species, but suspension feeding polychaetes domi-
E-mail address: [email protected] (M.H. Taylor). nate biomass (Tarazona et al., 1991).
0079-6611/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.pocean.2008.10.006
M.H. Taylor et al. / Progress in Oceanography 79 (2008) 336–351 337
The HCS experiences ‘natural’ climate variations on seasonal, scallops proliferate to nearly replace polychaetes as the main ben-
interannual, and decadal time scales, all of which affect the sys- thic consumer of plankton and detritus (Fig. 2).
tem’s productivity and fisheries. Off Peru the strongest interannual During both of the strong ENs of 1982/83 and 1997/98, A. pur-
perturbation is the warm ‘‘El Niño” (EN) phase of the El Niño puratus became the principal target of the diving fishery, which
Southern Oscillation (ENSO). ENs last about 18 months and occur experienced ‘‘gold rush” conditions with high catches and enor-
irregularly every 3–5 years. During EN a Kelvin wave travels east- mous revenues (Wolff, 1987, 1988, 1994; Wolff and Mendo,
wards on the equator across the Pacific then north and south along 2000; Mendo and Wolff, 2002). The fishing effort increased mainly
the South American coast, where it depresses the normally shallow due to migration of fishers from other areas. Catches largely reflect
thermocline and a raises sea level (Pennington et al., 2006). actual changes in the scallop population. Other high-price species
Although Peruvian coastal upwelling continues during EN, water associated with the scallop habitat are octopus and crab. Crab
upwells from above the thermocline and is thus nutrient poor (Bar- catches decreased during the 1997/98 EN, but octopus landings in-
ber and Chavez, 1983). As a result, the area of ‘productive habitat’ creased nearly 5-fold. Pelagic predatory fish migrated towards the
supported by coastal upwelling is greatly reduced in area (>1.0 mg coast during EN, such that catch of the line and net fishing fleet in-
chla m3; Nixon and Thomas, 2001), as is overall primary produc- creased by about 2.5 times.
tion (Carr, 2002). This reduction in production at the base of the While we have a good basic understanding of the main changes
food web negatively impacts many pelagic coastal species (Tam to benthic communities associated with EN in Peru, trophic model-
et al., 2008; Taylor et al., 2008). ing of the effects of EN has not yet been employed. Here we de-
EN can also produce significant positive faunal changes in ben- scribe and model the Independence Bay ecosystem from an
thic habitats, mainly as a result of increased oxygen levels (Arntz energy flow perspective. First, we compare steady-state trophic
et al., 1991). This is especially the case in shallow depths, where models of the system for the upwelling and El Nino conditions,
faunal density, biomass, species richness, and diversity can all in- and secondly we explore the drivers of these changes (trophic vs.
crease during EN (Tarazona et al., 1988). Several species from off- environmental) using dynamic simulations with performance
shore, equatorial, and subtropical coastal areas also migrate to measured against time series data of changing biomass. In partic-
the Peruvian coast during EN, such as swimming crabs and penaeid ular, we address the following questions: (i) Are the positive im-
shrimps (Arntz et al., 1991). In Independence Bay the resident scal- pacts observed in the shallow benthic community during EN
lop Argopecten purpuratus experiences much higher recruitment (increase in species richness, and diversity) also reflected in the en-
and growth during EN. Past El Nino densities have reached up to tire ecosystem through indicators of system maturity? (ii) How is
8 kg m2 and densities of 129 adult scallops m2 (Wolff, 1987; the system reorganized during EN? (iii) What insight can be gained
Arntz and Tarazona, 1990), which is about 50 times the normal le- into the management of the fishery during EN? (iv) Can changes in
vel. Yearly surveys of the macrobenthos of Independence Bay biomass and productivities be explained by direct responses to
(Fig. 1) conducted by the Instituto del Mar del Perú (IMARPE) have warming, or to trophic interactions? (v) What is the trophic effect
also observed EN biomass decreases in several functional groups of the increased scallop biomass, the reduced primary production
(e.g. macroalgae, benthic detritivores, herbivorous gastropods, (through biomass decrease of phytoplankton as well as macroal-
predatory gastropods, portunid crabs, and polychaetes); while gae), and the reduced crab predator biomass on the system?
Fig. 1. Map of the Peruvian coast and the study site, Independence Bay. Macrobenthic fauna sampling stations are indicated by circles for 1996 (n = 223) and triangles for
1998 (n = 252). The 30 m depth isocline is indicated by a dashed line.
338 M.H. Taylor et al. / Progress in Oceanography 79 (2008) 336–351
600
Small carnivores
500 Sea urchins
Biomass (ww gm2)

Sea stars
400 Scallops
Predatory gastropods
300 Predatory crabs
Polychaetes
200 Misc. filter-feeder
Macroalgae
100
Herbivorous gastropods
Benthic detritivores
0
1996 1997 1998 1999
Fig. 2. Biomass changes of benthic macrofauna observed from 1995 to 1999 (IMARPE). Boxes indicate model periods.
2. Materials and methods The model also permits the assessment of ecosystem dynamics un-
der different scenarios of use or environmental change (http://
2.1. Study area www.ecopath.org). Ecopath links the production and consumption
of all trophically connected groups within the model ecosystem, as
Independence Bay (14.238°S, 76.194°W) is located approxi- Pi ¼ Y i þ Bi M2i þ Ei þ BAi þ Pi ð1 EEi Þ ð1Þ
mately 250 km southeast of Lima (Fig. 1). The bay is broadly open
to the coastal ocean on either side of ‘Isla La Vieja’. Conditions where Pi is the total production rate of (i), Yi is the total fishery catch
within the bay thus largely reflect the nearshore Peruvian upwell- rate of (i), Bi the biomass of the group (wet weight), Ei the net migra-
ing system, characterized by low surface temperatures (14–18 °C) tion rate (emigration–immigration), M2i is the total predation rate
and high nutrient levels. Bottom (<30 m) oxygen concentrations for group (i), BA is the biomass accumulation rate for (i). P *
i i
averaging 3.5 mg L1 during normal upwelling conditions, but in (1 EEi) is the ‘other mortality’ rate (M0i), where EE is the ‘‘Eco-
the deeper central part of the bay, low oxygen conditions prevail trophic efficiency” and is the proportion of the group’s production
(<1.0 mg L1, >30 m) and benthic macrobiota is not abundant and that is consumed by higher trophic levels or is taken by the fishery
microbial processes dominate. During a strong EN, temperatures (for further information, see Christensen et al., 2000). In order to en-
increase as much as 10 °C and oxygen conditions are improved at sure mass balance between the groups, a second master equation is
the lower depths. Artisanal fisheries include a diving fishery using used:
hookah and compressor, which operates around the bay’s rocky Consumption ¼ production þ respiration þ unassimilated food
and soft-bottom habitats less than 30 m, and a gillnet and line fish-
ery that targets larger littoral and pelagic fish species. Energy flow in the model requires definition of the diet for all
consumers, which determines the fraction of each functional group
2.2. Model definition which will serve as food of the other groups. This diet matrix is fur-
ther used in the calculation of the trophic level of each group:
The two steady-state trophic models of Independence Bay were X
TLj ¼ 1 þ TLi DC ij ð2Þ
constructed for the soft-bottom habitats of <30 m depth that fringe
the bay, covering about 38% of the total bay area (65.8 km2 out of a where DCij is the fraction of prey (i), in the diet of the predator (j).
total of 172 km2; Fig. 1). This area was selected for the following The trophic level of the predator TLj is calculated as the mean tro-
P
reasons: (i) importance in overall bay macrobiota biomass, (ii) phic level of its prey ( TLi * DCij) plus 1.0. Primary producers and
availability of data, and (iii) it encompasses the main activities of detritus groups are assigned a trophic level of 1.0.
the artisanal fishery. Model periods are for 1996, representing a
‘normal’ upwelling year, and 1998, representing EN (end of the 2.4. Input parameters
1997/98 event; Fig. 2). The models were constructed with 20 func-
tional groups including detritus, two primary producers (phyto- Input parameters, detailed below, are derived from a number of
plankton and macroalgae), zooplankton, six benthic primary sources which are listed in Table 2. Input values for 1996 and 1998
consumers (polychaetes, scallops, sea urchins, herbivorous gastro- steady-state models can be found in 3.
pods, benthic detritivores, and miscellaneous filter-feeders), five
benthic carnivores (predatory gastropods, small carnivores, preda- 2.4.1. Biomass
tory crabs, sea stars, and octopus), three fish (littoral fish, small pe- Benthic macrofauna biomass was from IMARPE surveys for the
lagic fish, and pelagic predatory fish), and two top predators periods 19–29th April, 1996 and 15–24th July, 1998. A total of 223
(marine mammals and seabirds). These functional groups were and 252 1 m2 quadrants were sampled during the two surveys,
designated according to ecological status – organisms within a respectively. All epifauna and infauna of the upper 5 cm of sedi-
group are characterized by similar diets, predators, productivities ment were collected by hand and placed in mesh bags of 5 mm
and individual body size (Table 1). mesh size. Organisms were later counted and weighed (for further
information on sampling, see Samamé et al., 1985; Yamashiro
2.3. Basic modeling approach et al., 1990). Groups of small epifauna (herbivorous gastropods,
benthic detritivores, scallops, small carnivores) and polychaetes
A mass-balance modeling approach was applied using the soft- were increased by 25% to correct for undersampling. Miscellaneous
ware Ecopath with Ecosim 5.0 (EwE) (Christensen and Pauly, 1992; filter-feeders (consisting mainly of infaunal bivalves) were in-
Walters et al., 1997), which quantifies trophic flows among func- creased by 100% to also correct for undersampling—much of this
tional groups within an ecosystem and also includes fishery catch. grouṕs biomass is found deeper than 5 cm. These biomass correc-
Table 1
Functional groups and representative species. Species listed are not exhaustive (small benthos groups show the most important species, representing >95% of biomass and/or
species averaging >1 g m2); **, groups/species not found/recorded in captures in 1998; *, groups/species found in 1998 but low in biomass; bold: groups/species not found/
recorded in captures in 1996.
Functional group Species

2. Macroalgae Rhodymenia sp.*, Macrocystis sp.*, Gigartina sp.*, Codium sp.**, Ulva sp.*,
Caulerpa sp., Lessonia nigrescens
4. Polychaetes Diopatra sp., Chaetopteridae
5. Scallops Argopecten purpuratus
6. Sea urchins Tetrapigus niger, Arbacia spatuligera, Arbacia sp., Loxechinus albus, Strongylocentrotus sp.
7. Herbivorous gastropods Crepipatella dilatata, Crepipatella sp., Tegula euryomphalus, Tegula atra, Tegula sp., Crucibulum sp.,
Aplysia sp., Mitrella sp.
8. Benthic detritivores Ophiuroidea*, Pagurus sp., Eurypanopeus sp.*, Taliepus marginatus **
9. Misc. filter-feeders Ascidians, Aulacomya ater, Glycimeris ovata, Actinia sp., Prothothaca thaca,
Sponges, Semele solida, Chama sp.
10. Predatory gastropods Bursa ventricosa, Bursa nana, Bursa sp., Thais chocolata, Thaididae sp., Priene rude, Cymatium
weigmani, Cymathidae sp., Argobuccinum sp.**, Sinum cymba
11. Small carnivores Oliva peruviana, Oliva sp., Nassarius dentifer, Nassarius gayi, Nassarius sp., Trophon sp.**, Crassilabrum crassilabrum,
Natica sp.**, Xantochorus sp., Solenosteria gatesi, Solenosteria sp.,
Polinices uber
12. Predatory crabs Cancer setosus, Cancer porteri, Cancer coronatus**, Cancer sp., Hepatus chilensis,,
Platyxanthus cockeri**, Callinectes arcuatus, Callinectes sp.
13. Sea stars Luidia bellonae, Luidia magallanica, Luidia sp., Asterina chilensis.**, Patiria chilensis,
Heliaster helianthus
14. Octopus Octopus mimus
15. Littoral fish Isacia conceptionis, Seriolella violacea, Paralabrax humeralis, Cheilodactylus variegatus,
Labrisomus philippii, Hemilutjanus macrophthalmos, Acanthistius pictus, Paralichthys adspersus,
Cynoscion analis, Sciaena deliciosa, Calamus brachysomus, Mugiloides chilensis,
Diplectrum conceptione, Chloroscombrus orqueta, Sphyraena ensis, S. idiastes,
Myliobatis peruvianus, Orthopristis chalceus, Mugil cephalus, Diplectrum conceptione,
Chloroscombrus orqueta, Sphyraena ensis, Sphyraena idiastes, Myliobatis peruvianus
16. Small pelagic fish Sardinops sagax sagax, Ethmidium maculatum, Trachinotus paitensis
17. Pelagic predatory fish Trachurus picturatus murphyi, Cilus gilberti, Scomber japonicus,
Sarda chiliensis chiliensis, Auxis rochei, Scomberomorus sierra
18. Marine mammals Otaria byronia, Arctocephalus australis
19. Seabirds Leucocarbo bougainvillii, Sula variegata, Pelecanus thagus
tions were based on complementary benthic evaluations con- ciations of functional groups with particular habitats, their catch
ducted by the authors. statistics were adjusted as follows: scallops and predatory crab
Estimates of phytoplankton biomass for the 1996 model were catches come only from the soft-bottom habitats of the model
taken from Peruvian coastal averages under ‘typical’ upwelling and thus did not need correction; fish groups, octopus, and miscel-
conditions (settled volume, 3.0 mL m3) (Rojas de Mendiola et al., laneous filter-feeders, primarily found in soft-bottom habitats,
1985) and EN conditions (Delgado and Villanueva, 1998; Villanu- were reduced by only 10% to correct catches associated with rocky
eva et al., 1998). EN phytoplankton values were increased slightly habitats. Conversely, catches of herbivorous gastropods, predatory
over coastal averages (+15%) in order to balance the model. Settled gastropods, and sea urchins were mainly associated with broken
volumes were converted to g m2 by assuming 1 mL = 1 g and then shell or rocky substrates, and were thus reduced by 80% (Table 3).
multiplying by an average depth for the model area of 15 m by
assuming a well-mixed water column. 2.4.3. Production/biomass (total mortality)
Information on zooplankton in Independence Bay is of qualita- Direct estimates of production to biomass ratios (P/B) or Total
tive nature only (Yamashiro et al., 1990); thus zooplankton bio- mortality (Z) existed for several benthic invertebrate groups in
mass was left open to be calculated by the steady-state model the model – scallops, predatory crabs, and sea stars. Other groups
assuming an Ecotrophic efficiency (EE) of 0.95. were estimated using empirical relationships from Brey (2001)
Biomass of mobile species such as octopus and fish were esti- taking into account taxonomic group, mean body size, temperature
mated from catch data by assuming that the fishery takes 50% of of habitat, feeding modes, and habitat type. In most cases this pro-
yearly produced biomass. Small pelagic fish are not a principal tar- vided realistic estimates; however, values for polychaetes and
get of the artisanal fishery and so catch estimates are likely poor misc. filter-feeders were increased to 1.0 based on other estimates
indicators of the available biomass. Small pelagic fish biomass from the literature (Table 3).
was thus left open to be calculated by the steady-state model P/B of phytoplankton was estimated using a modified Eppley
assuming an EE of 0.95 (Table 3). curve (Eppley, 1972) as described by Brush et al. (2002):
G ¼ Gmax f LTLIM NUTLIM ð3Þ
2.4.2. Catches
1
Estimates of catch were derived from IMARPE catch statistics where G = realized daily growth rate (d )(base e), f is the fraction
for the artisanal fishery from the two main landing sites for Inde- of the day during which there is light, and LTLIM and NUTLIM are
pendence Bay – San Andres and Laguna Grande. Unfortunately, dimensionless ratios from 0 to 1 which describe light and nutrient
landings data do not identify habitat of capture, so that it was nec- limitation of growth, respectively (Kremer and Nixon, 1977). Gmax,
essary to estimate the relative sizes of the bay’s habitats in the as given by Eppley (1972) describes an exponentially-shaped enve-
model (ca. 10% rocky, 90% soft-bottom) and correct for the fact that lope for growth rates of phytoplankton under culture conditions
most rocky habitat catches are made outside the model area (ca. without light or nutrient limitation (as recalculated by Brush
10 greater than within the model). Taking into account the asso- et al., 2002):
340
Table 2
Sources of input data. IE = IMARPE benthic macrofauna evaluation, EM = empirical model (Brey, 2001), EO = Ecopath output, GU = guess estimate, IC = iterative consumption routine (for opportunistic feeding; described herein),
IS = IMARPE landings statistics.
Functional Biomass – Bi Production rate – Consumption rate Conversion Ecotrophic Catches – Yi Diet composition
group/ (t km2) Pi/Bi (y1) – Qi/Bi (y1) efficiency – GEi efficiency (t km2 y1) – DC
parameter – EEi
1. Phytoplankton GU based on GU based on modified Eppley curve (Eppley, – – EO – –
Rojas de 1972; Brush et al., 2002)
Mendiola et al.
(1985),
Delgado and
M.H. Taylor et al. / Progress in Oceanography 79 (2008) 336–351

Villanueva
(1998)
2. Macroalgae IE GU based on Macchiavello et al. (1987) – – EO – –
3. Zooplankton EO GU based on Mendoza (1993), Hutchings et al. GU adapted from Polovina and EO GU – GU
(1995) Ow (1985)
4. Polychaetes IE GU based on Martin and Grémare (1997) EO GU EO – GU
5. Scallops IE Mendo et al. (1987), Stotz and Gonzalez (1997) GU based on Wolff (1994) EO EO IS GU based on Rouillon et al. (2002)
6. Sea urchins IE EM EO GU EO IS GU
7. Herbivorous IE EM EO GU 0.3 based on Mann (1982) EO IS GU
gastropods
8. Benthic IE EM EO GU EO – GU
detritivores
9. Misc. filter- IE GU based on Wolff (1994) EO GU EO IS GU
feeders
10. Predatory IE EO GE based on Huebner and GU 0.3 based on Huebner and EO IS GU, IC
gastropods Edwards (1981) Edwards (1981)
11. Small IE EM EO GU EO – GU partially based on Keen (1972) for
carnivores gastropod spp., IC
12. Predatory IE Wolff and Soto (1992) Lang (2000), Wolff and Soto EO EO IS GU based on Leon and Stotz (2004), IC
crabs (1992)
13. Sea stars IE Ortiz and Wolff (2002) EO GU EO – GU, IC
14. Octopus GU based on EO Wolf and Perez (1992), Vega and Wolf and Perez (1992), Vega and EO IS GU, IC
catch data Mendo (2002) Mendo (2002)
15. Littoral fish GU based on GU 1.2 based on Wolff (1994) EO GU EO IS GU based on FISHBASE (2006)
catch data
16. Small pelagic EO GU EO GU 0.1 based on Moloney et al. GU IS GU based on FISHBASE (2006)
fish (2005)
17. Pelagic GU based on GU 0.85 based on Jarre et al. (1991) EO GU 0.1 based on Moloney et al. EO IS GU based on FISHBASE (2006)
predatory fish catch data (2005)
18. Marine GU GU based on Jarre et al. (1991) EO GU EO – GU
mammals
19. Seabirds GU GU based on Moloney et al. (2005) EO GU based on Moloney et al. EO – GU
(2005)
20. Detritus EO – – – – – –
Gmax ¼ 0:97 e0:0633T ð4Þ

Input-output parameters for steady-state models of Independence Bay in 1996 and 1998 after application of the Ecoranger resampling routine. In bold, Ecopath calculated parameters. Bi = Biomass, Pi/Bi = production rate, Qi/
350.911
35.527
0.311
0.515
0.366
0.821
0.733
0.684
0.223
6.737
0.742
1.871
0.306
1.036
1.057
0.500
0.002
0.000
0.000
0.000
1998 where T = water temperature. The ‘normal’ upwelling phytoplank-
ton production 1996 assumed a mean temperature of 16 °C with
2.155
1.284
1.891
86.734
0.542
0.954
0.339
0.722
0.836
0.949
0.824
0.595
1.825
0.555
1.750
0.093
40.800
0.000
0.000
0.000
1996
50% light (from self-shading) and 0% nutrient limitation factors,

M2i
while the EN condition of 1998 assumed a mean temperature of

26 °C with 80% light and 50% nutrient limitation factors. Calculated
15.261
11.217
1.248
1.772
0.731
0.284
0.267
0.511
1.854
0.288
0.466
3.240
0.063
0.506
0.106
0.056
0.036
0.006
0.100
0.000
1998
P/B ratios were high (245 and 365 for 1996 and 1998, respectively)
yet the value of total production for the 1996 model in terms of car-
bon, i.e. 800 g C m2 yr1, using a wet weight:C conversion of
168.495
13.685
0.357
0.196
0.153
0.322
0.151
0.248
0.306
0.750
0.201
0.101
5.027
0.028
0.052
0.073
0.097
0.063
0.034
0.000
14.25:1 from Brown (1991), is conservative with respect to other
1996
M0i
estimates for the Peruvian coastal system under upwelling condi-

tions, i.e. >1000 and >1500 g C m2 yr1 from Walsh (1981) and
0.542
0.168
0.113
0.584
0.378
2.212
0.013
0.160
0.002
0.008
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
1998
Chavez and Barber (1985), respectively. P/B values for other groups
are taken from the literature (Table 2).
Bi = consumption rate, EEi = ecotrophic efficiency, GEi = gross efficiency or conversion efficiency (Pi/Qi), Fi = fishing mortality, M0i = non-predation mortality, M2i = predation mortality.
0.316
0.184
0.123
2.377
0.551
0.017
0.047
0.051
0.430
0.007
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
1996
Fi
2.4.4. Consumption and Conversion efficiency

Direct estimates of consumption rates (Q/B) were available for a
1.935
1.838
1.673
235.421
0.146
3.153
4.349
0.195
0.012
0.206
0.002
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
few of the benthic invertebrate groups (octopus, scallops, preda-
1998
Catch (t y1)
tory gastropods, and predatory crabs). For most other groups, ra-
tios of Conversion efficiency (GE) or the ratio between Production
1.458
1.389
1.326
3.417
0.177
0.749
0.977
0.298
0.584
2.230
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
1996
and Consumption (P/Q) were applied (Tables 2 and 3).
2.4.5. Diet matrices

0.266
0.241
0.215
0.156
0.225
0.253
0.213
0.213
0.389
0.111
0.121
0.290
0.207
0.309
0.110
0.003
0.001
1998
Direct diet studies for Independence Bay are limited and thus
–
–
–
general knowledge from literature was used in the construction

0.261
0.157
0.219
0.213
0.333
0.182
0.349
0.181
0.213
0.426
0.125
0.160
0.210
0.093
0.097
0.002
0.001
1996
of diet matrices (Table 2). Initial attempts to balance the 1996

GEi
–
–
model resulted in insufficient production of many smaller epifau-

nal herbivore and detritivore invertebrate groups (scallops, sea
0.958
0.483
0.375
0.299
0.458
0.563
0.943
0.801
0.994
0.640
0.865
0.191
0.304
0.614
0.951
0.449
0.000
0.000
0.916
0.867
1998
urchins, herbivorous gastropods, benthic detritivores, and misc. fil-

ter-feeders) to meet the initial consumption values of the carnivo-
0.340
0.187
0.136
0.603
0.806
0.949
0.788
0.845
0.949
0.805
0.918
0.930
0.860
0.846
0.846
0.679
0.000
0.000
0.967
0.890
1996
rous benthic invertebrate groups (predatory gastropods, small

EEi
carnivores, predatory crabs, sea stars, and octopus). As macroinver-

tebrate groups are described to be rather unselective and opportu-
4.844
2.891
3.793
5.287
4.762
3.705
3.446
7.710
14.789
12.361
10.281
17.957
38.278
52.151
4.549
145.755
9.092
nistic feeders, limited more by their modes of feeding (Wilson and

1998
Parkes, 1998), diet proportions were adjusted to reflect both pred-

–
–
–
Qi/Bi (y1)
atory consumption rates and the available production of prey

5.611
2.589
2.778
5.425
4.859
4.952
11.441
10.426
20.868
3.254
49.087
7.928
62.560
175.677
4.731
9.889
10.037
groups. This was accomplished by iteratively distributing the prey

1996
production to the predators in weekly consumption increments,

–
–
–
assuming unselectivity. When the production of a single prey

1.407
17.954
38.767
366.172
1.337
2.191
0.987
1.139
2.168
0.734
0.845
1.042
2.305
0.650
1.101
0.790
4.809
0.036
0.100
group was completely utilized, the following iteration would con-

1998
sider only those prey not fully utilized. Base values of detritus feed-
–
ing were assumed and the calculated diets resulted in high

Pi/Bi (y1)
1.653
255.228
1.576
2.165
0.899
0.551
0.925
0.989
0.897
4.878
1.939
45.827
0.692
0.771
1.018
1.307
15.840
0.101
0.034
proportions of polychaetes in their diets – reflecting their high bio-

1996
mass and production in the benthic system in 1996. The 1998 sit-
–
uation was less problematic due to a reduction of carnivorous

benthic invertebrate biomass as well as an increase in scallop bio-
29.425
23.144
24.816
8.656
45.927
11.314
12.111
5.952
10.955
7.595
11.516
20.286
1.425
0.353
14.870
434.504
11.040
0.010
0.009
mass as prey (see Section 3). Assuming that scallops would be fa-
1998
vored prey, their proportion in diet was set high (60–75%) and
–
Bi (t km2)
the remaining diets were calculated as above (Table 4). Diets for
5.869
28.270
7.925
70.679
9.974
51.398
324.892
1.774
0.315
25.244
82.134
27.781
11.567
7.049
1.360
69.204
28.104
0.052
0.056
fish species were obtained from FishBase (Froese and Pauly,

1996
2006) and were adjusted to the fish groups based on relative spe-
–
cies contribution from recorded catches.

1998
Trophic Level
1.00
1.00
2.26
2.06
2.00
2.10
2.00
2.00
2.22
2.98
2.99
3.09
3.03
3.15
2.99
2.26
3.26
3.39
3.33
1.00
2.5. Addressing parameter uncertainty

1996
1.00
2.23
2.06
2.00
2.10
2.00
2.00
2.24
2.93
2.96
3.35
3.11
3.57
2.86
2.24
3.24
3.45
3.30
1.00
1.00
The balanced steady-state model for 1996 was subjected to the

EwE resampling routine Ecoranger (Christensen and Walters, 2004)
Functional group/parameter
7. Herbivorous gastropods
in order to assess the probability distributions of the input param-

17. Pelagic predatory fish
10. Predatory gastropods
eters. Using a Monte Carlo approach, the routine drew a set of ran-
8. Benthic detritivores
18. Marine mammals

16. Small pelagic fish
9. Misc. filter-feeders
11. Small carnivores
dom input variables from normal distributions for each basic

12. Predatory crabs
1. Phytoplankton
parameter, and all resulting combinations that satisfied mass-bal-

15. Littoral fish
3. Zooplankton
4. Polychaetes
6. Sea urchins
2. Macroalgae
13. Sea stars
anced constraints were recorded. Originally we allowed the

19. Seabirds
14. Octopus
20. Detritus
5. Scallops
routine to use confidence intervals as derived from a pedigree of

Table 3
the data sources, where highest confidence is placed in locally-de-

rived data; however, the initial results often gave parameter values
342
Table 4
Diet matrices for steady-state trophic models of Independence Bay for 1996 and 1998 after application of the Ecoranger resampling routine (values of 0.000 indicates a proportion of <0.0005).
Prey/predator Model 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
1. Phytoplankton 1996 0.702 0.293 0.787 0.709 0.802
1998 0.698 0.301 0.821 0.721 0.795
2. Macroalgae 1996 0.811 0.783 0.191 0.235
1998 0.808 0.801 0.226 0.255
3. Zooplankton 1996 0.190 0.051 0.195 0.002 0.513 0.198 0.470
1998 0.208 0.047 0.175 0.002 0.340 0.205 0.521
4. Polychaetes 1996 0.090 0.398 0.462 0.292 0.384 0.207
1998 0.091 0.052 0.074 0.027 0.039 0.094
M.H. Taylor et al. / Progress in Oceanography 79 (2008) 336–351

5. Scallops 1996 0.018 0.011 0.011 0.014 0.065
1998 0.746 0.613 0.765 0.758 0.743
6. Sea urchins 1996 0.030 0.032
1998 0.059 0.034
7. Herbivorousgastropods 1996 0.046 0.037 0.032 0.033 0.069
1998 0.026 0.031 0.016 0.017 0.033 0.019
8. Benthic detritivores 1996 0.159 0.149 0.094 0.106 0.194 0.003
1998 0.055 0.072 0.028 0.034 0.056 0.015
9. Misc. filter feeders 1996 0.196 0.164 0.134 0.164 0.220
1998 0.049 0.056 0.025 0.029 0.030
10. Predatory gastropods 1996 0.108 0.145 0.255 0.001
1998 0.033 0.037 0.050 0.031
11. Small carnivores 1996 0.020 0.015 0.015 0.014 0.045 0.000
1998 0.021 0.030 0.012 0.016 0.027 0.024
12. Predatory crabs 1996 0.191 0.001
1998 0.033 0.028
13. Sea stars 1996 0.025 0.000
1998 0.031 0.093
14. Octopods 1996 0.153
1998 0.060
15. Littoral fish 1996 0.001 0.251 0.093
1998 0.026 0.098 0.100
16. Small pelagic fish 1996 0.001 0.530 0.699 0.907
1998 0.015 0.479 0.847 0.900
17. Pelagic predatory fish 1996 0.050
1998 0.055
18. Marine mammals 1996
1998
19. Seabirds 1996
1998
20. Detritus 1996 0.108 0.655 0.213 0.099 0.217 0.809 0.096 0.164 0.129 0.097 0.109 0.038
1998 0.094 0.653 0.179 0.101 0.199 0.774 0.104 0.051 0.063 0.032 0.036 0.059
outside of reasonable biological constraints (e.g. high conversion and is predicted to be higher in more mature ecosystems (Ula-
efficiencies, high cannibalism) and thus we decided to fix all confi- nowicz, 1986). The difference between development capacity and
dence intervals at 20% variation as was similarly done by Arias- ascendancy (C–A) is the system overhead (U) and gives a measure
González et al. (1997). We allowed resampling until 10,000 runs of the system’s ‘strength in reserve’ from which it can draw to meet
passed the selection criteria. The ‘best’ run was then chosen as that perturbations (Ulanowicz, 1986).
with the smallest sum of square residuals between the input
parameters and the mean value of all successful runs (for more 2.6.4. Fishery
information, see Christensen et al., 2000). Other statistics allow for the assessment of the fishery activity
such as its Gross efficiency (catch/net PP), mean trophic level of
2.6. Outputs/system statistics the catch, and primary production needed to sustain the fishery.
Statistics for comparison of the two system states fall under the 2.7. Simulating transition from upwelling to El Niño state
categories of community energetics, cycling indices, and system
organization. Comparisons of the ‘health’ and maturity of the two The simulation runs conducted for this study with EwE calcu-
system states drew on statistics from all three areas. Further gen- late biomass changes through time by solving the set of differential
eral descriptive statistics from the calculated outputs of the models equations:
included: (i) total throughput (T) – measure of the total sum of hX i X X
flows within the system and indicates the ‘size’ or activity of the dBi =dt ¼ g i k
Q ki ðtÞ Q ðtÞ M0i Bi
j ij
F if ðtÞBi ð5Þ
system; (ii) contributions to T from different types of flows – con-
For species or functional groups i = 1,. . ., n. The first sum repre-
sumption, export, respiration and flows to detritus; (iii) breakdown
sents the food-consumption rate, Q, summed over prey types k of
of biomass and flows from different components of the system –
species i, and gi represents the growth efficiency (proportion of
pelagic vs. benthic biomass and production; and (iv) changes in
food intake converted into production). The second sum represents
feeding modes – Herbivory: detritivory ratios.
the predation loss rates over predators j of i. M0i represents the
instantaneous natural mortality rate due to factors other than
2.6.1. Community energetics modelled predation. The final sum represents the instantaneous
Several indices of community energetics allowed for the com- fishing mortality rate, F, as a sum of fishing components caused
parison of ecological succession and relative maturity according by fishing fleets f.
to Odum (1969) and include: (i) total primary production (PP) to The Qij are calculated by assuming that the Bi are divided into
total respiration (R) ratio (PP/R); (ii) biomass (B) supported by total vulnerable and invulnerable components (Walters et al., 1997),
primary production (PP/B); (iii) biomass supported by total and it is the flux rates vij and v0ij that move biomass into the vulner-
throughput (B/T); and iv) energy transfer efficiency (TE) between able and invulnerable pools, respectively. This assumption leads to
discrete trophic levels. the rate equation:
aij ðtÞvij ðtÞBi Bj
2.6.2. Cycling indices Q ij ¼ ð6Þ
vij ðtÞ þ v0ij þ aij ðtÞBj
The Finn’s cycling index (FCI) (Finn, 1976) is calculated as Tc/T,
where Tc is the amount of system flows that are recycled compared where the vij and v0ij parameters represent rates of behavioral ex-
to the total system throughput, T. According to Odum (1969) recy- change between vulnerable and invulnerable states and aij repre-
cling increases in more mature and less stressed systems. sents rate of effective search by predator j for prey type i. The
exact setting of the vij, remains uncertain, but the modeling soft-
2.6.3. Growth and development indices ware allows for adjusting the vulnerabilities by a fitting procedure
Global measurements of system organization are calculated through which the sum of squares between observed and simulated
according to a network analysis based on flows among elements (log) biomasses are minimized (see Walters et al., 1997). In EwE, the
in the system as defined by Ulanowicz (1986). Indices include vulnerabilities for each predator–prey interaction can be explored
the aforementioned throughput (T), along with a measure of ascen- by the user and settings will determine if control is top-down
dancy (A), and development capacity (C). Ascendancy incorporates (i.e., Lotka-Volterra; >2.0), bottom-up (i.e., donor-driven; <2.0), or
both size and organization of flows into a single measure and is cal- intermediate (2.0). We applied this fitting routine with our time
culated as throughput (T) multiplied by mutual information (I), series, and the computed vulnerabilities were then discussed in
which concerns the diversity and evenness of flows between com- the light of possible control mechanisms operating in the
partments (Baird et al., 1998). Development capacity is the theo- ecosystem.
retical upper limit to ascendancy and thus the dimensionless A/C As input for simulations of the ecosystem response to ENSO we
ratio allows for a comparable measure of ecosystem development used catch per unit of effort (CPUE) time series for the fishery
Table 5
Biomass data for model groups derived from IMARPE benthic surveys in Independence Bay (1996, 1997, 1998, and 1999). Longer time series (1996–2003) were calculated from
estimates of catch per unit effort (CPUE). Relative CPUE changes were used to reconstruct the longer time series relative to the 1996 starting values from the steady-state model.
Year/groups Pp-1 Ma-2 Po-4 Sc-5 Su-6 Hg-7 Bd-8 Mf-9 Pg-10 Sc-11 Pc-12 Ss-13 Oc-14 Lf-15 Ppf-17
1996 51.4 69.2 324.9 7.0 7.9 25.2 70.7 82.1 28.1 10.0 27.8 11.6 0.3 1.8 1.4
1997 28.6 56.6 224.2 28.5 7.4 16.5 24.2 37.7 14.5 10.6 31.4 19.6 0.7 1.5 1.8
1998 28.6 7.6 43.5 564.2 10.9 6.7 13.8 8.2 9.8 6.8 4.5 20.1 0.2 0.2 2.7
1999 51.4 31.1 0.2 233.3 11.7 17.1 27.8 26.7 49.2 25.2 13.9 32.3 0.1 1.7 1.3
2000 51.4 120.6 29.8 0.1 2.4 0.3
2001 51.4 16.1 73.8 0.1 2.4 9.3
2002 51.4 2.7 41.6 0.1 3.5 3.5
2003 51.4 3.7 39.2 0.1 3.8 3.9
resources for the period 1996–2003 (including the EN year 1998) 3. Summary statistics are presented in Table 6. The ‘size’, as mea-
as proxies for stock biomass, together with biomass data obtained sured by the total system throughput (T), indicates that the 1996
from the benthic surveys done by IMARPE for the years 1996, 1997, state was larger than 1998 (34,208 vs. 24,827 t km2 yr1) mainly
1998 and 1999 (Table 5). as a function of higher primary production. Contributions to T from
To distinguish between trophic and non-trophic effects on func- different types of flows indicate that the EN state is characterized by
tional group biomass changes, we forced biomass changes of sev- higher absolute and relative flows due to consumption (11,918
eral functional groups (drivers) in the model to measure their t km2 yr1 and 48.0% of T) and respiration (7097 t km2 yr1 and
impact. Drivers included biomass changes in 4 highly-variable 28.6% of T) and lower absolute and relative flows into detritus
functional groups whose abundances are known to be at least par- (14.8% of T) and as exports (8.6% of T). These results indicate better
tially controlled by non-trophic environmental changes associated utilization of primary production through increased consumption
with ENSO: phytoplankton (PP), macroalgae (MA), predatory crabs and decreased losses to detritus as is reflected by the increased EE
(C) and scallops (S). We successively forced the biomass changes of values for phytoplankton and detritus compartments. The overall
these groups for the simulated time period of 8 years (1996–2003) ratio of herbivory to detritivory feeding decreased slightly during
and recorded the changes in fit as calculated by the sum of squares 1998 (6.54 and 5.22 for 1996 and 1998, respectively). Ratios be-
between the predicted and observed estimates. tween pelagic and benthic biomass and production were similar
An initial exploration of the dynamics using the default preda- for both 1996 and 1998 states with the benthic system dominating
tor–prey vulnerability settings for all interactions either decreased in terms of biomass (pelagic/benthic biomass ratios equal 0.13 and
the fit of the simulation or made only small improvements. Thus, 0.14 for 1996 and 1998, respectively) while the pelagic components
we decided to first introduce all four drivers in combination and al- accounted for most of the production (pelagic/benthic production
lowed EwE to search for the best predator–prey vulnerability set- ratios equal 8.46 and 7.79 for 1996 and 1998, respectively). Besides
tings. Using these optimized vulnerability settings we again major changes in primary production between the two periods,
addressed the importance of each driver through single or com- which greatly impacted T, the overall biomasses of trophic levels
bined introduction to force the model through time. II and above are virtually unchanged despite significant changes
to several individual functional groups.
3. Results
3.2. Community energetics
3.1. General descriptive
Several statistics on community energetics point to EN condi-
Initial parameters of the balanced model can be found on the tions as being of a higher ‘maturity’ than normal conditions. The
Pangaea website (Taylor et al., 2007a, b). The Ecoranger resampling primary production to total respiration ratio (PP/R) came closer
routine resulted in balanced models in 0.75% and 2.20% of the runs to the proposed value of 1.0 for mature systems (Odum, 1969)
for the 1996 (normal conditions) and 1998 (EN conditions) models, (2.979 in 1996; 1.302 in 1998). Total primary production to bio-
respectively. The ‘best’ fitting model parameters are shown in Table mass (PP/B) and biomass to total throughput (B/T) ratios indicated
Table 6
System statistics, cycling indices, and informational indices for the two modeled periods of Independence Bay. Changes in values from the 1996 state to the 1998 state are given as
a percent; values in brackets are in percent of total system throughput.
Summary statistics 1996 1998 % Change

2 1
Sum of all consumption (t km yr ) 8389 (24.5%) 11,919 (48.0%) 42.1
Sum of all exports (t km2 yr1) 9444 (27.6%) 2145 (8.6%) 77.3
Sum of all respiratory flows (t km2 yr1) 4772 (14.0%) 7097 (28.6%) 48.7
Sum of all flows into detritus (t km2 yr1) 11603 (33.9%) 3666 (14.8%) 68.4
Total system throughput (t km2 yr1) 34208 24827 27.4
Sum of all production (t km2 yr1) 16133 11610 28.0
Calculated total net primary production (t km2 yr1) 14214 9242 35.0
Net system production (t km2 yr1) 9442 2146 77.3
Total biomass (excluding detritus) (t km2) 754 674 10.6
Pelagic/benthic biomass 0.13 0.14 15.6
Pelagic/benthic production 8.46 7.79 8.0
Connectance index 0.222 0.224 0.9
System omnivory index 0.169 0.122 27.8
Herbivory/detritivory 6.54 5.22 20.2
Fishing
Total catches (t km2 yr1) 12.605 248.930 1874.9
Mean trophic level of the catch 2.73 2.05 24.9
Gross efficiency (catch/net PP) 0.001 0.027 2936.5
PP required/catch 29.39 9.26 68.5
PP required/total PP (%) 1.43 17.85 1148.3
Community energetics
Total primary production/total respiration 2.979 1.302 56.3
Total primary production/total biomass 18.861 13.715 27.3
Total biomass/total throughput 0.022 0.027 22.7
Cycling indices
Finn’s cycling index (% of total throughput) 5.11 8.88 73.8
Predatory cycling index (% of throughput w/o detritus) 9.07 5.14 43.3
System development
System overhead/capacity (%) 67.0 72.5 1.2
Ascendancy/capacity (%) 33.0 27.5 16.7
TE = 9.6% TE = 6.7%
V 4.3%
V 4.6% IV 5.6%
IV 8.3%
III 8.3%
III 13.1%
II 6.6%
II 8.1%
t/km2/year 1996 1998
Fig. 3. Modified Lindeman pyramids of flows for steady-state models of Independence Bay. Transfer efficiencies are given for discrete trophic levels. Mean transfer efficiency
is the geometric mean of trophic levels II–IV.
that the 1998 state could support a higher relative biomass per combined with the decreased primary productivity, resulted in a
unit of primary production and total throughput. On the contrary, value of 18% of total primary production needed to sustain the fish-
mean transfer efficiency (TE) was higher for the 1996 state (Fig. 3) ery – an 11-fold increase from 1996.
due in part to a high utilization of herbivore and detritivore pro-
duction by predatory invertebrates, as well as higher cannibalism, 3.6. Functional group responses to forcing scenarios
and can be observed in the high EE values for these groups (groups
5–14, Table 3). This ‘bottleneck’ of flows did not occur in 1998 due The results of the forcing of trophic driver biomasses on the
both to a decrease in predator biomass and an increase in primary dynamics of remaining functional groups are presented for four
consumer biomass due to the proliferation of scallops. As TE can scenarios in the following paragraphs.
only be calculated for consumer groups, and Ecopath does not
quantify solar energy input to producer compartments, mean TE 3.6.1. Scenario 1 (S1): Decrease in primary production during EN (due
reflects the geometric mean of trophic levels II–IV only. Thus, the to lack of nutrient upwelling)
decrease in TE occurred despite an overall improvement in other As shown by Fig. 4, a decrease in phytoplankton and macroalgae
holistic community energetic indices in 1998; specifically, a higher biomass during EN (1997/98) resulted in decreases in polychaetes,
utilization of primary production and detritus. misc. filter-feeders and herbivorous gastropods. A slightly lagged
response is also seen in predatory gastropods, which decreased
3.3. Cycling indices in biomass. While the single addition of the macroalgae driver de-
creased SS more than did that of the phytoplankton driver (8.1%
A higher degree of cycling, as indicated by the Finn’s cycling infor macroalgae; 2.7% for phytoplankton), the average change in
dex, was calculated for the EN period (5.11% for 1996; 8.88% for combination with other drivers was greater from the phytoplank-
1998). Again, the higher utilization of primary production and ton driver at 2.8% (Fig. 5).
detritus was mainly responsible for this result. Removing this
influence is possible with the related Predator cycling index, which 3.6.2. Scenario 2 (S2): Decrease in predatory crab biomass during EN
showed that the 1996 state had more cycling at the higher trophic (due to temperature stress causing mortality and migration to deeper
levels (9.07% for 1996; 5.14% for 1998). waters)
The application of this driver resulted in a small increase in bio-
3.4. Growth and development indices mass of the groups sea stars and small carnivores as a result of the
reduced crab biomass (Fig. 4). The application of the predatory crab
The ascendancy to development capacity ratio (A/C) was slightly driver resulted in an average change of 4.8% in SS (Fig. 5).
higher during normal upwelling conditions in 1996 (33.0% for
1996; 27.5% for 1998) and indicates that this state shows more 3.6.3. Scenario 3 (S3): Increase in scallop biomass during EN (due to
maturity (i.e. higher total flows and predictability of flows). increased recruitment and growth)
Fig. 4 shows the functional group responses to the increased
3.5. Fishery Scallop biomass during the EN warming, which included: (1) abun-
dance increases in predatory gastropods, small carnivores, octopus,
The boom of A. purpuratus during EN was mostly responsible for sea stars, and (2) abundance decreases in the groups polychaetes,
the more than 18-fold increase in total catches for the model area, herb. gastropods, benthic detritivores, and misc. filter-feeders.
to 248.9 t km2 yr1. Pelagic predatory fish catches also increased The model also predicts an increase in predatory crab biomass,
about 7-fold, and as a result the model back-calculated a higher which is contrary to the observed decrease, supporting the obser-
small pelagic fish biomass for 1998. The gross efficiency (catch/ vation that EN warming likely induced a non-trophically mediated
net PP) of the fishery increased 25-fold and the primary production mass mortality and emigration of crabs to deeper, cooler waters
required per unit of catch decreased, due mainly to the lower tro- (Arntz and Fahrbach, 1991). Despite some improvements, the aver-
phic level of the scallop (mean TL of catch – 2.74 and 2.05 for 1996 age change from the application of the scallop driver was an in-
and 1998, respectively). The variable nature of the diving effort in crease of 1.8% in SS (Fig. 5), indicating a decrease in fit.
response to changing resource abundances also played an impor-
tant role. As the catch of scallops mainly drove the changes in ef- 3.6.4. Scenario 4 (S4): Combined forcing of all four drivers (scallops,
fort, scallops show fairly similar fishing mortality (F) values for phytoplankton, macroalgae, and predatory crabs)
the two periods, while other groups that were reduced in biomass The previously mentioned improvements from each driver sum
during 1998 show higher F values (misc. filter-feeders and preda- upto explain the dynamics in the majority of groups (Fig. 4). In most
tory gastropods) (Table 3). Overall, the expansion of the fishery, cases, the forced dynamics are similar to a dominating individually
Phytoplankton Macroalgae Polychaetes
Scallops Sea urchins Herbivorous gastropods
Benthic detritivores Misc. filter feeders Predatory gastropods
Small carnivores Predatory crabs Sea stars
Octopods Littoral fish Pelagic predatory fish
Force Phytoplankton, Macroalgae (S1)

Force Predatory crabs (S2)
Force Scallops (S3)
Force all drivers (S4)
Base values
Fig. 4. Simulated versus observed biomass changes. All simulations consider changes in fishing effort (fishing and diving). Simulation trajectories are shown for each of the
three scenarios (S1, bottom-up effect of reduced primary production – ‘‘Force Phytoplankton, Macroalgae”; S2, top-down effect of reduced benthic predation – ‘‘Force
predatory crabs”; and S3, effect of scallop proliferation – ‘‘Force scallops”) plus a combination of all four drivers applied together (S4, ‘‘Force all drivers”).
forced driver; however, the dynamics of small carnivores and sea 34,208 t km2 yr1) is higher than has been observed for other
stars are improved by the combined application of all four drivers. coastal zones along the Pacific coast, specifically, Gulfo Dulce, Costa
Rica (T = 1404) and Tongoy Bay, Chile (T = 20,835 t km2 yr1)
3.7. Vulnerability estimates (Wolff, 1994; Wolff et al., 1996), due mainly to its high primary
production. When our results are compared to models of specific
Table 7 summarizes the vulnerabilities computed for Scenario habitats in Tongoy Bay as constructed by Ortiz and Wolff (2002),
4. v-values <1.2 were considered bottom-up control (BU), between the sand-gravel habitat is most similar in terms of total throughput
1.2 and 2 (mixed control, MX) and above 2 top-down control (TD). (T = 33,579 t km2 yr1). This type of substrate is typical of Inde-
Accordingly, top-down control is suggested for: (i) predatory pendence Bay and is associated with strong currents where oxygen
gastropods on polychaetes, benthic detritivores and misc. filter- concentrations are increased through mixing and circulation and
feeders; (ii) predatory crabs on scallops; and (iii) sea stars on pred- permit higher macrofaunal biomass. Similar values of production,
atory gastropods. energy flows to detritus, respiration, and exports are also observed
Bottom-up control configurations are more dominant and are between this habitat in Tongoy Bay and the model of Independence
suggested for: (i) polychaetes to predatory crabs; (ii) scallops to Bay under upwelling conditions.
predatory gastropods and octopus; (iii) primary producers and While our estimate of Total throughput is not directly compara-
zooplankton prey to fish groups; and (iv) littoral fish and small pe- ble to models that use differing units to describe flows (e.g. dry
lagic fish to marine mammals and seabirds. weight or carbon units), we can compare the proportions of differ-
ent types of flows. Flow to detritus in Independence Bay during
4. Discussion 1996 (33.9%) is similar to that of Tongoy (29%) as well as several
US bay systems; e.g. Narragansett Bay (33%), Delaware Bay (30%),
4.1. Summary statistics, flow structure and maturity and Chesapeake Bay (27%) (Monaco and Ulanowicz, 1997). How-
ever, only the models of the South American bays calculated high
The total energy throughput of the Independence Bay ecosys- proportions of exports as well (29–34% vs. 7–10% for US bays). Part
tem under normal upwelling conditions (1996 model; T = of the difference may be attributable to higher exchange rates/low
10.0 month (October 1998) compared to normal levels of around 750.

Fishers also shifted effort to almost exclusively target scallops,
7.5 yet other soft-bottom species were also taken. Octopus (Octopus
mimus) is a particularly favored resource due to high price, and also
5.0
increases in biomass during EN (Arntz et al., 1988). O. mimus
2.5 growth and reproduction have been shown to increase under war-
mer conditions (Cortez et al., 1999) and embryonic development
(ave. and range)
% change in SS
0.0 time is also greatly accelerated under EN-like conditions in the lab-
oratory (Warnke, 1999). The increased catch of octopus is thus
-2.5
likely supported by increased production. Catch of pelagic preda-
-5.0 tory fish also increased, which may be explained through the over-
all shrinkage of the upwelling zone during EN and the subsequent
-7.5 intrusion of oceanic waters, which several predatory fish species
are associated with (e.g. Scomber japanicus, Sarda chilensis, and
-10.0
Scomberomerus sierra). This movement may be further related to
-12.5 the pursuit of prey, as anchovy stocks were observed to both con-
centrate near the coast and then retreat southward to the latitudes
s
n
ab near Independence Bay as recorded by acoustic surveys (Ñiquen

e
to
ps
ga
cr
nk
lo
al
and Bouchon, 2004).

la
y
al
ro
or
op
Sc
ac
at
yt
The expansion of the diving fishery during EN is also observed

ed
M
Ph
Pr
through much higher indices of gross efficiency (catch/net PP) such

Drivers
that 18% of total primary production required to sustain the fish-
Fig. 5. Percent changes to sum of squares, SS, of the 1996–2003 simulation after the ery. This value is lower than the value (25.1%) calculated by Pauly
forcing of biomass changes of several functional groups ‘drivers’. Drivers were and Christensen (1995) for upwelling systems, and may reflect low
applied in all possible sequences and combinations and SS was corrected for mean trophic level of the fishery (2.05). Nevertheless, for an arti-
artificial improvements caused by the fitting of the driver’s dynamics. Average
change (bar) and range (line) are displayed. Negative values (i.e. decrease in SS)
sanal fishery, it shows a remarkable efficiency of harvest during
indicate an improvement in fit. EN. On the contrary, the value for normal conditions is extremely
low at 1.4%, and illustrates the low level of exploitation of the fish-
ery during normal upwelling periods, likely due to the low abun-
residency time of water in relatively open bays like Independence dance of higher priced species like scallops and octopus. As a
and Tongoy, resulting in more export of production (Rybarczyk result, Independence Bay fishers are moving towards a combina-
et al., 2003); however, the high degree of primary production going tion of fishing and culture of scallops to maintain income levels be-
unutilized and remaining in the sediments may be more typical of tween EN ‘‘boom times”.
upwelling systems. Nixon (1982) showed that there is a highly positive correlation
The dynamic nature of the artisanal fishery in response to between primary production and fishery yield in coastal lagoons,
changes in resources helps maintain the ecosystem’s efficiency in yet Independence Bay catches are highest during the low primary
the face of reduced predation pressure. In response to the scallop production characteristic of EN. While the fish catches also in-
boom during EN, fishers migrated to Independence Bay. A main creased during EN mainly due to immigrations of fish towards
proportion of these migrant fishers were from Sechura Bay in the the coast, the catch of benthic resources increased the most. In-
north of Peru (6 °S), where the largest fishery for scallops is nor- creased oxygen concentration during EN has been suggested as
mally found. These fishers were mainly involved in the diving fish- important in the proliferation of benthic species (Arntz and Fahr-
ery, which increased in effort by 170% in 1998 compared to the bach, 1991). Overall consumption of primary production by several
previous year. Peak diving effort reached 4932 boat trips per primary consumers (i.e. scallops, herbivorous gastropods, and ben-
Table 7
Vulnerabilities calculated by EwE with the application of all four drivers (phytoplankton, macroalgae, scallops, predatory crabs). BU = Bottom-up control (vulnerability 2.0),
TD = top-down control (vulnerability 2.0), MX = mixed/intermediate control (vulnerability values between 1.2 and 2.0).
Prey/predator 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
1 Phytoplankton BU MX MX MX BU
2 Macroalgae MX MX MX BU
3 Zooplankton TD MX MX MX BU BU BU
4 Polychaetes MX TD MX BU MX BU
5 Scallops BU MX TD MX BU
6 Sea urchins MX MX
7 Herbivorous gastropods MX MX MX MX MX
8 Benthic detritivores TD MX MX MX BU MX
9 Misc. filter-feeders TD MX BU MX BU
10 Predatory gastropods BU TD BU MX
11 Small carnivores MX MX MX MX MX MX
12 Predatory crabs MX MX
13 Sea stars MX MX
14 Octopus TD
15 Littoral fish MX BU BU
16 Small pelagic fish MX BU BU BU
17 Pelagic predatory fish MX
18 Marine mammals
19 Seabirds
20 Detritus MX BU MX MX MX MX MX BU MX MX MX BU
thic detritivores) increased in order to sustain their increased bio- It does appear that the higher trophic levels were more severely
masses – as calculated from in situ or empirically-based estimates. impacted by EN in Independence Bay. Furthermore, the most sig-
As a result, primary production during EN appears to be almost nificant short intense loop would be the cycle through detritus,
completely consumed in Independence Bay, although several which increases during EN and results in the higher Finn’s cycling
assumptions were made concerning the levels of primary produc- index (FCI). As phytoplankton and macroalgae production were re-
tion. However, near-complete use of primary production during EN duced during EN (9247 t m2 y1 during EN; 14,214 t m2 y1 dur-
may be indicated by the clear, tropical-like water and decreases in ing normal conditions), and consumption of detritus by scallops
benthic detrital material observed during EN. Based on the model, actually increased, the proportion of recycled detritus is signifi-
recruitment and production increases of scallops account for this cantly higher in 1998.
result, as scallops consume 58% of phytoplankton production dur- This result depends on the decreased primary production dur-
ing EN. Wolff et al. (2007) found that the increase in scallops was ing EN. To illustrate, we can increase the primary production in
likely a non-trophic effect resulting from increased larval survival the EN model to the levels of 1996 in order to observe a less-biased
in warmer temperatures. This recruitment success combined with comparison (Fig. 6). The EN Finn’s cycling index (FCI) decreased to
increased oxygen concentrations is likely the main cause of the in- a slightly lower value than 1996. However, the relative Ascendancy
crease in fisheries yield. (A/C) and Overhead (U/C) increase and decrease, respectively, but
Indicators of system maturity show some contradictions – some not to the levels of 1996. Ascendancy is both a function of total
suggest normal conditions are more mature and developed while throughput (T) and system development (i.e. average mutual infor-
others for the EN state (1998). From a community energetics point mation, I), and while the increase in primary production brings T to
of view, the EN state is able to support a similar biomass compared a similar level as 1996, the EN state still shows lower development,
to 1996 despite lower primary production (PP/B ratio) and total I. Under this scenario, EN would appear as of lower maturity de-
throughput (B/T ratio), and thus the system’s primary production spite increased overall community energetics.
to respiration ratio is closer to the value of 1.0 predicted for mature Our models do not include information on the microbial loop,
and efficient systems (Odum, 1969). Similarly, an increased Finn’s which is undoubtedly an important component of ecosystem func-
cycling index is observed during EN due to a better utilization of tioning in Independence Bay. Energy flow through bacteria is likely
primary production and detritus by the primary consumers enhanced during the warm, oxygen-rich conditions of EN. While
(mainly scallops). These larger energy flows at lower trophic levels bacterial functional groups are often removed in other models be-
offset the negative impacts of EN at the higher trophic levels. Dur- cause their high flows overshadow other groups (Christensen,
ing EN the transfer efficiencies (TE) of higher trophic levels are de- 1995), they may be of particular importance in our understanding
creased and contribute to an overall lower mean TE, due to the of benthic processes in the Peruvian upwelling system. Thus, future
negative impact of EN on benthic predatory groups (predatory gas- research plans to investigate these important energy pathways for
tropods, small carnivores, and especially predatory crabs). These use in future models.
impacts are also observed through a decreased predatory cycling A community analysis for Independence Bay conducted by
index and Finn’s mean path length during EN, indicating poorer cy- Wolff and Mendo (2002) indicated that benthic diversity and even-
cling and transfer of energy in the higher trophic levels of the food ness increased during EN. An initial attempt to model the trophic
web. changes also showed maintenance of flow structure during EN.
Relative ascendancy (A/C) indicates slightly more ecosystem The authors proposed that this adjustment to abiotic changes
development and maturity during 1996 (33.0% in 1996; 27.5% in might indicate that EN is a condition to which the benthic commu-
EN). Related is the percent overhead (U/C), which indicates that nity has adapted during evolution. This hypothesis is supported by
the less mature EN state may be better able to withstand perturba- the present study’s results, yet may best apply to lower trophic lev-
tion. Baird et al. (1991) found a similar discrepancy when compar- els that responded quicker to the perturbation. High trophic level
ing A/C to FCI in several marine ecosystems, where a negative benthic predatory groups have been observed to recover quite
correlation between indices was observed even though both quickly (e.g. predatory crabs) after EN, but this is likely due to tem-
should have increased with system maturity. They suggested that porary emigrations to deeper waters rather than system adjust-
the discrepancy may lie in the fact that stress frequently impacts ment. In this respect we wonder if the post-EN ecosystem, with
higher-level species more than lower-level species. As a result, higher primary production, higher residual scallop biomass, and
the release of standing biomass of higher trophic levels can be ta- a return of predatory groups, might not show higher flows or more
ken up through increased recycling via ‘‘short intense loops”. maturity than the pre-EN state. If so, there may be some long-term
Fig. 6. Simulation of increasing primary production in the 1998 model to normal upwelling levels (1996 model values for macroalgae and phytoplankton) on indices of
Ascendancy (A/C), overhead (U/C), mutual information (I), and Finn’s cycling (FCI). Reference values for the 1996 model’s indices are given for comparison.
positive impact from EN that may foster the idea that it is an inte- in the Ecopath Pedigree Index, Peru system, 0.638; IB, 0.597;
gral part of HCS dynamics (Arntz and Valdivia, 1985b). scales between 0 and 1 with highest values for direct measure-
ments of the same system). Furthermore, the less-mobile nature
4.2. Bottom-up and top-down effects of the benthic organisms in Independence Bay may prevent
avoidance of deleterious conditions, thus making the effects of per-
When the model for normal conditions was forced from below turbations more pronounced. While the computed vulnerabilities
with a reduction in primary producer biomass (phytoplankton and for Independence Bay seem plausible, they should nevertheless
macroalgae to EN levels) the response confirms some of the EN- be considered with caution since the time series available for the
associated changes observed in functional group biomasses: misc. present study was quite short. In future years we will be able to
filter-feeders and herbivorous gastropods strongly decrease, and extend the data set over longer periods and may be able to confirm
polychaetes benthic detritivores which also decrease somewhat. some of the estimated vulnerabilities.
Oddly, macroalgae, when used as a single model driver, better ex- Generally, bottom-up control dominated the fit for the pelagic
plain the observed ecosystem changes (lowering SS) than components such as energetic flow from plankton to small pelagic
phytoplankton. fish to the higher predators marine mammals and seabirds. Littoral
Forcing a reduction in predatory crab biomass (release of top- fish also provided an important bottom-up link between benthic
down control), favors sea stars and small carnivores (competing production (both macroalgae and invertebrate) to marine mam-
predators), yet the modeled responses of other groups of the sys- mals. Top-down control was more important in the benthic com-
tem is insignificant. ponents of the system. This may be expected given the high
Neither EN triggered changes in the bottom-up (phytoplankton Ecological efficiencies calculated for many benthic primary con-
and macroalgae), nor the top-down (predatory crabs) forcing im- sumers during the normal upwelling year of 1996, due to their high
prove the fit of scallop dynamics, suggesting that trophic linkages utilization by higher trophic levels. In addition, the fact that only
of scallops to their food and predators do not cause proliferation 0.75% of the Ecoranger runs for 1996 resulted in a balanced model
of the scallop stock—an important if negative finding, especially (as compared to 2.20% in the 1998 model) illustrates the tightly
since predatory crabs are well known scallop predators and their coupled flows to the benthic predatory groups, which restricted
biomass reduction during the EN warming has been previously re- the parameter possibilities for the starting 1996 steady-state
lated to the scallop proliferation (Wolff and Alarcon, 1993; Wolff model.
and Mendo, 2000). A top-down configuration was fit for the predatory crab to scal-
While the scallop outburst during EN changed the entire char- lop interaction. This is possible during normal upwelling periods as
acter of the ecosystem, its inclusion as a model driver did not im- the crabs Cancer setosus and Cancer porteri are the dominant con-
prove the overall fit of the simulation considerably. This may in sumers of benthic production; however, the crab decrease during
part be due to lags in the population responses of several func- EN is not evidently responsible for the scallop outburst. Further-
tional groups as compared to the reference data; however, the sim- more, this vulnerability setting must be taken with caution as
ulation correctly predicts the direction of response for a number of the biomass fluctuations of both groups were forced through time
functional groups (positive: predatory gastropods, small carni- and thus the result is likely an artifact. Top-down configurations
vores, octopus, sea stars; Negative: polychaetes, herb. gastropods, between predatory snails and several of its prey (polychaetes, ben-
and misc. filter-feeders), supporting the central role of the scallop thic detritivores and misc. filter-feeders) help to explain these prey
in the Independence Bay ecosystem as prey for several consumer decreases after the EN period. But this result too must be taken
groups and as a competitor for other filter-feeders. It is likely that with caution as competitive interactions with scallops for space
the more immediate decreases in several competing primary con- may have also attributed to their declines.
sumer groups may be due to the negative effects of competition for The finding that the abundance of scallop and other Indepen-
space, as the scallop banks became so thick in parts as to obscure dence Bay filter-feeders exerts bottom-up control on predators’
the sea floor with several layers of scallops. abundance appears plausible and the finding is not new (Wolff
The scallop outburst apparently is caused by non-trophic effects and Alarcon, 1993). Despite the negative effects of EN on several
(i.e. temperature mediated recruitment). However, once the scal- higher-level benthic predators, the increased scallop biomass after
lops proliferated, the model suggests great changes to energy flow EN apparently supported the recovery of predatory gastropods,
within the system. During EN scallops proliferated and the biomass small carnivores, predatory crabs, and sea stars, all of which show
of primary producers and predatory crabs decreased, affecting higher post-EN biomasses than in 1996. Furthermore, the (possibly
other groups in our simulations (scenario 4) (Fig. 4). normal) bottom-up control of scallops and other filter-feeders by
When forcing by the diving and finfish fishery is removed in our phytoplankton under upwelling conditions may indeed be inverted
simulations, the simulated biomass trajectories of the functional during EN, when scallops are estimated to have consumed 58% of
groups were almost identical to those of scenario 4, suggesting that phytoplankton production alone. A similar role has been identified
fishing plays a very limited role in the dynamics of the system. This for the introduced Manila clam Tapes philippinarum in the Venice
may be explained by the fact that the diving fishery targeted Lagoon system (Pranovi et al., 2003). Furthermore, it was sug-
mainly scallops and its increase in catch rate was about propor- gested that this strong top-down control of phytoplankton by T.
tional to the scallop biomass increase; and the changes in finfish philippinarum may be responsible for the system not returning to
fishing rate were small over the whole period. a phytoplankton-based trophic web (Libralato et al., 2004),
although this seems unlikely in Independence Bay given the con-
4.3. Vulnerabilities stant refreshment of productive waters that enter from outside
the bay.
It is important to emphasize that the manipulation of the ‘driv-
ers’ did not improve the fit of the simulation without first allowing 4.4. Conclusions
for a fitting of vulnerabilities. This is contrary to the findings of
a similar exploration of the larger Peruvian Upwelling system Overall, it appears that the energy flow structure in Indepen-
(Taylor et al., 2008) whereby even default vulnerability values dence Bay is more or less maintained during El Niño despite nega-
reproduced many important dynamics. This may be due to higher tive impacts at higher benthic trophic levels. In particular, the
data quality in steady-state model (Tam et al., 2008) (as reflected proliferation of the scallop A. purpuratus apparently maintains
the energy flow within Independence Bay despite the reduction in Brey, T., 2001. Population dynamics in benthic invertebrates. Available from:
<http://www.awi-bremerhaven.de/Benthic/Ecosystem/FoodWeb/Handbook/
primary production. While some alleviation of top-down predation
main.html>.
pressure may be felt by benthic primary consumers through the Brown, P.C., Painting, S.J., Cochrane, K.L., 1991. Estimates of phytoplankton and
temperature–mediated decreases of crabs, the overall bottom-up bacterial biomass production in the northern and southern Benguela
affects of reduced primary production (macroalgae and phyto- ecosystems. South African Journal of Marine Science 11, 537–564.
Brush, M.J., Brawley, J.W., Nixon, S.W., Kremer, J.N., 2002. Modeling phytoplankton
plankton) appears to have reduced the biomass of several func- production: Problems with the Eppley curve and an empirical alternative.
tional groups. As seen for many areas along the Peruvian coast, Marine Ecology Progress Series 238, 31–45.
Independence Bay becomes more tropical during EN. Under these Carr, M.E., 2002. Estimation of potential productivity in eastern boundary currents
using remote sensing. Deep-Sea Research II 49, 59–80.
conditions the system utilizes most of the (reduced) phytoplank- Chavez, F., Barber, R.T., 1985. La productividad de las aguas frente a la costa del
ton production so that exports of primary production to detritus Perú. Boletín ERFEN 15, 9–13.
are greatly reduced. While the El Niño state appears to show some Christensen, V., 1995. Ecosystem maturity – towards quantification. Ecological
Modelling 77, 3–32.
higher efficiency in overall energetics, the structure and develop- Christensen, V., Pauly, D., 1992. ECOPATH II – a software for balancing steady-state
ment appears impacted. models and calculating network characteristics. Ecological Modelling 61, 169–
The rapid response and adaptation of the artisanal fishery to EN 185.
Christensen, V., Walters, C.J., 2004. Ecopath with Ecosim: methods, capabilities and
also increases the system’s efficiency; however, this increased fish- limitations. Ecological Modelling 172, 109–139.
ing pressure may have added stress to negatively impacted func- Christensen, V., Walters, C.J., Pauly, D., 2000. Ecopath with Ecosim Version 4, Help
tional groups through higher fishing mortality. A management systemÓ.
Cortez, T., Gonzalez, A.F., Guerra, A., 1999. Growth of Octopus mimus (Cephalopoda,
plan that allows for the newly recruited A. purpuratus population
Octopodidae) in wild populations. Fisheries Research 42, 31–39.
to fully grow and develop may not only reap higher monetary gains Delgado, E., Villanueva, P., 1998. Peruvian coastal phytoplankton community during
as suggested by Wolff and Mendo (2000), but may also enhance Cruise RV Humboldt 9803–05 from Tumbes to Tacna. Informe. Instituto del Mar
post-El Niño system through facilitation of the recovery of benthic del Peru, Callao. pp. 114–120.
Eppley, R.W., 1972. Temperature and phytoplankton growth in the sea. Fish Bulletin
predatory groups. 70, 1063–1085.
Finn, J.T., 1976. Measures of ecosystem structure and function derived from analysis
Acknowledgements of flows. Journal of Theoretical Biology 56, 363–380.
Froese, R., Pauly, D., 2006. FishBase. Available from: <www.fishbase.org>.
Huebner, J.D., Edwards, D.C., 1981. Energy Budget of the Predatory Maine Gastropod
The authors are grateful for the support and assistance from the Polinices duplicatus. Marine Biology 61, 221–226.
following: Dr. Villy Christensen of the Fisheries Centre, University Hutchings, L., Verheye, H.M., Mitchell-Innes, B.A., Peterson, W.T., Huggett, J.A.,
Painting, S.J., 1995. Copepod production in the southern Benguela system. ICES
of British Columbia, for his helpful advice regarding the use of Eco- Journal of Marine Science 52, 439–455.
ranger routine within Ecopath with Ecosim; Dr. Tom Brey of the Jarre, A., Muck, P., Pauly, D., 1991. Two approaches for modelling fish stock
Alfred Wegener Institute for Polar and Marine Research (AWI) for interactions in the Peruvian upwelling ecosystem. ICES Marine Science
Symposia 193, 178–184.
helpful discussions regarding benthic invertebrate energetics and Keen, M.A., 1972. Sea Shells of Tropical West America – Marine Mollusks from Baja
for the use of his somatic production models; Dr. Carl Walters for California to Peru. Stanford University Press, Stanford, California.
the use of the Ecosim software. This study was financed and con- Kremer, J.N., Nixon, S.W., 1977. A Coastal Marine Ecosystem. Simulation and
Analysis. Springer-Verlag, New York, USA.
ducted in the frame of the EU-project CENSOR (Climate variability
Lang, M., 2000. Populationsstruktur und Konsumverhaltung der decapoden Krebse
and El Nino Southern Oscillation: Impacts for natural resources and Cancer polydon und Cancer porteri in der Independencia Bucht, Peru. Diplome
management, Contract 511071) and is CENSOR publication 0050. thesis. Bremen University, Bremen, unpublished.
Leon, R.I., Stotz, W.B., 2004. Diet and prey selection dynamics of Cancer polyodon in
three different habitat types in Tongoy Bay, Chile. Journal of the Marine
References Biological Association of the United Kingdom 84, 751–756.
Libralato, S., Pranovi, F., Raicevich, S., Da Ponte, F., Giovanardi, O., Pastres, R.,
Arias-González, J.E., Delesalle, B., Salvat, B., Galzin, R., 1997. Trophic functioning of Torricelli, P., Mainardi, D., 2004. Ecological stages of the Venice Lagoon analysed
the Tiahura reef sector, Moorea Island, French Polynesia. Coral Reefs 16, 231– using landing time series data. Journal of Marine Systems 51, 331–344.
246. Macchiavello, J., Fonck, E., Edding, M., 1987. Antecedentes y perspectivas del Cultivo
Arntz, W., Valdivia, E., 1985a. Incidencia del fenómeno ‘El Niño’ sobre los mariscos de Gracilaria en Coquimbo. In: Arana, P. (Ed.), Manejo y Desarrollo Pesquero.
en el litoral peruano. In: Arntz, W., Landa, A., Tarazona, J. (Eds.), ‘El Niño’ y su Universidad Catolica de Valparaíso, Valparaiso, Chile, pp. 206–214.
impacto en la fauna marina. Instituto del Mar del Perú, Callao, Peru, pp. 91–101. Mann, K.H., 1982. Ecology of Coastal Waters. A Systems Approach. University of
Arntz, W., Valdivia, E., 1985b. Visión integral del problema ‘El Niño’: introducción. California Press, Berkeley, USA.
In: Arntz, W., Landa, A., Tarazona, J. (Eds.), ‘El Niño’ y su impacto en la fauna Martin, D., Grémare, A., 1997. Secondary production of Capitella sp. (Polychaeta:
marina. Instituto del Mar del Perú, Callao, Peru, pp. 5–10. Capitellidae) inhabiting different organically enriched environments. Scientia
Arntz, W., Tarazona, J., 1990. Effects of El Niño on benthos, fish and fisheries off the Marina 61, 99–109.
South American Pacific coast. In: Glynn, P.W. (Ed.), Global ecological Mendo, J., Wolff, M., 2002. Pesqueria y manejo de la concha de abanico (Argopecten
consequences of the 1982–83 El Niño-Southern Oscillation. Elsevier purpuratus) en la Bahia de Independencia. In: Mendo, J., Wolff, M. (Eds.), Bases
Oceanography Series, pp. 323–360. ecologicas y socioeconomicas para el manejo de los recursos vivos de la Reserva
Arntz, W.E., Fahrbach, E., 1991. El Nino-Klimaexperiment der Natur. Birkhaeuser National de Paracas. Universidad Agraria La Molina, Lima, pp. 188–194.
Verlag, Basel, Switzerland. Mendo, J., Valdivieso, V., Yamashiro, C., Jurado, E., Morón, O., Rubio, J., 1987.
Arntz, W.E., Valdivia, E., Zeballos, J., 1988. Impact of El Nino 1982–83 on the Evaluación de la población de concha de abanico (Argopecten purpuratus) en la
commercially exploited invertebrates (mariscos) of the Peruvian shore. Bahía Independencia, Pisco, Peru, 17 de enero – 4 de febrero de 1987 Informe
Meeresforschung 32, 3–22. del Instituto del Mar del Perú N°94, Callao, Peru.
Arntz, W.E., Tarazona, J., Gallardo, V.A., Flores, L.A., Salzwedel, H., 1991. Benthos Mendoza, J.J., 1993. A preliminary biomass budget for the northeastern Venezuela
communities in oxygen deficient shelf and upper slope areas of the Peruvian shelf ecosystem. In: Christensen, V., Pauly, D. (Eds.), Trophic Models of Aquatic
and Chilean Pacific coast, and changes caused by El Niño. Geological Society 58, Ecosystems, vol. 26. ICLARM Conference Proceedings, pp. 285–297.
131–154. Moloney, C.L., Jarre, A., Arancibia, H., Bozec, Y.-M., Neira, S., Jean-Paul Roux, J.-P.,
Arntz, W.E., Gallardo, V.A., Gutíerrez, D., Isla, E., Levin, L.A., Mendo, J., Neira, C., Shannon, L.J., 2005. Comparing the Benguela and Humboldt marine upwelling
Rowe, G.T., Tarazona, J., Wolf, M., 2006. El Niño and similar perturbation effects ecosystems with indicators derived from inter-calibrated models. ICES Journal
on the benthos of the Humboldt, California, and Benguela Current upwelling of Marine Science 62, 493–502.
ecosystems. Advances in Geosciences 6, 243–265. Monaco, M.E., Ulanowicz, R.E., 1997. Comparative ecosystem trophic structure of
Baird, D., McGlade, J.M., Ulanowicz, R.E., 1991. The comparative ecology of six three US mid-Atlantic estuaries. Marine Ecology Progress Series 161, 239–254.
marine ecosystems. Philosophical Transactions: Biological Sciences 333, 15–29. Ñiquen, M., Bouchon, M., 2004. Impact of El Niño events on pelagic fisheries in
Baird, D., Luczkovich, J., Christian, R.R., 1998. Assessment of Spatial and Temporal Peruvian waters. Deep-Sea Research Part II-Topical Studies in Oceanography 51,
Variability in Ecosystem Attributes of the St Marks National Wildlife Refuge, 563–574.
Apalachee Bay, Florida. Estuarine Coastal and Shelf Science 47, 329–349. Nixon, S.W., 1982. Nutrient dynamics, primary production and fisheries yields of
Barber, R.T., Chavez, F.P., 1983. Biological consequences of El Nino. Science 222, lagoons. Oceanologica Acta Special edition, 357–371.
1203–1210. Nixon, S., Thomas, A., 2001. On the size of the Peru upwelling ecosystem. Deep-Sea
Barber, R.T., Chavez, F.P., Kogelschatz, J.E., 1985. Biological effects of El Nino. Boletín Research I 48, 2521–2528.
ERFEN 14, 3–29. Odum, E.P., 1969. The strategy of ecosystem development. Science 104, 262–270.
Ortiz, M., Wolff, M., 2002. Spatially explicit trophic modelling of a harvested benthic Vega, R., Mendo, J., 2002. Consumo de alimento y crecimiento del pulpo Octopus
ecosystem in Tongoy Bay (central northern Chile). Aquatic Conservation: spp. alimentado con Argopecten purpuratus en la Bahía de Paracas, Pisco. In:
Marine and Freshwater Ecosystems 12, 601–618. Mendo, J., Wolff, M. (Eds.), Memorias I Jornada ‘‘Bases Ecológicas y
Pauly, D., Christensen, V., 1995. Primary production required to sustain global Socioeconómicas para el Manejo de los Recursos Vivos de la Reserva Nacional
fisheries. Nature 374, 255–257. de Paracas”. Universidad Nacional Agraria La Molina, Lima, pp. 212–220.
Pennington, J.T., Mahoney, K.L., Kuwahara, V.S., Kolber, D.D., Calienes, R., Chavez, Villanueva, P., Fernandez, C., Sanchez, S., 1998. Biomasa planctonica como
F.P., 2006. Primary production in the eastern tropical Pacific: a review. Progress alimento disponible durante el crucero BIC Humboldt y BIC Jose Olaya
in Oceanography 69, 285–317. Balandra 9808–09 de Paita a Los Paltos (Tacna). Informe del Instituto del
Polovina, J.J., Ow, M.D., 1985. An approach to estimating an ecosystem box model. Mar del Perú 141, 49–54.
Fishery Bulletin 83, 457–460. Walsh, J.J., 1981. A carbon budget for overfishing off Peru. Nature 290, 300–304.
Pranovi, F., Libralato, S., Raicevich, S., Granzotto, A., Pastres, R., Giovanardi, O., 2003. Walters, C., Christensen, V., Pauly, D., 1997. Structuring dynamic models of
Mechanical clam dredging in Venice lagoon: ecosystem effects evaluated with a exploited ecosystems from trophic mass-balance assessments. Reviews in
trophic mass-balance model. Marine Biology 143, 393–403. Fish Biology and Fisheries 7, 139–172.
Rojas de Mendiola, B., Gómez, O., Ochoa, N., 1985. Efectos del fenómeno El Niño Warnke, K., 1999. Observations on Embryonic Development of Octopus mimus
sobre el fitoplancton. In: Arntz, W., Landa, A., Tarazona, J. (Eds.), El Niño: su (Mollusca: Cephalopoda) from Northern Chile. The Veliger 42, 211–217.
impacto en la fauna marina. Instituto del Mar del Perú, Callao, Peru. Wilson, J.G., Parkes, A., 1998. Network analysis of the energy flow through the
Rouillon, G., Mendo, J., Ochoa, N., 2002. Fitoplankton en el contenido estomacal de Dublin ecosystem. Biology and Environment: Proceedings of the Royal Irish
Argopecten purpuratus (Mollusca, Bivalvia) suspendida a differentes Academy 98B, 179–190.
profundidades en Bahía Independencia. In: Mendo, J., Wolff, M. (Eds.), Wolff, M., 1987. Population dynamics of the Peruvian scallop Argopecten
Memorias I Jornada ‘‘Bases Ecológicas y Socioeconómicas para el Manejo de purpuratus during the El Nino phenomenon of 1983. Canadian Journal of
los Recursos Vivos de la Reserva Nacional de Paracas”. Universidad Nacional Fisheries and Aquatic Sciences 44, 1684–1691.
Agraria La Molina, Lima, pp. 60–67. Wolff, M., 1988. Spawning and recruitment in the Peruvian scallop Argopecten
Rybarczyk, H., Elkaim, B., Ochs, L., Loquet, N., 2003. Analysis of the trophic network purpuratus. Marine Ecology Progress Series 42, 213–217.
of a macrotidal ecosystem: the Bay of Somme (Eastern Channel). Estuarine. Wolff, M., 1994. A trophic model for Tongoy Bay – a system exposed to suspended
Coastal and Shelf Science 58, 405–421. scallop culture (northern Chile). Journal of Experimental Marine Biology and
Samamé, M., Benites, C., Valdivieso, V., Mendez, M., Yamashiro, C., Moron, O., 1985. Ecology 182, 149–168.
Evaluacón del recurso concha de abanico (Argopecten purpuratus) en la Bahía Wolff, M., Alarcon, E., 1993. Structure of a scallop Argopecten purpuratus (Lamarck,
Independencia, Pisco, en Octubre–Noviembre 1985. Instituto del Mar del Perú. 1819)-dominated subtidal macro-invertebrate assemblage in northern Chile.
Stotz, W.B., González, S.A., 1997. Abundance, growth, and production of the sea Journal of Shellfish Research 12, 295–304.
scallop Argopecten purpuratus (Lamarck 1819): bases for sustainable Wolff, M., Mendo, J., 2000. Management of the Peruvian bay scallop (Argopecten
exploitation of natural scallop beds in north-central Chile. Fisheries Research purpuratus) metapopulation with regard to environmental change. Aquatic
32, 173–183. Conservation: Marine and Freshwater Ecosystems 10, 117–126.
Tam, J., Taylor, M.H., Blaskovic, V., Espinoza, P., Ballón, R.M., Díaz, E., Wosnitza- Wolff, M., Mendo, J., 2002. Upwelling is the Disturbance, not El Niño: Insights from
Mendo, C., Argüelles, J., Purca, S., Ayón, P., Quipuzcoa, L., Gutiérrez, D., Goya, E., Modelling Community Organization and Flow Structure. Investigaciones
Ochoa, N., Wolff, M., 2008. Trophic flows in the Northern Humboldt Current Marinas 30, 166–167.
Ecosystem, Part 1: comparing 1995-96 and 1997–98. Progress in Oceanography Wolff, M., Perez, H., 1992. Population dynamics, food consumption and gross
79, 352–365. conversion efficiency of Octopus mimus Gould, from Antofagasta (northern
Tarazona, J., Salzwedel, H., Arntz, W., 1988. Oscillations of macrobenthos in shallow Chile). ICES Council Meetings Papers. ICES, Copenhagen, Denmark. p. 12.
waters of the Peruvian central coast induced by El Nino 1982–83. Journal of Wolff, M., Soto, A., 1992. Population dynamics of Cancer polyodon in La Herradura
Marine Research 46, 593–611. Bay, northern Chile. Marine Ecology Progress Series 85, 70–81.
Tarazona, J., Canahuire, E., Salzwedel, H., Jeri, T., Arntz, W., Cid, L., 1991. Wolff, M., Hartmann, H.J., Koch, V., 1996. A pilot trophic model for Golfo Dulce, a
Macrozoobenthos in two shallow areas of the Peruvian Upwelling ecosystem. fjord-like tropical embayment, Costa Rica. Revista de biologia tropical 44, 215–
Taylor, M.H., Wolff, M., Mendo, J., Yamashiro, C., 2007a. Diet matrix for the steady- 231.
state models of Independence Bay for the years 1996 and 1998 before Wolff, M., Wosnitza-Mendo, C., Mendo, J., 2003. The Humboldt current-trends in
application of the Ecoranger resampling routine. PANGAEA, Available from: exploitation, protection and research. In: Hempel, G., Sherman, K. (Eds.), Large
<http://www.pangaea.de>. Marine Ecosystems of the World. Elsevier, pp. 279–309.
Taylor, M.H., Wolff, M., Mendo, J., Yamashiro, C., 2007b. Input–output parameters Wolff, M., Taylor, M., Mendo, J., Yamashiro, C., 2007. A catch forecast model for the
for the steady-state models of Independence Bay for the years 1996 and 1998 Peruvian scallop (Argopecten purpuratus) based on estimators of spawning stock
before application of the Ecoranger resampling routine. PANGAEA, Available and settlement rate. Ecological Modelling 209, 333–341. doi:10.1016/
from: <http://www.pangaea.de>. j.ecolmodel.2007.07.013.
Taylor, M.H., Tam, J., Blaskovic, V., Espinoza, P., Ballón, R.M., Wosnitza-mendo, C., Yamashiro, C., Rubio, J., Jurado, E., Auza, E., Maldonado, M., Ayon, P.,
Argüelles, J., Díaz, E., Purca, S., Ochoa, N., Ayón, P., Goya, E., Quipuzcoa, L., Antonietti, E., 1990. Evaluación de la población de Concha de Abanico
Gutiérrez, D., Wolff, M., 2008. Trophic flows in the Northern Humboldt Current (Argopecten purpuratus) en la Bahía de Independencia, Pisco, Perú 20 de
Ecosystem, Part 2: Elucidating mechanisms of ecosystem change over an ENSO febrero–04 de marzo de 1988. Informe del Instituto del Mar del Perú
cycle by simulating changes in low trophic level dynamics. Progress in N°98, Callao, Peru.
Oceanography 79, 366–378. Zuta, S., Tsukayama, I., Villanueva, R., 1983. El ambiente marino y las fluctuaciones
Ulanowicz, R.E., 1986. Growth and development: ecosystem phenomenology. de las principales poblaciones pelágicas de la costa peruana. FAO Fisheries
Springer-Verlag, New York. Report 291, 179–253.

Progress in Oceanography: Marc H. Taylor, Matthias Wolff, Jaime Mendo, Carmen Yamashiro

Uploaded by

Copyright:

Available Formats

Progress in Oceanography: Marc H. Taylor, Matthias Wolff, Jaime Mendo, Carmen Yamashiro

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Progress in Oceanography: Marc H. Taylor, Matthias Wolff, Jaime Mendo, Carmen Yamashiro

Uploaded by

Copyright:

Available Formats

View metadata, citation and similar papers at core.ac.

uk brought to you by CORE

Progress in Oceanography 79 (2008) 336–351

Contents lists available at ScienceDirect

500 Sea urchins

Biomass (ww gm2)

Functional group Species

M.H. Taylor et al. / Progress in Oceanography 79 (2008) 336–351

Gmax ¼ 0:97  e0:0633T ð4Þ

50% light (from self-shading) and 0% nutrient limitation factors,

while the EN condition of 1998 assumed a mean temperature of

estimates for the Peruvian coastal system under upwelling condi-

2.4.4. Consumption and Conversion efﬁciency

and Consumption (P/Q) were applied (Tables 2 and 3).

2.4.5. Diet matrices

general knowledge from literature was used in the construction

of diet matrices (Table 2). Initial attempts to balance the 1996

model resulted in insufﬁcient production of many smaller epifau-

urchins, herbivorous gastropods, benthic detritivores, and misc. ﬁl-

rous benthic invertebrate groups (predatory gastropods, small

carnivores, predatory crabs, sea stars, and octopus). As macroinver-

nistic feeders, limited more by their modes of feeding (Wilson and

Parkes, 1998), diet proportions were adjusted to reﬂect both pred-

atory consumption rates and the available production of prey

groups. This was accomplished by iteratively distributing the prey

production to the predators in weekly consumption increments,

assuming unselectivity. When the production of a single prey

group was completely utilized, the following iteration would con-

ing were assumed and the calculated diets resulted in high

proportions of polychaetes in their diets – reﬂecting their high bio-

uation was less problematic due to a reduction of carnivorous

ﬁsh species were obtained from FishBase (Froese and Pauly,

cies contribution from recorded catches.

2.5. Addressing parameter uncertainty

The balanced steady-state model for 1996 was subjected to the

in order to assess the probability distributions of the input param-

18. Marine mammals

11. Small carnivores

dom input variables from normal distributions for each basic

parameter, and all resulting combinations that satisﬁed mass-bal-

13. Sea stars

anced constraints were recorded. Originally we allowed the

routine to use conﬁdence intervals as derived from a pedigree of

the data sources, where highest conﬁdence is placed in locally-de-

M.H. Taylor et al. / Progress in Oceanography 79 (2008) 336–351

Summary statistics 1996 1998 % Change

t/km2/year 1996 1998

Phytoplankton Macroalgae Polychaetes

Scallops Sea urchins Herbivorous gastropods

Benthic detritivores Misc. filter feeders Predatory gastropods

Small carnivores Predatory crabs Sea stars

Octopods Littoral fish Pelagic predatory fish

Force Phytoplankton, Macroalgae (S1)

10.0 month (October 1998) compared to normal levels of around 750.

ab near Independence Bay as recorded by acoustic surveys (Ñiquen

and Bouchon, 2004).

The expansion of the diving ﬁshery during EN is also observed

through much higher indices of gross efﬁciency (catch/net PP) such

Gmax ¼ 0:97 e0:0633T ð4Þ