universidade de lisboa faculdade de ciÊncias …

UNIVERSIDADE DE LISBOA

FACULDADE DE CIÊNCIAS

DEPARTAMENTO DE BIOLOGIA ANIMAL

VERTICAL DYNAMICS OF PLANKTONIC

COMMUNITIES AT SOFALA BANK, MOZAMBIQUE

Miguel Albuquerque da Costa Leal

MESTRADO EM ECOLOGIA MARINHA

2009

UNIVERSIDADE DE LISBOA

FACULDADE DE CIÊNCIAS

DEPARTAMENTO DE BIOLOGIA ANIMAL

VERTICAL DYNAMICS OF PLANKTONIC

COMMUNITIES AT SOFALA BANK, MOZAMBIQUE

Dissertação orientada pelos:

Professor Doutor José Paula

Professora Doutora Vanda Brotas

Miguel Albuquerque da Costa Leal

MESTRADO EM ECOLOGIA MARINHA

2009

i

Acknowledgements

To all people that in some way contributed to this work, I express my sincere gratitude, especially to:

Professor Doutor José Paula, for taking a chance on me and for all the support, supervision,

confidence, friendship and transfer of knowledge, but most of all for his joviality and all the joyfull

moments, which turned this hard and serious work into a delightful and pleasant learning process.

Professora Doutora Vanda Brotas, for having accepted to guide this thesis and for the support,

recommendations and companionship, for believing in me and also for spending some of her

precious (and little) time with me.

Carolina Sá and Rafael Mendes, without whom part of this work would never be possible, for all the

trust, patience, and time spent on laboratory, as well for the friendship, support and enthusiasm

during this work, even in the hardest moments.

All people onboard the R/V “Dr. Fridtjof” Nansen, for the collaboration during sample period, namely

Jens‐Otto Krakstad and Diana Zaera, for the nice discussions and help with paperwork, and also to

Carlos Bento, Domingos Zacarias e Zé Chamuse for their very important help on the sampling

process.

Sónia Nordez and Maurício Lipassula, for their help processing plankton samples in Lisbon; to James

Mwaluma, for the larval fish identification; and to Paulo Oliveira for his expertise on current data

interpretation.

All the people of the LMG, namely Ricardo Mendes, Gil Penha‐Lopes and Laura Antão, for the

friendship and nice chats when difficult choices had to be made.

Bruno Jesus, for his patience in teaching me the world of R and statistics, which definitively was very

important for the statistical and graphical analysis.

All my friends, both from the University and Ponto Zero, for all the joy and happiness and also for

encouraging and cheering me up in my worst days. Special thanks to my truly friend Vasco Mota, for

ii

all the articles he sent me from the Netherlands and, most of all, for his friendship and nice

discussions.

My Family, for the unconditional support, concern, care, recommendations and patience. Special

thanks to my mother, father and sister for accepting those (long) periods in which I’m not present,

for teaching me the importance of organization at work, discipline and timings, but essentially for

what I am today; to my uncle João Martins for the nice chats, holidays and gym classes that were

essential to my welfare and mental sanity; to my uncle João Caraça, for the conversations that gave

me a scientific view of this world; and also to my uncle Luis Leal, for my initiation to biology and

kindle my interest in become a marine biologist, and to the discussions and stories about the natural

world, which are the reason for my interest in biology and curiosity to explore and understand our

natural world.

At last, but not, by any means, less important, to Marta, my source of motivation to be a better

person and a hard working scientist, for all the love, care, patience and trust, for providing me

everyday with a smile in my face and for making me understand how much this life is worth living.

Thank you for being here!

iii

Resumo

As zonas costeiras são ecossistemas de grande importância para as actividades humanas, onde se

desenrolam complexos processos naturais com elevada variabilidade espacial e temporal. Destes

processos, com particular importância para a geração de recursos, destacam‐se os fenómenos de

maré e as correntes costeiras, que, entre outros factores, estão fortemente associados às variações

das propriedades físico‐químicas da massa de água costeira (p.e. temperatura, salinidade e

nutrientes). Consequentemente, esta variabilidade influencia as diferentes comunidades que estão

na base dos ecossistemas marinhos, como as comunidades fito‐ e zooplanctónicas, assim como as

larvas de peixes e de invertebrados marinhos. É de salientar que, no que respeita aos processos bio‐

ecológicos das larvas de peixes e de invertebrados marinhos, a variação na abundância e padrões de

distribuição na zona costeira está fortemente dependente de fenómenos como os ventos e correntes

que, entre outros factores, condicionam a sobrevivência e, consequentemente, a taxa de

recrutamento para as populações de organismos costeiros.

O presente estudo decorreu no Banco de Sofala (Moçambique), uma extensa plataforma

continental existente na zona costeira situada em frente ao delta do Rio Zambeze (um dos maiores

rios de África), onde ocorrem as mais importantes actividades pesqueiras de Moçambique. Tendo em

consideração a importância económica do Banco de Sofala e a grande influência do Rio Zambeze no

funcionamento dos processes ecológicos costeiros, os objectivos desta investigação foram

direccionados para as variações existentes ao nível hidrológico e biológico que suportam esta intensa

actividade pesqueira. Aspectos como as propriedades físico‐químicas da massa de água e correntes

costeiras foram estudados conjuntamente com a distribuição e abundância de nutrientes,

comunidades fitoplanctónicas (estudados através de assinaturas de pigmentos analisadas por HPLC),

zooplâncton e larvas de peixe (especificamente da espécie Herklotsichthys quadrimaculatus),

decorrentes no Banco de Sofala. Os trabalhos desenvolvidos foram especialmente direccionados

para a distribuição horizontal na plataforma continental (baseada na amostragem de 11 estações

radiais distribuídas ao longo de 3 transectos perpendiculares à costa), e para a dinâmica vertical

destes processos ao longo de um ciclo de 48 horas (com uma periodicidade de amostragem de 2 h

numa estação fixa). O processo de amostragem decorreu ao longo de 3 dias e foi desenvolvido no

decorrer de uma missão oceanográfica realizada a bordo do N/O “Dr. Fridtjof Nansen”, integrada

num projecto direccionado para o estudo dos ecossistemas costeiros de Moçambique.

iv

A variação horizontal da maioria dos parâmetros analisados revelou a existência de gradientes

costa – largo e de descargas do rio Zambeze, evidenciadas pela de intrusão de águas estuarinas na

zona costeira. A distribuição superficial das propriedades físico‐químicas analisadas evidenciou a

presença de uma massa de água menos salina e mais quente na zona central, associada

provavelmente a descargas estuarinas, enquanto que a distribuição das concentrações de nutrientes

só evidenciou gradientes perpendiculares à costa para os silicatos, dado que a distribuição de

fosfatos foi independente da distância à costa e a concentração de nitratos foi sempre menor que o

limite de detecção do método. Relativamente à distribuição superficial de fitoplâncton, foi observada

maior biomassa numa das estações mais costeira, visto que foi provavelmente a primeira a ser

influenciada pelas descargas do rio Zambeze e, consequentemente, onde maior concentração de

nutrientes ficou primeiramente disponível para o fitoplâncton. O único grupo de fitoplâncton com

diferente distribuição relativamente à biomassa fitoplanctónica foi o das Cianobactérias, para o qual

foram observadas maiores concentrações pigmentares na massa de água mais oceânica.

No que respeita a dinâmica vertical estudada ao longo de 48 horas, as correntes de maré

revelaram ser um factor crucial para as variações hidrológicas decorrentes na coluna de água,

nomeadamente para o transporte horizontal e vertical da biomassa fitoplanctónica e para a variação

da disponibilidade de nutrientes. As concentrações de nutrientes obtidas estão de acordo com

resultados obtidos em outras zonas costeiras oligotróficas, contudo os rácios de nutrientes obtidos

(N:P e N:Si) foram extremamente baixos e fortemente influenciados pelas reduzidas concentrações

de nitratos + nitritos (N). As concentrações de N obtidas foram anormalmente baixas relativamente a

muitos ecossistemas costeiros, contudo semelhantes a alguns estudos obtidos no Oceano Índico,

nomeadamente na Austrália e no Quénia, onde o N desempenha um papel crucial como factor

limitante da biomassa fitoplanctónica. Para além destas concentrações evidenciarem o estado

oligrotrófico do Banco de Sofala durante o período de amostragem, sugerem ainda a existência de

reduzidas descargas estuarinas para a zona costeira, o que é típico da época seca durante a qual foi

realizado este estudo.

As maiores concentrações de pigmentos fitoplanctónicos foram observadas nas maiores

profundidades amostradas (30 e 40 m), onde também foram registadas maiores concentrações de N.

Durante o decorrer do estudo, a comunidade fitoplanctónica foi dominada por microflagelados, mais

concretamente por microalgas do grupo Prymnesiophyceae, o que está de acordo com outros

estudos em regiões oligotróficas onde a dominância de microflagelados, e não de microplâncton

(diatomáceas e dinoflagelados), é habitualmente observada. De acordo com os resultados obtidos,

toda a comunidade fitoplânctónica distribuiu‐se de modo similar na coluna de água ao longo do

v

espaço de tempo considerado, com a excepção das Cianobactérias que apresentaram uma

distribuição vertical mais superficial.

Relativamente à distribuição do zooplâncton, foram observadas elevadas densidades junto ao

fundo (30/40 m), onde as maiores concentrações de fitoplâncton foram também verificadas. Durante

o período nocturno foram registadas maiores densidades de zooplâncton nas camadas de água mais

superficiais (< 10 m) relativamente aos períodos nocturnos, o que evidencia os processos de

migração vertical típicos das comunidades zooplanctónicas onde a variação da intensidade luminosa

desempenha um papel crucial. Contudo, o movimento vertical observado envolveu apenas uma

pequena parte da abundância total de zooplâncton, visto que as maiores densidades observadas à

superfície durante a noite foram geralmente inferiores às densidades observadas no estratos mais

profundos (20/30 e 30/40 m).

Relativamente à investigação direccionada especificamente para as larvas de Herklotsichthys

quadrimaculatus, os resultados obtidos evidenciaram o padrão típico de migração vertical com

periodicidade diária, realizando movimentos para a superfície e para o fundo durante o anoitecer e

amanhecer, respectivamente. Relativamente à variação desta dinâmica vertical em função das

variações ontogenéticas, os resultados obtidos evidenciam diferenças na posição vertical das larvas

em função do período do dia e do estádio de desenvolvimento. As larvas em estádios de

desenvolvimento mais avançados dominaram a massa de água superficial da coluna de água

(neuston) durante a noite, enquanto que as larvas mais pequenas estiveram distribuídas na camada

de água subjacente (0 – 20 m). Durante o dia, a quantidade de larvas presente nos estratos

amostrados foi quase nula, o que evidencia a agregação das larvas de H. quadrimaculatus junto ao

fundo (40 – 50 m), onde a penetração luminosa é muito reduzida e, consequentemente, o risco de

mortalidade associado a predadores visuais é baixo.

A investigação realizada realça a importância dos processos físico‐químicos nas zonas costeiras e

o seu papel na dinâmica vertical das comunidades planctónicas. Este caso de estudo exemplifica os

processos base que ocorrem num ecossistema costeiro tropical onde as marés e as descargas

estuarinas são factores importantes na dinâmica das comunidades biológicas presentes no plâncton.

Os resultados obtidos são então um exemplo dos ecossistemas costeiros do Oceano Índico Ocidental,

onde a dinâmica das comunidades planctónicas é ainda pouco conhecida e compreendida.

Palavras‐chave: Pigmentos fitoplanctónicos; Zooplâncton; Estádios larvares; Herklostichthys

quadrimaculatus; Migração vertical; Moçambique

vii

Abstract

Coastal ecosystems are largely influenced by the interaction of several factors operating at

various temporal and spatial scales, specifically those responsible for primary and secondary

production processes that modulate marine resources. Hydrological processes (e.g. tides and coastal

currents), nutrients availability, phytoplankton groups, zooplankton and larval fish (specifically the

clupeid Herklotsichths quadrimaculatus) abundance and distribution were investigated at the Sofala

Bank (Mozambique), with special emphasis on their horizontal distribution and vertical dynamics.

Horizontal distribution has shown onshore‐offshore gradients in all analysed parameters, as well as

inshore waters intrusion probably related to Zambezi River delta runoff. Tidal currents were

responsible for major hydrological vertical variations and for horizontal and vertical advection of

phytoplankton biomass in the surface and deepest layers, respectively. Nutrient concentrations were

typical from oligotrophic regions, and nutrient ratios were strongly influenced by depleted nitrite +

nitrate concentrations, suggesting low estuarine discharges typical from the dry season. Both

phytoplankton pigments and zooplankton were found mainly near bottom (40 m deep), despite the

latter displayed vertical migrations triggered by light variations. Phytoplankton community was

dominated by microflagellates, specifically prymnesiophyceans, and vertical distribution changes

were similar for the whole community. Cyanobacteria was the only phytoplankton group that

displayed a different vertical distribution pattern, mainly concentrated at mid water column depths

(10 – 20 m). Herklotsichthys quadrimaculatus larvae displayed typical diel vertical migrations and

were mainly distributed in the upper water column (0 – 20 m) during the night and almost absent

from all sampled strata during the day. Larger larvae dominated the neuston layer during the night

while during daylight periods remained close to the bottom. This investigation enhances the

importance of physico‐chemical phenomena determining the planktonic communities vertical

dynamics at a tropical coastal ecosystem of the Western Indian Ocean, where planktonic dynamics

are still poorly described and understood.

Keywords: Phytoplankton pigments; Zooplankton; Fish larvae; Herklostichthys quadrimaculatus;

Diel vertical migration; Mozambique

Table of contents

Acknowledgements i Resumo iii Abstract vii

CHAPTER 1 1

General introduction 3 References 8

CHAPTER 2 11

Distribution and vertical dynamics of planktonic communities at Sofala Bank, Mozambique 13 Abstract 13 Introduction 14 Material and Methods 16 Results 18 Discussion 30 Conclusion 34 References 36

CHAPTER 3 41

Vertical dynamics of the gold‐spot herring (Herklotsichthys quadrimaculatus) larvae at Sofala Bank,

Mozambique 43 Abstract 43 Introduction 43 Material and Methods 45 Results 46 Discussion 50 References 52

CHAPTER 4 55

Final considerations 57

Chapter 1

General Introduction

3

General introduction

Coastal ecosystems and planktonic communities

Coastal ecosystems are among the most biologically productive environments in the world,

providing important services to human populations. Their existence at the interface between the

terrestrial and marine environment exposes them to a wide variety of human and natural stressors

occurring at different spatial and time scales, which makes the coastal regions complex and dynamic

ecosystems (Kennedy et al., 2002).

The natural stressors occurring in the coastal marine environments are typically physical and

chemical processes that differently affect biological communities, hence their population dynamics.

Some of the most important physical processes are winds, tides and coastal currents, which promote

water mass advection, thus spatial and temporal variability on water column physico‐chemical

properties (Epifanio and Garvine, 2001; Lutjeharms, 2006; Queiroga et al., 2007). The interactions

between physical and chemical variations induce high heterogeneity on coastal shelf waters,

specifically near river deltas, where surface and river runoff together with tidal currents are

responsible for strong hydrological gradients and cyclic variations (e.g. temperature and salinity),

clines vertical oscillation, both water column stratification and vertical mixing processes, and also

large nutrient inputs (Verity et al., 1993; Epifanio and Garvine, 2001; Wawrik et al., 2004; Silva et al.,

2008).

Nutrient loadings into coastal ocean zones are from various sources and frequently associated to

human activities. Surface and river runoff, estuarine discharges and exchanges with the open ocean

are commonly the source and sink of major nutrients inputs, but continental shelf sediments and

respiration associated with the sediment surface photosynthesis are also thought to be important

nutrient sources and recycling (Wawrik et al., 2004; Wassmann and Olli, 2005, Giraud et al., 2008).

Additionally, potentially important sources of nutrients are atmospheric deposition and groundwater

discharges. All nutrient inputs to coastal zones, together with the diverse physical processes,

differently disturb coastal biological communities, and the most directly influenced are possibly from

the lower trophic levels (e.g. phyto‐ and zooplankton) (Cottingham, 1999; Tyrrell, 1999; Anderson et

al., 2002; Kennedy et al., 2002).

Changes on physico‐chemical water properties may be reflected in the dynamics of populations

Chapter 1

4

thriving in coastal systems, strongly determining their abundance and distribution variations (Paula

et al., 1998; Cottingham, 1999; Calbet et al., 2001; Elliott et al., 2007). The small scale changes on

planktonic communities distribution patterns makes them particularly sensitive to hydrological

variations, presenting a great variety of both spatial and temporal changes, respectively from thin

water column layers to large‐scale gradients (e.g. onshore‐offshore, north‐south) and from hours to

seasons (Verity et al., 1993; Paula et al., 1998; Gibson, 2001; Walther et al., 2002; Balkis, 2003;

Brunet and Lizon, 2003; Queiroga et al., 2007; Criales‐Hernández, 2008). Accordingly, these

variations are often related to environmental drivers, particularly the preference or avoidance to

particular physico‐chemical properties and hydrological processes, such as temperature, salinity,

oxygen, light penetration, turbulence, currents and nutrients availability.

Spatial and temporal variations of several planktonic organisms (e.g. phytoplankton, zooplankton

and invertebrate and fish larvae), specifically their vertical distribution, have been widely studied in

different coastal ecosystems, due to their importance to organic matter flows, primary and

secondary production, larval recruitment processes and population connectivity, among others (e.g.

Steinberg et al., 2002; Brunet and Lizon, 2003; Fiksen et al., 2007; Leis, 2007; Queiroga et al., 2007).

Plankton vertical dynamics

Vertical migration is a commonly observed phenomenon in many species of freshwater and

marine plankton and occurs at a variety of periods, including diel, semi‐diel and also tidal periods

(Ohman, 1990; Hill, 1991; Epifanio and Garvine, 2001; Brunet and Lizon, 2003; Fiksen et al., 2007;

Queiroga et al., 2007). The explanations for vertical migration may include predation avoidance,

bioenergetic advantages and regulation of horizontal position. The hypothesis of a predator

avoidance mechanism is generally accepted, as vertical migration often results when prey

populations are under intense, selective pressures from visually dependent predators (e.g. Zaret and

Suffern, 1976; Irigoien et al., 2004). However, some contradictory cases show that predation is not

the only adaptive reason for vertical migration, and the attempt to provide only one hypothesis to

account for all observed vertical migrations represent a simplistic outlook on natural selection. Other

hypothesis is vertical migration as a bioenergetic advantage, resulting of a compromise between

reducing energy expenditure and mortality, taking advantage of feeding opportunities (e.g. Ohman,

1990; Fiksen et al., 2007). The other common hypothesis to planktonic vertical migration is the

horizontal position regulation. These communities have very limited capacity for horizontal

movements, thus position control by vertical migration potentially provides a means to affect

transport to specific locations or to promote retention in particular areas (Hill, 1991; Leis, 2007;

Queiroga et al., 2007). The integration of several hydrological and biological processes, such as tidal


5

currents and behaviour, respectively, regulate dispersal, supply to coastal habitats and recruitment

of several planktonic organisms, specifically larval stages of fish and crustacean species (Hill, 1991;

Epifanio and Garvine, 2001; Queiroga et al., 2006; Santos et al., 2006). Taking into account the

several theories and the diversity of organisms displaying vertical movements there should be

variability among vertical migrations patterns, since the same vertical behaviour does not produce

the same optimum compromise between reduced mortality and energetic cost for different groups,

species and stages (Ohman, 1990; Epifanio and Garvine, 2001; Irigoien et al., 2004; Leis, 2007).

Furthermore, the mortality risks for each organism are not the same because of the different

selectivity of predators.

Different vertical distribution and diel vertical migration (DVM) can be found among plankton

communities, from phyto‐ to zooplankton and also invertebrate and fish larvae. Phytoplankton

vertical distributions may differ between day and night, indicating either migration or differences in

production/mortality rates (Hajdu et al., 2007; Brunet et al., 2008). Accordingly, the vertical

distribution differences found on previous studies are mainly triggered by light requirements and

nutrient availability in different water column layers. However, sinking cells and water exchange

ascribable to tidal currents and vertical mixing processes can easily bias results, since phytoplankton

cells are easily advected (Brunet and Lizon, 2003; Hajdu et al., 2007). Furthermore, additional

difficulties arise when taking into account phytoplankton − zooplankton interactions, specifically

grazing and nutrient regeneration, which affect phytoplankton abundance estimations (Vertiy et al.,

1993; Vrede et al., 1999). Zooplankton itself also displays vertical movements, both to predate and

avoid predators (Zaret and Suffern, 1976; Ohman, 1990). Accordingly, both normal and reverse DVM

are found, strongly determined respectively by diurnal or nocturnal predators, and also by prey

distribution. Besides phyto‐ and zooplankton, invertebrate and fish species commonly have

planktonic larval stages, which also display DVM (Epifanio and Garvine, 2001; Hare and Govoni, 2005;

Fiksen et al., 2007; Queiroga et al., 2007). Due to the importance of larval fish ecology to recruitment

success, thus to enhance fish stocks, special attention is given to larval fish vertical migration.

Larval fish vertical dynamics

The fact that most of coastal fish species have a pelagic larval stage has important implications for

the dynamics of fish populations and fisheries management (Leis, 2007). Management of coastal

fishes must incorporate the scales over which their populations are connected due to dispersal

processes and connectivity, therefore it is important to integrate coastal currents knowledge with

Chapter 1

6

larval behaviour, specifically vertical migration, to improve our understanding of larval survival,

growth and dispersal (Fiksen et al., 2007).

Larval fish vertical migration is strongly determined by several factors (e.g. predator/prey

interactions, bioenergetics and mortality, dispersal processes and connectivity) and much of the

existent knowledge was obtained from studies focusing clupeid species, since its importance to

fisheries and human populations (e.g. Reid, 2001; Santos et al., 2006; Voss et al., 2007). Clupeid larval

stages usually feed on several planktonic species and are strongly predated by visual predators,

namely adult fishes (Aksnes et al., 2004; Fiksen et al., 2007). Therefore, during day and night

different depth distributions are usually observed, in order to minimize mortality rates without

ignoring feeding requirements. Furthermore, according to the larval stage, different vertical positions

are observed, since larvae depend upon different coastal currents to disperse and recruit (Hare and

Govoni, 2005; Fiksen et al., 2007). Once more, different DVM patterns are observed for different

species and larval stages. For example, Hare and Govoni (2005), for species that moved inshore or

remained on the shelf, found that larvae were in deeper positions in the water column than larval

from species that were exported from the shelf. Santos et al. (2006) found that sardine larvae

vertical position was strongly associated to coastal currents, retaining the larvae in high food

availability layers. The knowledge of how much energy planktonic species, specifically fish larvae,

invest on predation risk and feeding opportunities help us to understand their ecology and life

history strategies, as indicators of how much do adults invest in larval stages hence, offspring quality

(Fiksen et al., 2007).

Work context

The work developed during this thesis constituted a special study as part of the oceanographic

campaign “Ecosystem Survey Mozambique 2007” of the wider Nansen EAF Programme

“Strengthening the Knowledge Base for and Implementing and Ecosystem Approach to Marine

Fisheries in Developing Countries”. Sampling was made onboard the R/V “Dr. Fridtjof Nansen” and

took place at the Sofala Bank, a wide shallow shelf influenced by the Zambezi River delta and other

estuaries, and where the most important national fisheries occur. Besides important surface and

river runoff discharges, strong tidal currents and currents from the Mozambique Channel have major

effects on hydrological and occurring ecological processes (Ridderinkhof et al., 2001; Lutjeharms,

2006).

Studies from Sofala Bank are very scarce and only cruise and fisheries reports can be found

(Bandeira et al., 2002, and references therein). It is known that this neritic region has high

phytoplankton biomass comparing to the greater part of the Mozambican oligotrophic waters,


7

mainly because of the nutrient input from the Zambezi river runoff, which enhances the system

productivity and supports local fisheries (Gammelsrød, 1992; Lutjeharms, 2006). Therefore, Sofala

Bank is an excellent case study to understand the nutrients, phytoplankton, zooplankton and larval

fish dynamics in relation to the coastal currents and physico‐chemical environmental variability.

Several studies identified tidal energy, light and diel variations as key factors regulating planktonic

communities vertical distribution and diel changes (Hill, 1991; Queiroga et al., 2007; Brunet et al.,

2008). At Sofala Bank, the tidal currents and estuarine discharges may strongly determine nutrients

availability and water column light penetration due to increased turbidity, thus affecting base

processes that modulate planktonic communities dynamics. Stratification of the water column,

strongly condition the effectiveness of primary production magnitude at the surface layer and may

act as a barrier to vertical movements of zooplankton (Criales‐Hernández et al., 2008). However, due

to the strong local tidal currents, vertical mixing might locally induce homogenization of the water

column. The investigation of tidal and diel variations and other regulating factors is therefore

important to understand fundamental ecological processes that act in generating primary

productivity and modulate resources supporting intense local fisheries activity.

Concerning larval fish dynamics, despite numerous worldwide studies on a great variety of

species, specifically clupeids, none focused on the tropical herring Herklotsichthys quadrimaculatus

(Rüppell 1837), also known as the gold‐spot herring. This species is a major component of tuna

baitfish (Lewis, 1990) and consumed by local coastal communities at the Eastern Africa (KMFRI,

1981). At Mozambique, H. quadrimaculatus is broadly present on the coastal zone, namely at the

Sofala Bank (INIPM, unpublished data). Knowledge of Herklotsichthys spp. larvae is very scarce

(Thorrold and Williams, 1989), and for H. quadrimaculatus from the Western Indian Ocean only

Harris and Cyrus (1999) described its high abundance throughout the year near Durban (South

Africa). The vertical distribution and behaviour is still unknown for H. quadrimaculatus and

information regarding ontogenic changes on depth distribution throughout diel variations are still

missing. Furthermore, the hydrological unique properties of Sofala Bank may change the influence of

typical factors regulating larvae vertical dynamics (e.g. light and currents variations). Tidal energy and

vertical mixing, as well as other local oceanographic features, could therefore have a major role

determining H. quadrimaculatus larvae vertical distribution and diel movements.

The scientific papers presented point toward a broadly understanding of the planktonic

communities vertical dynamics. The first one (submitted to Estuarine, Coastal and Shelf Science)

Chapter 1

8

investigates the horizontal and vertical oceanographic and planktonic processes at the Sofala Bank

shelf waters, specifically aimed to study currents, physico‐chemical water properties (temperature,

salinity and nutrients), phytoplankton pigments as chemotaxonomic markers and zooplankton. The

second paper (submitted to African Journal of Marine Science) is a specific study of fish larvae

vertical dynamics, aimed to study the vertical distribution movements of H. quadrimaculatus larvae,

its relation to several hydrological and biological factors (e.g. light variation, vertical mixing and

zooplankton abundance) and also vertical distribution changes according to larval ontogeny.

References

Aksnes, D.L., Nejstgaard, J., Soedberg, E., Sørnes, T., 2004. Optical control of fish and zooplankton populations. Limnology Oceanography 49, 233‐238.

Anderson, D.M., Glibert, P.M., Burkholder, J.M., 2002. Harmful Algal Blooms and Eutrophication: Nutrient Sources, Composition, and Consequences. Estuaries 25, 704‐726.

Balkis, N., 2003. Seasonal variations in the phytoplankton and nutrient dynamics in the neritic water of Büyükçekmece Bay, Sea of Marmara. Journal of Plankton Research 25, 703‐717.

Bandeira, S.O., Silva, R.P., Paula, J., Macia, A., Hernroth, L., Guissamulo, A.T., Gove, D.Z., 2002. Marine Biological Research in Mozambique: Past, Present and Future. AMBIO 31, 606‐609.

Brunet, C., Lizon, F., 2003. Tidal and diel periodicities of size‐fractionated phytoplankton pigment signatures at an offshore station in the southeastern English Channel. Estuarine, Coast and Shelf Science 56, 833‐843.

Brunet, C., Gasotti, R., Vantrepotte, V., 2008. Phytoplankton diel and vertical variability in photobiological responses at a coastal station in the Mediterranean Sea. Journal of Plankton Research 30, 645‐654.

Calbet, A., Garrido, S., Saiz, E., Alcaraz, M., Duarte, C.M., 2001. Annual zooplankton succession in coastal NW Mediterranean waters: the importance of the smaller size fractions. Journal of Plankton Research 23, 319‐331.

Cottingham, K.L., 1999. Nutrients and zooplankton as multiple stressors of phytoplankton communities: Evidence from size structure. Limnology Oceanography 44, 810‐827.

Criales‐Hernández, M.I., Schwamborn, R., Graco, M., Ayón, P., Hirche, H.J., Wolf, M., 2008. Zooplankton vertical distribution and migration off Central Peru in relation to the oxygen minimum layer. Helgoland Marine Research 62, S85‐S100

Elliott, M., Burdon, D., Hemingway, K.L., Apitz, S.E., 2007. Estuarine, coastal and marine ecosystem restoration: Confusing management and science – A revision of concepts. Estuarine, Coastal and Shelf Science 74, 349‐366.

Epifanio, C.E., Garvine, R.W., 2001. Larval Transport on the Atlantic Continental Shelf of North America: a Review. Estuarine, Coastal and Shelf Science 52, 51‐77.

Fisken, Ø., Jørgensen, C., Kristiansen, T., Vikebø, F., Huse, G., 2007. Linking behavioural ecology and oceanography: larval behaviour determines growth, mortality and dispersal. Marine Ecology Progress Series 347, 195—205.

Gammelsrød,T., 1992. Variation in Shrimp Abundance on the Sofala Bank, Mozambique, and its


9

Relation to the Zambezi River Runoff. Estuarine, Coastal and Shelf Science 35, 91‐103.

Gibson, R.N., 2001. Go with the flow: tidal migration in marine animals. Hydrobiologia 503, 153‐161.

Giraud, X., Quéré, C.L., Cunha, L.C., 2008. Importance of coastal nutrient supply for global ocean biogeochemistry. Global Biogeochemical Cycles, 22, GB2025.

Hajdu, S., Höglander, H., Larsson, U., 2007. Phytoplankton vertical distributions and composition in Baltic Sea cyanobacterial blooms. Harmful Algae 6, 189‐205.

Hare, J.A., Govoni, J.J., 2005. Comparison of average larval fish vertical distributions among species exhibiting different transport pathways on the southeast United States continental shelf. Fisheries Bulletin 103, 728—736.

Harris, S.A., Cyrus, D.P., 1999. Composition, abundance and seasonality of fish larvae in the mouth of Durban harbour, Kwazulu‐Natal, South Africa. Sout African Journal of Marine Science 21, 19—39.

Hill, A.E., 1991. Vertical migration in tidal currents. Marine Ecology Progress Series 75, 39‐54.

Irigoien, X., Conway, D.V.P., Harris, R.P, 2004. Flexible diel vertical migration behaviour of zooplankton in the Irish Sea. Marine Ecology Progress Series 267, 85‐97.

K.M.F.R.I., 1981. Aquatic Resources of Kenya. Proc. of the Workshop of Kenya Marine and Fisheries Research Institute, July 13‐19.

Kennedy, V.S., Twilley, R.R., Kleypas, J.A., Cowan, J.H., Hare, S.R., 2002. Coastal and marine ecosystems & Global climate change – Potential Effects on U.S. Resources. PEW Center on Global Climate Change.

Leis, J.M, 2007. Behaviour as input for modelling dispersal of fish larvae: behaviour, biogeography, hydrodynamics, ontogeny, physiology and phylogeny meet hydrography. Marine Ecology Progress Series 347, 185‐193.

Lewis, A.D., 1990. Tropical south Pacific tuna baitfisheries. In: Blaber, S.J.M., Copland, J.W. (Editors) Tuna baitfish in the Indo‐Pacific region. Aust. Coun. Int. agr. Res. Proc. 30, 10—21.

Lutjeharms, J.R.E., 2006. The costal oceans of south‐eastern Africa. In: A.R. Robinson, K.H. Brink (Editors) The Sea Vol 14. Harvard University Press, Cambridge, MA, pp. 783‐834.

Ohman, M.D., 1990. The Demographic Benefits of Diel Vertical Migration by Zooplankton. Ecological Monographs 60, 257‐281.

Paula, J., Pinto, I., Guambe, I., Monteiro, S., Gove, D., Guerreiro, J., 1998. Seasonal cycle of planktonic communities at Inhaca Island, southern Mozambique. Journal of Plankton Research 20, 2165‐2178.

Queiroga, H., Almeida, M.J., Alpuim, T., Flores, A.A.V., Francisco, S., Gonzàlez‐Gordillo, I., Miranda, A.I., Silva, I., Paula, J., 2006. Tide and wind control of megalopal supply to estuarine crab populations on the Portuguese west coast. Marine Ecology Progress Series 307, 21‐36.

Queiroga, H., Cruz, T., dos Santos, A., Dubert, J., González‐Gordillo, J.I., Paula, J., Peliz, A., Santos, A.M.P., 2007. Oceanographic and behavioural processes affecting invertebrate larval dispersal and supply in the western Iberia upwelling ecosystem. Progress in Oceanography 74, 174‐191.

Reid, D.G., 2001. SEFOS – shelf edge fisheries and oceanography studies: an overview. Fisheries Research 50, 1—15.

Ridderinkhof, H., Lutheharms, J.R.E., Ruijter, W.P.M., 2001. A research cruise to investigate the Mozambique Current. South African Journal of Science 97, 461‐464.

Chapter 1

10

Santos, A.M.P., Ré, P., Dos Santos, A., Péliz, A., 2006. Vertical distribution of the European sardine (Sardina plichardus) larvae and its implications for their survival. Journal of Plankton Research 28, 523—532.

Silva, A., Mendes, C.R., Palma, S., Brotas, V., 2008. Short‐time scale variation of phytoplankton succession in Lisbon bay (Portugal) as revealed by microscopy cell counts and HPLC pigment analysis. Estuarine, Coastal and Shelf Science 79, 230‐238.

Steinberg, D.K., Goldthwait, S.A., Hanself, D.A., 2002. Zooplankton vertical migration and the active transport of dissolved organic and inorganic nitrogen in the Sargasso Sea. Deep‐Sea Research 49, 1445‐1461.

Thorrold, S.R., Williams, D.M., 1989. Analysis of Otolith Microstructure to Determine Growth Histories in Larval Cohorts of a Tropical Herring (Herklotsichthys castelanaui). Canadian Journal Fisheries Aquatic Sciences 46, 1615—1624.

Tyrrell, T., 1999. The relative influences of nitrogen and phosphorus on oceanic primary production. Nature 400, 525‐531.

Verity, P.G., Yoder, J.A., Bishop, S.S., Nelson, J.R., Craven, D.B., Blanton, J.O., Robertson, C.Y., Tronzo, C.R., 1993. Composition, productivity and nutrient chemistry of a coastal ocean planktonic food web. Continental Shelf Research 13, 741‐776.

Voss, R., Schmidt, J.O., Schnack, D., 2007. Vertical distribution of Baltic sprat larvae: changes in patterns of diel migration?. ICES Journal Marine Science, 64.

Vrede,K., Vrede, T., Isaksson, A., Karlsson, A., 1999. Effects of nutrients (phosphorus, nitrogen, and carbon) and zooplankton on bacterioplankton and phytoplankton – a seasonal study. Limnology Oceanography 44, 1616‐1624.

Walther, G.‐R., Post, E., Convey, P., Menzel, A., Parmesan, C., Beebee, T.J.C., Fromentin, J.‐M., Hoegh.Guldberg, O., Bairlein, F., 2002. Ecological responses to recent climate change. Nature 416, 389‐395.

Wassmann, P., Olli, K., 2005. Drainage basin nutrient inputs and eutrophication: an integrated approach. University of Tromsø, Norway, 325 pp.

Wawrik, B., Paul, J.H., Bronk, D.A., John, D., Gray, M., 2004. High rates of ammonium recycling drive phytoplankton productivity in the offshore Mississippi River plume. Aquatic Microbial Ecology 35, 175‐184.

Zaret, T.M., Suffern, J.S., 1976. Vertical migration in zooplankton as a predator avoidance mechanism. Limnology Oceanography 21, 804‐813.

Chapter 2

Distribution and vertical dynamics of planktonic communities at Sofala Bank, Mozambique

13

Distribution and vertical dynamics of planktonic communities at

Sofala Bank, Mozambique

M. C. Leal1, C. Sá2, S. Nordez3, V. Brotas2, J. Paula1

1 Centro de Oceanografia, Laboratório Marítimo da Guia, Faculdade de Ciências da Universidade de Lisboa, Av. Nª Senhora do Cabo, 939, 2759‐374 Cascais, Portugal

2 Centro de Oceanografia, Instituto de Oceanografia, Faculdade de Ciências da Universidade de Lisboa, Campo Grande, 1749‐016 Lisboa, Portugal

3 Instituto de Investigação Pesqueira de Moçambique, Av. Mao Tse‐Tung nº 389, Maputo, C. P. 4603 Moçambique

Abstract

Coastal ecosystem processes are largely influenced by the interaction of different factors

operating at various temporal and spatial scales, specifically those responsible for primary

production patterns that modulate zooplankton and subsequent trophic levels. Hydrological

processes, such as tidal cycles and coastal currents, nutrients availability, phytoplankton groups

(studied through algal pigment signatures analysed by HPLC) and zooplankton abundance and

distribution were investigated at the Sofala Bank (Mozambique), with special emphasis on their

horizontal distribution and vertical dynamics (48‐hour). Horizontal distribution has shown

onshore‐offshore gradients in all analysed parameters, as well as inshore waters intrusion

probably related to Zambezi River delta runoff. Tidal currents were responsible for major

hydrological vertical variations and for horizontal and vertical advection of phytoplankton

biomass in the surface and deepest layers, respectively. Nutrient concentrations were typical

from oligotrophic regions, and nutrient ratios were strongly influenced by depleted nitrite +

nitrate concentrations, suggesting low estuarine discharges typical from the dry season. The very

low N:P ratio obtained, suggests strong nitrogen limitation to phytoplankton communities,

supporting the low phytoplankton abundance observed. Both phytoplankton pigments and

zooplankton were found mainly near bottom (40 m deep), despite the latter displayed vertical

migrations triggered by light variations. Phytoplankton community was dominated by

microflagellates, specifically prymnesiophyceans, and behaved as a whole, except Cyanobacteria

that displayed vertical distribution movements different from other phytoplankton groups, being

mainly concentrated at mid water column depths (10 – 20 m). This investigation enhances

Chapter 2

14

physico‐chemical phenomena and their importance determining the planktonic communities

vertical dynamics at a tropical coastal ecosystem of the Western Indian Ocean, where planktonic

dynamics are still poorly described and understood.

Keywords: Phytoplankton; Pigments; Zooplankton; Vertical dynamics; Nutrient deficiency;

Mozambique

Introduction

Coastal shelf waters are ecosystems of great human and ecological interest where complex

processes occur. The interaction of physical (e.g. coastal currents, upwelling, tides, advection),

chemical (variable chemical properties including nutrient inputs) and ecological (e.g. biological

production and its dynamics, prey/predator interactions) processes induce high spatial variability

on the water column over different time scales (Brunet and Lizon, 2003; Lutjeharms, 2006;

Queiroga et al., 2007). This variability determines the abundance and structure of different

biological communities present in coastal waters, specifically phyto‐ and zooplankton as oceanic

food chain lower levels. Zooplankton abundance and distribution is often related to

predator/prey interactions, as a prey for both planktivorous fishes and some planktonic larvae,

and as consumers of phytoplankton organisms (e.g. González‐Gordillo and Rodríguez, 2003;

Santos et al., 2006). Phytoplankton is therefore regulated by zooplankton and fish herbivory, but

also by nutrients availability, mainly nitrogen, phosphate and silica (Vanni, 1987; Svensson and

Stenson, 1991; Garnier and Cugier, 2004). The relation between the different nutrient ratios and

phytoplankton in coastal systems has been widely studied, and nitrogen generally plays a key role

limiting phytoplankton growth thus strongly determining its community structure and

chemotaxonomic pigment composition (Hecky and Kilham, 1988; Tyrrell, 1999).

The present study took place at the Sofala Bank (Mozambique), a wide shallow shelf

influenced by the Zambezi River delta and other estuaries, and where important fisheries occur.

Besides important river runoff discharges, strong tidal currents and currents from the

Mozambique Channel have major effects on hydrological and occurring ecological processes

(Ridderinkhof et al., 2001; Lutjeharms, 2006). Even though the high importance of this coastal

ecosystem for Mozambican fisheries and economy, studies from Sofala Bank are very scarce and

only cruise and fisheries reports can be found (Bandeira et al., 2002, and references therein). This

neritic region has high phytoplankton biomass comparing to the greater part of the Mozambican

oligotrophic waters, mainly because of the nutrient input from the Zambezi river runoff, which is


15

closely related to rainfall events (Lutjeharms, 2006). Hence, despite that hydrological and

planktonic dynamics of several temperate coastal ecosystems are already studied (e.g. Verity et

al., 1993; Balkis, 2003; Sabetta et al., 2008; Silva et al., 2008), Sofala Bank is an excellent case

study to understand the nutrients, phytoplankton and zooplankton dynamics in relation to the

physico‐chemical environmental variability of a tropical coastal ecosystem from the Western

Indian Ocean, where scientific knowledge is very scarce.

Several studies identified tidal energy, light and diel variations as key factors changing neritic

planktonic vertical distribution (Hill, 1991; Queiroga et al., 2007; Brunet et al., 2008). At Sofala

Bank, the strong tidal currents and estuarine discharges may strongly determine nutrients

availability and water column light penetration due to increased turbidity, thus affecting base

processes that modulate planktonic communities dynamics. Stratification of the water column,

specifically the pycnocline depth and its strength, condition the effectiveness of primary

production magnitude at the surface layer and may act as a barrier to vertical movements of

zooplankton (Criales‐Hernández et al., 2008). However, due to the strong local tidal currents it is

probable that vertical mixing locally induces homogenization of the water column. The

investigation of tidal and diel variations and other regulating factors is therefore important to

understand fundamental ecological processes that act in generating high primary productivity and

modulate resources supporting intense local fisheries activity.

This study was a special study as part of the 2007 campaign “Strengthening the Knowledge

Base for and Implementing and Ecosystem Approach to Marine Fisheries in Developing Countries”

of the Nansen EAF Programme, and its main goal was to investigate the horizontal and vertical

oceanographic and planktonic processes in the Sofala Bank shelf waters, specifically the

horizontal distribution and vertical temporal variation of (1) hydrological processes and

parameters, (2) nutrients concentration, (3) phytopigments composition and abundance and (4)

zooplankton abundance along the water column.

Chapter 2

16

Material and Methods

Sampling

The study area was located at the Sofala Bank, in front of the Zambezi river delta. Sampling

took place on board the R/V Dr. Fridtjof Nansen from the 6th to the 8th December 2007, covering

in the first 24 hours a total of 11 stations distributed along three transects (Fig. 1), from a bottom

depth of 24 to 374 m, and the outermost stations separated by 60 km. Their position was set as to

characterize the shelf section where the fixed station sampling was carried out in the following

48‐hour period. The fixed station was positioned over a bottom depth of 50 m and samples were

taken every 2 h.

Fig. 1. Radial and fixed stations location. Both radial () and fixed stations () are presented and the arrows indicate sampling order.

Current magnitude and direction were measured with a hull‐mounted Acoustic Doppler

Current Profiler (ADCP) and transparency (m) using a Sechi Disk. At every radial and fixed station,

CTD profiles were conducted and temperature, salinity and fluorescence (used as a proxy of

chlorophyll a) data recorded at every 1 m depth interval. Water samples were taken at five

predetermined depths (5, 10, 20, 30 and 40 m), except for shallower stations, using a rosette

equipped with Niskin bottles. Immediately after collection, two replicates of 100 ml were

collected and stored frozen for posterior nutrient analysis (nitrate + nitrite, phosphate and

silicate). For HPLC pigment analysis, two litres of seawater were immediately filtered through

glass fibre filters (25 mm φ, 0.7 µm pore ‐ Whatman GF/F) and filters were kept frozen in the dark

for posterior analysis.


17

Depth‐stratified zooplankton samples (0‐5, 5‐10, 10‐20, 20‐30 and 30‐40 m) were collected

with a multinet (Midi model, 0.5 x 0.5 m mouth size, Hydro‐bios) with 405 µm mesh size and

towed at ~2 knots for 2 min, sampling on oblique hauls in each stratum. Flow rate was monitored

by a flowmeter mounted in the mouth of the aperture, and each sample represented

approximately 40 m3 of water filtered. A neuston net (0.2 x 1.0 m mouth size) with the same mesh

size and a flowmeter mounted was towed horizontally at similar speed and time, sampling the

upper 20 cm of the water column. All zooplankton samples were preserved in 4% borax‐buffered

formaldehyde, prepared using seawater.

Laboratory procedures

All the laboratory procedures were done at the Oceanography Centre of University of Lisbon,

Portugal. Colorimetric analyses, with a Tecator FIAstarTM 5000 Analyser, were performed to

determine nutrient concentrations. Nitrite (NO2‐), nitrate (NO3

‐), phosphates (PO43‐, hereafter as

P) and silicates (Si(OH)4, hereafter as Si) were respectively determined according to Grasshoff

(1976), Bendshneider and Robison (1952), Murphey and Riley (1962), and Faning and Pilson

(1973). Since water properties from this region are typically oligotrophic (Lutjeharms, 2006), the

nitrite and nitrate sum was used (NO2‐ + NO3

‐, hereafter as N).

The identification and abundance of phytoplankton functional groups can be achieved by high

performance liquid chromatography (HPLC) analytical technique, which is increasingly in use as it

is a less time consuming method in relation to microscopy phytoplankton identification and

counting. HPLC quantifies chemotaxonomic pigments allowing to estimate the contribution of

phytoplankton groups to chlorophyll a (Chl a) using photosynthetic marker pigments, such as

alloxanthin for cryptophytes, 19’‐hexanoyloxyfucoxanthin for prymnesiophyceans, and other less

specific biomarkers such as fucoxanthin for diatoms (also present in chrysophytes) and zeaxanthin

for cyanobacteria (also present in green algae), among others (Jeffrey et al., 1997). In order to

extract photosynthetic pigments, frozen filters were disrupted with 2 ml of 95% cold‐buffered

methanol (2% ammonium acetate) for 30 min at ‐20º C in the dark. Samples were sonicated for 1

min in the beginning of the extraction period and then centrifuged at 4000 rpm for 15 min, at 4º

C. Extracts were filtered (Millipore membrane filters, 0.2 µm) immediately before injection in the

HPLC to remove cell and filter debris. Pigment extracts were analysed using a Shimadzu HPLC

comprised of a solvent delivery module (IC‐10ADVP) with system controller (SCL‐10AVP) and a

photodiode array (SPD‐M10ADVP). The chromatographic separation of pigments was achieved

Chapter 2

18

using the method described in Zapata et al. (2000), which uses a monomeric OS C8 column and a

mobile phase constituted by two solutions: methanol:acetonitrile:aqueous pyridine and,

acetonitrile:acetone; a flow rate of 1 mL min‐1 and a run duration of 40 min. Pigments were

identified by comparison of retention times and absorption spectra with pure crystalline

standards.

Biovolume, using sedimentation volumes with a conical jar and 24 h settling time, was

measured to assess zooplankton abundance, estimated through planktonic organisms larger than

405 µm. Large gelatinous organisms (e.g. jelly fish) were removed because their significant

buoyancy makes the method less precise (Postel et al., 2000).

Statistical analysis

In order to test for cyclic phenomena in the variation of the different parameters and their co‐

variations, autocorrelations and linear cross‐correlations were performed. Significant

autocorrelations with a 2 h lag were analysed considering the 2‐hour sampling periodicity.

Spearman correlation and Student’s t‐test were applied. The study of phytoplankton community

was done through the analysis of chemotaxonomic pigments and their ratios to Chl a.

Furthermore, the Fp pigment index (Claustre, 1994) was calculated in order to identify the trophic

status of this marine ecosystem. The Fp pigment index is given by:

Fp = (Σ fucoxanthin + Σ peridinin) x (Σ fucoxanthin + Σ peridinin + Σ 19’‐hexanoyloxyfucoxanthin

+ Σ 19’‐butanoyloxyfucoxanthin + Σ zeaxanthin + Σ chlorophyll b + Σ alloxanthin) ‐1

All the statistical analyses were carried out using R (R Development Core Team, 2008) while

maps for display horizontal distributions were processed using Ocean Data View (Schlitzer, 2008).

Results

Horizontal distribution patterns

The horizontal variations of hydrological data evidenced the onshore‐offshore gradients (Fig.

2A and B). Despite the complex pattern of temperature and salinity distribution, the lowest

temperatures were found in the most onshore stations and generally lower salinities were

associated to warmer temperatures. However, the more coastal water mass showed higher

salinity and lower temperature that the central water mass. Examining the vertical sections of the


19

northern and southern transects (Fig. 3), a central less saline water mass with higher temperature

was observed only at the surface layers, until 20 m deep. Despite the halocline of this structure

was noticeable, vertical temperature profile was stratified along the water column.

Fig. 2. Surface distribution of salinity (A) temperature (B, ºC), phosphates (C, µmol l‐1), silicates (D, µmol l‐1), total chlorophyll a (E, fluorescence data) and zeaxanthin (F, µg l‐1).

Chapter 2

20

Fig. 3. Vertical cross‐section of the northern (A) and southern (B) transect perpendicular to shore, as regards to salinity (1), temperature (2, ºC) and total chlorophyll a (3, fluorescence data) variation.

Concerning nutrients concentration, P was higher in the northern transect and Si

concentration was higher near coast (Fig. 2C and D), while N was always under method detection

limit (< 0.005 µmol l‐1). Shelf vertical cross sections of P concentration presented no vertical trend,

while for S higher concentrations on surface onshore water mass were observed (data not

shown). Phytoplankton pigments presented very low concentrations in almost every radial

station, with a distribution pattern similar to fluorescence data, where the southern and closest

stations to the coast presented higher biomass signal, with a lowering gradient perpendicular to

coast (Fig. 2E), and no different pigment composition at the central water mass. Vertical sections

of the outermost transects displayed different phytoplankton distributions given by fluorescence

signal (Fig. 3), with higher biomass on surface water mass nearer to coast on the southern

transect, while in the northern transect higher values were obtain near the bottom, where a

discontinuity occurs. Zeaxanthin concentration was the only pigment with a different spatial

distribution, since higher concentrations were observed in the most northern and offshore

stations (Fig. 2F). Zooplankton total abundance was generally higher in the northern transect (Fig.

4), apart from the most offshore station where total biovolume was extremely low, which was

also observed in the other transects. The majority of night sampled stations (two northern

transects) presented higher total zooplankton abundances than day stations.


21

Fig. 4. Horizontal distribution of zooplankton water column total abundance.

Vertical dynamics

Current direction and magnitude measurements were very similar throughout water column

and results observed were in accordance with tidal variation (Table 1). Current direction varied

mainly between South and West, respectively during flooding and ebbing tides, and speed was

significantly higher during flood tides (t value = ‐ 4.7042, p < 0.01), with average current of 252

and 202 mm s‐1, respectively for flood and ebb tides. Minimum velocity values observed were

coincident with predicted tides at Chinde, near the Zambezi delta (Fig. 5), thus showing no

significant delay to the sampling area. Current direction was similar in all sampling depths

throughout the 48‐hour sampling period and the residual currents were generally over 150 mm s‐1

from South. However, current speed presented vertical variation without apparent consistence,

i.e., in some periods higher values were at the 16 m deep while in others at 25 m. It should be

noted that the ADCP only registered below the 16 m depth layer, thus results do not show the

water mass behaviour above that depth. Water transparency varied mainly between 10 and 20 m

deep, presenting no significant periodicity throughout sampling period.

35.5 36.0 36.5 37.0 37.5

-19.6

-19.4

-19.2

-19.0

-18.8

-18.6

-18.4

Biovolume

Longitude

Latitude

1.6

3.2

4.8

6.4

8

ml.l !1

Chapter 2

22

Table 1. Auto‐ and cross‐correlations of relevant hydrobiological parameters and their significance

Parameter Periodicity (h) Significance Depth (m) (Cross ‐ ) correlated with Temperature 6 (opp.) Tide variation 20, 30*, 40* Current direction** (opp. phase; r = 0.681) Salinity 6 (opp.) Tide variation 20 Current direction** (r = 0.681) Fluoroscence 6 (opp.) Tide variation 30, 40 ‐ Current magnitude 12 Tidal cycle 0 – 30** ‐ Current direction 12 Tidal cycle All depths*** N* (r = 0.363) NO2

‐ + NO3‐ 6 (opp.) Tide variation 40 Si*** (r = 0.533)

Si(OH)4 6 (opp.) Tide variation 20* P* (r = 0.369) diadinoxanthin (30 m: r = 0.289; 40 m: r = 0.213) fucoxanthin (30 m: r = 0.366; 40 m: r = 0.468)

Zeaxanthin 4 Water mass 20** ‐ Zooplankton 12 (opp.) Diel cycle All** (except 10/20 m) ‐

* p < 0.1, ** p < 0.05; *** p < 0.01; opp., opposed phase


23

Fig. 5. Variation of 1‐h average currents direction (a, degrees) and magnitude (b, mm s‐1) at several depths ( ‐ 16 m, ‐ 25 m, ‐ 34 m). Upper and downward arrows indicate predicted high and low tides, respectively; shaded rectangles indicate night periods.

Chapter 2

24

Fig. 6. Temperature (a, ºC) and salinity (b) variation during 48‐hour cycle. Upper and downward arrows

indicate predicted high and low tides, respectively; shaded rectangles indicate night periods.

The water column was vertically stratified below the thermocline during sampling period, as

seemed clearly from salinity data (Fig. 6). Thermocline and halocline depths oscillated between 15

and 20 m for most of the cycle, and both temperature and salinity followed the same trend with a

6‐hour opposed‐phase periodicity (Table 1). During the low tide period it was generally observed

less saline waters and warmer temperatures down to greater depths in the water column (Fig. 6).

Throughout the last period of the 48‐hour cycle the water column physical properties did not

show the same dynamics, specifically in the end of the second day where an intrusion of less

saline waters was observed deeper (around 30 m).

Nutrient concentrations ranged from 0 to 2.01 µmol l‐1 for N, 1.65 to 2.94 µmol l‐1 for P and

4.70 to 15.70 µmol l‐1 for Si (Fig. 7). N concentration was higher near the bottom during low tides,

which is consistent with a significant 6‐hour opposed‐phase periodicity and to the strong

correlation found with current direction at the 40 m depth layer (Table 1). Despite Si

concentration only showed tidal periodicity at shallower depths, it was positively correlated to N

and P at 40 m deep (Table 1). Nutrient ratios N:P and N:Si were calculated and ranged both from 0


25

to 0.41 and 0.15, respectively. These ratios were strongly determined by N concentration that was

generally under the method detection limit.

Fig. 7. Nutrient concentrations (µmol l‐1) during the 48‐hour cycle (a ‐ N;b ‐ P;c ‐ Si). Upper and downward arrows indicate predicted high and low tides, respectively; shaded rectangles indicate night periods.

Chapter 2

26

Fig. 8. Pigment ratios during the 48‐hour cycle (left graphic: ‐ chlorophyll c3, ‐ chlorophyll c1+c2, ‐ zeaxanthin ‐ pheophorbide a; right graphic: ‐ 19’‐hexanoyloxyfucoxanthin, ‐ chlorophyll b, ‐ 19’‐butanoyloxyfucoxanthin, ‐ fucoxanthin). Upper and downward arrows indicate predicted high and low tides, respectively; shaded rectangles indicate night periods.

! ! !

!

! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !

Index

0

1

2

3

! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !

Index

0

1

2

3

! ! ! ! ! ! ! ! ! ! ! ! ! ! !

!

! ! ! ! !

!

! !

Index

0

1

2

3

! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !

!

!

!

Index

0

1

2

3

!

!!

!

!

! ! !!

!

!

!

!

!

!

!

!

! ! ! !

!

!!

Index

0

1

2

3

17 21 1 5 9 13 17 21 1 5 9 13

Index

Index

Index

Index

Index

17 21 1 5 9 13 17 21 1 5 9 13

Local Hour (GMT + 2)

Pig

me

nt

rati

os

5 m

10 m

20 m

30 m

40 m


27

Deep chlorophyll maximum (DCM) was close to the bottom (around 40 m deep) and

fluorescence data from 30 and 40 m presented a 6‐hour opposed‐phase periodicity variation

(Table 1). HPLC analysis revealed a total of 13 pigments, generally presenting low concentrations

(Table 2). Chl a concentration was very low throughout the study (between 0.003 and 0.202 µg l‐1)

and the highest values were normally found at 40 m deep. The most abundant pigments were

19’‐hexanoyloxyfucoxanthin, chlorophyll b and pheophorbide a, located mainly between 20 and

40 m deep. Other detected pigments, such as alloxanthin, diadinoxanthin, peridinin and

prasinoxanthin, were very scarce, and higher concentrations were also observed at 40 m deep

(Table 2). Vertical dynamics of pigment ratios is shown in Fig. 8. The only pigment with a different

vertical distribution pattern was zeaxanthin, indicator pigment of cyanobacteria, which presented

low concentrations at deeper strata (30 and 40 m) and higher at the upper ones (t value = ‐

5.3773, p < 0.01). Concerning the relation between pigments and nutrients, apart from zeaxanthin

and peridinin, all pigments were correlated to N at the deeper strata (30 and 40 m), while

correlations to P, when significant, were always negative. Si was positively correlated to

diadinoxanthin and fucoxanthin at the deeper stratum (Table 1). As regards to the trophic status

obtained though the Fp pigment index, it was calculated for each sampled vertical profile (every 2

h), and results obtained were on average 0.16 (sd = 0.09) for the 48‐hour sampling period.

Zooplankton biovolume ranged from 0.04 to 3.16 ml l‐1, with maximum values near the bottom

during daylight and near the surface layers at night periods (Fig. 9). During dark periods the

biovolume at all strata presented similar values denoting an even distribution of organisms

throughout the water column vertical profile, while during day zooplankton abundance was very

different along the sampled strata. Significant 12‐hour cyclic variations were observed for

zooplankton at both shallower and deeper samples (Table 1). Nonetheless, the maximum

zooplankton higher density was consistently at the bottom layers, a similar pattern when

compared to pigments vertical distribution.

Chapter 2

28

Table 2. HPLC pigment mean concentrations (µg l‐1) at 40 m deep and their associated phytoplankton classes (Jeffrey et al. 1997)

Pigments Concentration (min ‐ max) % Occurrence Chlorophyll a 0.080 (0.013‐0.189) 41 A proxy of total algae biomass Chlorophyll c1, c2 0.024 (0.000 – 0.071) 12.3 Diatoms, prymnesiophytes, crysophytes, dinoflagellates Chlorophyll c3 0.062 (0.000 – 0.145) 31.8 Crysophytes, prymnesiophytes Chlorophyll b 0.029 (0.000 – 0.079) 14.9 Chlorophytes, euglenophytes, prasinophytes Total chlorophylls 0. 195 (0.013 – 0.484) 100 Fucoxanthin 0.054 (0.000 – 0.210) 27.1 Diatoms, prymnesiophytes, crysophytes Peridinin 0.003 (0.000 – 0.143) 1.5 Dinoflagellates Diadinoxanthin 0.008 (0.000 – 0.021) 4.0 Diatoms, prymensiophytes, crysophytes, dinoflagellates 19’‐hexanoyloxyfucoxanthin 0.086 (0.000 – 0.154) 43.2 Prymnesiophytes Alloxanthin 0.001 (0.000 – 0.002) 0.5 Cryptophytes Prasinoxanthin 0.012 (0.000 – 0.028) 6.0 Prasinophytes Zeaxanthin 0.007 (0.000 – 0.023) 3.5 Cyanobacteria, chlorophytes 19’‐butanoyloxyfucoxanthin 0.028 (0.000 – 0.066) 14.1 Crysophytes, prymnesiophytes Total carotenoids 0.199 (0.000 – 0.647) 100 Pheophorbide a 0.091 (0.000 – 0.197) Zooplankton grazing


29

Fig. 9. Zooplankton abundance (ml l‐1) at sampled strata ( ‐ neuston, ‐ 0/5 m, ‐ 5/10 m, ‐ 10/20 m, ‐ 20/30 m, ‐ 30/40 m) during the 48‐hour cycle. Upper and downward arrows indicate predicted high and low tides, respectively; shaded rectangles indicate night periods.

Chapter 2

30

Discussion

Hydrology and nutrients

The Sofala Bank is a Mozambican wide shelf region with strong tidal currents and great

hydrological variability, which is largely determined by the Zambezi delta discharge and also

mesoscale oceanographic features of the Mozambique Channel (Ridderinkhof et al., 2001;

Lutjeharms, 2006). Surface distribution of hydrological parameters, combined with tidal movements

during the 48‐hour period, showed the complexity of this shelf region, and the inshore‐offshore

gradient revealed this influence of both the Zambezi delta runoff and heterogeneity of

oceanographic currents on the mid continental shelf waters. During the sampled period, the

estuarine water intrusion was not significantly affecting this shelf region, since salinity here obtained

near the coast was around 35, which is considerably higher than salinity values similar to 20,

described by Lutjeharms (2006) for the Sofala Bank region during the wet season. The period from

January to March is when higher rainfalls are observed and greater effects of estuarine discharges

influence this shelf region, thereby increasing nutrients inputs and decreasing light penetration due

to dispersal of the turbid estuarine plumes. This study was carried out in December, at the end of dry

season in the area, before the beginning of the rainy season. Accordingly, minor estuarine runoff

effects were observed, explaining the high salinities and water transparency observed. A central

water mass with different water properties from the onshore and offshore analysed coastal region

was evidenced by vertical cross‐sections perpendicular to shore (Fig. 3). The existence of this water

mass could be a marker of an earlier estuarine discharge event, since the distinct lower salinity and

its only presence at the surface (< 20 m). However, evidence of this particular structure was not

observed through nutrient concentrations, and only Si concentrations evidenced a gradient

perpendicular to shore.

Diel hydrological variations also showed a degree of river influence through lower salinity in the

upper water column layers, modulated by tidal currents as well as other oceanographic currents. The

lower salinity values observed until greater depths at the end of the sampling period (Fig. 6) could be

explained by vertical mixing phenomena, since during that period the tidal currents were slightly

different for analysed depths, and the surface water mass seemed to be homogenised down to

greater depths. Tidal currents also affected nutrient dynamics, observed by N and Si concentrations

periodicity at certain depths. These trends were probably associated to sediment resuspension, and

not to estuarine intrusion, as a source of nutrient loadings, since none of nutrient periodicities were

detected near surface, where less dense estuarine waters were present. Nitrogen source is typically


31

from river effluents (Verity et al., 1993; Wawrik et al., 2004), however, the N distribution here

observed suggests vertical resuspension as its source ascribable to tidal energy, which could also

explain the correlation amongst nutrients concentrations at 40 m deep.

Nutrient concentrations obtained were typical from oligotrophic waters (Kromkamp et al., 1997;

Tyrrell, 1999; Giraud et al., 2008) and the extremely low concentrations of N in relation to P and Si

should be carefully discussed. During the 48‐hour cycle P concentration was on average 2.253 µmol l‐

1, which are high concentrations when compared to other coastal zones studies (e.g. Balkis, 2003;

Wawrik et al., 2004; Sabetta et al., 2008) though comparable concentrations to other Western Indian

Ocean studies (e.g. Paula et al., 1998; Lugomela et al., 2001). Regarding Si, average concentration

was 8.857 µmol l‐1, considerably higher than P and N, in accordance with other studies at the

Western Indian Ocean (Paula et al., 1998; Barlow et al., 2007, 2008). The unusual N:P and N:Si ratios

observed were strongly determined by the extremely low N concentrations. Several studies have also

found such low ratios. Wawrick et al. (2004) studied the nutrient dynamic in the Mississippi River

plume and obtained nutrient ratios of N:P ~ 2 and N:Si ~ 0.2 at non‐plume stations, while Burford et

al. (1995) obtained N:P ratios generally less than 4, ranging from 0.1 to 20, in the Gulf of Carpentaria

(Australia). For the Western Indian Ocean, in particular, Kitheka et al. (1996), reporting results on

nutrient dynamics at the Kenya coast, observed mean N:P ratios ~1 during the flood tide. Mengesha

et al. (1999) also observed very low N concentrations ranging from < 0.1 to 0.41 µmol l‐1, even lower

concentrations than observed at Sofala Bank. These authors identified ammonia as the major

inorganic nitrogenous nutrient, representing on average 72% of the total dissolved inorganic

nitrogen concentrations. In the present study ammonia concentration was not measured, however it

is unlikely that if this component was taken into account the N:P ratio would change significantly. For

N at oligotrophic waters, Lalli and Parsons (1977) indicated values of 0.01 to 0.1 uM as constants of

semi‐saturation. Therefore, as discussed by several authors (Burford et al., 1995; Tyrrell, 1999; Balkis,

2003; Wawrick et al., 2004; Howarth and Marino, 2006), it seems clear the limiting role of N for

Sofala Bank during our study, thus influencing phytoplankton abundance and composition.

Phytoplankton

Sofala Bank is known as one of the most productive shelf regions of Mozambique (Lutjeharms,

2006; Barlow et al., 2008), however the horizontal distribution of chl a revealed very low

phytoplankton biomass, apart from the nearer to coast area where the effects of river runoff

nutrient loadings undergo first. Currents describe south‐western directions throughout the tidal

Chapter 2

32

wave rotation, therefore the estuarine plume was most likely moving southwards and major

estuarine effluents were enhancing phytoplankton biomass at this southern nearer to coast area.

This hypothesis is confirmed by vertical cross‐section of the southern transect, where phytoplankton

biomass was higher at the surface waters nearer to shore. On the other hand, the vertical profiles of

northern stations display higher phytoplankton biomass near the bottom, where a discontinuity

occurs. This bottom discontinuity probably promotes the resuspension of sediment ascribed to tidal

direction and magnitude variation, supporting nutrients availability that enhanced phytoplankton

biomass.

Apart from distance to shore and associated environmental gradients, phytoplankton

communities’ structure and abundance are also determined by hydrological vertical variations

(Balkis, 2003; Barlow et al., 2007; Sabetta et al., 2008). During the 48‐hour cycle the majority of

phytoplankton biomass was near the bottom, probably because of nutrient availability and high light

penetration, as DCM was always below the pycnocline depth. Nutrient availability often determines

primary production vertical patterns and phytoplankton communities’ structure and abundance

(Verity et al., 1993; Bouman et al., 2003). The high planktonic concentrations observed near the 30

and 40 m layers are explained through the availability of N near the bottom and its role limiting

phytoplankton biomass. As regards to light penetration, the exposure to high light radiation induce

photodamage and physiological stress in phytoplankton cells, even though the existence of some

photoprotective pigments (Brunet et al., 2008). Therefore, phytoplankton depth regulation will

depend upon nutrients availability and light penetration, in order to lead to a better exploration of

the ecosystem resources by phytoplankton cells. Moreover, biological responses to environmental

conditions seem to be simultaneously depending on the group and size of phytoplankton species

(Brunet and Lizon, 2003).

Microflagellates phytoplankton cells (e.g. prymnesiophyceans) were the most abundant group

observed at Sofala Bank, which is in agreement with Claustre (1994), Cortés et al. (2001) and

Dandonneau et al. (2006), among others, who pointed out the preference of prochlorophytes and

small flagellates to be most adapted to survive in impoverished oligotrophic environments. The

pigment ratios > 1 obtained indicates pigment concentrations higher than Chl a, which is in

accordance to other phytoplankton communities from distinct oligotrophic environments (e.g. North

Atlantic and Mediterranean Sea) where phytoplanktonic cells contain more accessory pigments that

Chl a (Claustre, 1994). The trophic status here identified by nutrients composition is in accordance to

the Fp pigment index calculated. The 0.16 mean pigment ratio obtained agrees with results from the

North Atlantic (0.06 ± 0.01) and Mediterranean (0.18 ± 0.01) oligotrophic regions (Claustre, 1994),

where picoplankton and nanoplankton are dominant.


33

The majority of detected pigments displayed similar vertical movements in accordance with tidal

variation and higher concentrations near the bottom. Despite that tidal energy was a key factor in

water mass physico‐chemical properties variation, the majority of phytoplankton groups temporal

variation revealed strong correlations to N, and also diatoms to Si. Low diatoms abundance was

probably related to high temperatures registered, which limits their growth (Verity et al., 1993;

Bouman et al., 2003; Barlow et al., 2008), while peridinin containing dinoflagellates were probably

limited by these oligotrophic waters.

Cyanobacteria were the only analysed group that seemed to have its maximum abundance at

surface layers above thermocline depth, where turbulence phenomena are less intense, which is in

accordance with other studies describing shallower distribution patterns of Cyanobacteria,

specifically the ubiquitous Synechococcus sp. throughout world oceans (Glover et al., 1988; Goericke

et al., 2000; Brunet et al., 2008). Cyanobacteria variation periodicity at 20 m deep was probably

related to the water mass cyclical movements, as seen by salinity and temperature variations at the

same depth Zeaxanthin presence is normally associated to Chlorophytes and Cyanobacteria,

however, since in this study there were no correlations between zeaxanthin and chlorophyll b

(marker pigment of Chlorophytes) and divinyl chlorophyll a was below detection limits, which is also

in accordance with Barlow et al. (2008) for inshore stations at Delagoa Bight (Mozambique),

zeaxanthin here detected was attributed to Cyanobacteria, which was also assumed in other

Western Indian Ocean studies (e.g. Barlow et al., 2007, 2008). Thereby, zeaxanthin horizontal

distribution here observed is in accordance to the preference for offshore waters also detected in

other Cyanobacteria studies (Hajdu et al, 2007; Barlow et al., 2008).

The determination of the vertical distribution and migration patterns of Cyanobacteria and other

phytoplankton groups are easily biased by sampling errors because horizontal advection of the water

mass, disruption of vertical structures by wind and patchiness, which, together with the

accumulation of phytoplankton groups in relatively thin layers, makes it difficult to resolve the real

vertical distribution (Hajdu et al., 2007). Some pigments registered different and amply distributions

along the 48‐hour cycle. Their presence at surface layers, specifically those detected at 5 m deep

could be related to horizontal transport from water mass nearer shore and not to phytoplankton

vertical movements. The advection of water mass occurs during low tides when water movements

are transversal, i.e., inshore ‐ offshore, contributing to the exportation of biomass from the coast.

Therefore, less saline surface waters transport allochthonous biomass from coastal waters where the

Chapter 2

34

chl a concentration is higher, suggesting an enhancement of the system productivity ascribable to

tidal energy (Brunet and Lizon, 2003).

Zooplankton

Zooplankton biovolume measurements gave rough estimations of zooplankton abundance with

no detailed information concerning zooplankton groups, however showed the trend of the

community as a whole. Zooplankton horizontal distribution was greater at the northern stations,

where estuarine influence and phytoplankton biomass were concomitantly lower, as it probably

suffered high grazing pressure. However, radial stations where zooplankton abundance was higher

were sampled during night, therefore giving a biased view of zooplankton distribution.

During the 48‐hour sampling, zooplankton abundance was higher at the upper strata during night

and at deeper strata during day. These results are in accordance with the typical diel vertical

migration (DVM) behaviour where light variation plays a key role. However, the deeper affinity of

zooplankton here verified could also be related to deeper DCM, since most of zooplankton organisms

are herbivorous. Zooplankton higher densities was often at layers deeper than 20 m and there was

no greater shallower abundances during the night, which might be related to energy acquisition and

phytoplankton deeper affinity, and to the avoidance of unnecessary risks related to energy

expenditure and predation threat (Zaret and Suffern, 1976; Lampert, 1989). The absence of

significant correlations of zooplankton to any hydrobiological parameter could be explained by

different feeding activity periodicities, linked to the different species diversity and/or developmental

stages (Daro, 1985). Pheophorbide a has been considered a proxy of zooplankton grazing (Barlow et

al., 1993; Brotas and Plante‐Cuny, 1998), and its concentration at shallower levels here observed

during night periods could be associated to zooplankton DVM. Pheophorbide a higher concentration

at 40 m deep could also be related to grazing as well to sediment resuspension and phytoplankton

degradation products, thus it was not possible to determine their source and the real impact of

zooplankton on phytoplankton community structure and abundance.

Conclusion

Low phytoplankton biomass and nutrient concentrations here obtained are in accordance with

typical oligotrophic regions, yet a productive shelf region such as the Sofala Bank should have rich

nutrient loadings and high phytoplankton biomass to support the strong fisheries levels that

characterize the area. The unexpected oligotrophic features observed were probably related to low


35

precipitation levels prior to and during the sampling period, therefore conditioned by low estuarine

discharges and associated low nutrients inputs. Even though nutrients flux from the Zambezi delta

may be significantly subsequent to rain, concentrations in coastal waters could also remain low due

to mixing of different water masses from the estuarine effluents and shelf waters (Verity et al.,

1993).

Nutrient dynamics here observed was strongly determined by N concentration, however data

collected and incomplete existent knowledge about Sofala Bank is clearly insufficient to a full

comprehension of the abnormal nutrient ratios obtained. Even though, it was possible to conclude

that phytoplankton biomass was strongly limited by N availability, which together with tidal

variations were important driving factors modulating this ecosystem vertical dynamics.

Phytoplankton abundance and community structure, given by chemotaxonomic pigments, was

dominated by prymnesiophyceans and other microflagellates, and vertical distribution changes on

pigments concentration were quite probably associated to water mass advection, which is an

important process to enhance offshore system productivity. Even though the horizontal hydrological

processes influencing phytoplankton, the highest pigment concentrations were observed near the

bottom layers (30 and 40 m) and not at surface layers. This phytoplankton deeper affinity was quite

probably influencing zooplankton vertical movements, since most of zooplankton organisms are

herbivorous and no greater shallow abundances during night were observed. However, it still

remains to be answered why some zooplankton organisms were migrating if their food was on the

bottom layers.

Few worldwide coastal studies are aimed to vertical dynamics of planktonic communities, from

hydrological processes and nutrients that modulate production processes to phyto‐ and zooplankton.

The present study contributes therefore to vertical dynamics knowledge on planktonic ecosystems,

specifically for tropical ecosystems were these processes are still poorly understood. However, to

achieve a better understanding, further studies should aim to seasonal variations. Special attention

should be given to nutrient dynamics in tropical ecosystems, in order to understand the

biogeochemical cycles occurring at tropical coastal regions.

Acknowledgements

This study was a Special Study as part of the research cruise ‘Ecosystem Survey Mozambique’ of

the wider program “Strengthening the knowledge base for and implementing an ecosystem

approach to marine fisheries in developing countries” from the Nansen Program, an EAF Project

Chapter 2

36

funded by the Norwegian government through NORAD, the Institute of Marine Research (Bergen,

Norway) and the United Nations through FAO. We thank IIP (Instituto de Investigação Pesqueira,

Mozambique) and its current director, Dr. Domingos Gove, for all support. We are grateful for the

productive assistance provided by the crew of R/V Dr. Fridtjof Nansen and the helpful support of J‐O.

Krakstad, D. Zaera and C. Bento as well as other fellow scientists from IIP onboard. The authors thank

C. R. Mendes for his contribution to nutrient and pigment analysis, M. Lipassula for his contribution

to nutrient and zooplankton processing and to P. Oliveira for his helpful contribution interpreting

ADCP data.

References

Balkis, N., 2003. Seasonal variations in the phytoplankton and nutrient dynamics in the neritic water of Büyükçekmece Bay, Sea of Marmara. Journal of Plankton Research 25, 703‐717

Bandeira, S.O., Silva, R.P., Paula, J., Macia, A., Hernroth, L., Guissamulo, A.T., Gove, D.Z., 2002. Marine Biological Research in Mozambique: Past, Present and Future. AMBIO 31, 606‐609

Barlow, R.G., Mantoura, R.F.C., Gough, M.A., Fileman, A.T., 1993. Phaeopigment distribution during the 1990 spring bloom in the Northeastern Atlantic. Deep‐Sea Research Part I 40, 2229‐2242

Barlow, R., Stuart, V., Lutz, V., Sessions, H., Sathyendranath, S., Platt, T., Kyewalyanga, M., Clementson, L., Fukasawa, M., Watanabe, S., Devred, E., 2007. Seasonal pigment patterns of surface phytoplankton in the subtropical southern hemisphere. Deep‐Sea Research Part I 54, 1697‐1703

Barlow, R., Kyewalyanga, M., Sessions, H., van der Berg, M., Morris, T., 2008. Phytoplankton pigments, functional type, and absorption properties in the Delagoa and Natal Bights of the Agulhas ecosystem. Estuarine, Coastal and Shelf Science 80, 201‐211

Bendshneider, K., Robison, N.J., 1952. A New spectrophotometric determination of nitrite in seawater. Journal of Marine Research 11, 87‐96

Bouman, H.A., Platt., T., Sathyendranath, S., Li, W.K.W., Stuart, V., Fuentes‐Yaco, C., Maass, H., Horne, E.P.W., Ulloa, O., Lutz, V., Kyewalyanga, M., 2003. Temperature as indicator of optical properties and community structure of marine phytoplankton: implications for remote sensing. Marine Ecology Progress Series 258, 19‐30

Brotas, V., Plante‐Cuny, M.R., 1998. Spatial and temporal patterns of microphytobenthic taxa of estuarine tidal flats in the Tagus Estuary (Portugal) using pigment analysis by HPLC. Marine Ecology Progress Series 171, 43‐57

Brunet, C., Lizon, F., 2003. Tidal and diel periodicities of size‐fractionated phytoplankton pigment signatures at an offshore station in the southeastern English Channel. Estuarine, Coast and Shelf Science 56, 833‐843

Brunet, C., Gasotti, R., Vantrepotte, V., 2008. Phytoplankton diel and vertical variability in photobiological responses at a coastal station in the Mediterranean Sea. Journal of Plankton Research 30, 645‐654


37

Burford, M.A., Rothlisberg, P. C., Wang, Y., 1995. Spatial and temporal distribution of tropical phytoplankton species and biomass in the Gulf of Carpentaria, Australia. Marine Ecology Progress Series 118, 255‐266

Claustre, H., 1994. The trophic status of various oceanic provinces as revealed by phytoplankton pigment signatures. Limnology Oceanography 39, 1206‐1210.

Cortés, M.Y., Bollmann, J., Thierstein, H.R., 2001. Coccolithophore ecology at the HOT station ALOHA, Hawaii. Deep‐Sea Research Part II 48, 1957‐1981.

Criales‐Hernández, M.I., Schwamborn, R., Graco, M., Ayón, P., Hirche, H.J., Wolf, M., 2008. Zooplankton vertical distribution and migration off Central Peru in relation to the oxygen minimum layer. Helgoland Marine Research 62, S85‐S100

Dandonneau, Y., Montel, Y., Blanchot, J., Giraudeau, J., Neveux, J., 2006. Temporal variability in phytoplankton pigments, picoplankton and coccolithophores along a transect through the North Atlantic and tropical southwester Pacific. Deep‐Sea Research I 53, 689‐712

Daro MH (1985) Feeding rhythms and vertical distribution of marine copepods. Bulletin of Marine Science 37, 487‐497

Fanning, K.A., Pilson, M.E.Q., 1973. On the spectrophotometric determination of dissolved silica in natural waters. Analytical Chemistry 45:136‐141

Garnier, J., Cugier, P., 2004. Drainage basin use and nutrient supply by rivers to the coastal zone. A modelling approach to the Seine river. In: P. Wassmann, K. Olli (Editors), Drainage basin nutrient inputs and eutrophication: an integrated approach. University of Tromsø, Norway, p 60‐87

Giraud, X., Quéré, C.L., Cunha, L.C., 2008. Importance of coastal nutrient supply for global ocean biogeochemistry. Global Biogeochemical Cycles, 22, GB2025

Glover, H.E., Prézelin, B.B., Campbell, L., Wyman, M., 1988. Pico‐ and ultraplankton Sargasso Sea communities: variability and comparative distributions of Synechococcus spp. and algae. Marine Ecology Progress Series 49, 127‐139

Goericke, R., Olson, R.J., Shalapyonok, A., 2000. A novel niche for Prochlorococcus sp. in low‐light suboxic environments in the Arabian Sea and the Eastern Tropical North Pacific. Deep‐Sea Research Part I 47, 1183‐1205

González‐Gordillo, J.I., Rodríguez, A., 2003. Comparative seasonal and spatial distribution of decapod larvae assemblages in three coastal zones off the south‐western Iberian Peninsula. Acta Oecologica 24, S219‐S233

Grasshoff, K., 1976. Methods of Seawater Analysis. Verlag Chimie, New York

Hajdu, S., Höglander, H., Larsson, U., 2007. Phytoplankton vertical distributions and composition in Baltic Sea cyanobacterial blooms. Harmful Algae 6, 189‐205

Hecky, R.E., Kilham, P., 1988. Nutrient limitation of phytoplankton in freshwater and marine environments: A review of recent evidence on the effects of enrichment. Limnology Oceanography 33, 796‐822

Hill, A.E., 1991. Vertical migration in tidal currents. Marine Ecology Progress Series 75, 39‐54.

Howarth, R.W., Marino, R., 2006. Nitrogen as the limiting nutrient for eutrophication in costal marine ecosystems: Evolving views over three decades. Limnology Oceanography 51, 364‐376

Chapter 2

38

Jeffrey, S.W., Mantoura, R.F.C., Wright, S.W., 1997. Phytoplankton Pigments in Oceanography. UNESCO, Paris, 661 pp.

Kitheka, J.U., Ohowa, B.O., Mwashote, B. M., Shimbira, W.S., Mwaluma, J. M., Kazungu, J. M., 1996. Water circulation dynamics, water column nutrients and plankton productivity in a well‐flushed tropical bay in Kenya. Journal of Sea Research 35, 257‐268

Kromkamp, J., Bie, M., Goosen, N., Peene, J., Rijswijk, P.V., Sinke, J., Duineveld, G.C.A., 1997. Primary production by phytoplankton along the Kenyan coast during the SE monsoon and November intermonsoon 1992, and the occurrence of Trichodesmium. Deep‐Sea Research Part II 44, 1195‐1212

Lampert, W., 1989. The adaptive significance of diel vertical migration of zooplankton. Functional Ecology 3, 21‐27

Lalli, C.M., Parsons, T.R., 1997. Biological Oceanography. An Introduction, Second Edition, The Open University, 314 pp.

Lugomela, C., Wallberg, P., Nielsen, T.G., 2001. Plankton composition and cycling of carbon during the rainy season in a tropical coastal ecosystem, Zanzibar, Tanzania. Journal of Plankton Research 23, 1121‐1136

Lutjeharms, J.R.E., 2006. The costal oceans of south‐eastern Africa. In: A.R. Robinson, K.H. Brink (Editors) The Sea Vol 14. Harvard University Press, Cambridge, MA, pp. 783‐834

Mengesha, S., Dehairs, F., Elskens, M., Goeyens, L., 1999. Phytoplankton Nitrogen Nutrition in the Western Indian Ocean: Ecophysiological Adaptations of Neritic and Oceanic Assemblages to Ammonium Supply. Estuarine, Coastal and Shelf Science 48, 589‐598

Moore, L.R., Goericke, R., Chisholm, S.W., 1995. Comparative physiology of Synechococcus and Prochlorococcus: influence of light and temperature on growth, pigments, fluorescence and absortive properties. Marine Ecology Progress Series 116, 259‐275

Murphy, J., Riley, J.P., 1962. A modified single solution method for the determination of phosphate in natural waters. Analytica Chimica Acta 27, 31‐36

Paula, J., Pinto, I., Guambe, I., Monteiro, S., Gove, D., Guerreiro, J., 1998. Seasonal cycle of planktonic communities at Inhaca Island, southern Mozambique. Journal of Plankton Research 20, 2165‐2178

Postel, L., Fock, H., Hagen, W., 2000. Biomass and abundance. In: R.P. Harris, P.H. Wiebe, J. Lenz, H.R. Skjoldal, M. Huntley (Editors) ICES Zooplankton methodology manual. Academic Press, London, pp. 83‐192

Queiroga, H., Cruz, T., dos Santos, A., Dubert, J., González‐Gordillo, J.I., Paula, J., Peliz, A., Santos, A.M.P., 2007. Oceanographic and behavioural processes affecting invertebrate larval dispersal and supply in the western Iberia upwelling ecosystem. Progress in Oceanography 74, 174‐191

R Development Core Team, 2008. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. WWW Page, http://www.R‐project.org

Ridderinkhof, H., Lutheharms, J.R.E., Ruijter, W.P.M., 2001. A research cruise to investigate the Mozambique Current. South African Journal of Science 97, 461‐464

Sabetta, L., Basset, A., Spezie, G., 2008. Marine phytoplankton size‐frequency distributions: Spatial patterns and decoding mechanisms. Estuarine, Coastal and Shelf Science 80, 181‐192

Santos, A.M.P., Ré, P., dos Santos, A., Peliz, A., 2006. Vertical distribution of the European sardine (Sardina pilchardus) larvae and its implications for their survival. Journal of Plankton Research 28, 523‐532

Schlitzer, R., 2008. Ocean Data View. WWW Page, http://odv.awi.de


39

Silva, A., Mendes, C.R., Palma, S., Brotas, V., 2008. Short‐time scale variation of phytoplankton succession in Lisbon bay (Portugal) as revealed by microscopy cell counts and HPLC pigment analysis. Estuarine, Coastal and Shelf Science 79, 230‐238

Svensson, J.E., Stenson, J.A.E., 1991. Herbivoran impact on phytoplankton community structure. Hydrobiologia 226, 71‐80

Tyrrell, T., 1999. The relative influences of nitrogen and phosphorus on oceanic primary production. Nature 400, 525‐531

Urbach, E., Scanlan, D.J., Distel, D.L., Waterbury, J.B., Chisholm, S.W., 1998. Rapid Diversification of Marine Picophytoplankton with Dissimilar Light‐Harvesting Structures Inferred from Sequences of Prochlorococcys and Synechococcus (Cyanobacteria). Journal of Molecular Evolution 46, 188‐201

Vanni, M.J., 1987. Effects of Nutrients and Zooplankton Size on the Structure of a Phytoplankton Community. Ecology 68, 624‐635

Verity, P.G., Yoder, J.A., Bishop, S.S., Nelson, J.R., Craven, D.B., Blanton, J.O., Robertson, C.Y., Tronzo, C.R., 1993. Composition, productivity and nutrient chemistry of a coastal ocean planktonic food web. Continental Shelf Research 13, 741‐776

Wawrik, B., Paul, J.H., Bronk, D.A., John, D., Gray, M., 2004. High rates of ammonium recycling drive phytoplankton productivity in the offshore Mississippi River plume. Aquatic Microbial Ecology 35, 175‐184

Zaret, T.M., Suffern, J.S., 1976. Vertical migration in zooplankton as a predator avoidance mechanism. Limnology Oceanography 21, 804‐813

Zapata, M., Rodríguez, F., Garrido, J.L., 2000. Separation of chlorophylls and carotenoids from marine phytoplankton: a new HPLC method using a reversed phase C8 column and pyridine‐containing mobile phases. Marine Ecology Progress Series 195, 29‐45

41

Chapter 3

Vertical dynamics of the gold‐spot herring (Herklotsichthys quadrimaculatus) larvae at Sofala Bank, Mozambique

43

Vertical dynamics of the gold‐spot herring (Herklotsichthys

quadrimaculatus) larvae at Sofala Bank, Mozambique

Miguel Costa Leal1, Tânia Costa Pereira2, Vanda Brotas3, José Paula1

1 Centro de Oceanografia, Laboratório Marítimo da Guia, Faculdade de Ciências da Universidade de Lisboa, Av. Na Senhora do Cabo, 939, 2759‐374 Cascais, Portugal

2 Instituto Nacional de Investigação Pesqueira de Moçambique, Av. Mao Tse‐Tung nº 389, Maputo, C. P. 4603 Moçambique

3 Centro de Oceanografia, Instituto de Oceanografia, Faculdade de Ciências da Universidade de Lisboa, Campo Grande, 1749‐016 Lisboa, Portugal

Abstract

Larval fish vertical movements are important to determine survival and dispersal processes, hence

recruitment variability. This study presents results of the vertical dynamics of the gold‐spot herring

(Herklotsichthys quadrimaculatus) larvae observed at Sofala Bank (Mozambique) throughout a 48‐

hour sampling where depth‐stratified zooplankton samples were taken every 2 h together with CTD

profiles at a fixed station. The vertical position according to ontogenic variation and its relation to

zooplankton abundance were also investigated. Larvae displayed diel vertical migrations and were

mainly distributed in the upper water column (0 – 20 m) during the night and almost absent of

sampled strata during the day. Larger larvae dominated the neuston layer during the night while

during daylight periods remained close to the bottom. Vertical dynamics of H. quadrimaculatus here

observed are discussed considering horizontal dispersal processes, ontogenic changes, feeding

activity upon zooplankton and predation risk.

Keywords: Fish larvae; diel vertical migration; ontogeny; predation risk

Introduction

Larval fish vertical dynamics is one of the key processes determining recruitment success (Fiksen

et al., 2007). Accordingly, the integration of larval behaviour with hydrological circulation patterns is

crucial to improve our knowledge on fish larvae growth, survival and dispersion. Furthermore,

Chapter 3

44

information on its vertical distribution is fundamental for efficient and reliable sampling for studies

on larval ecology (Yamashita et al., 1985), dispersal processes (Hare and Govoni, 2005; Voss et al.,

2007) and stock estimation (Reid, 2001).

Larvae active depth selection has been proven to be advantageous for the avoidance of adverse

physical conditions (e.g. temperature, light, turbulence), optimization of horizontal distribution and

attraction and escaping respectively to high prey and predator density (Olla and Davis, 1990; Heath,

1992; Hare and Govoni, 2005). These features have been widely studied for a number of fish larvae,

including clupeids. Blaxter (1973) studied the effect of light intensity on the vertical movements of

herring and concluded that larvae tended to aggregate close to surface at night, which has also been

described by Voss et al. (2007) and Santos et al. (2006), respectively for sprat and sardine larvae.

These latter studies also investigated changes on diel vertical movements throughout ontogeny and

observed surface night samples dominated by larger larvae, while smaller larvae were widely spread

through water column with minor vertical movements. These patterns were explained by predator

avoidance, feeding activity and preference for hydrological water mass properties. However, other

studies also aimed to clupeid larvae reported the opposite trend, observing larger larvae at deeper

layers during night and few noticeable vertical movements in response to light (Heath et al., 1988;

Olivar et al., 2001; Stenevik et al., 2001). Accordingly, predator avoidance, feeding activity and

preference for hydrological water mass properties were once again pointed out as causative factors

for the pattern described. Vertical behaviour thus seem to vary according to species and to local

hydrographic processes, since light response, prey selectivity, reproduction period and sinking rates

are different amongst species, and unique hydrographic features of each costal zone differently

affect larvae behaviour, which is adapted to maximize recruitment success (Heath et al., 1988; Hare

and Govoni, 2005; Santos et al., 2006; Fisken et al., 2007; Voss et al., 2007).

Despite the numerous studies on several clupeid larvae, none focused on the tropical herring

Herklotsichthys quadrimaculatus (Rüppell 1837), also known as the gold‐spot herring, which is a

major component of tuna baitfish (Lewis, 1990) and consumed by local coastal communities at the

Eastern Africa (KMFRI, 1981). Adult forms are fast growing, feed selectively on high‐calorie

zooplankton taxa and spawn continuously after maturing until dying (Milton et al., 1994). At

Mozambique H. quadrimaculatus is broadly present on the coastal zone, namely at the Sofala Bank

(INIPM, unpublished data), a widely shallow shelf in front of the Zambezi River mouth where the

most important national fisheries occur. Knowledge of Herklotsichthys spp. larvae is very scarce

(Thorrold and Williams, 1989) and for H. quadrimaculatus from the Western Indian Ocean only Harris

and Cyrus (1999) described its high abundance throughout the year near Durban (South Africa). The

vertical distribution and behaviour is still unknown for H. quadrimaculatus and information regarding


45

ontogenic changes on depth distribution throughout diel variations are still missing. Furthermore,

the hydrological unique properties of Sofala Bank, such as strong tidal currents and effects of

estuarine discharge (Lutjeharms, 2006), may change the influence of typical factors regulating larvae

vertical dynamics. Tidal energy and vertical mixing, as well as other local oceanographic features,

could therefore have a major role determining H. quadrimaculatus larvae vertical distribution and

diel movements.

This study was a special study as part of the 2007 Mozambique Ecosystem Survey of the Nansen

EAF Programme, and its main goal was to examine the vertical distribution and diel movements of H.

quadrimaculatus larvae and its variation according to larval size. Furthermore, the relation to several

hydrological and biological factors, such as light variation, vertical mixing and zooplankton

abundance, were also investigated based on a vertical stratified sampling for 48 h.

Material and Methods

Sampling took place onboard the R/V “Dr. Fridtjof Nansen” from the 7th to the 9th December 2007

at the Sofala Bank, specifically near the Zambezi River mouth were river runoff effluents and tidal

currents strongly determine occurring ecological processes (Lutjeharms, 2006). A fixed station was

carried out for 48 h over a bottom depth of 50 m (Fig. 1) and every 2 h transparency was measured

using a Sechi Disk and CTD profiles were conducted recording density, temperature and salinity data

every 1 m depth interval. Detailed and integrated analysis of currents, physico‐chemical (salinity,

temperature, nutrients) and biological (phytoplankton pigments, zooplankton) data are described in

Leal et al. (unpublished data). Depth‐stratified zooplankton samples were taken using a multinet

(Midi model, 0.5 x 0.5 m mouth size, Hydro‐bios) with 405 µm mesh size and towed at ~2 knots,

sampling on oblique hauls for 2 min in each stratum (0‐5, 5‐10, 10‐20, 20‐30 and 30‐40 m), and a

neuston net (0.2 x 1.0 m mouth size) with the same mesh size towed horizontally at similar speed

and time, sampling the upper 20 cm of the water column. Both Hydro‐bios and neuston nets had

flowmeters mounted to monitor flow. All zooplankton samples were preserved in ~4% borax‐

buffered formaldehyde, prepared using seawater. Herring H. quadrimaculatus larvae were

subsequently sorted, counted and measured (standard length) to the nearest 0.1 mm. Lavae were

identified according to Leis and Rennis (1983) and Leis and Carson‐Ewart (2004).

Chapter 3

46

Fig. 1. Location of the sampling station during the 48‐hour cycle.

Zooplankton abundance was measured by means of biovolume using sedimentation conical jars,

for 24 h settling time. Large gelatinous organisms (e.g. jelly fish) were removed because its significant

buoyancy makes the method less precise (Postel et al., 2000). Zooplankton abundance weighted

mean depth (WMD) of was calculated according to Worthington (1931):

WMD = (∑nidi)/(∑ni)

where ni is the zooplankton biovolume at mean depth di of each stratum.

In order to test variation periodicities of hydrological and biological data and their co‐variations,

autocorrelations and linear cross‐correlations were performed. Significant autocorrelations with a 2

h lag were carefully analysed considering the sampling periodicity. Spearman correlation, Student’s t‐

test and one‐way ANOVA were also applied. All the statistical analysis and graphics were carried out

using the software R (R Development Core Team, 2008).

Results

Water column was vertically stratified for physico‐chemical properties during the sampling

period, with intense cline depths varying between 10 and 30 m depth, and a similar variation pattern

throughout the 48‐hours was observed for density, temperature and salinity (Fig. 2). Concerning

depth variation, the upper 10‐20 m of the water column were very homogeneous and presented low

density and salinity and higher temperature, while near bottom higher density and salinity and lower

temperatures were observed. Throughout the sampling period, physio‐chemical vertical variations


47

were very similar. At 20 m deep, density, temperature and salinity presented the same trend with a

6‐hour opposed‐phase periodicity and at 30 and 40 m deep only for salinity and temperature

significant periodicities were observed (Table 1). These 6‐hour opposed‐phase periodicities

correspond to a complete cycle variation of 12‐hours, i.e., every 12 hours the same pattern is

repeated. During the 48‐hour cycle these hydrological properties did not show the same dynamics,

specifically in the end of the second day where clines depth were considerably deeper (until 30 m)

than in former sampling period (Fig. 2). Water transparency varied mainly between 10 and 20 m

deep and no significant periodicity was detected.

Fig. 2. Water column (a) density σt (kg m‐3), (b) temperature (ºC) and (c) salinity variation throughout the 48‐

hour cycle. Dashed rectangles indicate nigh periods and upward and downward arrows indicate high and low tides, respectively.

Chapter 3

48

Table 1. Auto‐ and cross‐correlations of relevant hydrobiological parameters and their significance.

Parameter Periodicity (h) Significance Depth (m) Temperature 6 (opp.) Tide variation 20, 30*, 40 Salinity 6 (opp.) Tide variation 20 Density 6 (opp.) Tide variation 20, 30*, 40* H. quadrimaculatus 12 (opp.) Diel cycle All depths* Zooplankton WMD 12 (opp.) Diel cycle All depths** * P < 0.1, ** P < 0.01; opp., opposed phase

Larvae total abundance in the water column was significantly different during day and night (t test

= 5.38, P < 0.001), with higher total abundances observed during night. However, the highest larval

concentrations were registered during dusk, since higher larvae abundance was observed in the first

night samples (Fig. 3). Zooplankton biovolume ranged from 0.04 to 3.16 ml l‐1 and maximum volumes

were observed near the bottom during daylight and near the surface layers during night samples.

Zooplankton WMD varied mainly between 10 and 30 m deep and followed the same trend registered

for H. quadrimaculatus larvae, with shallower mean depths during night and deeper during day (Fig.

3). Both zooplankton WMD and larval fish total abundance in the water column presented significant

12‐hour periodicity (Table 1) and were significantly negatively correlated to each other (r = ‐0.65, P <

0.001), i.e., when zooplankton WMD was lower (shallower) larval abundance was higher. Larval and

zooplankton abundance were positively correlated during the sampling period at upper strata

(neuston: r = 0.58, P < 0.002; 0/5 m: r = 0.51, P < 0.05; 5/10 m: r = 0.42, P < 0.05), while at deeper

strata (below 20 m deep) there were no statistically significant correlations.

Fig. 3. Distribution of total water column abundance of H. quadrimaculatus larvae (bars) and zooplankton WMD variation () during the 48‐hour cycle. Dashed rectangles indicate nigh periods and upward and downward arrows indicate high and low tides, respectively.

seq(0.5, 23.5)

1/(

(bvo

l_w

md

* 2

)/4

0)

17 19 21 23 1 3 5 7 9 11 13 15 17 19 21 23 1 3 5 7 9 11 13 15

0

10000

20000

30000

40000

40

30

20

10

0

Local hour (GMT + 2)

Ab

un

da

nce

(larvae.1

00

m!3)

Zo

op

lan

kto

n W

MD

(m

)


49

Larval fish distribution lengths varied between 2.99 and 22.89 mm and were significantly different

during day and night (ANOVA, F = 201.45, P < 0.001) as well as throughout water column at different

sampling strata (ANOVA, F = 208.25, P < 0.001). However, the only stratum significantly different

from all other was the neuston layer (Tukey HSD, P < 0.05). Larger larvae were only observed during

the night and showed a clear preference for the neuston layer while smaller larvae tended to

concentrate mainly at middle strata during daylight samples (between 10 and 30 m deep) and at

shallower depths during the night (between 0 and 20 m, except the neuston layer) (Fig. 4). The

analysis of the length composition revealed a displacement for small larval sizes with increasing

depths mainly during the night, since during day larval fish abundance was residual in sampled strata.

Fig. 4. Larval length composition by depth strata and day/night periods. Dawn and dusk times were approximately at 05:00 and 17:30 hours. Black and white columns represent pre‐ and postflexion larvae, respectively.

Chapter 3

50

Discussion

Sofala Bank is a Mozambican wide shelf region characterized by strong tidal currents and large

influence of estuarine runoff effects, which determine hydrological features of this coastal area

(Lutjeharms, 2006). The variation of both tides and estuarine runoff were proven to be important

factors determining water mass physico‐chemical properties that influence planktonic communities

vertical dynamics (Leal et al., unpublished data). These vertical and temporal variations of water

mass properties, once more shown and shortly analysed in this study, were strongly regulated by the

interaction of both tidal cycle and Zambezi delta estuarine discharge, illustrated by the significant

tidal periodicities observed for salinity, temperature and density, as well by their vertical gradient,

illustrated by the upper layer of warmer and less saline water. The vertical variation of water mass

properties, specifically water mass stratification, has been showed to affect larval fish vertical

behaviour. In order to avoid turbulence and clines, both upward (Olla and Davis, 1990; Coombs et al.,

2001) and downward (Heath et al., 1988) displacement of the larvae vertical centre of distribution

have been registered according to species analysed. However, the results here obtained concerning

vertical movements appeared to be only determined by diel light variations with no apparent

avoidance for clines neither preference or avoidance for specific temperature or salinity, also shown

by Gray and Kingsford (2003).

Larval fish vertical distribution here observed points toward a concentration in the deepest layers

(below 40 m deep) during the daylight period, possibly because at such depth the light penetration

was very scarce, thus reducing predation risk by visual predators. During dusk an upward migration

was observed with higher total abundances during the first night hours, suggesting that H.

quadrimaculatus larvae were following a critical light threshold in order to feed upon zooplankton,

since they are visual predators (Milton et al., 1994), without greatly increase the predation risk. The

significant relationship between larvae and zooplankton abundance here observed supports this

upward migration of H. quadrimaculatus larvae associated to feeding activity. The scarce presence of

larvae during daylight samples justifies the absence of correlation to zooplankton abundance at

deepest strata, since larvae probably aggregated close to the bottom (below 40 m deep) while

zooplankton also moved to deeper strata but remained at sampled depths (above 40 m deep).

The relation between feeding activity and vertical movements has shown to be determinant for

larval survival, whereas ontogenic changes throughout larval growth are important to decrease

predation risk (see Fiksen et al., 2007). In this study larger larvae displayed a wider depth distribution

and a more active migration to surface during nigh than smaller ones, probably because larger larvae

are more able to swim and actively chase zooplankton in order to feed. Therefore, swimming,


51

feeding activity and their balance with predation risk possibly determined the ontogenic vertical

distribution here observed. This biological pattern was also observed in other studies (Santos et al.,

2006; Voss et al., 2007), where larger larvae were almost absent during day and during night

appeared at the upper water column layers.

Beyond vertical behaviour, water mass stratification has also implications on fish larvae horizontal

migrations, hence recruitment success (Fisken et al., 2007). Hare and Govoni (2005) after studying

the vertical distribution of fish larvae from various marine systems (e.g. shelf‐resident, estuarine‐

dependent, etc.) where tidal currents were a dominant factor, concluded that fish larvae depth

distribution is closely connected to larval transport, diel cycles and water mass gradients (e.g.

turbulence, temperature and salinity). Even though vertical distribution of H. quadrimaculatus larvae

observed at Sofala Bank showed no relation to tides, probably because its vertical movements were

mostly determined by diel changes and not to tide variation, the deeper position of larvae during day

is consistent with species that remained in the shelf region and avoid to move offshore (Hare and

Govoni, 2005). Possibly, larvae were trying to maintain this onshore position since its high

productivity and prey availability, remaining near bottom during day and displaying night upward

migrations upon feeding requirements. The knowledge of larvae vertical dynamic is crucial to

understand larval survival and dispersion hence recruitment success and variability of small pelagic

fish species. As regards Sofala Bank, recruitment success and variability of H. quadrimaculatus larvae

is thus closely linked to coastal hydrological processes and to estuarine discharges, which enhance

system productivity hence prey availability.

In conclusion, diel vertical movements of H. quadrimaculatus larvae here observed were triggered

by light variation and probably related to feeding requirements. Ontogenic changes were also an

important factor determining larvae depth regulation. The results obtained contribute to the

knowledge of the vertical distribution movements of gold‐spot herring larvae, which is important

information for future studies on larvae dispersal and survival. Besides the obvious gaps on the

current knowledge of the basic ecology of this species, such as distribution, growth, feeding habits

and reproduction, also the larval movements and their dependence on coastal processes should be

investigated as base mechanisms that modulate survival and recruitment.

Acknowledgements

This study was a Special Study as part of the research cruise ‘Ecosystem Survey Mozambique’ of

the wider program “Strengthening the knowledge base for and implementing an ecosystem

Chapter 3

52

approach to marine fisheries in developing countries” from the Nansen Program, an EAF Project

funded by the Norwegian government through NORAD, the Institute of Marine Research (Bergen,

Norway) and the United Nations through FAO. We thank IIP (Instituto Nacional de Investigação

Pesqueira, Mozambique) and its current director, Dr. D. Gove, to P. Afonso and E. André for all

support, and to J. Mwaluma (KMFRI, Kenya) for his help identifying fish larvae. We are grateful for

the productive assistance provided by the crew of R/V “Dr. Fridtjof Nansen” and the helpful support

of J‐O. Krakstad and C. Bento, as well as other fellow scientists from IIP onboard.

References

Blaxter, J.H.S., 1973. Monitoring the vertical movements and light responses of herring and plaice larvae. Journal Marine Biological Assessment UK 53, 635—647.

Coombs, S.H., Morgans, D., Halliday, N.C., 2001. Seasonal and ontogenetic changes in the vertical distribution of eggs and larvae of mackerel (Scomber scombrus L.) and horse mackerel (Trachurus trachurus L.). Fisheries Research 50, 27—40.

Fisken, Ø., Jørgensen, C., Kristiansen, T., Vikebø, F., Huse, G., 2007. Linking behavioural ecology and oceanography: larval behaviour determines growth, mortality and dispersal. Marine Ecology Progress Series 347, 195—205.

Gray, C.A., Kingsford, M.J., 2003. Variability in thermocline depth and strength, and the relationships with vertical distributions of fish larvae and mesozooplankton in dynamic coastal waters. Marine Ecology Progress Series 247, 211—224.

Hare, J.A., Govoni, J.J., 2005. Comparison of average larval fish vertical distributions among species exhibiting different transport pathways on the southeast United States continental shelf. Fisheries Bulletin 103, 728—736.

Harris, S.A., Cyrus, D.P., 1999. Composition, abundance and seasonality of fish larvae in the mouth of Durban harbour, Kwazulu‐Natal, South Africa. South African Journal Marine Science 21, 19—39.

Heath, M.R., Henderson, E.W., Baird, D.L., 1988. Vertical distribution of herring larvae in relation to physical mixing and illumination. Marine Ecology Progress Series 47, 211—228.

Heath, M., 1992. Field investigations of the early life stages of marine fish. Advances Marine Biology 28, 1—174.

K.M.F.R.I., 1981. Aquatic Resources of Kenya. Proc. of the Workshop of Kenya Marine and Fisheries Research Institute, July 13‐19.

Leis, J.M., Rennis, D.S., 1983. The Larvae of Indo‐Pacific Coral Reef Fishes: a Guide to Identification. New South Wales University Press, Sydney.

Leis, J.M., Carson‐Ewart, B.M., 2004. The larvae of Indo‐Pacific coastal fishes: An identification guide to marine fish larvae. Brill, The Nestherlands.

Lewis, A.D., 1990. Tropical south Pacific tuna baitfisheries. In: Blaber, S.J.M., Copland, J.W. (Editors) Tuna baitfish in the Indo‐Pacific region. Aust. Coun. Int. agr. Res. Proc. 30, 10—21.


53

Lutjeharms, J.R.E., 2006. The costal oceans of south‐eastern Africa. In: Robinson, A.R., Brink, K.H. (Editors) The Sea. Vol. 14. Harvard University Press, Cambridge.

Milton, D.A., Blaber, S.J.M., Rawlinson, N.J.F., 1994. Diet, prey selection and their energetic relationship to reproduction in the tropical herring Herklotsichthys quadrimaculatus in Kiribati, Central Pacific. Marine Ecology Progress Series 103, 239—250.

Olivar, M.P., Salat, J., Palomera, I., 2001. Comparative study of spatial distribution patterns of the early stages of anchovy and pilchard in the NW Mediterranean Sea. Marine Ecology Progress Series 217, 111—120.

Olla, B.L., Davis, M.W., 1990. Effects of physical factors on the vertical distribution of larval walleye Pollock Theragra chalcogramma under controlled laboratory conditions. Marine Ecology Progress Series 63, 105—112.

Postel, L., Fock, H., Hagen, W., 2000. Biomass and abundance. In: Harris, R.P., Wiebe, P.H., Lenz, J., Skjoldal, H.R., Huntley, M. (Editors) ICES Zooplankton methodology manual. Academic Press, London.

R Development Core Team, 2008. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. www.R‐project.org

Reid, D.G., 2001. SEFOS – shelf edge fisheries and oceanography studies: an overview. Fisheries Research 50, 1—15.

Santos, A.M.P., Ré, P., Dos Santos, A., Péliz, A., 2006. Vertical distribution of the European sardine (Sardina plichardus) larvae and its implications for their survival. Journal Plankton Research 28, 523—532.

Stenevik, E.K., Sundby, S., Cloete, R., 2001. Influence of buoyancy and vertical distribution of sardine Sardinops sagax eggs and larvae on their transport in the Northern Benguela ecosystem. South African Journal Marine Science 23, 85—97.

Thorrold, S.R., Williams, D.M., 1989. Analysis of Otolith Microstructure to Determine Growth Histories in Larval Cohorts of a Tropical Herring (Herklotsichthys castelanaui). Canadian Journal Fisheries Aquatic Science 46, 1615—1624.

Voss, R., Schmidt, J.O., Schnack, D., 2007. Vertical distribution of Baltic sprat larvae: changes in patterns of diel migration?. ICES Journal Marine Science 64.

Worthington, E.B., 1931. Vertical movements of freshwater macroplankton. Internationale Revue Gesamten Hydrobiologie 25, 394—436.

Yamashita, Y., Kitagawa, D., Aoyama, T., 1985. Diel Vertical Migration and Feeding Rhythm of the Larvae of the Japanese Sand‐eel Ammodytes personatus. Bulletin Japonese Society Scientific Fisheries 51, 1—5.

Chapter 4

Final Considerations

57

Final considerations

Planktonic communities are of great importance to coastal ecosystems functioning and

generation of marine resources, and constitute the coastal and oceanic lower trophic levels.

Furthermore, primary and secondary production processes occurring at coastal regions are highly

variable and strongly determined by hydrological processes. Such variability occurs at different

spatial and time scales, and depending on the analysed frame different hydrological and biological

patterns will be observed.

In the present study both spatial and temporal patterns were observed in a coastal tropical

ecosystem. Respectively, the onshore‐offshore gradients and the time evolution of the physico‐

chemical and planktonic vertical dynamics were investigated. Since this study was performed at a

large continental shelf (Sofala Bank) in front of a river mouth (Zambezi river), tides and estuarine

discharges were important factors modulating hydrological and biological variations, both at spatial

and temporal frames. Despite the sampling period occurred during the end of the dry season, the

estuarine discharges influence was observed at the physico‐chemical water properties, specifically in

the salinity measurements. Nutrients availability (specifically nitrates + nitrites), which is usually

associated to estuarine discharges, was concluded to be a major factor limiting phytoplankton

abundance and community structure. The nutrient ratios obtained (N:P and N:Si) were extremely

low, thus supporting the hypothesis of strong nitrogen deficiency. The higher nitrogen concentration

near the bottom, ascribable to sediment resuspension, influenced phytoplankton vertical

distribution, which was mainly distributed between 30 and 40 m deep. Since most of zooplankton

organisms are herbivorous, higher zooplankton densities were also found at the deeper sampled

layers and no major shallow densities during night were observed. Nevertheless, diel differences on

zooplankton abundances in different strata were still observed, according to typical diel vertical

migration patterns (shallower and deeper distribution during night and day, respectively).

Phytoplankton community structure was dominated by microflagellates, since the most abundant

pigment was 19’‐hexanoyloxyfucoxathin, a marker pigment for prymnesiophyceans. Other relatively

abundant pigments were fucoxanthin and 19’‐butanoyloxyfucoxanthin, both existent in

prymnesiophycean and also in other algae groups. The phytoplankton community dominance by

microflagellates is typical from oligotrophic regions, whereas eutrophic nutrient rich region are

usually dominated by microplankton, specifically diatoms and dinoflagellates. This approach was

Chapter 4

58

given by the calculated Fp pigment index, which was very low and similar to other oligotrophic

marine regions. This trophic status evaluation, from phytoplankton pigments composition, thus

confirmed the nutrient approach.

As regards to Herklotsichthys quadrimaculatus larvae vertical dynamics, results obtained were in

accordance to previous clupeid larvae studies. Larvae displayed diel vertical migrations, occurring

near the surface during night samples and migrating towards the bottom during daylight. The

association of this pattern to light penetration and zooplankton total abundance evidenced an

upward migration in order to feed upon zooplankton, and downward migration to reduce mortality,

ascribed to avoidance of predation from visual predators. However, changes on vertical distribution

movements were observed when considering larval ontogeny. Results point towards a dominance of

the neuston layer by larger larvae and a vertical displacement for smaller larvae with increasing

depths, mainly during nigh samples. During daylight, almost no larvae were collected at sampled

strata (between 0 and 40 m), which suggested that larvae were aggregated near the bottom, where

light penetration was very scarce, thereby greatly reducing predation risk and increasing survival

rates.

The work here developed was important to improve the scientific knowledge regarding the Sofala

Bank, concerning hydrological and planktonic communities variations. Moreover, it was crucial to

understand the strong relationship between the base hydrological processes and the primary and

secondary production, and also that ecological studies, specifically the ones aimed to coastal

ecosystems and plankton, should have a community and ecosystem based approach in order to

broadly understand their functioning and variation patterns. Only with such base knowledge, it will

be possible correctly develop management tools to sustainably explore the marine resources,

promoting the ecosystem integrity and biodiversity.

universidade de lisboa faculdade de ciÊncias …

Documents