University of Groningen the Biology and Impacts of Oreochromis

Home , Lake Tanganyika sardine, Mormyrus longirostris, Oreochromis macrochir, Oreochromis mortimeri, Oreochromis variabilis, Pharyngochromis

University of Groningen

The biology and impacts of Oreochromis niloticus and Limnothrissa miodon introduced in Lake Kariba Chifamba, Chiyedza Portia

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below. Document Version Publisher's PDF, also known as Version of record

Publication date: 2019

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA): Chifamba, C. P. (2019). The biology and impacts of Oreochromis niloticus and Limnothrissa miodon introduced in Lake Kariba. Rijksuniversiteit Groningen.

Copyright Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

The publication may also be distributed here under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license. More information can be found on the University of Groningen website: https://www.rug.nl/library/open-access/self-archiving-pure/taverne- amendment.

Take-down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.

Download date: 04-10-2021

The biology and impacts of Oreochromis niloticus and Limnothrissa miodon introduced in Lake Kariba

Portia C. Chifamba

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 1

The research presented in this thesis was conducted at the University Lake Kariba Research Station and the Department of Biological Sciences of the University of Zimbabwe and at the Lake Kariba Fisheries Research Institute, Zimbabwe, according to the requirements of the Graduate School of Science, Faculty of Science and Engineering, Institute of Evolutionary Life Sciences (GELIFES), University of Groningen.

This research was funded by Nuffic grant (grant number NFP-PhD.11/ 858) awarded to Portia Chifamba, International Foundation for Science (IFS) Grant awarded to Portia Chifamba (grant number A/3159-1, Tonolli Memorial Fund Fellowship of the International Society of Limnology (SIL), and University of Zimbabwe Research Grant. The printing was supported by the University of Groningen (RUG).

The preferred citation for this thesis is: Chifamba PC (2017) The biology and impacts of Oreochromis niloticus and Limnothrissa miodon introduced in Lake Kariba. PhD thesis, University of Groningen, Groningen, The Netherlands.

Cover design: Portia C. Chifamba & Jan H. Wanink Lay‐out: Jan H. Wanink & Portia C. Chifamba Figures: Portia C. Chifamba Pictures including cover: Portia C. Chifamba, Jan H. Wanink, social media

Printed by: Ipskamp Printing, Enschede, The Netherlands

ISBN: 978‐94‐034‐1472‐0 ISBN: 978‐94‐034‐1471‐3 (electronic version)

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 2

The biology and impacts of Oreochromis niloticus and Limnothrissa miodon introduced in Lake Kariba

Proefschrift

ter verkrijging van de graad van doctor aan de Rijksuniversiteit Groningen op gezag van de rector magnificus prof. dr. E. Sterken en volgens besluit van het College voor Promoties.

De openbare verdediging zal plaatsvinden op

maandag 4 maart 2019 om 16.15 uur

door

Chiyedza Portia Chifamba

geboren op 2 augustus 1963 te Guruve, Zimbabwe

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 3 Promotores Prof. dr. H. Olff Prof. dr. B.D.H.K. Eriksson

Beoordelingscommissie Prof. dr. ir. C. Both Prof. dr. E. van Donk Prof. dr. P.F.M. Verdonschot

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 4 This thesis is dedicated to my parents who made it all possible.

Oh, ik ben zo blij

Urombo by Jane Eugenia Chifamba

Urombo Urombo Urombo Urombo

Mwanasikana muka utarire Nyika yaugere igungwa rebvura Nyatotarira rinoda kukunyudza Vakomana venyika vanokunyengedza Vanokufurira unyangadze vakuseke Uri chirombe, chifuza, chinzenza chamakoko

Urombo Urombo Urombo Urombo

Ambuya vangu vakafa, havakasiya mari Nambuya vako vakafa, havakasiya mari Taivakirwe chikoro chevanasikana Vaidzidzira kuchengeta nhaka dzavapwere vavo Nzvimbo dzacho dzotinetsza Hwangova urombo

O urombo. O urombo. O urombo. O urombo

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 5

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 6

Table of Contents

Chapter 1 Introduction 9

PART I BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS Chapter 2 Replacement of the indigenous Oreochromis mortimeri by the 25 invader Oreochromis niloticus in the Southern-African Lake Kariba: in relation to differences in their reproductive potential Chapter 3 Growth rates of alien Oreochromis niloticus and indigenous 41 Oreochromis mortimeri in Lake Kariba, Zimbabwe Chapter 4 Diet overlap between a native and an invasive tilapia species in 57 the Southern-African Lake Kariba Chapter 5 Comparative aggression and dominance of Oreochromis 77 niloticus (Linnaeus, 1758) and Oreochromis mortimeri (Trewavas, 1966) from paired contest in aquaria

PART II BIOLOGY AND IMPACTS OF LIMNOTHRISSA MIODON Chapter 6 Developing a sustainable pelagic fishery in an African reservoir: 93 trends in the catches of the introduced freshwater sardine Limnothrissa miodon and associated species in Lake Kariba, Zimbabwe Chapter 7 Growth of the freshwater sardine, Limnothrissa miodon 117 (Boulenger 1906) estimated from diurnal increments in otoliths

Chapter 8 Synthesis 139

References 165 Authors affiliations and addresses 193 List of publications 197 Summary 201 Samenvatting (Dutch summary) 207 Curriculum Vitae 213 Acknowledgements 217

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 7

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 8

Introduction Chapter 1 Portia C. Chifamba

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 9

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 10 Introduction

World-wide introduction of exotic fish species in waterbodies has had both beneficial 1 and adverse ecological, social and economic outcomes of varying magnitude (Reynolds & Greboval 1988; Welcomme 1988, Balirwa 1992; Witte et al. 1995). Thus, information on impacts of introduced species and how well they have adapted to their new conditions, is essential when considering management options both to protect indigenous species and to enhance fisheries. In addition, information on the ecological impacts can enhance our understanding of ecological processes caused by introductions. Differences in the modes of introduction might have some bearing on the type of impacts arising from the introduction. Fish introductions are either deliberate or planned, in order to fill a vacant niche or accidental, as a result of escapees invading an already occupied niche (De Silva & Sirisena 1987; Welcomme 1988; Balirwa 1992; van Zwieten et al. 2011). The deliberate introduction of a large predator, Nile perch (Lates niloticus), has well established negative ecological impacts, having caused the extinction of many haplochromine species from Lake Victoria. At the same time, Nile perch brings about large economic benefits from harvesting (Reynolds & Greboval 1988; Witte et al. 1995; Twongo 1995). Nile tilapia, Oreochromis niloticus (Linnaeus 1758), was also deliberately introduced into Lake Victoria. Oreochromis niloticus is today one of the three commercially important species caught in the lake but has caused the disappearance of some native tilapias (Goudswaard et al. 2002; Njiru et al. 2005). The introduction of O. mossambicus increased fish catches in reservoirs in Sri Lanka (De Silva & Sirisena 1987), just as the deliberate introduction of Limno- thrissa miodon in Lake Kivu (de Iongh et al. 1995). Lake Kariba provides a typical case study both for planned introductions of fish into supposedly open ecological niches as well as accidental, unplanned introductions into occupied niches. This was possible as the creation of Lake Kariba, a reservoir in the Zambezi River, created a completely new lacustrine ecosystem whose physical attributes such as oxygen concentration, water depth and distance from the lake margin differed profoundly from the lotic system to which the native riverine species were adapted (Jackson et al. 1988). This new, complex matrix not only presented economic opportunities for fisheries development and fish farming but also raised ecological challenges. First, the newly created and therefore vacant pelagic niche was filled by a freshwater sardine, Limnothrissa miodon (Boulenger 1906). Limnothrissa miodon is native to Lake Tanganyika and was introduced to improve fish production (Bell- Cross & Bell-Cross 1971). Physical and environmental conditions in Lake Kariba were unlike those in Lake Tanganyika, and therefore changes in L. miodon’s biological characteristics were anticipated. For example, L. miodon in Lake Kariba was considered stunted (Marshall 1987a) until studies on growth rate using otoliths showed that the fish in Lake Kariba follows the same growth trajectory as in the

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 11 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

native Lake Tanganyika (Chifamba 1992). However, patterns appear to suggest that L. miodon may have inhibited the expansion of a small fish (Brycinus lateralis) native to the Upper Zambezi and affected the habits of the predatory tigerfish Hydro- cynus vittatus (Woodward 1974). Since its introduction, the L. miodon population has been subjected to considerable fishing pressure which may have induced evolution of its life history parameters. Therefore, further research on L. miodon is essential to determine its adaptation to its new environment, its interaction with native fish species and factors affecting the biology of this fish. Secondly, ponds on the Lake Kariba shore provided opportunities for the farming of O. niloticus which resulted in escapees entering the lake and invading the niche already occupied by the congeneric native O. mortimeri (Chifamba 1998, 2006). This negative interaction of the exotic and native species thus presented an opportunity to study factors that confer competitive advantage to the invader by comparing growth, diet and aggression in these two species. Research on the biology of O. niloticus and L. miodon in Lake Kariba would determine adaptations of these introduced species to their new environment. Both introduced fish species are important in the Lake Kariba fisheries, L. miodon as the main catch of the pelagic fishery and O. niloticus as one of the important species in the artisanal fishery. Hence, the objective of introducing L. miodon into Lake Kariba to improve fish production was realised. However, catches have declined since the beginning of the fishery, resulting in economic losses and a need to improve the management of the fishery. Informed management of the fish resource is needed to ensure sustainable fisheries. Such knowledge is currently needed for explaining changes in the productivity of the sardine industry that is thought to have crashed as a result of overexploitation or environmental changes.

The aim of this thesis is to establish the degree of suitability of the ecosystem created by the Lake Kariba dam to the introduced species Oreochromis niloticus and Limnothrissa miodon, to identify factors that may have caused O. niloticus to displace the native species Oreochromis mortimeri, and to investigate potential factors causing the decline of the L. miodon fishery.

The environment Lake Kariba physical characteristics Lake Kariba was formed on the middle Zambezi River at 485 m altitude in 1958 and was then the world’s largest reservoir (5820 km2 at maximum storage) (Coche 1974). The reservoir became full in 1963 and the weight of water was so large that it increased

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 12 Introduction 1

Figure 1.1 Map of Lake Kariba showing the Lake basins.

the seismic activity in the area. In 1963 alone, five tremors above 5.0 on the Richter scale were experienced (Tumbare & Sakala 2000). Lake Kariba has a length of 277 km, width of 32 km at its widest point, mean depth of 29 m, and a maximum depth of 120 m. The Zambezi River contributes most (77%) to the water volume in the lake, whilst other rivers contribute 16% and rainfall 7% (Balon & Coche 1974). The dam was constructed primarily for hydro-electricity generation. Hence, the bulk of the water is lost through hydro-electricity turbines and 14% by evaporation. Retention time of the water is about 3 - 4 years and the lake level experiences an annual change of 1 - 5 m, resulting from inflowing floods and drawdowns through turbines and spillage through the sluice gates (Karenge & Kolding 1995). The lake is separated into five basins, marked by chains of islands and narrows (Figure 1.1). The uppermost two lake basins, Mlibizi and Binga, are riverine due to the influence of the Zambezi River. They are flushed out in May by the Zambezi River floods and thereby assume turnover characteristics earlier than the other three basins, which are truly lacustrine and have temperature-induced turnover. This river- lake environment gradient has a profound effect on the fish species composition (Begg 1974).

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 13 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

Rainfall and temperature The Kariba area has one rainy season, with most rain falling in December and January. Annual rainfall is between 400 and 800 mm depending on location (Begg 1970). The amount of rainfall affects lake level and drawdown. Lake Kariba is monomictic and has a thermal stratification from September to early June. Mixing or turnover usually takes place in July each year. Stratification begins around September, with a thermocline at around 15 m depth, which gradually moves down to around 35 m at the time of turnover. Stratification prevents deep water in the hypolimnion from mixing with the epilimnion (Begg 1970). Lake Kariba is a warm lake with a surface water temperature of 28 to 30 0C and a hypolimnion of about 22 0C when stratified. When mixed, the whole water column is 22 0C (Marshall 2012a). The mean maximum air temperature in Kariba has steadily risen from about 33.1 0C in 1968, to 35.5 0C in 1998 (Magadza 2010). As a result of the temperature increase, the depth of the thermocline decreased from 10 to 15 m in 1986 – 1987, to a consistent 5-m depth in 2007 – 2008 (Ndebele-Murisa et al. 2014). A shallow thermocline likely means a reduction in the optimal habitat for phytoplankton and fish. Cochrane (1978) found a correlation between the catches of L. miodon and water volume above the thermocline, therefore the recent change of thermocline depth may have affected L. miodon abundance and catches negatively. Temperature and hydrological factors (rainfall, riverflow and lake level) are correlated with L. miodon catches through nutrients brought in by the rivers and the effect of temperature on phytoplankton and zooplankton on production (Chifamba 2000).

Phyto- and zooplankton productivity During the stratification period, the hypolimnion becomes depleted of oxygen and the epilimnion of nutrients due to photosynthesis. Turnover increases the availability of nutrients in the epilimnion and the euphotic zone, increasing phytoplankton production (Ramberg, 1987; Masundire, 1989). Hence, turnover, rainfall and river mouths are associated with increased plankton production (Magadza, 1980; Ramberg, 1987; Masundire, 1992, 1994; Cronberg, 1997). Temperature therefore mediates nutrient cycling in the lake and is an important driver of productivity in the lake. Temperature also drives seasonal and annual variation in phytoplankton and zooplankton communities. Cyanobacteria (blue-green algae) dominated from December to May and diatoms from June to September, whilst 60% of the annual biomass consisted of cyanobacteria in 1982 – 1983 (Ramberg 1987). Chlorophyceae dominated during periods of relatively low temperature compared to Cyanophyceae. Comparable results were found in a laboratory study by Sibanda (2003). The growth rates (% increase in number of cells per day) of Chlorophyceae in the laboratory declined at water temperatures above 25 0C, becoming negative above 28 0C. At the same time, the growth rate of Cyanophyceae increased almost exponentially up to 34 0C, resulting in a transition temperature from Chlorophycea to Cyanophy-

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 14 Introduction

ceae domination of about 28 0C (Magadza 2011). In Lake Kariba, that has warmed by a mean of 1.54 0C between 1965 and 1990, the mean epilimnion summer temperature reached this transition limit in 1987 (Magadza 2011). 1 The rise in water temperature in the lake has indeed caused a change in the phytoplankton towards a community dominated by cyanobacteria (Magadza 2011). Due to toxicity and morphology, cyanobacteria are poor food for zooplankton relative to small chlorophytes and flagellates (Wilson et al. 2006). This is reflected in the shift in the zooplankton from large- (Calanoida; Daphnidae) to small-bodied (Bosmina; Cyclopoida) species, and a strong reduction in abundance, that accompanied the transition to cyanobacteria in the lake. Being a zooplankton feeder, L. miodon was expected to be negatively affected by these changes. Though a causal relation could not be established, it is remarkable that a decline in the catches of L. miodon started shortly after the mean epilimnion summer temperature exceeded 28 0C (Magadza 2011). This thesis evaluates changes in individual growth rate of L. miodon as a possible mechanism by which temperature can affect L. miodon production. The shift in the phytoplankton composition is also of interest in relation to competition in the inshore planktivorous fish, the endemic Oreochromis mortimeri and introduced O. niloticus. The question is whether such a shift in the plankton community would then favour O. niloticus, which is known to feed and digest cyanobacteria (Moriarty 1973). Thus, to understand the recent changes in the fish populations in Lake Kariba, it is important to study the diet composition of both introduced (e.g. O. niloticus) and native species (e.g. the potentially competing indigenous planktivorous species, O. mortimeri).

The Fish Fish species composition changes At the future Kariba Dam site in the Zambezi River, which Jackson (1960) described as a sandbank river with little vegetation, Hydrocynus vittatus, Distichodus sp., Barbus sp. and Labeo sp. were dominant, while cichlids such as O. mortimeri and some small fish species were rare. The fishes were subjected to a seasonal period of flooding, when food and shelter were plenty, and a dry season when flow was low and remaining pools in the river small, thus providing little food and shelter (Jackson 1960). It could be anticipated that the transformation from a riverine to a lacustrine ecosystem would create new conditions that would alter fish distribution patterns. Initially, nutrients were high after impoundment from the decomposing submerged vegetation and leaching from the soil. The new lake was characterised by a high productivity of algae, the invasive water fern, Salvinia molesta and fish. Nutrients decreased with time as the lake matured. Macrophyte beds developed in the inshore area, increasing habitat diversity, shelter and food, supporting an increase in fish species such as Tilapia sparrmanii.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 15 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

Surveys done before the creation of Lake Kariba found between 28 and 31 species of fish in the Middle Zambezi River (Jackson 1961; Harding 1964; Bell-Cross1965). In the late 1990s, when the reservoir was filled, 45 species were reported by Marshall (2006) and 50 by Songore & Kolding (2003). Species such as Brycinus lateralis, Barbus poechii, Labeo lunatus, Oreochromis andersonii, Sargochromis giardi, Sargochromis carlottae, Serranochromis macrocephalus and Serranochromis robustus which before the creation of the dam were restricted to the Upper Zambezi River, were captured in the lake (Balon 1974). However, of these, only B. lateralis and S. macrocephalus were caught in a lake-wide survey in 2006 (Zengeya & Marshall 2008). Fish species in the new water body revealed preference for diverse conditions. Many riverine species, especially tilapias, prefer still-water pools and marshes in a river (Jackson 1966). These fishes found the lake’s stable and stagnant environment favourable and thrived particularly in the most lacustrine Sengwa, Bumi and Sanyati basins (Figure 1.1) (Begg 1974). In these basins, the cichlid fishes, mainly Oreo- chromis mortimeri, Sargochromis condringtoni and Tilapia rendalli, made up between 64.1 and 96.2% of the catch in 1968 to 1970. In an unpublished report from 1959, only 0.75% of the fish caught were O. mortimeri, while in 1962, as the lake filled, the contribution had increased to 35% of the catch (Kenmuir 1984; Jackson et al. 1988). For these species, the lake environment mimicked the period of plenty in a flooded river when food and shelter were plenty resulting in high survival and reproduction, and consequently high catches. The genera that prefer flowing water and were abundant in the river before impoundment, Hydrocynus, Distichodus, Barbus and Labeo, became dominant only in the more riverine uppermost two lake basins and in the estuaries of inflowing streams (Jackson 1960; Begg 1974). The rheophilic species, Chiloglanis neumanii, Opsaridium zambezense (and possibly also Leptoglanis rotundiceps) are now confined to the tributaries of the two uppermost basins (Balon 1974). All the former river fishes are restricted to the inshore shallow (< 15m depth) area, except Clarias gariepinus, Mormyrus longirostris and Syndontis zambezensi that can live in water down to 30 m depth when the water is well oxygenated (Jackson 1960; Coke 1968; Sanyanga et al. 1995). The inshore fish species are the basis of a gill-net artisanal fishery that started in 1958 and 1962 in Zambia and Zimbabwe, respectively. Between 1964 and 1972, three species dominated the inshore catches: the predator H. vittatus and two cichlids, Oreochromis mortimeri and Serranochromis condringtoni (Karenge & Kolding 1995). The Shannon diversity index for the Lakeside fish increased between 1972 and 1990 as a result of natural introduction from the upper Zambezi and tributaries, as well as fish introductions (Karenge & Kolding 1995). To monitor the inshore fish population, the Lake Kariba Fisheries Research Institute (LKFRI) in Zimbabwe established in the 1960s a Lakeside Experimental Sampling Programme where gill-

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 16 Introduction

net nets are set routinely at a site named Lakeside, situated on the shores of Kariba town. Data from this programme were used in my research. 1 Introduction of fish species Although at least five species are known to have been introduced into Lake Kariba, only the clupeid L. miodon and the tilapiine cichlid O. niloticus have been successfully established and are now widespread throughout the lake. Oreochromis macrochir is found only sporadically; on average about 10 specimens per year were caught out of 5 – 10 000 total number sampled in Zimbabwe at Lakeside between 1974 and 2001 and only three specimens in 1985, 1992 and 1996 in the Zambian Experimental Gillnet Survey (Kolding et al. 2003; Marshall 2006). The introduced Micropterus salmoides was caught only once and has probably not established a viable population. Two of the introduced species, Tilapia rendalli and Serranochromis robustus, may have invaded the lake naturally and the former might have been in the system pre-impoundment (Kolding et al. 2003). Limnothrissa miodon (sardine), a zooplanktivorous pelagic freshwater clupeid, was deliberately introduced into Lake Kariba from Lake Tanganyika from 1967 to 1968 in order to fill a vacant pelagic niche and increase fish production (Bell-Cross & Bell-Cross 1971). About 30% of the lake is shallower than 17 m and only the shallow area less than 20 m is used by most of the indigenous Zambezi River fishes, because they are not adapted to a pelagic environment (Begg 1970; Coke 1968). By 1969, there was evidence that the sardine had become established (Kenmuir 1971). The introduction is considered a success because L. miodon has the largest single fish stock in Lake Kariba. Annual commercial catches landed reached a maximum of about 31 000 tonnes in 1990, and a minimum of 15 000 t in 2003 (Kinadjian 2012). From the beginning of the fishery, the Lake Kariba Fisheries Research Institute in Zimbabwe and the Department of Fisheries in Zambia collected data on the catches and fishing effort. Data collected by the two institutions and from other research, show that catches of sardines vary seasonally and annually (Marshall 1988b, 2012a). Each year catches usually reflect two peaks that differ in magnitude. A small peak occurs during April – May and a larger peak during August – September. The exotic Nile tilapia (Oreochromis niloticus) was first caught in gillnets set routinely at Lakeside in 1993 (Chifamba 1998). Up to August 1994, O. niloticus were only caught close to the fish farms where they may have escaped. Even then, the abundance of O. niloticus was low, constituting a mere 0.4% of the catch by mass and 0.17 % by numbers. Judging from the range of fish size caught (3 – 30 cm and 1 – 1 069 g) and the presence of both sexes in the sample, these fish were by then already established in the lake. The farming of O. niloticus on the shores of Lake Kariba is responsible for the introduction of this species into the lake. Oreochromis niloticus was selected for farming in Kariba because it is widely used in aquaculture all over the world. This

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 17 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

is due to its superior growth rate and adaptation to aquaculture condition, compared to other cichlids (Philippart & Ruwet 1982; Blow & Leonard 2007). Prior to this fish introduction, there was no assessment of its suitability and of the probable impacts.

Impact of fish introductions Oreochromis niloticus Since the introduction of O niloticus resulted from escapees from fish farms, it was not carefully monitored (Chifamba 1998). However, with the introduction of O. niloticus, the once abundant O. mortimeri, an endemic to the Zambezi system, declined and disappeared in many parts of Lake Kariba (Chifamba 2006; Zengeya & Marshall 2008). Several factors may have contributed to this species displacement, and a key aim of this thesis is to evaluate the major potential causes. Competition is a potential driver of local species displacement (MacArthur & Levins 1967). Because O. mortimeri and O. niloticus show similarities in their diets and reproduction strategies (Chifamba 1998; Mhlanga 2000; Marshall 2011), competition between these species can be expected. Both species feed on algae and organic detritus, plant material, insects and zooplankton, varying with availability (Lowe-McConnell 1958; Moriarty 1973). In both species, the male constructs a large nest in an arena, which it defends (Jubb 1974; Marshall 2011. Therefore, competition for food, nesting and nursery space may have stimulated antagonistic behaviour. Aggression is one mechanism that O. niloticus may have used to displace native O. mortimeri in Lake Kariba. Life history trade-offs are reported to be strong determinants of competitive abilities, under both stable and changing ecological conditions (Lancaster et al. 2017). A higher growth rate or an ultimate large size of O. niloticus could be another mechanism to displace O. mortimeri. Many studies have shown that large fish tend to have a larger number of eggs (Schemske 1974, Baglin & Hill 1977, Schenck & Whiteside 1977, Bagenal 1978; Wanink & Witte 2000; Barneche et al. 2018). High growth rates contribute to fitness when large size has benefits such as higher fecundity and reduced mortality. Fast growth would therefore result in a larger number of eggs at an earlier fish age. In addition, fast growth may also reduce the time an animal spends at a vulnerable size because smaller animals tend to be more vulnerable to predation (Sutherland 1996). Hence, a comparison of growth rate and maximum size of the introduced O. niloticus and the native O. mortimeri will give an indication of potential relative competitiveness of these species.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 18 Introduction

In this thesis I evaluate the potential for competition between Oreochromis mortimeri and Oreochromis niloticus, by comparing their reproductive effort, 1 aggression levels, growth rates and diet. To compare their diet, I analysed the stomach contents of the two species and estimated the Schoener similarity index (Schoener 1970) and the Pianka overlap coefficient (Pianka 1973) to assess the level of potential competition between the two species. Fish of known size and weight were aged by counting annual increments on scales, opercula and otoliths, and the results used to estimate the growth parameters. The mean length of each age group in the fish sample was estimated and, together with the growth parameters, used to determine which species grows faster and thus can confer size advantage in a contest. In order to assess the relative aggression of the two species, pairwise contests were setup in an aquarium. The number of aggressive acts such as ‘biting’ and ‘ramming’ were used to score aggression. Aggression indicates which of the two species is likely to be outcompeted in the event of a contest arising during competition for a resource such as food and a breeding site. Reproductive effort was estimated using the monthly proportion of fish in the samples of both species, that were ready to release gametes (eggs or milt). Having a higher reproductive effort may help a species to outcompete the other by increasing its own population rapidly. The monthly proportions of breeding fish were correlated to rainfall and temperature, both known to trigger reproduction (Clark et al. 2005; Taranger et al. 2010; Quintana et al. 2004), to find out if the influence of those abiotic on the two species is different.

Limnothrissa miodon Limnothrissa miodon was deliberately introduced into Lake Kariba to utilize the plankton production in the newly formed pelagic area. This may have prevented population expansion of the native zooplanktivore Brycinus lateralis (Woodward 1974). Early catches of L. miodon from the open water contained 20.5% B. lateralis, showing the capability of the latter species to expand from the Upper Zambezi River and fill the vacant pelagic niche in the lake. Therefore, the expansion of B. lateralis into the pelagic area may have been prohibited by competition from L. miodon (Marshall 1991). The introduction of L. miodon also affected the habits of the native tigerfish (Hydrocynus vittatus). This predator soon added L. miodon to its diet and began to inhabit open water in pursuit of its new prey, where it occurred in the developing sardine fishery (Cochrane 1976; Marshall 1987b, 1991). From April 1969 to March 1970, only 1.5% of the stomach content of H. vittatus consisted of L. miodon, whereas from April 1970 to March 1971 the amount rose to 41.4% and remained high thereafter

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 19 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

(Kenmuir 1973; Mhlanga 2003). Hydrocynus vittatus was an important bycatch species from the beginning of the L. miodon fishery. Between 1973 and 1975 the catches of H. vittatus increased with increasing catches of L. miodon (Cochrane 1976). A decline in catch per unit effort (CPUE) of less than one year old H. vittatus between 1974 and 1986 correlated with a decrease in CPUE of L. miodon (Marshall 1987b). Already before its introduction into Lake Kariba it was known that, in Lake Tanganyika, L. miodon is not an obligate pelagic species but that it inhabits the inshore area for a substantial part of its life (Poll 1953; Matthes 1968). In the inshore area L. miodon competes with small cichlids, Brycinus sp. and Barilius sp.

In this thesis, I investigate the causes of the decline in the pelagic catches by analysing the relationships between Limnothrissa miodon catches, fishing effort, and several key environmental variables: air temperature and the hydrological factors, rainfall, river flow, and lake level.

Fishing on Lake Kariba Although power generation is the most important economic function of Lake Kariba, it is also the largest source of fish in the country. Two fisheries evolved, one based on the introduced L. miodon, operating in deep/pelagic water, and an inshore fishery. During 1994, the major economic activities on and around Lake Kariba, combined for Zambia and Zimbabwe, generated revenue of about 124 million USD, of which 54% was from power generation and 37.9% from the fisheries (Tumbare 2000). The pelagic fishery is semi industrial and of greater value than the artisanal inshore fishery, contributing 33.9% against 4.0% of the overall revenue and landing 28 423 and 2 473 metric tonnes of fish, respectively. Dried sardines are an important source of protein, particularly in the rural areas, because of the long shelf life of the dried fish. However, while the sardine catches rose with fishing effort at the beginning of the fishery, they have steadily declined since 1990 (Magadza 2011).

In this thesis, the impact of the sardine fishery on the inshore fish species, through the capture of juveniles, will be explored by analysing the bycatch. I also assess the relationship between Limnothrissa miodon and Hydrocynus vittatus, to elucidate the response of H. vittatus to changes in the population of L. miodon.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 20 Introduction

Objectives of the thesis PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS 1 In the first part of the thesis, I evaluate the contribution of the increase in Oreochromis niloticus to the decline of its native congener Oreochromis mortimeri, a species endemic to the Middle Zambezi River. First, to determine the suitability of the new environment for O. niloticus, I estimate its growth parameters and compare them to those of the same species in other water bodies. This is because environmental suitability can be judged by individual growth rate and the maximum size the fish attains in that environment, as compared to other environments. Such information can be used for the estimation of fishable biomass. Furthermore, I look at possible interactions that could have contributed to the displacement of O. mortimeri, by comparing the growth rates, the degree of diet overlap, and aggression levels in O. niloticus and O. mortimeri. Fast growth may confer competitive advantage to the introduced O. niloticus if this translates to more surviving offspring. A large diet overlap can result in a strong competition for food resources (Hanson & Leggett 1985). Higher aggression levels offer advantage in the form of access to better nesting and brooding sites (Philippart & Ruwet 1982; Seppänen et al. 2009).

PART II – BIOLOGY AND IMPACTS OF LIMNOTHRISSA MIODON In the second part of the thesis, I evaluate the declining catches of the freshwater sardine Limnothrissa miodon, by investigating growth rate and age at first maturity as potential causes of the decline. Other potential causes are overfishing and environmental changes. These are assessed by correlating the catches with fishing effort, air temperature, rainfall, river flow, and lake level. Information on growth rate and age at maturity should reveal whether the fishery catches too small, immature fish. Limnothrissa miodon in Lake Kariba is considered stunted, and the bulk of the catches consists of small fish (< 6 cm total length) compared to those in the fisheries on Lakes Tanganyika and Kivu (Marshall 1987a). In an exploited fish population, environmental and fishing effort simultaneously affect the fished population. I use multiple regression analysis to determine the relative contribution of environmental factors and fishing effort to the declining catches. I also estimate the Maximum Sustainable Yield from two fisheries models. That provides a guideline for the level of fishing effort, to be used together with considerations such as environmental variation and the biology of the target species, in order to achieve a sustainable fishery on Limnothrissa miodon in Lake Kariba.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 21 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

Outline of the chapters

PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS  Chapter 2 describes the displacement of the indigenous tilapia Oreochromis mortimeri by the introduced Oreochromis niloticus, and it explores the possible role of interspecific differences in reproductive potential. Monthly variation in the gonadal activity of the two species is presented in relation to rainfall and temperature.  In Chapter 3, growth rates of O. niloticus and O. mortimeri are determined and compared. Ages are estimated from reading scales, opercula and otoliths. Ages from these body parts as well as ages from three time periods are compared.  Chapter 4 shows the results of stomach content analyses performed on O. niloticus and O. mortimeri. Diet overlap was estimated to inform on the degree of competition between the two species.  Chapter 5 deals with the comparison of aggression levels and dominance of O. niloticus and O. mortimeri, observed in paired contests in an aquarium. The working assumption is that a high aggression level will infer competitive advantage.

PART II – BIOLOGY AND IMPACTS OF LIMNOTHRISSA MIODON  In Chapter 6, catch trends of Limnothrissa miodon are evaluated to determine to what extent they can be explained by fishing effort and temperature and how they relate to the catches of their predator, Hydrocynus vittatus. The occurrence of large sized sardines in the inshore waters is discussed in terms of feeding and breeding. Impact of the sardine fishery on other fish species is explored by analysing incidental catches.  Growth of L. miodon and spatial differences therein are evaluated in Chapter 7. Fish were aged using daily increments in otoliths. Deposition rate of the increments was validated using electron microscopy. Implications of age and size at first maturity for fishery management are discussed.

SYNTHESIS  Chapter 8 gives an integrated discussion of the results from all the previous chapters. Here the achieved answers to the research questions are evaluated.

Oreochromis niloticus (left; from Boulenger 1907) and Limnothrissa miodon (from Poll 1952).

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 22 PART I

BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 23

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 24 Replacement of the indigenous Oreochromis mortimeri by the invader Oreochromis niloticus in Chapter 2 the Southern‐African Lake Kariba: in relation to differences in their reproductive potential

Portia C. Chifamba Han Olff

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 25

Abstract

The Nile tilapia Oreochromis niloticus is one of the worst invaders in African lakes and rivers and can cause a massive decline of abundance and diversity of other cichlid species but is also valued for its economic importance in aqua-

culture. However, the invasion process and extend of environmental impacts are

not well understood. Similarity of ecological niche of Oreochromis niloticus,

introduced into Lake Kariba in the 1990s, placed it in direct completion with an endemic congeneric cichlid, Oreochromis mortimeri. This makes the study of the joint development of their populations (based on catches) both ecological and economically interesting. The resulting patterns in catches found were explained by studying differences in the reproductive potential of the two species. Oreo- chromis niloticus was found to have a lower length at maturity (17.63 ± 0.70 cm) than O. mortimeri (19.19 ± 0.99 cm). In addition, the proportion of mature fish with ripe gonads was always higher for O. niloticus compared to O. mortimeri, indicating a higher reproductive potential of the former. The introduction of O. niloticus in Lake Kariba appears not to have improved the fish catches because

the combined catches of the two species remained constant during the period of

analysis. This displacement of an endemic species and driving it into the critically endangered species category, demonstrates how unplanned introductions can pose a threat to biodiversity.

A female Nile tilapia partly eaten by an African fish eagle. An impression of the reproductive potential can be obtained by relating the size of the tiny ripe eggs (scattered inside and outside the fish) to the size of the buccal cavity.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 26 Reproductive potential and replacement of an indigenous congener

Introduction Nile tilapia (Oreochromis niloticus) is an economically important species for African freshwater aquaculture and fisheries (Brummett & Williams 2000), and for that reason has been introduced (planned or unplanned) in many rivers and lakes throughout the 2 continent outside its native range (Schwanck 1995; Zengeya et al. 2011). However, in some ecosystems, for example in Lake Victoria, this has led to an ecological catastrophe, where the native congenerics were reduced in their abundance (Balirwa 1992; Bbole et al. 2014; Deines et al. 2014). The mechanism of this exclusion is not clear due to the high cichlid diversity of the large East-African lakes, which makes it difficult to study pair-wise species interactions. For this, simpler systems with less species can provide insight. Lake Kariba is a man-made lake in the Gwembe rift valley, formed by the construction of a dam across the Zambezi River at the Kariba Gorge in 1958. The lake is bordered by Zambia on the north and Zimbabwe on the south and at the time of its construction it was the world largest man-made lake (5 820 km2) (Coche 1974). The formation of Lake Kariba converted an existing riverine ecosystem into a lacustrine one and it affected the fish species composition and abundance. Oreochromis mortimeri (Trewavas 1983), one of the native riverine fish species in Lake Kariba, is a cichlid endemic to the middle Zambezi river system from the Victoria Falls to Cahora Bassa Gorge (Marshall 2011). Before the Kariba dam was built, catches of O. mortimeri in the river were small, compared to those of other species. Following impoundment, catches of O. mortimeri in the Lake Kariba reservoir increased rapidly, making up 38% of the inshore catches in the 1970s (Kenmuir 1984). Although the catches decreased during the 1980s, O. mortimeri was still important to the fishery (Karenge & Kolding 1995). Oreochromis niloticus (Linnaeus, 1758) invaded Lake Kariba from fish farms along the shore, first appearing in fisheries catches in 1989, and subsequently being caught throughout the whole lake (Chifamba 1998; Zengeya & Marshall 2008). Its introduction was followed by a reduction in the catches of O. mortimeri (Chifamba 2006). An analysis of the catches taken from the Sanyati basin of Lake Kariba showed a spatial and temporal gradient in the proportion of O. niloticus to O. mortimeri (Chifamba 2006). The natural distribution of O. niloticus includes the Nile basin, Rift Valley lakes and some West African rivers (Skelton 2001). In locations where it is introduced, Oreochromis niloticus tends to displace other fish (Balirwa 1992; Marshall 1999). It displaced Oreochromis esculentus (Graham) and Oreochromis variabilis (Boulenger) in Lake Victoria and native cyprinids in Lake Luhondo (Balirwa 1992). In Lake Chivero, Zimbabwe, it replaced the previously introduced tilapia Oreochromis macrochir and later accounted for about 95 % of the catches from that lake (Marshall

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 27 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

1999; Tiki 2011). Oreochromis niloticus may adversely affect resident fish species through competition for food, hybridisation and competition for space (Zengeya & Marshall 2007; Weyl 2008; Zengeya et al. 2011). To explain the competitive success of O. niloticus over O. mortemeri, it is important to assess the reproductive potential of different species. Techniques for assessing this, include the determination of the onset of maturity, the duration of the breeding season, the fecundity and the fraction of spawners (Murua et al. 2003). In both O. niloticus and O. mortimeri, breeding takes place throughout the whole year and their breeding peaks are associated with the rains (Lowe-McConnell 1982; Donnelly 1969). In this study, we describe the trends in the population of O. niloticus after it was introduced and the resulting change in the population of the indigenous O. mortimeri, using catch data. We try to explain the differences by comparing the reproductive potential of the two cichlids using the length at maturity and the proportion of mature fish throughout the year. The impact of the introduction on the catches was assessed by analysing the trend in the combined catches of O. niloticus and O. mortimeri and the size of fish in the catches.

Materials and Method Site description Lake Kariba is separated into five basins, marked by chains of islands and narrows. The data used in this study was collected in the Sanyati basin, closest to the dam. The Kariba area has one rainy season (November to April) and receives between 400 and 800 mm of rainfall annually (Begg 1970). The Zambezi River floods reach Lake Kariba in March – May, causing the lake level to rise until June or July, followed by a decline until November. Lake Kariba is a warm lake with a surface water temperature of between 28 and 30 0C, and a hypolimnion temperature of about 22 0C when stratified. When mixed (turnover between June and July), the value for the whole water column is 22 0C (Marshall 2012b). Maximum temperatures are recorded in October and November. They decrease gradually during the rains to a minimum in June and July (Balon & Coche 1974).

Data collection Fish catch data was obtained from a routine gill net sampling programme at Lakeside in the Sanyati basin of the lake, conducted by the Lake Kariba Fisheries Research Institute. Since the 1960s, at least one fleet of gill nets was set overnight, at Lakeside, every fortnight, resulting in a long time series for this site, interrupted only in times of socio-economic crisis. A fleet of gill nets consists of twelve joined 45 m long panels, of stretched mesh sizes 38, 51, 64, 76, 89, 102, 114, 127, 140, 152, 165 and 178 mm. Each fish caught was identified, its standard length (SL; to the nearest mm), mass (to the nearest gram) measured and the sex and stage of gonad development

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 28 Reproductive potential and replacement of an indigenous congener

determined. Both the standard and total length (TL) were measured in 2005, so that the results obtained in this study could be compared with those from studies that only used total length. Artisanal fishery data was obtained from Nyamhunga, Dandawa, Mudzimu and Nematombo fishing villages from December 2003 to October 2004. Only species name, weight and length of the fish were recorded for the artisanal catch. 2

Trends in the catches Trends in the catch per set of O. niloticus and O. mortimeri were compared to assess the impact of the introduction of the former on the latter species. A Loess smoother, set at a sampling proportion of 0.2 and at weights from a Gaussian density function, was used for smoothing the trends plotted using Sigma Plot 12 software.

Sex and length at first maturity Each specimen was dissected to determine sex and the stage of development of the gonads. Seven stages of gonad development (Ricker 1968; Witte & van Densen 1995; Table 1) were used. Classification of the ovary developmental stages was based on the magnitude of gonad distension, and on the size, yolk content and colour of the ova. For testis, the magnitude of gonad distension and colour were used. Fish too spoilt to be sexed were excluded from the analysis of maturity and gonadal stage. All fish with gonads at the inactive stage were considered immature and all the other stages beyond this as mature. Error in classification may occur when fish recovering from spawning are indistinguishable from earlier stages (Witte & van Densen 1995). If these fish are included, the 50% maturity level can be overestimated and the percentage of mature fish cannot reach 100%, because some females at resting stage are considered immature. This source of error can result in erroneous classification and should be considered when estimating and interpreting length at maturity. Length at maturity is the length at which 50% of the individuals in a length class are mature. It was determined by fitting a sigmoid function to the percentage of male and female mature fish in 1-cm length classes, using a Sigma Plot 12 sigmoid equation with the asymptote set at 100% (mature fish in the population). Percentage sexual maturity was described by a sigmoid function:

PL = 100 / (1+ exp (- (L - L50) / b))

where PL is the percentage of mature fish at length L, L50 = length at sexual maturity, and b determines the steepness of the curve.

Seasonality of breeding The breeding seasonality was deduced from changes in the proportion of mature fish that were in the ripe, ripe running and spent stages of gonad development (Table 2.1).

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 29 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

Table 2.1 The stages of gonad development used to identify mature and the ripe Oreochromis niloticus and O. mortimeri.

Gonad stage Code Condition

Immature I Young fish which have not yet reproduced, characterized by very small gonads. Inactive/active IA Gonads bigger than in inactive fish, about half their ripe size; translucent eggs visible with magnification. Active A Testis and ova opaque, reddish with blood capillaries; eggs visible to the naked eye as whitish and granular. Active/ripe AR Gonads 2/3 of their ripe size; ovaries white reddish; individual eggs easy to see without magnification. Ripe R Sexual organ fills the cavity; eggs completely formed; milt is white, but not extruded under light pressure. Ripe/running RR Sexual products are extruded when light pressure is applied. Spent S Sexual products extruded, aperture inflamed; gonads like deflated sacs; residual eggs/sperm may be present.

Fish with gonads in this stage of development were considered breeding. In order to investigate the environmental triggers of breeding, the monthly proportions of breeding fish were correlated with monthly rainfall and air temperature in the Kariba area using Spearman’s rank correlation in SPSS 17.0.

Results A total of 1 169 O. niloticus and 1 822 O. mortimeri were used in the analysis. The minimum and maximum length of O. niloticus was 7.5 and 62 cm, and that of O. mortimeri 7.8 and 51.8 cm, respectively.

Trends in the catches Ever since O. niloticus appeared in the Lakeside experimental gill net fishing programme catches in 1993, the catches of this species increased, whilst those of O. mortimeri declined. The latter species disappeared in 2006, except for an occasional catch (Figure 2.1a). About 50% of the variation in the catches of O. mortimeri could be explained by the variation in O. niloticus (Figure 2.1b). The catches of O. niloticus rose to a peak in 2000, remaining stable thereafter with a mean catch of 2.8 kg per set. Combined catches of these two species show the replacement of the O. mortimeri by O. niloticus with the trend in catches strongly fluctuating between years but without a clear trend in the total catch during the period of displacement from 1992 to 2012 (Figure 2.2).

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 30 Reproductive potential and replacement of an indigenous congener

40 a) O. niloticus O. mortimeri O. niloticus O. mortimeri 30

2 20

Catch per set (Kg) 10

0 1960 1970 1980 1990 2000 2010 2020 Year 6 b) catches exponential model

4 (Kg) Figure 2.1 a) Trends in the catch per set of nets and (with the trend shown as a Loess

O. mortimeriO. 2 smoother) b) regression of Oreochromis niloticus and O. mortimeri catches from the Lakeside experimental gill‐

0 netting site from1968 to 2012. 0246810 O. niloticus (Kg)

Figure 2.2 Combined catches 10 of Oreochromis niloticus and O. mortimeri

(Kg/set) O. niloticus O. mortimeri per set of gill 8 nets per night from 1992 to 2012.

mortimeri O. 6 + 4

O. niloticus 2

Total catch of 0 1995 2000 2005 2010

Year 31

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 31 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

Gillnet size selection The mesh size of the net that resulted in the largest catch of fish was higher for O. niloticus than for O. mortimeri (Figure 2.3). Peak catch for O. niloticus over the whole fishing period was reached with the 152-mm mesh at a total number of 194 fish, with a mass of 181.8 kg (Figure 2.3a). The mean length of O. niloticus caught in the 152-mm mesh was 29.3 cm. In comparison, O. mortimeri catches peaked in the 114-mm mesh nets and the mean length caught in this mesh was 23.4 cm (Figure 2.3b). The net with the maximum catch is above the legal smallest mesh size (102 mm) allowed in Lake Kariba.

400 40 a) O. niloticus

300 30 ) cm (

th g 200 20

len Mean Total number/Mass (kg) number/Mass Total 100 10

8000 400 b) O. mortimeri

Mesh size (mm) vs Nilo no fish wt 600 Mesh size (mm) vs Nilo wt kg total 30 Mesh size (mm) vs Nilo mean length (cm)

400 20

(cm) length Mean (kg) number/Mass Total 200 10

0 0 20 40 60 80 100 120 140 160 180 200 Mesh size (mm) Total number of fish Total mass (kg) Mean length (cm) Figure 2.3 Total number, mass and mean length of a) Oreochromis niloticus and b) O. mortimeri caught at Lakeside in gill nets of different mesh sizes from 1992 to 2012. Error bars for mean length are standard deviations.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 32 Reproductive potential and replacement of an indigenous congener

The size distributions of these two species caught in the experimental gill nets are shown in Figure 2.4 and compared with the length frequency of O. niloticus from the artisanal fishery. For the experimental gill nets, the modal catch of O. niloticus is higher and size distribution wider than that of O. mortimeri. The catches from the fishing villages reflect that their nets had a larger mesh size than those used in the 2 experimental gill-netting programme (Figure 2.4). To enable size comparison of our fish, measured in standard length, with reported values of total length, the following function that describes the relationship between standard length and total length was derived:

Standard length (SL) = 0.838 total length (TL) – 0.2408 (R2 = 0.99; n =25)

300 a) O. niloticus 250

200 150

Frequency 100

0 b) O. mortimeri 300 y 200

Frequenc 100

0 4 yrs 600 c) Artisanal fishery 5 yrs 500

400 3 yrs 300

Frequency 2 yrs 200

100

0 0 10 20 30 40 50 60 Length (cm)

Figure 2.4 Length frequency of a) Oreochromis niloticus and b) O. mortimeri caught in the gill nets at Lakeside and c) O. niloticus caught in gill nets at Nyamhunga, Dandawa, Mudzimu and Nema‐ tombo artisanal fishing villages from December 2003 to October 2004. Ages are from scale readings (Chapter 3 – Chifamba & Videler 2014).

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 33 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

100

60 ity ur t a M

% 40

20 Length at first maturity

0 0204060 Standard length (cm) Figure 2.5 Length at first maturity for Oreochromis niloticus (○) and O. mortimeri ( ) calculated using Lakeside data from 1992 to 2012.

Length at first maturity The smallest mature O. niloticus female caught was 8.0 cm, versus 13 cm for the smallest mature male. Figure 2.5 shows that the length at 50% maturity was smaller for O. niloticus than for O. mortimeri. Length at 50% maturity for O. niloticus was 17.63 cm, obtained from the sigmoidal equation below (R2 = 0.95):

% mature fish = 100 / (1 + exp (- (fish length - 17.6266) / 3.3132))

There was little temporary change in the length at maturity. During the 1993 – 2002 and 2003 – 2012 time periods, the length at maturity was on average 17.91 cm and 18.28 cm respectively. The model explains 80% (R2 = 0.80) of the variation in the data from the first period, compared to 95% for the second period and 95% for the aggregated data from both periods (Table 2.2). The smallest mature O. mortimeri were of almost the same size in females and males, being 11.5 and 11.0 cm, respectively. The length at first maturity of O. mortimeri was 19.19 cm for the whole data set. The sigmoidal function below describes the relationship between the percentages of mature fish in each age group and fish length (R2 = 0.93):

% mature fish = 100 (1 + exp (- (fish length - 19.1884) / 4.2728))

The length at first maturity for the period 1993 – 2002 (17.77 cm) was less than that for the 2003 – 2012 period (18.52 cm) though the estimate for the first period was estimated with greater error than that for the second and the whole data set (Table 2.2).

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 34 Reproductive potential and replacement of an indigenous congener

Table 2.2 Results from fitting the sigmoidal function to the proportion (%) of mature fish at a given length to estimate the length at first maturity of Oreochromis niloticus and O. mortimeri using data collected from 1992 to 1993. (CI = 95% confidence interval calculated from the standard error (SE)). Period Length (cm) SE CI R2 F p n 2 O. niloticus 1992 ‐ 2002 17.91 0.86 ± 1.69 0.80 152.0 0.0001 41 2003 ‐ 2012 18.28 0.34 ± 0.66 0.95 578.5 0.0001 38 1992 ‐ 2012 17.62 0.36 ± 0.70 0.95 775.8 0.0001 47 O. mortimeri 1992 ‐ 2002 17.77 0.39 ± 0.77 0.91 339.5 0.0001 32 2003 ‐ 2012 18.52 0.32 ± 0.63 0.95 546.4 0.0001 28 1992 ‐ 2012 19.18 0.50 ± 0.99 0.93 399.2 0.0001 34

The combined data show a slightly higher length at first maturity for O. mortimeri compared to O. niloticus. The length at first maturity of O. mortimeri is closer to the size where the highest catches (mean length = 23.4 cm, caught in the 114-mm meshed nets) are made than that of O. niloticus (mean length = 29.3 cm, caught in 152-mm meshed nets) (Figures 2.3 and 2.4).

Seasonal trend in gonad activity The analysis of the annual trend in gonad activity shows that the proportion for O. niloticus with ripe and beyond stages of gonad development was always higher than that for O. mortimeri (Figure 2.6). The trend in females is more distinct compared to that in males. In both species, breeding takes place throughout the year. Breeding of female O. niloticus peaks in December and is lowest in May. In O. mortimeri the peak is in November and the lowest in June. The annual breeding pattern followed broadly the annual variation in rainfall and temperature that is shown in Figure 2.7. Figure 2.8 shows the relationships between rainfall and minimum temperature, as well as the proportion of breeding for both species. For O. niloticus, the minimum temperature (rs = 0.84; p = 0.001; n = 12) is significantly correlated to the breeding pattern, whilst rainfall (rs = 0.50; p = 0.098; n = 12) is not (Table 2.3). Both minimum temperature (rs = 0.80; p = 0.002; n = 12) and rainfall (rs = 0.82; p = 0.001; n = 12) are significantly correlated to the proportion of breeding O. mortimeri, with rainfall showing a higher correlation.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 35 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

70 O. niloticus female 60 O. niloticus male O. mortimeri female O. mortimeri male 50

30 Figure 2.6 Annual variation in the proportion of fish classified as ripe 20 (ripe, ripe/running, and spent) 10 stage of gonad development Proportion of ripe gonads (%) gonads ripe of Proportion in Oreochromis niloticus and 0 O. mortimeri.

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

40 80

30 C) )

mm ( 25 40 Figure 2.7 Mean month‐ ly rainfall ( ), minimum, Rainfall

Temperature ( Temperature 20 mean and maximum air 20 temperature (○) in Kariba 15 measured between 1992 and 2008. (Data from Zambezi River Authority). 10 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Figure 2.8 Relationship between the proportion of ripe gonads of Oreochromis niloticus (○) and O. mortimeri (●) and a) rainfall and b) minimum temperature.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 36 Reproductive potential and replacement of an indigenous congener

Table 2.3 Spearman rank correlation coefficients (rs) and significance levels for the relationships among monthly percentage of ripe and running Oreochromis niloticus, O. mortimeri, temperature (minimum, maximum and mean) and rainfall (significance: * = 0.05; ** = 0.01).

O. morti Min temp Max temp Mean temp Rainfall O. niloticus 0.66* 0.84** 0.76** 0.84** 0.50 2 O. mortimeri 0.80** 0.60* 0.75** 0.82** Min temp 0.92** 0.99** 0.66* Max temp 0.94** 0.39 Mean temp 0.62*

Discussion We found that the introduced Oreochromis niloticus replaced O. mortimeri in the inshore area of the Sanyati basin of Lake Kariba. The results suggest that this is caused by a higher reproductive potential of O. niloticus compared to O. mortimeri. Specifically, O. niloticus has a higher proportion of reproductively active individuals in the population and reaches maturity earlier. Furthermore, the breeding of O. niloticus is correlated with temperature only, whereas rainfall seems to be a more important trigger of gonad maturation in O. mortimeri. Even so, breeding takes place throughout the year in both species and sexes. The gillnet mesh with the maximum catch was higher for O. niloticus than for O. mortimeri. The total catch of both species together did not show a clear trend during the study, despite the competitive displacement that we observed. The introduction of Oreochromis niloticus in waterbodies is often associated with the disappearance of indigenous fish species (Weyl 2008; Balirwa 1992; Marshall 1999) whilst it is unclear if it really improves fisheries opportunities. Introduced Oreochromis niloticus displaced Oreochromis esculentus (Graham) and Oreochromis variabilis (Boulenger) in Lake Victoria and some native cyprinids in Lake Luhondo (Balirwa 1992). In Lake Chivero (Zimbabwe), O. niloticus replaced another introduced species (Oreochromis macrochir) and accounts for about 95% of the catches from that lake (Marshall 1999; Tiki 2011). Oreochromis niloticus is therefore an invasive fish species that has displaced O. mortimeri in Lake Kariba and other species in other waterbodies, through mechanisms that may include reproductive superiority of O. niloticus. In the present study, the replacement of O. mortimeri by O. niloticus did not lead to a change in the total catch, suggesting competitive displacement of a species by its functional equivalent. Hence, there was little added value of the introduction for the fishery. Successful invaders appear to be characterised by traits such a tolerance to a wide range of environmental conditions, fast growth and reproductive strategies to facilitate the establishment and expansion of their populations (Ruesink 2005; 37

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 37 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

Cucherousset et al. 2009; Deacon et al. 2011). The size and age at maturity of O. niloticus has been reported to vary among African lakes in adaptation to the environment (Lowe-McConnell 1982; Duponchelle & Panfili 1998). Maturation occurred at a total length of 39 (32 SL), 28 (23 SL) and 17 (14 SL) cm for both sexes in Lake Turkana, Lake George and Lake Kijanebalo, respectively, and at 14 (11 SL) and 12 (10 SL) cm for males and females, respectively, in Lake Albert. In Lake Nyamusingiri and Lake Kyasanduka, two crater lakes in western Uganda, the size at maturity was small (12 – 15 cm TL in both lakes) (Bwanika et al. 2004). Length at maturity in Lake Victoria was 30.81 (26 SL) cm total length for females and 34.5 (29 SL) cm TL for males (Njiru et al. 2006b). In Lake Ayame, Côte d’Ivoire where Oreochromis niloticus was stunted, it matured in the first year (10 months old), at 11.6 – 13.5 cm standard length (Duponchelle & Legendre 2000). The size at maturity that O. niloticus obtained in Lake Kariba, is in the middle of the range observed in different African lakes and is reached when the fish are about two years old (Chapter 3 – Chifamba & Videler 2014). Based on the reported age-specific survival and fecundity combinations that maximize fitness (Hutchings 1993), we consider the medium size at maturity of O. niloticus in Lake Kariba to be the optimum for maximizing reproductive effort. The relatively early onset of maturity of O. niloticus in Lake Kariba would give it a competitive advantage over O. mortimeri, which matures a little later. In this study, we found that the reproductive capacity of O. niloticus is enhanced by having a higher proportion of individuals breeding throughout the whole year than O. mortimeri. This may have made it possible for O. niloticus to increase its population faster than O. mortimeri. The duration of the reproductive season, and the spawning fraction are aspects of reproduction that are frequently used in assessing the reproductive potential of fish species (Murua et al. 2003). In some studies, the success of O. niloticus as an invader has been attributed to its high fecundity compared to the displaced species (Lowe-McConnell 1982; Duponchelle & Panfili 1998; Njiru et al. 2006b). This confirms the importance of the seasonal dynamics of reproduction as a component in invasion success. Breeding all year round in addition to a breeding peak in a specific season is typical of O. niloticus, both within its natural habitat and where it was introduced, a trait that might contribute to its success as an invader (Peterson et al. 2004; Njiru et al. 2006b). Oreochromis niloticus spawns throughout the year with a single peak associated with rainfall at high latitudes and two peaks in the equatorial region (Lowe- McConnell 1982; Peterson et al. 2004; Njiru et al. 2006a). In Lake Victoria, the breeding peak was between December and June, whilst in Cairo (Egypt) the main breeding peak was from April to mid-May (Trewavas 1983; Njiru et al. 2006a). Peterson et al. (2004) documented a year-round reproduction of Nile tilapia in the coastal watershed of the Mississippi, with one peak in spring (March to May) and another in late summer (August to September), though McDonald et al. (2007) recognized one period from November through to July in the same system. Breeding

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 38 Reproductive potential and replacement of an indigenous congener

of O. mortimeri (described as O. mossambicus in the paper; Trewavas (1983) seems to also occur throughout the year, with a peak in October – December (Donnelly 1969), though in our study the peak was in November – February and the lowest breeding in June – July. Hence, the breeding patterns in O. niloticus and O. mortimeri are remarkably similar, with differences only in the timing of the breeding peaks. 2 The results of this study suggest that gonad maturation and breeding in O. mortimeri is triggered by both temperature and rainfall, although the correlation with rainfall is much stronger than that with temperature. In contrast, rainfall shows an insignificant correlation with gonad maturation in O. niloticus. This difference in environmental triggers may have consequences for the reproductive success, an issue that needs further exploration. Furthermore, other studies confirm that breeding patterns in fish are often triggered by environmental factors (Hyder 1970; Brummett 1995). For example, Duponchelle et al. (1999) found that seasonal changes in temperature, rainfall, day length, chlorophyll-a concentration and water level often correspond with changes in the annual spawning cycle. They conclude that the annual periodicity of O. niloticus reproduction is influenced by the ephemerides cycle. The gonadosomatic index was found to be lowest during winter (May – August) in Tilapia rendalli and Oreochromis mossambicus in Lake Chicamba, a subtropical lake in central Mozambique (Weyl & Hecht 1998). Introduced fish are associated with both ecological costs and economic benefits such as those observed in Lake Victoria following the introduction of O. niloticus and Nile perch, Lates niloticus (Balirwa 1992; Witte et al. 1992; Njiru et al. 2008b). The introduction of O. niloticus and L. niloticus in Lake Victoria resulted in increased fish catches, at the expense of the native fish species (Njiru et al. 2008b). In Lake Kariba, such an improvement of the catches was not observed. Differences in environmental conditions of these two lakes may explain the difference in impact of O. niloticus. For example, Lake Kariba is an oligotrophic lake with a low primary production (0.1 – 1.7 g C2 / h) compared to Lake Victoria (3.3 – 3.5 g C2 / h) (Ndebele- Murisa et al. 2010). Another consideration is that O. niloticus in Lake Victoria switched to an omnivorous diet, including midge larvae and shrimp, which became abundant after the Nile perch eradicated their specialized predators, several species of haplochromine cichlids (Wanink 2005). Thus, O. niloticus in Lake Victoria might have taken a suddenly available open niche. In Lake Kariba, O. niloticus took over an occupied niche from which it displaced the native O. mortimeri, hence the lack of increase in the tilapia catches. The introduction of O. niloticus in Lake Kariba may have resulted in the use of a larger mesh size. As the largest catch in the experimental gill nets set at Lakeside occurred at a larger size in O. niloticus than in O. mortimeri, larger meshed nets are needed to maximize the catches of the fishery for O. niloticus than those of the fishery for O. mortimeri. There are some indications that the mesh size used by the tilapia fishery in Lake Kariba increased after the invasion of Nile tilapia (PC Chifamba, 39

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 39 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

personal observation; Anonymous fisherman, personal communication). Such a change in the mesh size has been reported for Lake Victoria, where the introduction of O. niloticus resulted in the return of mesh sizes as large as 127 mm (5 inch). This large mesh size had not been used since the native tilapia stock was overfished and the catches consisted of small-sized fish only (Njiru et al. 2006a). In addition to increasing the declining catches of the native tilapias, returning the use of large mesh sizes was one of the aims of the introduction of O. niloticus into Lake Victoria. There is a need to monitor and analyse the catches and the gear used in the inshore or artisanal fishery of Lake Kariba, to ascertain if this change in gear persists. Oreochromis niloticus has established populations in many rivers in Southern Africa (Kafue, Nata, Runde-Save, Buzi and Limpopo). The species has the potential to successfully invade the Cunene, the Okavango and the Upper Zambezi rivers, because of its tolerance to broad climatic conditions (Schwanck 1995; Zengeya et al. 2015). Oreochromis mortimeri, that was displaced by O. niloticus in the Zambezi river, is in the Critically Endangered biodiversity category (Darwall et al. 2009). Two other congeners, Oreochromis andersonii and O. macrochir are endemic to the Cunene, the Okavango, and the Upper Zambezi. Therefore, the introduction of O. niloticus into these river systems is a potential threat and a conservation issue in need of attention. More stringent monitoring and enforcement of measures to prevent the introduction of this species into these systems is recommended, especially because our study shows that it may not always benefit fisheries opportunities.

Acknowledgements The Lakeside experimental gill-net data used in the analysis was kindly provided by the Lake Kariba Fisheries Research Institute. We are indebted to the Zambezi River Authority for sharing the Kariba temperature and rainfall data used in this paper. Field work was supported by a grant from the International Foundation for Science (IFS) A/3159-1.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 40 Growth rates of alien Oreochromis niloticus and indigenous Oreochromis mortimeri in Lake Kariba, Zimbabwe

Portia C. Chifamba Chapter 3 John J. Videler

Published in 2014 African Journal of Aquatic Science 39: 167‐176

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 41

Abstract Growth rates of indigenous Oreochromis mortimeri and alien Oreochromis niloticus from Lake Kariba were estimated from samples collected in 1997–2000, 2003–2005 and 2010–2011. Growth zones on scales and otoliths of O. niloticus and on the otoliths and opercula of O. mortimeri were deposited annually.

Age estimates obtained from otoliths, scales and opercula were similar for both species. Length at age was described by the von Bertalanffy equations:

L(t) = 32.4 (1 – exp (−0.25 (t + 2.3))) and L(t) = 30.2 (1 – exp (−0.23 (t + 3.73)))

for O. niloticus and O. mortimeri respectively, in 1997–2000, and by

L(t) = 44.6 (1 – exp (−0.29 (t − 0.05))) and L(t) = 36.8 (1 – exp (−0.64 (t − 0.73)))

respectively, in 2003–2005.

There were more older (≥10 years) age classes and lower L∞ in the 1997–2000 dataset for both species than in the later sampling periods. Mean length at age of O. niloticus was higher than that of O. mortimeri for the 1997–2000 dataset, but similar for the 2003–2005 dataset.

This study supports the hypothesis that, in Lake Kariba, the alien O. niloticus has a higher growth rate than O. mortimeri, and infers its competitive advantage over the indigenous species.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 42 Growth rates of alien and indigenous congeners

Introduction The original fauna of the Zambezi River included Oreochromis mortimeri (Trewavas 1966), a species endemic to the middle Zambezi Valley from the Victoria Falls to Cahora Bassa Gorge (Marshall 2011). Before Kariba dam was built, catches of O. mortimeri in the river were small, compared to those of other species. Following impoundment, which began in 1958, catches of O. mortimeri in Lake Kariba in- 3 creased rapidly, making up 38% of the inshore catches in the 1970s (Kenmuir 1984). Although catches decreased during the 1980s, it has remained important in the fishery (Karenge & Kolding 1995). Oreochromis niloticus (Linnaeus 1758) invaded Lake Kariba from fish farms at Kariba town, first appearing in fisheries catches in 1989, and subsequently being caught throughout the whole lake (Thys van den Audenaerde 1994; Chifamba 1998; Zengeya & Marshall 2008). Its introduction was followed by a reduction in the catches of O. mortimeri (Chifamba 2006). The natural distribution of O. niloticus includes the Nile basin, Rift Valley lakes and some West African rivers (Skelton 2001). It has the potential to affect native fish species adversely through competition for food, hybridisation and competition for space (Zengeya & Marshall 2007; Weyl 2008; Zengeya et al. 2011, 2013). Many studies have shown that large fish tend to have more eggs than small fish (Bagenal 1978). Fast growth would therefore result in a larger number of eggs at an earlier age. In addition, fast growth may also reduce the time an animal spends at a vulnerable size, since smaller animals tend to be more vulnerable to predation (Sutherland 1996). Hence, a comparison of the growth and maximum size of introduced O. niloticus and native O. mortimeri will give an indication of potential relative competitiveness of each of these species. The use of growth zones deposited in calcified structures is an accepted method for ageing fish. Even in the tropics there are periodic depositions that can be used for age determination (Zekeria et al. 2006; Bwanika et al. 2007). Validation is, however, necessary to ascertain the periodicity of deposition of growth zones (Campana 2001). Studies on fish growth show that prior selection of a growth model (which is often the von Bertalanffy model) is inappropriate because fish growth trajectories vary with species and growth stage (Araya & Cubillos 2006; Enberg et al. 2008). The alternative is to apply the strength of evidence concept, where the model that fits the observations best is selected (Stewart & Hughes 2005; Katsanevakis & Maravelias 2008). The models are compared and ranked using Akaike’s information criterion; a procedure comprehensively described in the literature (Burnham et al. 2011). Despite its 23-year presence in Lake Kariba, little is known about the biology of O. niloticus. Estimates of its growth rates in various African water bodies differ

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 43 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

considerably (Getabu 1992; Bwanika et al. 2007) and thus an understanding of its growth performance in Lake Kariba was needed. The objective of this study was to compare the growth rates of O. niloticus and O. mortimeri by answering the following questions. How reliable are scales, otoliths and opercula in ageing the two species? Which model best explains the observed length-at-age data? Are there differences in the growth parameters and the mean length at age of O. niloticus and O. mortimeri?

Materials and methods Collection of samples The study was carried out on the Zimbabwean side of the Sanyati Basin of Lake Kariba. Fish samples were collected using multi-mesh fleets of gillnets, each comprising twelve 40 m panels of 38, 51, 64, 76, 89, 102, 114, 127, 140, 152, 165 and 178 mm stretched mesh. Samples were collected during three periods: 1997 – 2000 (n = 50 O. niloticus, 75 O. mortimeri); 2003 – 2005 (n = 113 O. niloticus, 30 O. mortimeri); and 2010 – 2011 (n = 52 O. niloticus). The datasets from these periods are hereafter referred to as the 1997, 2003 and 2010 datasets, respectively. Standard length (SL, mm) and sex of each fish were recorded, and the two sagittal otoliths, the opercula and 6 – 8 scales were collected from between the dorsal fin and the lateral line, in line with the pectoral fin. The opercula were boiled for 10 minutes and air-dried, after which the flesh was scraped off. Otoliths and scales were cleaned with water and air-dried.

Otolith, operculum and scale reading All three structures showed, in most cases, concentric patterns of alternating translucent and opaque zones. The zones on opercula were easier to observe without magnification. Scales were read on a scale reader using transmitted light. Otoliths were read whole on a microscope slide, with their glycerine-wetted surface illumi- nated, and examined against a black background. Structures not clearly showing translucent and opaque zones were discarded.

Validation Several methods are used for validation of annual growth increments, one of which is edge analysis, as used by Winker et al. (2010). Edge analysis was used for validating the periodicity of band deposition. The type of zone forming the outer margin of each structure was noted. Each type of body structure was examined independently so that judgement on one did not influence that on the others. Changes in the monthly proportion of the fish with each band type were used to deduce the pattern of band deposition. Two experienced readers made independent counts of the bands, and ages were assigned by consensus. Counts using the opercula, otoliths and scales were carried

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 44 Growth rates of alien and indigenous congeners

out on separate days to prevent counts on one body part influencing those on others, in order to obtain independent datasets for each of the structures used. Not all three structures from every specimen could always be used.

Comparison of ages from scales, opercula and otoliths Symmetry in the ages obtained from the three body structures was assessed using Bowker’s test (Hoenig et al. 1995). In this method a Chi-square (χ2) test is used to determine if the number of fish assigned age k using method 1 and m using method 3 2 differ from the number assigned age m using method 1 and k using method 2. Only those fish that could be aged using all three body structures were included in the analyses.

Selection of growth model Four commonly-used growth models – the von Bertalanffy function (VBGM), the Gompertz model (Gompertz, 1825), a logistic model and a power function – were fitted to all sets of length-at-age data. The VBGM model has been widely used to study growth in fish, including O. niloticus (Getabu 1992; Zekeria et al. 2006; Bwanika et al. 2007), and was therefore considered useful for comparing current results with published findings. The equations for the four models respectively are:

VBGM: L(t) = L∞ (1 – exp (– k1 (t – t0)))

Gompertz: L(t) = L∞ exp (–1 / k2) exp (– k2 (t – t1))

Logistic: L(t) = L∞ / (1 + exp (– k3 (t – t2))) b Power: L(t) = a0 + a1 t

where L(t) = length at age t (cm), L∞ = asymptotic length (cm) and t = age (years) at –1 capture. In the VBGM, k1 is a relative growth parameter (y ) and t0 is the age at

which individuals would have been at zero-length. For the Gompertz model, k2 is –1 the rate of exponential decrease in the relative growth rate (y ) and t1 = 1 ∕ k2 ln λ,

where λ is the theoretical initial growth rate at zero-age. In the logistic model, k3 is –1 relative growth rate (y ) and t2 the inflection point of the sigmoidal curve. In the

power model, a0 is the intercept or the length at zero-age, whilst a1 is the scaling factor or the rate at which the length increases with time, and b the exponent describing the rate of change in the relationship between length and age. Model selection and the estimation of model selection uncertainty were based on the information theory approach (Anderson et al. 2000; Burnham & Anderson 2002; Burnham et al. 2011), with the most parsimonious model for each year, species and ageing structure chosen with the lowest small sample biased-adjusted Akaike information criterion (AICc), where:

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 45 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

RSS AICc = n

�� where n is the numberln of� 𝒏𝒏data� ��points (observations),𝒏𝒏�� RSS is the residual sums of squares and K is the number of parameters in the model.

The model-averaged asymptotic length ∞ for species, year and age-ageing method was calculated through multi-model inference outlined in Johnson & Omland (2004). 𝑳𝑳� The parameters and their variances were estimated by the least-squares method computed using the Levenberg-Marquardt method of the SPSS (v. 19) statistical software.

Comparison between species and years Mean length-at-age data were used to compare growth of O. niloticus with that of O. mortimeri, as well as temporal differences within each species. The mean length at age was the average length of fish assigned the same age. For each species and dataset, a series of ages and their mean lengths was generated. Differences between the species from the three test periods, as well as annual differences within species, were tested using ANOVA.

Results The proportion of fish scales, opercula and otoliths with an opaque marginal zone increased from September, the start of the warm and wet season, reaching maximum values in May. Subsequently, the proportion decreased gradually towards values close to zero between June and August. In that period, during the cool, dry season, most structures showed translucent margins (Figure 3.1). Thus, an annual growth cycle consisted of an opaque and a translucent zone, which together were regarded as one annulus that was used to age the fish. Whole scales, otoliths and opercula of 215 O. niloticus and 105 O. mortimeri were analysed. The whole otoliths were more difficult to read, so 28% of the otoliths from each species that could not be read were discarded.

Figure 3.1 Monthly proportions of Oreochromis niloticus scales with an opaque band on their edge, plotted from the 1997‐ 2000 dataset.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 46 Growth rates of alien and indigenous congeners

Comparison of age estimates obtained from scales, otoliths and opercula The numbers of fish assigned to particular ages using a pair of body structures are presented in Tables 3.1 and 3.2. Ages read on scales and opercula of O. niloticus agreed more often for fish aged 1 to 3 years than for older fish (Table 3.1a). Ages assigned to O. mortimeri using scales and opercula differed more in older fish than in younger fish (Table 3.2a). In both species, otolith ages differed most from those from the other body structures (Tables 3.1b–c and 3.2b–c). There was no significant 3 Table 3.1 Numbers of Oreochromis niloticus caught in Lake Kariba in 1997‐2000 assigned to particular ages (y) using (a) scales and opercula, (b) scales and otoliths, and (c) opercula and otoliths. Numbers where ages agree in bold type.

Operculum age (a) Scale age 1 2 3 4 5 6 7 8 9 10 1 6 2 3 4 3 1 2 1 4 2 0 5 2 0 1 1 6 0 1 0 0 7 1 1 3 2 1 8 1 1 0 1 2 9 1 0 1 1 10 1 0 Otolith age (b) Scale age 1 2 3 4 5 6 7 8 9 10 1 4 1 1 2 2 3 1 1 3 2 1 1 0 1 4 0 1 0 5 1 1 0 6 0 1 7 1 0 2 8 1 0 9 2 0 10 1 0 Otolith age (c) Operculum age 1 2 3 4 5 6 7 8 9 10 1 4 3 2 2 1 2 0 1 3 2 3 0 4 1 1 1 5 0 1 1 1 6 1 0 1 7 0 0 1 8 1 1 1 9 1 0 10 1 0

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 47 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

Table 3.2 Numbers of Oreochromis mortimeri caught in Lake Kariba in 1997‐2000 assigned to particular ages (y) using (a) scales and opercula, (b) scales and otoliths, and (c) opercula and otoliths. Numbers where ages agree in bold type.

Operculum age (a) Scale age 1 2 3 4 5 6 7 8 9 10 11 12 13 1 0 2 2 0 2 1 3 1 0 2 1 1 4 1 1 3 1 0 5 0 1 0 1 1 0 1 6 2 0 1 2 0 1 1 1 7 5 1 1 0 0 0 8 0 1 2 1 0 0 0 9 0 1 1 0 1 1 10 1 3 2 0 0 11 0 0 12 0 13 1 0 Otolith age (b) Scale age 1 2 3 4 5 6 7 8 9 10 11 12 13 1 0 1 3 2 1 2 3 2 1 1 4 0 1 2 1 0 1 5 0 0 0 1 0 1 1 6 1 0 2 1 1 2 0 7 2 2 0 1 0 8 0 1 0 9 1 0 2 0 1 10 0 1 1 0 0 11 0 0 12 0 0 0 13 1 1 0 Otolith age (c) Operculum age 1 2 3 4 5 6 7 8 9 10 11 12 13 1 0 2 2 1 0 3 1 0 1 4 0 1 2 1 1 5 1 1 1 0 0 0 1 6 0 1 2 1 0 1 0 7 1 2 0 0 0 1 0 8 2 0 1 1 0 9 0 0 1 0 2 0 10 0 0 1 0 0 0 11 0 2 0 0 1 12 1 1 1 1 0 13 0

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 48 Growth rates of alien and indigenous congeners

difference (p > 0.05) in any of the χ2 tests for symmetry, hence there was no statistical difference in the ages from the three body structures (Table 3.3).

Table 3.3 Chi‐square, df and p values from test of systematic agreement in ages of Oreochromis niloticus and O. mortimeri caught in Lake Kariba in 1997‐2000 read from various pairs of body structures (scales, opercula and otoliths).

2 Body structures used χ df p 3 O. niloticus Scales vs opercula 16 14 0.31 Scales vs otoliths 8 10 0.66 Opercula vs otoliths 15 14 0.38 O. mortimeri Scales vs opercula 21 21 0.44 Scales vs otoliths 25 21 0.25 Opercula vs otoliths 31 27 0.26

Table 3.4 R2 and AICc values for von Bertalanffy (VBGM), Gompertz, logistic and power growth models fitted to Oreochromis niloticus and O. mortimeri length‐at‐age data from scales, otoliths and opercula ages. Best models in bold type.

Growth model Fish species and dataset Body structure n VBGM Gompertz Logistic Power R2 O. niloticus 1997 Scales 0.857 0.854 0.851 0.864 46 Otoliths 0.792 0.791 0.900 0.793 33 Opercula 0.853 0.852 0.850 0.858 44 O. niloticus 2003 Scales 0.582 0.582 0.581 0.581 111 O. niloticus 2010 Scales 0.550 0.550 0.550 0.530 52 O. mortimeri 1997 Scales 0.590 0.584 0.578 0.493 73 Otoliths 0.696 0.698 0.699 0.689 40 Opercula 0.489 0.490 0.491 0.480 55 O. mortimeri 2003 Scales 0.493 0.501 0.504 0.478 30 AICc O. niloticus 1997 Scales 28.1 28.5 28.9 27.1 46 Otoliths 28.0 28.1 28.1 27.9 33 Opercula 27.7 28.0 28.2 27.1 44 O. niloticus 2003 Scales 174.3 174.3 174.4 174.4 111 O. niloticus 2010 Scales 54.3 54.3 54.1 54.8 52 O. mortimeri 1997 Scales 69.5 69.9 70.4 76.2 73 Otoliths 27.5 27.4 27.3 27.9 40 Opercula 52.9 52.8 52.8 53.3 55 O. mortimeri 2003 Scales 51.5 51.3 51.3 51.9 30

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 49 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

Growth model selection The differences among the R2 and AICc values were small, but the power and logistic models explained the data better than the VBGM and the Gompertz models (Table 3.4). Considering the O. niloticus 1997 dataset and the AICc values, the power

Table 3.5 Parameters from the asymptotic models VBGM (k = k1; t = t0), Gompertz (k = k2; t = t1),

Logistic (k = k3; t = t2), the 95% lower and upper limits and average model L∞.

Fish species and dataset Body structure Model L∞ k t O. niloticus 1997 Scales VBGM 32.4 ± 3.4 0.25 ± 0.14 ‐2.30 ± 1.54 Gompertz 31.7 ± 2.7 0.31 ± 0.14 ‐4.56 ± 3.82 Logistic 31.3 ± 2.3 0.38 ± 0.15 0.11 ± 0.52 Average 31.6 Otoliths VBGM 36.7 ± 15.7 0.15 ± 0.21 ‐4.04 ± 4.41 Gompertz 34.7 ± 10.3 0.21 ± 0.21 ‐8.70 ± 12.66 Logistic 33.5 ± 7.7 0.28 ± 0.23 ‐0.05 ± 1.13 Average 34.9 Opercula VBGM 32.3 ± 3.7 0.27 ± 0.17 ‐2.31 ± 1.67 Gompertz 31.7 ± 3.0 0.34 ± 0.18 ‐4.22 ± 3.93 Logistic 31.3 ± 2.5 0.41 ± 0.19 ‐0.12 ± 0.51 Average 32.2 O. niloticus 2003 Scales VBGM 44.6 ± 17.7 0.29 ± 0.25 0.05 ± 0.65 Gompertz 38.1 ± 7.9 0.59 ± 0.30 0.46 ± 1.17 Logistic 35.5 ± 5.2 0.89 ± 0.35 1.86 ± 0.41 Average 36.3 O. niloticus 2010 Scales VBGM 37.8 ± 10.3 0.29 ± 0.32 ‐0.75 ± 2.46 Gompertz 36.4 ± 7.3 1.24 ± 0.73 0.40 ± 0.34 Logistic 35.5 ± 5.7 0.52 ± 0.37 1.31 ± 0.72 Average 36.5 O. mortimeri 1997 Scales VBGM 30.2 ± 4.2 0.23 ± 0.27 ‐3.73 ± 5.77 Gompertz 30.4 ± 4.5 0.23 ± 0.26 ‐9.19 ± 16.25 Logistic 30.6 ± 4.8 0.24 ± 0.26 ‐2.17 ± 3.42 Average 30.8 Otoliths VBGM 34.0 ± 10.2 0.15 ± 0.19 ‐4.14 ± 5.3 Gompertz 32.9 ± 7.4 0.20 ± 0.19 ‐9.52 ± 13.95 Logistic 32.1 ± 5.8 0.26 ± 0.20 ‐0.27 ± 1.18 Average 33.0 Opercula VBGM 30.2 ± 4.0 0.20± 0.17 ‐3.49 ± 4.04 Gompertz 29.8 ± 3.3 0.25 ± 0.19 ‐7.36 ± 9.54 Logistic 29.5 ± 2.8 0.30± 0.20 ‐0.61 ± 1.95 Average 29.6 O. mortimeri 2003 Scales VBGM 36.8 ± 14.6 0.64 ± 0.73 0.73 ± 0.72 Gompertz 33.6 ± 8.8 1.17 ± 1.13 1.59 ± 0.93 Logistic 32.3 ± 7.2 1.68 ± 1.60 1.75 ± 0.43 Average 34.4

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 50 Growth rates of alien and indigenous congeners

Figure 3.2 Growth curves for (a) Oreochromis niloticus and (b) O. mortimeri from scales, opercula and otoliths collected in the 1997‐2000 period from Lake Kariba. Power model fitted in each case.

model was the best model for all body structures. The Gompertz and logistic models were best for the 2003 and 2010 datasets, respectively, using the AICc. For O. mortimeri, the best model for the scales data was the VBGM, whilst the logistic model was best for all the other datasets. With the current data, no one model could be used for both species, or for the same species, all times.

The VBGM model almost always resulted in a higher estimate of L∞ than the logistic and Gompertz models (Table 3.5). Otoliths of both species had the highest

L∞. The estimates of L∞ from the VBGM were 32.4, 36.7 and 32.3 cm for the scales, otoliths and opercula of O. niloticus, and 30.2, 34.0 and 30.2 cm for O. mortimeri, respectively, for the 1997 dataset. The growth curves from the body structures were

comparable for each species (Figure 3.2). The L∞ showed temporal changes, being

lowest for the 1997 dataset and highest for the 2003 dataset. The VBGM model L∞ of O. mortimeri was 30.2 and 36.8 cm for the 1997 and 2003 datasets, respectively. Values for O. niloticus were 32.4, 44.6 and 37.8 cm for the 1997, 2003 and 2010 datasets, respectively. Length at age was described by the von Bertalanffy equations L(t) = 32.4 (1 – exp (−0.25 (t + 2.3))) and L(t) = 30.2 (1 – exp (−0.23 (t + 3.73))) for O. niloticus and O. mortimeri, respectively, in 1997–2000, and by L(t) = 44.6 (1 – exp (−0.29 (t − 0.05))) and L(t) = 36.8 (1 – exp (−0.64 (t − 0.73))) in 2003 and 2004. The earlier samples contained older fish (≥10 years) of both species and smaller O. niloticus compared to the later years (Figures 3.3 and 3.4). In the 1997 dataset, Oreochromis niloticus had higher mean lengths at age for most age groups than O. mortimeri (Table 3.6). At ages 2 and 7, the size of O. niloticus was significantly larger than that of O mortimeri (p < 0.05, ANOVA). The sample size of both these age groups was ≥5, whereas it was smaller in other groups. There were significantly older O. mortimeri (mean 6.4 years) than O. niloticus (mean = 4.7) in the 1997 dataset (p < 0.05, ANOVA). In the 2003 dataset, the mean age was 3 years for both species, signifying a decrease in age compared to the 1997 dataset. 51

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 51 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

Figure 3.3 Lengths at age of Oreochromis niloticus and O. mortimeri from Lake Kariba for the 1997‐2000, 2003‐2005 and 2010‐2011 datasets. Power model fitted in each case.

Figure 3.4 Length frequencies of catches of Oreochromis niloticus and O. mortimeri from Lake Kariba for the periods 1997‐2000 and 2003‐2005. Numbers in bars = number of specimens.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 52 Growth rates of alien and indigenous congeners

Table 3.6 Mean lengths of Oreochromis niloticus and O. mortimeri at a given age for the 1997, 2003, 2011 and combined datasets. No O. mortimeri were caught in 2011.

Mean length (cm) Age O. niloticus O. mortimeri (y) 1997 2003 2011 All 1997 2003 2011 All 1 17.5 11.2 13.2 5.7 7.4 6.6 2 21.9 18.6 22.5 19.3 19.8 19.4 19.6 3 23.9 26.7 24.2 26.3 25.6 28.9 27.6 3 4 25.0 29.7 28.6 29.0 25.8 30.4 27.0 5 26.9 34.1 31.4 32.0 25.2 34.5 26.2 6 27.8 40.5 32.3 32.0 26.7 26.7 7 28.5 32.3 28.5 26.4 26.4 8 30.5 34.9 30.5 28.9 28.9 9 30.6 30.6 27.8 27.8 10 31.7 31.7 30.1 30.1 13 30.3 30.3

Discussion Periodicity in opaque and translucent zone deposition Annual periodicity in the deposition of alternating zones in body structures of O. mortimeri and O. niloticus was confirmed (Figure 3.1). Annual periodicity may be typical of Lake Kariba fish, where Hydrocynus vittatus and Sargochromis codringtonii also showed annual periodicity (Balon 1972; Moyo & Fernando 2000). Scales of Oreochromis andersoni and Pseudocrenilabrus philander investigated by serial harvesting from ponds in Chilanga, Zambia (Balon & Chadwick 1974), confirmed annual deposits of zones in scales of O. andersoni. Using scales and otoliths, Booth et al. (1995) also confirmed annual deposition of zones in O. andersonii from the Okavango Delta, Botswana. Marginal zone analysis was used to validate annual band formation in O. niloticus, Oreochromis macrochir and O. mossambicus (Booth & Merron 1996; Weyl & Hecht 1998; Gómez-Márquez et al. 2008; Grammer et al. 2012). In contrast, biannual bands were found in O. niloticus from Lakes Awassa, Ethiopia, and Nabugabo, Uganda (Bwanika et al. 2007).

Comparison of ageing methods The differences in the ages obtained using different body structures varied with the age of the fish. This is a common problem in ageing fish, as estimates of age from different body structures often give different results in older fish (Burnham-Curtis & Bronte 1996; Maceina & Sammons 2006). This tendency is indicative of difficulties in reading ages as the fish grow, which introduces ageing error. From age 6, otoliths gave lower age estimates than scales, suggesting greater difficulty in reading ages

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 53 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

from otoliths or an increase in the number of zones on scales. The number of unreadable otoliths was indicative of problems in using whole otoliths. Otolith readability can be improved by sectioning or grinding prior to reading, enabling the ageing of older fish, compared to whole otoliths (Campana 1984; Dwyer et al. 2003). Taylor & Weyl (2012) found that sectioned otoliths of Micropterus salmoides were more readable and gave more precise age readings than scales. Sectioned otoliths of O. andersoni gave more reliable readings than scales (Booth et al. 1995). Therefore, future studies and monitoring programmes should use sectioned otoliths to reduce the bias in age estimates.

Growth model selection Based on the AICc, the logistic and power models were the best growth models for both O. niloticus and O mortimeri, supporting the fact that the VBGM is not necessarily the best growth model for fish. Even so, the VBGM can be the best in some cases where growth models for different fish species were compared using AICc (Katsanevakis 2006; Alp et al. 2011). Support for the VBGM was reported by Alp et al. (2011) for the European catfish Silurus glanis. As in our study, Katsanevakis (2006) and Katsanevakis & Maravelias (2008) found that other growth models fitted some fish species better and that the VBGM was not the best model. The Gompertz growth model fitted the data of Yellowfin tuna Thunnus albacares and Sandbar shark Carcharhinus plumbeus best, whereas the VBGM was best for Striped seabream Lithognathus mormyrus, and the generalised VBGM for the Rougheye rockfish Sebastes aleutianus (Katsanevakis 2006). These results should be interpreted not only in mathematical terms but also in biological terms. The VBGM is used in fisheries models and may therefore be the most applicable, but not the best-fitting, model. In addition, there was no significant difference in the growth parameters of O. niloticus and O. mortimeri, judging from the confidence intervals, which were wide because very few large fish were caught.

Growth comparison Results obtained from scales were comparable to those of past studies on O. mortimeri from Lake Kariba, in which scales were also used (Krupka 1974). In Krupka’s study the largest O. mortimeri measured 39.0 cm SL and weighed 2.9 kg at 8 years old. One-year-old fish from different locations in Lake Kariba measured between 6.8 and 9.2 cm SL, whilst fish that were 5 years old measured between 32.8 and 33.7 cm SL. Krupka’s (1974) results were more similar to those from the 2003 than to those from the 1997 dataset in the present study. Mhlanga (1998) reported a maximum standard length of 33 cm and an asymptotic length of 34.7 cm, using length-based methods. Based on the growth rate (k) and L∞, the growth rate of O. niloticus in Lake Kariba was comparable to its growth rates in other African water bodies (Getabu 1992; Njiru et al. 2006b). Even so, the maximum size attained by this species in different water 54

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 54 Growth rates of alien and indigenous congeners

bodies varied, and lower growth rates and maximum sizes for it have been recorded (Trewavas 1983). The local environment seems to determine the growth and maximum size attained. In Lake Turkana a maximum size of 64 cm TL was recorded by Lowe McConnell (1958). Growth was faster in Lake Nabugabo, where O. niloticus had a relatively energy-rich omnivorous diet, compared to that in Lake Wamala, where they fed on phytoplankton (Bwanika et al. 2007). Oreochromis niloticus of over 4 years of age from Lake Nabugabo were, on average, 10 cm larger than those from Lake Wamala. 3 The relatively faster growth of O. niloticus indicates that this species may have had a competitive advantage over O. mortimeri in Lake Kariba. The large maximum size and fast growth of O. niloticus may increase their reproductive capacity, since large fish carry more eggs (Schenck & Whiteside 1977; Njiru et al. 2006b). In addition, when competing for resources such as food and nesting sites, large size is an advantage (Cutts et al. 1999; Mikheev et al. 2005). The temporal differences in the growth rates of O. mortimeri and O. niloticus observed in this study are interesting and could have arisen from a number of causes, including changes in fishing pressure, as well as environmental changes. The disappearance of the older individuals in the 2003 dataset suggests that heavy fishing pressure may have contributed to the temporal changes. In the 1997 dataset, the oldest O. niloticus and O. mortimeri were 10 and 13 years old, respectively, but in the 2003 dataset the maximum ages had reduced to 5 and 6 years for O. niloticus and O. mortimeri, respectively. Bwanika et al. (2004) suggested that fishing pressure was the cause of the small mean size of O. niloticus in two lakes in Uganda, contrary to the increase in length observed in this study. Additional studies are needed to explore further the findings of the present study, which was based on small sample sizes. The question of possible fishing-induced age and size structure of the fishery, as suggested by this study, needs to be thoroughly investigated.

Acknowledgements We are indebted to the International Foundation for Science (IFS) for a grant for this research. Lake Kariba Fisheries Research Institute granted us permission to extract body parts from fish caught in their Lakeside Gill-net Sampling Programme. This work was carried out at the University Lake Kariba Research Station.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 55

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 56 Diet overlap between a native and an invasive tilapia species in the Sou thern‐African Lake Kariba

Portia Chiyedza Chifamba Britas Klemens Eriksson Chapter 4

Submitted for publication African Journal of Ecology

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 57

Abstract

Nile tilapia (Oreochromis niloticus L.) is a widely used aquaculture species in African lakes, where it often has established and displaced native species. However, it is not clear if the ecological impact of the Nile tilapia is triggered by species competition or changed food-web interactions. In Lake Kariba, escaped Nile tilapia has displaced the native Kariba tilapia (Oreochromis mortimeri Trewavas 1966), but the ecological interactions between the species are not well known. To investigate if food competition is a potential driver of the displacement, the extent of diet overlap of the two tilapia species was studied. Benthic pennate diatoms were the dominating food source for all individuals studied, contributing 83% to the volume of the diet for Nile tilapia and 78% for the Kariba tilapia. Most tilapia individuals consumed both detritus (ca 80%) and sand (ca 70 %). Similarity indices demonstrated a strong diet overlap between the two tilapia species, and their relative protein digestion efficiency was also similar. Thus, the different tilapia species almost had an identical food niche, indicating that there is a significant potential for food competition being a contributing factor for the decline of the native Kariba tilapia after the introduction of Nile tilapia.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 58 Diet overlap between native and invasive congeners

Introduction Oreochromis niloticus L. (Nile tilapia) is a cichlid fish endemic to the Nile River, which is widely used in aquaculture all over the world due to its superior growth rate and adaptation to aquaculture conditions compared to other cichlids (Phillippart & Ruwet 1982; Blow & Leonard 2007). Today, Nile tilapia has been established into many African lakes, either by deliberate introduction or by escaping from fish farms (Ogutu-Ohwayo 1990; Weyl 2008; Chifamba 1998). In connection with establishment and subsequent increases in population sizes of the Nile tilapia, many native species have declined or totally disappeared (Ogutu-Ohwayo 1990; de Vos et al. 4 1990). The cause of the decrease in native fish fauna has been attributed to competitive exclusion by the introduced species, or to overfishing from a new and more intensive fisheries that have developed around the newly established species (Ogutu-Ohwayo 1990); Canonico et al. 2005; Njiru et al. 2006a). In this paper, we investigate if increased competition from the introduced Nile tilapia may have contributed to the decline of the native Kariba tilapia (Oreochromis mortimeri) in Lake Kariba by documenting the extent of their diet overlap. Lake Kariba, on the Zambezi River, was created in 1958 to provide hydroelectricity, though other economic activities that include fisheries were anticipated. To facilitate the development of a fishery, some tracts of land were cleared of trees before inundation (van der Lingen 1973). Between 1964 and 1972, three fish species dominated the catches, a predator, Hydrocynus vittatus (Castelnau 1861) and two cichlids, Oreochromis mortimeri (Trewavas 1966) and Serranochromis condringtoni (Karenge & Kolding 1995), both endemic to the Middle Zambezi River. In 1962, O. mortimeri contributed 35% to the artisanal fishery catches from Sinazongwe area. Later in 1974, it formed 38% of catches from Sanyati to Sengwa basins (Kenmuir 1984). Aquaculture activity in Kariba grew from small farms in 1980s to include a comer- cial scale cage culture production with a fish processing factory in 1990s. Kariba fish farmers preferred growing O. niloticus and soon escapees from the farms on the shores entered the lake. By 1993 the species had become established (Chifamba, 1998). With the introduction of O. niloticus, the once abundant O. mortimeri declined and disappeared from many parts of the Lake Kariba (Chifamba 2006; Zengeya & Marshall 2008). A similar effect on indigenous fish was observed in other waterbodies where O. niloticus has been introduced (Balirwa 1992; de Vos et al. 1990). In Lake Victoria, it displaced Oreochromis esculentus (Graham) and Oreochromis variabilis (Boulenger) and native cyprinids in Lake Luhondo (Balirwa 1992; de Vos et al. 1990). This negative effect has been attributed to competition for resources such as food. Earlier investigations of the diets of O. niloticus and that of O. mortimeri in Lake Kariba showed that the two species utilize similar food, but they did not quantify the

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 59 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

composition of different food items (Chifamba 1998; Mhlanga 2000). Oreochromis niloticus feeds on phytoplankton in Lake George and Turkana, though macrophytes are consumed in certain waters (Trewavas 1983). The percentage ‘Frequency of Occurrence’ (FO) used in these studies gave an indication of the food spectrum utilized by the two species. The method has the advantage that it is quick and requires minimum apparatus. However, the method does not give any information on the amount of the diet item consumed (Hyslop 1980), which is needed to calculate diet overlap. This study estimates the quantity of consumed food items using the ‘Numerical Relative abundance method’ (NM) and ‘Volumetric Relative abundance method’ (VM) (Windell 1968; Hyslop 1980). The NM emphasises the importance of the small food item taken in large numbers, excludes categories that cannot be enume- rated due to mastication and the digestive process and those that do not occur in discrete units such as macroalgae and detritus (Hyslop 1980). It has been suggested that over-estimation may be compensated by faster digestion and evacuation of small prey compared to large ones (Ahlbeck et al. 2012). The volume method gives a better indication of the nutritional contribution of each food item to the diet and reveals any specialization whilst the VM emphasises the large rare food items. To minimize the bias of each method, we use the Index of Relative importance (IRI) that combines the FO, NM and VM (Pinka et al. 1971). Diet overlap between O. niloticus and O. mortimeri, not before estimated, is then obtained using a modified percent similarity index, D (Schoener 1970) and the overlap coefficient Q (Pianka 1973). The ability of different diet overlap indices to detect simulated distribution was compared by Linton et al. (1981) who found that only Schoener’s D index was accurate over a large range of diet overlap. Most studies of diet-overlap in fish employ Schoener’s index (Khallaf & Alne-ne-ei 1987; Bacheler et al. 2004, Kahilainen et al. 2005) but diets of cichlids fishes in Lake Chivero, Zimbabwe were calculated using Pianka’s Q index (Zengeya &Marshall 2007). The aim of the present work is to evaluate the potential for food competition between the exotic O. niloticus and the indigenous O. mortimeri in Lake Kariba, by quantifying and comparing their diet composition, diet overlap, and digestion efficiency.

Materials and methods Fish were caught from the littoral zone of Lake Kariba up to a depth of 13 m, using an electrofisher (Smith-Root Inc, Type VI-A) during daytime in July and August 2003 and at night using gillnets in October 2003, December 2004 and in February and June 2005 (Figure 4.1). The gill-nets were set perpendicular to the shoreline, at depths ranging from half a metre down to 13 m deep. Each gill-net fleet comprised of twelve 40-m panels of 38, 51, 64, 76, 89, 102, 114, 127, 140, 152, 165 and 178 mm

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 60 Diet overlap between native and invasive congeners

Figure 4.1 Skematic map of the Lake Kariba Sanyati basin showing the areas where fish for the gut analysis were obtained (dark grey areas).

stretched mesh sizes. Fish used for estimation of the digestion efficiency were caught in the Nyaodza River bay by artisanal gill-net fishermen of Nyaodza Fishing Village (Figure 4.1). Standard length and body mass were recorded for each fish, after which the stomach was removed and stored in 70% alcohol, except for those to be used for determination of digestion efficiency, which were stored on ice until further processing.

Diet composition In the laboratory, the stomach contents were removed and placed in a measuring cylinder and diluted with distilled water and then placed in a counting chamber. A good dilution was when individual food items were clearly visible and could be measured. The counting chambers were 24 mm in diameter and 4 or 10 ml in volume. The stomach contents were allowed to settle for 24 hours before counting on an Olympus CK40 inverted microscope, at a magnification of × 400. Material that could not be identified as any organism and consisting of what appeared to be organic matter was classified as detritus, and inorganic granules as sand. The proportion of sand and detritus in the sample was estimated visually and the sample then examined to identify the different taxa. All food items within a minimum of ten microscope fields of view (1.52 mm2) were identified to genera using various references

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 61 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

(Needham & Needham 1962; Pesez & Pesez 1977; John 2000; Komárková-Legnerová & Cronberg 1994; Janse van Vuuren et al. 2006). Each food item was then scored against the list of taxa. The length and width/diameter of each food item was measured using a calibrated eyepiece graticule. For calibration, the size of each eyepiece graticule unit was measured on a slide micro-meter. The volume of each food item was estimated using equations of nearest shape and using the relationships between biovolume and the length, width and diameter of size groups of different species found in literature (Hillebrand et al. 1999; Olenina et al. 2006) and unpublished (Ronald Bijkerk, personal communication). The detritus and sand were not included in the volumetric and numerical estimates. For the analysis of diets, food items were placed into food categories according to genera. The stomach contents were analysed using three classic methods; percentage frequency of occurrence (FO), numerical relative abundance (NM) and the volumetric relative abundance methods (VM) (Hyslop 1980). The FO is the percentage of stomachs containing a particular food category out of the total number of stomachs analysed. The numerical method expresses the relative abundance of a particular food category of the total volume of all food categories. To prevent stomachs with many food items from influencing the estimate more than those with less food items, the mean proportion of each food category for each fish was calculated separately and then averaged over all the stomachs from each species. The VM was calculated in the same way as the numeric method. An index of relative importance (IRI) was calculated for each food item by the formula IRI = (%NM + %VM) × (%FO) (Pinkas et al. 1971). IRI was expressed as a proportion of the sum of the IRI values of all prey items (%IRI). The most important food item has the highest IRI. The %IRI ranges from 0 to 100% where zero indicates no overlap and 100% complete overlap. The proportion of the detritus and sand in the stomachs of O. niloticus and O. mortimeri were tested for similarity using ANOVA in SPSS version 18 statistical package. The proportion was arcsine transformed because percentage or proportions form a binomial rather than a normal distribution.

Diet overlap Diet overlap between O. niloticus and O. mortimeri was estimated using a modified similarity index D (Schoener 1970) and the overlap coefficient Q (Pianka 1973):

� ∑�𝑷𝑷�� 𝑷𝑷�� ∑ / 𝑷𝑷 �𝑷𝑷 /� � � ��∑ 𝑷𝑷�� ∑ 𝑷𝑷�� where n = number of food categories, Pij = proportion (% by volume) of food category

‘i’ in the diet of species ‘j’ and P = proportion (% by volume) of food category ‘i’ in

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 62 Diet overlap between native and invasive congeners

the diet of species ‘k’. Values of 0 or 1 for D or Q indicate zero, or full overlap respectively. The degree of overlap in diet between two species is considered significant when the index is greater than 0.60 (Wallace 1981). The similarity of the diet was further checked by plotting the proportions of one fish against another. The goodness of fit of the trendline (R2 value) was used to test difference in the diets. The Reduced Major Axis Regression (RMA) was used because both variables (the diets) were measures with error (Smith 2009). Contrary to the univariate regression where the errors are minimized along the vertical or y- axis, in RMA errors in both the y- and x-axis are minimized using the product of the y and x distance of the observation from the trend line. The trend line was fitted 4 using SPSS version 18 statistical package.

Relative digestion efficiency The stomach and hindgut contents were used to determine the relative protein digestive efficiency. Bowen (1981) and De Silva (1985) reported that in O. mossambicus assimilation does not occur in the stomach, so the stomach contents were used to represent the food ingested. The hindgut content was taken to represent faecal material. The stomach was cut open to remove contents and those of the hindgut were withdrawn from the fish by stripping with a needle and then dried at 70 °C to constant mass. Crude protein was determined using the Kjeldahl method (Bremner & Malvaney 1982). The digestive coefficient (DC), modified from Blackburn and Southgate (1981), for each fish was calculated as:

DC = (S – F) / S × 100

where DC is the digestive coefficient, S the amount (mg) of protein per mg of stomach contents and F is the amount (mg) of protein per mg of faeces.

Results A total of 51 O. niloticus and 48 O. mortimeri containing food items were used in the diet comparison. Of the stomachs collected using gill-nets, 17 out of 34 and 12 out of 28 O. niloticus and O. mortimeri stomachs, respectively, were empty, as compared to only two (n = 35) and zero (n = 32) from electrofishing, respectively. Because of the small sample sizes, the fish from both gears were pooled for the diet analysis. The mean stomach fullness was 70.9 ± 9.4 and 73.4 ± 9.8 for O. niloticus and O. mortimeri (mean ± SD), respectively, and the difference between species was not significant. Length of O. niloticus caught by electrofishing varied between 3.0 and 10.5 cm (mean = 6.4cm, SD = 0.6) and those by gillnet between 9.0 and 32.7 cm (mean = 21.6 cm, SD = 3.4). The range of O. mortimeri caught by electrofishing was 2.5 to 18.5 cm (mean = 6.5 cm, SD = 1.2) and those by gillnets 7.5 to 33.7 cm

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 63 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

(mean = 24.7, SD = 7.1). Four fish caught by electrofishing could not be identified as either O. niloticus or O mortimeri and were excluded from the analysis.

Diet composition Most O. niloticus (76.5%) and O. mortimeri (83.3%) consumed detritus. Almost as many O. niloticus (72.5%) as O. mortimeri (70.5%) had sand in their stomachs. Macrophytes were found in 37.3 and 31.3% of O. niloticus and O. mortimeri stomachs, respectively. The proportion of detritus was 37.1 ± 8.5 and 33.2 ± 8.5, while the proportion sand was 19.0 ± 6.9 and 16.6 ± 5.6 in O. niloticus and O. mortimeri stomachs, respectively (mean ± SD). There was no significant difference in the average

proportion of detritus (F1,79 = 2.21, p = 0.14, ANOVA) and sand (F1,79 = 0.05, p = 0.82, ANOVA) in the stomach content of both fish. One hundred and eight genera were identified among the food items, many of which were diatoms (Table 4.1).

Table 4.1 Diet class and genera of O. mortimeri and O. niloticus as percentage frequency of occur‐ rence (FO), numerical relative abundance (NM), volumetric relative abundance methods (VM) and index of relative importance (IRI). Class values are highlighted. (x = observed only; O. mortimeri stomachs: n = 48; O. niloticus stomachs: n = 51).

FO NM VM IRI FO NM VM IRI Class/Genera O. mortimeri O. niloticus Arcellinida 2.1 Difflugia x Asplanchnidae 2.1 Asplanchna 2.1 0 0 0 0 0 0 0 Bacillariophyceae 100 65 58 73 100 68 67 77 Achnanthes 10 0 0 0 12 0 0 0 Achnanthidium 10 0 0 0 6 0 0 0 Amphora 29 1 1 1 24 1 2 1 Asteroniella 0 0 0 0 10 0 0 0 Brachysira 4 0 0 0 6 0 0 0 Caloneis 0 0 0 0 8 1 1 0 Cocconeis 17 1 0 0 8 0 0 0 Cymatopleura 31 1 4 1 14 1 2 0 Cymbella 75 10 8 12 73 9 7 11 Diatoma 88 18 2 15 71 15 2 12 Encyonema 54 4 4 4 41 4 4 3 Entomoneis 0 0 0 0 4 0 0 0 Epithemia 50 2 1 1 26 1 0 0 Eunotia 21 1 0 0 10 0 0 0 Fallacia 2 0 0 0 2 0 0 0 Fragilaria 71 7 1 5 69 8 1 6 Gomphonema 92 7 2 7 86 6 2 7 Gyrosigma 13 0 1 0 20 1 3 1

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 64 Diet overlap between native and invasive congeners

FO NM VM IRI FO NM VM IRI Class/Genera O. mortimeri O. niloticus Navicula 60 4 2 3 65 6 3 6 Neidium 2 0 0 0 0 0 0 0 Nitzschia 13 0 0 0 16 1 0 0 Pinnularia 67 2 15 10 73 4 14 13 Rhopalodia 35 1 8 3 45 1 13 6 Sellaphora 2 0 0 0 0 0 0 0 Stauroneis 2 0 0 0 2 0 0 0 Surirella 69 6 9 9 43 7 9 7 Tryblionella 33 1 0 0 35 2 4 2 4 Brachionida 4 Keratella x Branchiopoda 4 Cladocera x Bosmina x Charophyceae 2 Chara x Chlorophyceae 56 4 6 3 49 3 5 2 Bulbochaete 0 0 0 0 8 0 0 0 Characium 0 0 0 0 4 0 0 0 Chlamydomonas 0 0 0 0 2 0 0 0 Chlorococcum 25 2 1 1 6 0 0 0 Coelastrum 4 0 0 0 0 0 0 0 Botrycoccus x Euastrum 0 0 0 0 2 0 0 0 Eudorina 0 0 0 0 4 0 0 0 Microspora 25 0 0 0 24 1 0 0 Oedogonium 23 0 5 1 16 1 4 1 Pediastrum 8 0 0 0 4 0 0 0 Scenedesmus 25 1 0 0 20 1 0 0 Ulothrix 6 0 0 0 6 0 0 0 Chrysophyceae 6 0 2 0 2 0 0 0 Mallomonas 6 0 2 0 2 0 0 0 Conjugatophyceae 48 2 2 1 39 3 2 1 Closterium 13 0 0 0 8 0 1 0 Cosmarium 27 0 0 0 24 0 0 0 Micrasterias 2 0 0 0 0 0 0 0 Mougeotia 8 0 0 0 20 1 1 0 Staurastrum 35 1 2 1 24 0 1 0 Zygnema 4 0 0 0 0 0 0 0 Coscinodiscophyceae 8 0 0 0 10 0 0 0 Melosira 6 0 0 0 8 0 0 0 Stephanodiscus 0 0 0 0 2 0 0 0 Cryptophyceae 0 0 0 0 4 0 0 0 Cryptomonas 0 0 0 0 4 0 0 0

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 65 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

FO NM VM IRI FO NM VM IRI Class/Genera O. mortimeri O. niloticus Cyanophyceae 50 6 7 4 59 7 6 5 Anabaena 27 2 1 1 33 2 1 1 Anabaena mat 2 0 0 0 0 0 0 0 Aphanizomenon 0 0 0 0 14 0 0 0 Aphanocapsa 8 0 1 0 0 0 0 0 Arthrospira 0 0 0 0 10 1 0 0 Calothrix 4 0 0 0 14 0 2 0 Chroococcus 6 0 0 0 6 0 0 0 Cyanodictyon 6 0 0 0 2 0 0 0 Cylindrospermum 4 0 0 0 4 0 0 0 Dactylococcopsis 2 1 0 0 0 0 0 0 Dichothrix 4 0 1 0 2 0 0 0 Gloeotrichia 6 0 0 0 8 0 0 0 Hydrocoleus 0 0 0 0 2 0 0 0 Hassalia 2 0 0 0 0 0 0 0 Johannesbaptistia 0 0 0 0 2 0 0 0 Leptopogon 0 0 0 0 2 0 0 0 Lyngbya 4 0 0 0 12 0 0 0 Merismopedia 10 0 0 0 10 0 0 0 Microcystis 23 1 1 0 8 1 0 0 Nodularia 6 0 0 0 8 0 0 0 Nostoc 4 0 0 0 4 0 0 0 Oscillatoria 10 0 3 0 10 0 2 0 Phormidium 4 0 0 0 8 1 0 0 Planktolyngbya 0 0 0 0 4 1 1 0 Planktothrix 8 0 0 0 20 0 0 0 Pseudanabaena 8 0 0 0 10 0 0 0 Rivularia 10 0 0 0 4 0 0 0 Snowella 2 0 0 0 4 0 0 0 Trichodesmium 0 0 0 0 2 0 0 0 Desmidiaceae 6 2 1 0 4 0 0 0 Microcoleus 6 2 1 0 0 0 0 0 Spaerozosma 0 0 0 0 4 0 0 0 Diatomaphyceae 60 4 3 2 47 2 2 1 Aulcoseira 44 1 1 1 28 1 1 1 Cyclotella 29 2 2 1 24 1 1 0 Dinophyceae 17 1 2 0 10 0 2 0 Ceratium 2 0 0 0 0 0 0 0 Gymnodinium 4 0 0 0 2 0 0 0 Peridinium 13 1 2 0 8 0 1 0 Peridinopsis 4 0 0 0 4 0 0 0 Euglenoidea 4 0 0 0 10 0 0 0 Euglena 4 0 0 0 8 0 0 0 Euglenophyceae 6 0 0 0 12 0 0 0

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 66 Diet overlap between native and invasive congeners

FO NM VM IRI FO NM VM IRI Class/Genera O. mortimeri O. niloticus Phacus 4 0 0 0 6 0 0 0 Strombomonas 2 0 0 0 0 0 0 0 Trachelomonas 4 0 0 0 4 0 0 0 Fragilariophyceae 94 13 17 16 94 13 13 14 Frustulia 81 5 6 8 59 6 4 6 Meridion 21 1 0 0 2 0 0 0 Synedra 85 7 11 13 69 6 9 11 Tabellaria 6 0 0 0 14 1 0 0 Maxillopoda 2 0 4 Copepoda x Nauplii 8 1 0 0 0 0 0 0 Synurophyceae 0 4 Synura 0 0 0 0 2 0 0 0 Oocystis 0 0 0 0 2 0 0 0 unknown Platyhelminthes 0 0 0 0 2 0 0 0 unknown 75 3 0 2 49 2 0 1 Xanthophyceae 0 0 0 0 4 1 0 0 Goniochloris 0 0 0 0 4 1 0 0 Zygnemophyceae 2 0 0 0 10 0 1 0 Spirogyra 4 0 0 0 10 0 1 0

Diatoms in the classes Bacillariophyceae, Diatomaphyceae and Fragilariophyceae occurred in all stomachs of both fish species and contributed most (83%) to the diet of O. niloticus by number and volume (Figure 4.2; Table 4.1) when excluding detritus, sand and macrophytes. In O. mortimeri, the contribution of diatoms was 82% by number and 78% by volume. For both O. niloticus and O. mortimeri the IRI of diatoms was the highest of all the other groups in the diet. The diatoms belonging to the genera Gomphonema, Cymbella, Pinnularia, Diatoma, Fragilaria, Synedra, Navicula and Frustulia occurred in more than half the stomachs of O. niloticus (Table 4.1). Surirella, Encyonema and Epithermia were also common diatom genera in the diet of O. mortimeri. Diatoma was the most numerous genus whilst Pinnularia contributed the highest to the proportion by volume in both species (Figures 4.3 and 4.4). Many food categories were rare and contributed little to the numbers and volume of the stomach contents. The most important taxa in the diet of O. niloticus according to the IRI were benthic epilithic genera Pinnularia (13%), Diatoma (12 %), Cymbella (11%), Synedra (11%) and Sirirella (7%). These taxa were also the top most important in the diet of O. mortimeri, though they differ in their order of importance. Diatoma (15%) was most important, followed by Synedra (13%), Cymbella (12%), Pinnularia (10%),

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 67 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

a) Frequency of occurance c) % Volume 100 100

80 80

60 60

40 % volume 40

20 20 % Frequency of occurance

0 0

b) % Number d) % Index of Relative importnce 100 100 O. niloticus O. niloticus O. mortimeri O. mortimeri 80 80

60 60

40 % Number 40

20 20 % Index of% Relative Index Importance

0 0 e e e e e e e e e e a e e e e e e e e e e e e e e a e e e e a a a a a a a a a a e a a a a a a a a a a a a a a e a a a a e e e e e e e e e e d e e e e e e e e e e e e e e id e e e e c c c c c c c c c c i c c c c c c c c c c c c c c c c c c o y y y y y y y y y o y y y y y y y y y y y a y y y y y y ia n h h h h h h h i h h n h h h h h h h h h h h h h h h h h p p p p p p p d p p e p p p p p p p p p p p id p p le p p p p i l o o o o o o o a o g o o o o o o o o o o o a o g o o o o i r t t m i i r s t c t n m n n i h r s c n n u n r h r a p m i u r t m o y a s p s m i t m a lo y s a s e a la l r i y a E le la n l r g i y e o E l l n e l h g r y e o D i e il h h r y t D i i h u d t g a n ju d g a n c j C D a g c C o C C D a u g C C n o C i u X g C n i a X g a n a y a n E r y o i D E r o i D B Z B c F Z C c F C s s o o C C Figure 4.2 The a) percentage frequency of occurrence (FO) b) % numerical relative abundance c) volumetric relative abundance and the d) % index of relative importance of the food classes consumed by O. niloticus and O. mortimeri.

Table 4.2 Percentage protein in the extract from the stomach and hindgut and the Apparent Digestion Coefficient (ADC) of O. mortimeri and O. niloticus (CI = 95% confidence interval).

% protein % protein Data type stomach hindgut ADC stomach hindgut ADC O. mortimeri O. niloticus Individual 30.3 28.4 6.2 27.7 23.4 15.6 Individual 21.6 15.8 26.7 22.1 8.4 62.0 Individual 22.1 3.5 84.2 22.6 16.8 26.0 Individual 43.1 21.2 50.8 38.7 19.0 50.9 Individual 36.6 25.9 29.2 21.9 11.9 45.7 Individual 39.1 16.4 57.9 37.5 15.3 59.3 Individual 35.8 21.9 38.6 48.9 29.8 39.2 Individual 17.4 14.1 19.0 23.8 11.3 52.6 Individual 19.7 7.5 61.9 10.0 7.1 28.8 Individual 12.4 3.5 71.9 Individual 16.0 10.8 32.4 Mean and CI 29.5 ± 7.3 17.2 ± 6.3 41.6 ± 18.7 25.6 ± 8.0 14.3 ± 5.1 44.0 ± 11.6

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 68 Diet overlap between native and invasive congeners

Figure 4.3 Relationship between numerical relative abundance (NM) and volumetric relative abundance methods (VM) of food categories occurring in more than 35 % of stomachs, consumed by a) O. niloticus and b) O. mortimeri.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 69 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

Figure 4.4 Relationship between the percentage frequency of occurrence (FO) and volumetric relative abundance methods of food categories occurring in more than 35 % of stomachs, consumed by a) O. niloticus and b) O. mortimeri.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 70 Diet overlap between native and invasive congeners

a) Frequency of occurance c) Volume method 100 25

80 20 i 60 15

40 10 O.mortimer O. mortimeri O.

20 5

0 0 0 20 40 60 80 100 0 5 10 15 20 25 O. niloticus O. niloticus 4 b) numerical method d) IRI 25 25

20 20

15 15

10 10 O. mortimeriO. O.mortimeri

5 5

0 0 0 5 10 15 20 25 0 5 10 15 20 25 O. niloticus O. niloticus

O. niloticus vs O. mortimeri isometric line linear regression Figure 4.5 Relationships between the importance of food items found in the stomachs of O. niloticus and O. mortimeri, as demonstrated by: a) the frequency of occurrence (R2 = 0.89, t = 28,1) b) the abundance (numerical method) (R2 = 0.92, t = 35.3) c) the volume (volumetric method) (R2 = 0.89, t = 29.2) and d) the index of relative importance (IRI) (R2 = 0.91, t = 32.5) (p < 0.0001 for all plots).

and Sirirella (9%). Zooplankton taxa such as copepods were rare and were only found in two stomachs, while the cladoceran Bosmina sp., the rotifer Keratella sp., an un- identified cladoceran and a flatworm (Platyhelminthes) were found only once. The importance of different food items based on relative frequency (FO), abundance (NM), volume (VM) and relative importance (IRI) all correlated strongly between O. niloticus and O. mortimeri (FO: R2 = 0.89, t = 28.1, p = 0.001; NM: R2 = 0.92, t = 35.3, p = 0.001; VM: R2 = 0.89, t = 29.2, p = 0.001; IRI: R2 = 0.91, t = 32.5, p = 0.001; Figure 4.5). Comparing the slope of the reduced major axis regression line of O. niloticus and O. mortimeri for the FO, NM, VM and IRI, shows that the differences in the diet were small (1.18, 1.10, 1.00 and 1.03 for FO, NM, VM and IRI, respectively).

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 71 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

Diet overlap and relative digestion efficiency The diets of Oreochromis niloticus and O. mortimeri overlapped significantly, with D and Q similarity indices of 0.75 and 0.95, respectively. These values are well above the biologically significant limit of 0.60 suggested by Wallace (1981). There was no significant difference in the amount of protein in the stomach and the intestine as well as in the digestion coefficient of the two species (stomach: t = 0.80, df = 18, p = 0.44; intestine: t = 0.83, df = 18, p = 0.42; digestion efficiency: t = 0.26, df = 18, p = 0.80). The mean quantities (± SD) of protein in extracts from the stomach and hindgut of O. niloticus were 25.6% ± 8.0 and 14.3% ± 10.8, respectively, whilst the mean DC was 44.0% ± 11.6 (Table 3.2). Comparable amounts were found in O. mortimeri (stomach: 29.5% ± 7.3; hindgut: 17.2% ± 13.7; DC: 41.6% ± 18.7).

Discussion Diet compositions of the introduced Oreochromis niloticus and the native O. mortimeri in Lake Kariba were very similar and dominated by benthic diatoms. Most O. niloticus (76.5%) and O. mortimeri (83.3%) consumed detritus. Both Schoener’s D (1970) and Pianka’s Q (1973) similarity indices, indicated an almost complete overlap in diet and the digestion efficiencies of the species were also highly comparable. This suggests that the species have very similar diet niches and that competition for food is likely if food resources are limited.

Diet composition Tilapiines are known to consume a wide range of food that include algae, detritus, macrophytes, zooplankton and aquatic insects; a typical generalist diet of Oreochromis niloticus (Lowe-McConnell 2000; Moriarty 1973; Trewavas 1983). Therefore, the diet of O. niloticus and O. mortimeri found in this study is typical. Confirming earlier findings in Lake Kariba, diatoms are common in the diet of both O. niloticus and O. mortimeri even though other algae species, macrophytes, zooplankton and detritus are consumed (Chifamba 1998). Matthes (unpublished) found that adults O. mortimeri in Lake Kariba mainly feed on algae gathered from substrate and that higher plants were incidentally ingested whereas juveniles (< 7 cm) were omnivorous and had a diet that included small invertebrates such as insect larvae and algae. There is not much published data on the diet of O. mortimeri, but from this study and the available literature its diet appears to be similar to that of O. niloticus, and mostly derived from scraping and suction at the bottom. Detritus nutrition is derived from the detritus itself, associated organisms which include epiphytic and benthic algae, and non-protein amino acids (Bowen 1982). Oreochromis mortimeri of different sizes feed by dredging the bottom rather than browsing (van der Lingen 1973). This feeding strategy may explain the high

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 72 Diet overlap between native and invasive congeners

proportion of stomachs that contained detritus, sand and benthic diatoms found in this and other studies. O. niloticus also ingests detritus and other benthic organisms (Peterson et al. 2006; Tudorancea et al. 1988). The most frequent items in the stomachs of O. niloticus caught in the Mississippi watershed were amorphous debris, detritus, sand grains and mud clumps (Peterson et al. 2006). Benthic diatoms such as Synedra, Cymbella, Neidium and Gomphonema blue greens were the main diet of juvenile O. niloticus (10 to 50 mm) in Lake Awassa, Ethiopia (Tudorancea et al. 1988). Adults (18 - 32 cm) in the same lake fed mostly on Chroococcus, Oscillatoria and Botryococcus (Getachew & Fernando 1989). Phytoplankton and periphyton are also important sources of energy and the diet of Oreochromis niloticus depends on what dietary items are available in a water body 4 (Huchette et al. 2000; Huchette & Beveridge 2003; Lowe-McConnell 1958; Trewavas 1983). In Lakes George and Turkana where phytoplankton is abundant, it is the dominant food compared to Lake Albert where the main food is epiphytic diatoms (Trewavas 1983). The diet of O. niloticus in Lake Kariba differs from that in Lake Chivero, Zimbabwe where the fish fed mainly on blue-greens (Ndebele 2003; Zengeya & Marshall 2007). Microcystis sp. and Melosira sp. were dominant, whilst Cyclotella and Pediastrum sp. were common (Ndebele 2003). Numerical dominance of the plankton community in Lake Chivero by Microcystis aeruginosa (64%) and Melosira sp. (19.3%) may have influenced the diet. Because of low nutrients concentrations in Lake Kariba, blue-green algal blooms were present only in rich rivers and estuaries (Cronberg 1997). A total of 40 genera of diatoms were found on the submerged macrophyte Valisneria aethiopica. The most abundant genera were Achnanthidium and Gomphonema, which made up 23.4 and 42.9% of the diatom count, respectively (Phiri et al. 2007). Diatoms are therefore an available food resource in Lake Kariba explaining their dominance in the diet of O. niloticus and O. mortimeri, as found in this and previous studies (Chifamba 1998). There is seasonality in algal composition with blue-greens dominating the algal community during the warm-rainy season and diatoms during the cold-dry season (Cronberg 1997). Significant seasonal difference in the abundance of Chroococcus and Oscillatoria was found in the stomach contents of O. niloticus caught in Lake Awasa Ethiopia (Getachew & Fernado 1989). Therefore, a seasonal analysis of diet might have revealed seasonality in the algal species in the diet. A diet shift attributed to environmental change was observed in O. niloticus in lakes Nabugabo and Victoria (Njiru et al. 2004; Bwanika et al. 2006). In these studies, omnivory was observed in lakes Nabugabo and Victoria where Nile perch (Lates niloticus) was introduced and had reduced the populations of haplochromine cichlids, and phytoplanktivory where L. niloticus was absent. Before L. niloticus was introduced in Lake Victoria, O. niloticus was herbivorous feeding mostly on algae but following the introduction of L. niloticus the diet diversified to include insects,

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 73 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

fish, algae and plant material. Hence, the diet of O. niloticus displays both spatial and temporal variation, a valuable trait for an invasive fish species. Different diet items eaten by O. mortimeri and O. niloticus require different feeding techniques for their acquisition (Trewavas 1983; Yamaoka 1991). Peri- phyton is obtained by scraping or rasping it from the substrate, and in the process large quantities of the substrate are also consumed (Yamaoka 1991). Oreochromis niloticus consumes phytoplankton by gulping water and collecting plankton by means of a mucus- trap mechanism (Trewavas 1983). Copepods as well as terrestrial and aquatic insects are caught by actively pursuing the prey whilst detritus is consumed in a pecking motion. Ability of Oreochromis spp to utilize a variety of feeding methods is related to the functional morphology of their mouths, that are intermediate in size compared to predominantly visually feeding cichlids and suction feeders (Beveridge & Baird 2000). The use of diverse feeding methods by Oreo- chromis niloticus has enabled diet plasticity, which is an important attribute in an invasive species.

Diet overlap Above a true overlap of 0.76 the Schoener index tends to underestimate the overlap, whilst the value from the Pianka equation is more accurate (Linton et al. 1981). This means that the higher overlap value from the Pianka equation may be closer to the true diet overlap of the two species. In Lake Chivero, both Oreochromis niloticus and O. macrochir fed mostly on blue-green algae (> 50%) and their diets, in all size classes, overlapped almost completely (Zengeya & Marshall 2007). Dietary overlap (Pianka 1973) coefficient for Lake Chivero was 0.98 (close to the value of 0.95 for O. niloticus and O. mortimeri in Lake Kariba), indicating an almost complete overlap in diet between the species. Having similar diet does not always prevent coexistence because factors such as feeding behaviour, feeding site and habitat may reduce competition (Yamaoka 1991). In its native habitat in a River Nile canal in the Egyptian delta, the diet of O. niloticus overlapped significantly with that of Tilapia zillii (Gervias), particularly in the age groups 10 – 19.9 cm (Khallaf & Alne-na-ei 1987). The main food for O. niloticus was macrophytes (97%) compared to T. zillii (92%). Competition between the species may be reduced by specialization of T. zillii on aquatic insects. The diet of three cichilids, Orochromis andersonii, O. macrochir and O. sparmanii, was dominated by vegetative detritus and yet these fish species cohabited the Upper Zambezi River and floodplain (Winemiller & Kelso-Winemiller 2003). Yamaoka (1991) reviewed studies where commensal and mutualistic relationships may have allowed coexistence of cichlid species in African Great Lakes. Mechanisms for sharing food in species recently been placed in the same environment may be lacking, leading to displacement of one of the species with competitive disadvantage.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 74 Diet overlap between native and invasive congeners

There are conditions under which competition is likely, such as the distribution of the food resource in space and time (Milinski & Parker 1991). Even though it has been suggested that a diet of detritus may not be limited in the environment, its quality has been shown to vary in space (Bowen 1979, 1980, 1981). Bowen (1979) found that juveniles of Sarotherodon mossambicus that feed in the shallow water in Lake Sibaya, South Africa, have access to detritus with higher protein level than adult fish that inhabit and feed in deep water. This leads to malnourishment of adults whilst the juveniles had good growth rates. Such spatial differences in resource introduce the possibility for contest for optimal feeding sites. Though the spatial distribution of food in Lake Kariba was not mapped, a spatial difference in the distribution of O. niloticus and O. mortimeri was observed indicating displacement 4 of the native species (Chifamba 2006).

Digestion efficiency Protein content and digestion efficiency could be a factor that confers competitive advantage to O. niloticus over O. mortimeri. It has been shown that protein content is the factor limiting the growth of herbivorous and detritivorous fish. Whilst the prey of carnivorous fish contain > 80% protein, the protein content in the food of tilapias ranges from < 1% to 50% only (Bowen 1980, 1982; Bowen et al. 1995). Protein content in the stomachs of O. niloticus (mean = 25.6% ± 8.0) and O. mortimeri (mean = 29.5% ± 7.3) measured in this study fell within this range. The mean digestive coefficient of proteins for O. niloticus in Kariba (44.0% ± 11.6) was lower than that found in Lake Chivero (62% ± 9.1) but that of O. macrochir in Lake Chivero was even lower (39.4% ± 9.1; Marufu & Chifamba 2013). Differences between the lakes may have arisen from the differences in food sources; whereas the difference between O. macrochir and O. niloticus may be attributed to the superior digestion capability of blue-green algae by O. niloticus (Moriarty 1973). This study found no significant difference in the DC of O. niloticus (44.0% ± 11.6) and O. mortimeri (41.6% ± 18.7). Therefore, differences in the species competitiveness is not likely to be due to protein content of the diet or the digestion efficiency. Furthermore, the stomach fullness was high for both O. niloticus (70.9% ± 9.4) and O. mortimeri (73.4% ± 9), which suggests that neither species has an inadequate diet as a result of feeding on similar food items. Oreochromis niloticus seems to have competitive superiority to congeneric species wherever it was introduced in Africa, and similarity in diet seems to be a common factor (Balirwa 1992; Zengeya & Marshall 2008). In Lake Victoria, it displaced Oreochromis esculentus (Graham) and Oreochromis variabilis (Boulenger), and other native cyprinids in Lake Luhondo (Balirwa 1992,). Wherever it was introduced in Zimbabwe, O. niloticus became the dominant cichlid species by replacing O. macrochir, O. mossambicus and O. mortimeri (Marshall 1999; Chifamba 2006). Even though similarity in diet is a common factor among these species it may

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 75 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

not be the only factor determining the outcome of the competition. Other traits such as fast growth rate and aggression levels might be of importance (Chapter 3 – Chifamba & Videler 2014; Chapter 5 – Chifamba & Mauru 2017).

Implication to fish introductions Diet similarity between O. mortimeri and O. niloticus demonstrates that the introduced O. niloticus does not occupy a new niche in the Lake Kariba ecosystem. Consequently, fish production was not enhanced by introducing this alien species. Therefore, we recommend that introduction of other fish species into Lake Kariba or O. niloticus into other waterbodies should be preceded by examination of niches that are not occupied by the native species to avoid their displacement. Since it is not possible to reverse the consequence of O. niloticus introduction, management must focus on the sustainable exploitation of this fish. In addition, refugia populations of O. mortimeri need to be identified and protected to prevent loss of biodiversity.

Acknowledgements We are grateful to the crew of MV Erika, the staff of University Lake Kariba Research Station and of Lake Kariba Fisheries Research Institute, especially the late Mr Mushaike and Mr Chisaka. We thank Mr Ronald Bijkerk for providing tables of plankton biovolumes used in estimating biovolume. Mr Edwin Tambara assisted in the processing of the fish samples. Collection of gillnet samples was funded by the International Foundation for Science (IFS) Grant A/3159-1. Electrofishing samples were obtained during a VLIR scientific expedition.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 76 Comparative aggression and dominance of Oreochromis niloticus (Linnaeus, 1758) and Oreochromis mortimeri (Trewavas, 1966) from paired contest in aquaria

Portia C. Chifamba Tendai Mauru Chapter 5

Published in 2017 Hydrobiologia 788: 193‐203

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 77

Abstract

Oreochromis niloticus was introduced in Lake Kariba, and it displaced the endemic Oreochromis mortimeri from many areas of the lake. Studying the inter-

action between the two species sheds light on the nature of the displacement process. The levels of aggression within and between the two species and the effect of relative size were studied in the laboratory. O. mortimeri attacked O. niloticus first in most of the encounters whether it was the bigger or smaller of the pair. In encounters where O. niloticus was smaller, 2 (8.3%) O. niloticus made the first bite and 8 (40.0%) made the first bite when O. niloticus was the

bigger of the pair. Over a 30-min encounter, O. mortimeri was dominant and delivered significantly more bites (7.79 ± 2.31 bites) than O. niloticus (4.53 ± 1.53 bites) (p = 0.03, t = 2.18). Unlike O. mortimeri, O. niloticus attacked first only when it was considerably bigger than the opponent. The association of large body size with higher aggression may mean that O. niloticus, which grows faster and larger than O. mortimeri, has size advantage. Therefore, interaction between the two species may be complex, and aggression may be just one of the factors that affect the interaction of these fish species.

Nests of Nile tilapia in a drained reservoir at a fish farm in Kariba.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 78 Aggression and dominance of native and invasive congeners in aquaria

Introduction A cichlid endemic to the Zambezi system, Oreochromis mortimeri (Trewavas 1966) became abundant in Lake Kariba soon after its formation on the Zambezi River in 1957 for the purpose of generating hydroelectricity. In an unpublished report (1959), only 0.75% of the fish caught were O. mortimeri, and by 1962, the catch had risen to 35% of the total (Kenmuir 1984). In the 1980s, accidental introductions of Oreo- chromis niloticus (Linnaeus 1758) through escapees from fish ponds in Kariba resulted in the establishment of O. niloticus in the lake by 1993 (Chifamba 1998). With the introduction of O. niloticus, the once abundant O. mortimeri (Trewavas 1966) declined and disappeared in many parts of the Lake Kariba (Zengeya & Marshall 2008). Therefore, it was essential to investigate the attributes that may have imparted competitive advantage to the exotic species, enabling it to dominate over 5 the native fish species. Various studies suggest that the displacement of an indigenous species by an introduced species may be associated with similarities between the exotic and indigenous fish species. Similarity in diet and reproductive behaviour can stimulate aggressive behaviour. Both species feed on algae, organic detritus, plant material, insects and zooplankton, varying with availability (Chifamba 1998; Mhlanga 2000). In both species, males construct large nests in arenas which they defend (Marshall 2011). A female visits the arena and, after courtship that lasts a few minutes, lays eggs in the nest and after fertilization takes the eggs into her mouth for brooding. Fry are released by the mother into nursery areas, in the shallow inshore waters. Therefore, O. niloticus and O. mortimeri share food as well as nesting and nursery areas, which are contestable resources. An introduced species can displace the native species through interspecific competition for space and food, intraguild predation and agonistic interaction (Taniguchi et al. 2002; Martin et al. 2010; Sanches et al. 2012; Kakareko et al. 2013). A native fish of the estuaries of the Gulf of Mexico, the redspotted sunfish, Lepomis miniatus (DS Jordan 1877), was displaced from the preferred structured habitats by the agonistic behaviour of the introduced O. niloticus, thus exposing it to higher predation. In Brazil, O. niloticus was also shown to be more aggressive than the native pearl cichlid, Geophagus brasiliensis (Qouy & Gaimard 1824), with which it has an overlapping ecological niche (Sanches et al. 2012). Relative aggressive tendencies of O. niloticus and O. mortimeri which were not known yet, might explain the competitive advantage of the former, and hence the need for the study. Dominance status in fish is associated with growth rate (Abbott& Dill 1989; Tiira et al. 2009). Given the same amount of food, ten of twelve dominant steelhead trout, Salmo gairdneri (Richardson 1836), grew faster than their paired subordinates (Abbott & Dill 1989). Tiira et al. (2009) found that dominance status affected growth

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 79 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

in brown trout, Salmo trutta (Linneaus 1758) populations, with individuals of the lowest ranks growing less compared to those of a higher rank. Further, Alvarenga & Volpato (1995) noted a significant association between some agonistic profiles and metabolism. Reproductive success is also higher in dominant fish because they occupy the best nesting sites which attract mates (Philippart & Ruwet 1982; Seppänen et al. 2009). For example, breeding males in African cichlids make and fiercely defend nests in an arena. The best nests are suitably positioned for visits by females and are occupied by dominant males (Phillippart & Ruwet 1982). Compared to subordinate males, dominant male O. niloticus had the highest gonadosomatic index and higher levels of gonadotropins hormones that trigger spermatogenesis, whereas the subordinate had reduced gene expression of key factors for steroid production (Pfemmig et al. 2012), thus compromising the latter’s reproductive capacity. Aggression and dominance could have thus conferred a competitive advantage to the invader, O. niloticus, enabling it to displace O. mortimeri. In their seminal paper on cichlid aggression behaviour, Baerends & Baerends-van Roon (1950) present ethograms for several species. The various acts of aggression generally include charging, chasing, biting and displaying. Aggression levels can be mild, consisting of threats or explicit, when the fish bite each other. In an established dominance relationship, the fish no longer participate in simultaneous or reciprocal threatening. Instead, the subordinate fish displays fleeing and escaping behaviour when approached by the dominant; the dominant chases and bites the subordinate (Miklosi et al. 1995; Oliveira & Almada 1996a). We hypothesized that aggressive behaviour is one mechanism O. niloticus used to displace the native species O. mortimeri in Lake Kariba. The study investigated the aggression interaction of the two species in order to evaluate the role of aggression and dominance on the competitive advantage to O. niloticus over O. mortimeri.

Materials and methods Live specimens of the two fish species, O. niloticus and O. mortimeri, were captured from Lake Kariba in 2005 and 2007 (more than 10 years after O. niloticus were known to have been established in the lake) and kept separately for more than a month in fish tanks, prior to the experiments. The water was aerated and maintained at temperature between 21.5 and 25.5 0C which was within the range of surface temperatures, between 21 and 28 0C, in Lake Kariba (Chifamba 2000). The fish were fed commercial pellets, to satiation, twice a day. Fish used in the experiments were sub-adults: 6.9 – 14.9 and 9.3 – 14.3 cm long for O. niloticus and O. mortimeri, respectively. In Lake Kariba, O. niloticus grows to an asymptotic length of 44.6 cm and O. mortimeri 36.8 cm (Chapter 3 – Chifamba

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 80 Aggression and dominance of native and invasive congeners in aquaria

&Videler 2014). Only those fish without breeding colouration were used in the experiment because breeding behaviour would contribute to observed agonistic behaviour. Males, for example, become more aggressive when breeding and defending a breeding site (Marshall 2011). Three sets of experiments were conducted in a glass aquarium measuring 60 × 20 × 30 cm.

Experiment 1: aggression assessment of O. mortimeri The objective was to assess the reaction of O. mortimeri to O. niloticus of different size and weight. The weight and total length were measured at the beginning of each trial, and pairs with matched size and weight were classified as same. The difference in weight and length of a pair was expressed as a proportion (%) of the length and weight of each fish of the pair. Four O. mortimeri were each exposed to twelve different O. niloticus of varying size and weight. Oreochromis niloticus used in the 5 experiment weighed 28 – 107 g and were 7.9 – 14.2 cm long, whilst the four O. mortimeri were 36 – 66 g and 10.5 – 14.3 cm, respectively. These fish were scooped out of the holding aquarium, one at a time, to obtain O. niloticus smaller than, bigger than, or the same size as the O. mortimeri with whom they were confronted. As such, there were pairs where O. niloticus was bigger than, smaller than, or the same size as the O. mortimeri. With each fish pair placed in the experimental tank at the same time, aggression was recorded for 10 min, starting from the time when the pair was placed in the aquarium. Acts of aggression used for scoring are similar to those identified by Baerends & Baerends-van Roon (1950). In this experiment, aggression was scored as the total number of attacks per fish, where a scored attack for each fish was any of the following: chasing, biting, mouth wrestling or side swiping. A bite was scored when the mouth of a fish touched that of the opponent irrespective of whether the fish was initiating or retaliating an offence.

Experiment 2: intraspecific aggression (biting) The objective of the second experiment was to assess the level of aggression within a species. Sixteen pairs of unsexed O. niloticus from 17.4 to 60.2 g (6.9 – 13.9 cm length) and twelve of O. mortimeri from 22.9 to 77.0 g (9.3 – 16.0 cm standard length), respectively, were used. The same procedure of staging the confrontation was followed as in Experiment 1, except that each fish was confronted with another of the same species. During the experiment, each individual fish in a pair could be distinguished by size and colouration. Scoring was based on the number of bites made by each fish from the onset of the encounter. The bites were recorded for 30 min at 1-min intervals during the encounter.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 81 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

Figure 5.1 The mean number of attacks by O. niloticus and O. mortimeri of different relative a) weights and b) lengths, and the mean number of attacks of O. niloticus and O. mortimeri in pairs where O. mortimeri was bigger, O. niloticus was bigger and were they were similar in c) weights and d) lengths. Bars are standard error of means.

Experiment 3: interspecific aggression biting and dominance The aggression level between O. niloticus and O. mortimeri was compared using 64 pairs, each containing both species. The length (cm) of each fish was recorded, and the sex determined by examining the genital papilla. Oreochromis niloticus used in the experiment weighed 35.8 – 70.4 g and were 10.0 – 12.5 cm long, respectively. The ranges of O. mortimeri were 40.1 – 67.8 g and 10.5 – 14.0 cm. The level of aggression was scored as the number of bites made by each fish, at 1-min intervals during the encounter. Similar to intraspecific aggression, scoring was based on only the number of bites because these were easy to distinguish without ambiguity compared to attacks. The first bite was taken to signify the start of the aggressive encounter. The first bite, the time it was executed, and the fish that made it were noted and recorded for all pairs. At the end of the 30 min, the dominant fish was also noted. As in studies by Oliveira & Almada (1996a) and Corrêa et al. (2003), dominance was defined as

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 82 Aggression and dominance of native and invasive congeners in aquaria

when, within the pair, the fish no longer attacked or threatened each other, or when the ‘subordinate’ fled when approached by the ‘dominant’.

Statistical analysis The mean number of attacks or bites, time of first bite and dominance of O. niloticus and O. mortimeri were tested for similarity using t-tests against a probability level of p = 0.05. The effect of species, sex and the difference in weight were statistically analysed simultaneously in a multivariable General Linear Model, using SPSS.

Results Experiment 1: aggression assessment of O. mortimeri Different trends in the number of attacks according to differences in relative weight and length were observed in the two species. Overall, O. mortimeri attacked signifi- 5 cantly more times than O. niloticus (p = 0.035, t = 2.19, n = 52). Both species attacked more when one species was bigger than the other in a pair (Figure 5.1). Also, the number of attacks increased as the difference in size decreased to zero in O. mortimeri, especially when compared according to weight rather than to length (Figure 5.1a, b). However, there were almost no attacks when O. niloticus was the smaller of the pair, and the attacks were highest when it was bigger (Figure 5.1c, d).

Experiment 2: intraspecific aggression (biting) The intraspecific aggression experiment revealed different levels of within-species aggression. The bites scored confirmed the higher aggression of O. mortimeri (Figure 5.2a, b). When statistical tests were conducted on results obtained from pairs of the same species, the mean number of bites of O. niloticus was again significantly lower than that of O. mortimeri (p = 0.025, t = -2.30, n = 56; Table 5.1).

Experiment 3: interspecific aggression (biting and dominance) Different aggression trends according to size and species were recorded in this experiment. Body size determined the number of bites with the larger individuals mostly at an advantage (Figure 5.2c, d). There was a tendency for O. niloticus biting only when it was the larger in the pair (Table 5.1). On the other hand, whether larger or smaller in the pair, O. mortimeri bit the opponent almost the same amount of times, whereas for O. niloticus, relative size mattered more. When larger, O. mortimeri bit an average of 7.25 versus 7.32 times when smaller; yet, O. niloticus, when larger, bit 7.17 times and only 2.31 times when smaller (Table 5.1). The mean number of bites by O. mortimeri was higher than that of O. niloticus even when O. niloticus was the bigger of the pair (Figure 5.3a). Disregarding differences in size, O. mortimeri attacked significantly more times than O. niloticus (p = 0.03, t = 2.18, n = 128).

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 83 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

Figure 5.2 Number of bites from O. niloticus and O. mortimeri against the relative difference in size, in experimental setup: a) O. niloticus alone, b) O. mortimeri alone, c) O. niloticus interspecific and d) O. mortimeri interspecific (Dashed lines denote % difference 10 units class means).

Table 5.1 Mean number of bites, number of pairs and 95% Confidence interval (CI) of bites from each species, in different pairing of species and size, in the intra‐ and interspecific experiments.

Species Species paired Relative size Mean bites n Lower CI Upper CI O. niloticus O. niloticus & O. niloticus Larger 5.75 16 3.46 7.92 Smaller 3.31 16 1.47 5.46 All 4.53 32 3.04 6.03 O. mortimeri O.mortimeri & O. mortimeri Larger 9.67 12 5.67 14.18 Smaller 5.92 12 3.70 8.50 All 7.79 24 5.52 10.50 O. mortimeri O. niloticus & O. mortimeri Larger 7.43 28 5.23 9.60 Smaller 2.25 36 0.38 4.12 All 4.52 64 3.99 6.04 O. mortimeri O. mortimeri & O. niloticus Larger 7.25 36 5.14 9.36 Smaller 7.32 28 4.46 10.80 All 7.28 64 5.61 8.96

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 84 Aggression and dominance of native and invasive congeners in aquaria

Figure 5.3 The mean number of attacks by O. niloticus (N) and O. mortimeri (M) in different sex combinations and where either O. mortimeri or O. niloticus was bigger: a) all data, b) O. niloticus is female and O. mortimeri is male, c) both are male and d O. niloticus is male and O. mortimeri is female. Bars are standard error of means.

In mixed sex contests, female O. niloticus bit O. mortimeri more than those in other combinations (Figure 5.3b, d). No similar difference occurred when an O. niloticus male was paired with an O. mortimeri female. When both species were male, the smaller O. niloticus bit O. mortimeri the fewest times (0.5 times). The mean time to first bite from O. niloticus was 13.1 min which was significantly higher than that of O. mortimeri at 9.6 min (p = 0.04; t = -2.12, n = 70). There was a delay in biting when the fish were about the same size compared to when the size difference was large (Figure 5.4). For O. niloticus, the time at first bite decreased exponentially with positive difference in fish weight (R2 = 0.546; p = 0.000). The other trends in Figure 5.4 are not significant, even though for O. mortimeri there is a similar downward trend in the time at first bite with positive difference in weight. Fewer O. niloticus than O. mortimeri made the first bite when paired with either a much smaller or much larger O. mortimeri. In encounters where O. niloticus was larger, eight (40%) made the first bite compared to two (8.3%) when O. niloticus was the smaller. In all the encounters, the worst injury was the loss of a few scales, which rarely occurred.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 85 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

Figure 5.4 The time elapsed from start of encounter at which each fish attacked against the percentage weight difference in the pair for a) O. niloticus and b) O. mortimeri. Lines are linear and exponential trends time at first bite for negative and positive differences in length or weight.

Attacks ended when the dominant fish chased the subordinate fish and swam round the whole aquarium tank, while the latter used a small section of the tank, hardly moving away from that position. Except in three cases, dominance was established by the end of the observation period. Figure 5.5 shows that most of the time O. mortimeri was dominant when it was bigger compared to O. niloticus. Eleven smaller O. mortimeri became dominant when compared to one smaller O. niloticus. The species, more so difference in weight, significantly affected the number of bites and dominance status (Table 5.2). Judging on the p-values, these differences were more pronounced for dominance compared to the number of bites.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 86 Aggression and dominance of native and invasive congeners in aquaria

Figure 5.5 The dominance status of a pair of O. niloticus and O. mortimeri of different weight for each encounter. The line represents similar weight below which O. niloticus is bigger and above smaller.

Table 5.2 Statistical results from the Generalized Linear Model to test whether difference in weight, sex or species significantly affected the number of bites and dominance.

Source Type III sum of squares df Mean square F p (95%) Corrected model Bites (30 min) 655.7 3 218.6 5.7 0.001 Dominance 15.1 3 5.0 10.1 0.000 Intercept Bites (30 min) 492.7 1 492.7 12.7 0.001 Dominance 2.4 1 2.4 4.8 0.030 Weight difference (%) Bites (30 min) 337.0 1 337.0 8.7 0.004 Dominance 9.5 1 9.5 19.1 0.000 Sex Bites (30 min) 43.5 1 43.5 1.1 0.291 Dominance 0.0 1 0.0 0.1 0.756 Species Bites (30 min) 163.9 1 163.9 4.2 0.042 Dominance 3.1 1 3.1 6.3 0.013

Discussion Oreochromis niloticus was found to be less aggressive than the native O. mortimeri, contrary to expectations based on that the former species displaced the latter from Lake Kariba. Taniguchi et al. (2002), Martin et al. (2010), Sanches et al. (2012) and Kakareko et al. (2013) explained how an introduced species can displace a native 87

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 87 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

species through interspecific competition for space and food, intraguild predation and agonistic interaction. For instance, a native fish of the estuaries of the Gulf of Mexico, redspotted sunfish (Lepomis miniatus), was displaced by the agonistic behaviour of the introduced O. niloticus from preferred structured habitats to exposed ones, where it might have suffered higher predation (Martin et al. 2010). The level of aggression displayed by O. niloticus may depend on the species it confronts or other factors such as size for advantage in an encounter. The level of aggression and dominance can also be predicted from size, with larger fish exhibiting more aggression and often becoming dominant as shown in this and other studies (Turner & Huntingford 1986; Neat et al. 1998a; Cutts et al. 1999). Larger opponents won 35 of the 38 encounters of Oreochromis mossambicus (WKH Peters 1952) pairs in the study by Turner & Huntingford (1986). They also reported a weak and negative relationship between relative body size and the intensity of a fight. In our study, there was a tendency for the number of bites to increase with increase in size asymmetry. The time of the first attack by O. niloticus was shorter at high size asymmetry than at the lower end, demonstrating the tendency for large fish to be aggressive. Neat et al. (1998b) demonstrated that high size asymmetry is associated with greater cost because a large fish has more strength and ability to inflict more harm or cost to a smaller opponent through dislodging of scales and high accumulation of lactic acid in the muscles. This means that it is more costly for the smaller fish to contest against a larger fish when size asymmetry is high and domination of the smaller fish more possible. It also means that O. niloticus which grows faster and larger than O. mortimeri (Chapter 3 – Chifamba & Videler 2014) may have size advantage. Faster growth means that O. niloticus at any given age has a size advantage which we have shown to be significantly associated with number of bites and dominance. Being large, aggressive and dominant can determine growth and ability to acquire quality territories (Koebele 1985; Abbott & Dill 1989; Cutts et al. 1999). Baras & Lucas (2010) found a significant, positive relationship between individual growth and aggression or boldness in a study of individual growth trajectories of sibling Brycon moorei. (Steindachner 1878) raised in isolation since egg stage. Growth in aggressive fish can be a result of high food intake and growth efficiency (Carline & Hall 1973; Li & Brocksen 1977; Ejike & Schreck 1980). Dominant O. niloticus males had higher expression of factors important for steroid production, and gonadotropin that triggers spermatogenesis (Pfennig et al. 2012). Being aggressive and dominant must mean that O. mortimeri had an advantage over O. niloticus through better access to food and territories, which are important for nest building and mating. Although investment in a nest was related to female mate choice in O. niloticus, deprivation of nest, achieved by not providing gravel substrate, did not affect mating success (Mendonça & Gonçalves-de-Freitas 2008). Furthermore, aggression and confrontation are not always advantageous because it comes at an energetic cost for both winners and losers (Neat et al. 1998b). An

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 88 Aggression and dominance of native and invasive congeners in aquaria

example of such cost was reported by Turner (1986) where dominant male O. mossambicus that held and defended territories in a tank, displayed less growth than the subordinate males. Therefore, high aggression may have an energetic cost for O. mortimeri. Sex and the stage in the development of gonads affect the level of aggression in cichlids (Oliveira & Almada 1996b; Neat et al. 1998a). Smaller size Tilapia zillii (Gervais 1848) were more aggressive when gonads were at an advanced stage of development (Neat et al. 1998a). Sexual dimorphism in aggression occurs in a closely related cichlid, O. mossambicus, where the males display more agonistic behaviour than the females (Oliveira & Almada 1996b). In nature, there is a general preference for the own species as shown in mangrove killifish, Kryptolebias marmoratus (Poey 1880) — preferential association and reduced aggression towards members of their own genotype, compared to members of a different genotype 5 (Edenbrow & Croft 2012). This may explain the sex difference observed in this study. The laboratory techniques used in this study of aggression limit the application of the findings to nature, due to the unnatural conditions the fish were subjected to during the experiment. Experimental studies using different techniques either in the field or in the laboratory, are essential in the acquisition of knowledge of fish behaviour (Rowland 1999). Laboratory experiments enable the manipulation of the independent variables to test hypotheses and to elucidate the cause and effect relationships as done in this study. This manipulation of the environmental conditions in the laboratory might affect the results. Hence, the results in this behaviour study might not directly apply in the field, and this needs to be borne in mind when interpreting the results. Direct field observations ought to be done to increase our understanding of the interactions of O. mortimeri and O. niloticus. Almeida & Grossman (2012) reviewed studies where the direct methods were used and recommended them for studying the interaction between invasive and native species.

Conclusion This research gives insight on the agonistic interaction between the introduced O. niloticus and the native O. mortimeri. Although O. niloticus is less aggressive, this seeming disadvantage may be compensated by being larger at a given age than O. mortimeri. Because O. niloticus was not dominant, aggression may not have been the most important factor in the displacement of O. mortimeri in Lake Kariba. The interaction between these two species is likely to be complex, involving a number of possible mechanisms that caused the disappearance of the native species that ought to be investigated. There is also need for research work on the effect of aggressive interaction on reproduction and feeding, hence fitness.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 89 PART I – BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS

Acknowledgements We would like express gratitude to Lake Kariba Fisheries Research Institute and the Department of Biological Sciences of the University of Zimbabwe in Harare where this research was carried out. We are also thankful to all who helped in the capture and keeping of the live fish, and in particular, the late Mr Joel Chisaka and Mrs Margaret Gariromo. Much thanks to Dr. J.H. Wanink, Prof. J.J. Videler, Prof. C.H.D. Magadza and Ms Audrey Chifamba, who provided invaluable assistance in the in preparation of this paper.

Compliance with ethical standards Ethical approval This research was approved by the Animal Research Ethics Committee in the Department of Livestock and Veterinary Services, Ministry of Agriculture, Mechanization and Irrigation Development. All procedures used in the research complied with the laws set out by the Animal Research Ethics Committee.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 90 PART II

BIOLOGY AND IMPACTS OF LIMNOTHRISSA MIODON

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 91

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 92 Developing a sustainable pelagic fishery in an African reservoir: trends in the catches of the introduced freshwater sardine Limnothrissa miodon and associated species in Lake Kariba, Zimbabwe

Portia C. Chifamba Han Olff

Chapter 6

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 93 Abstract The formation of reservoirs in former tropical riverine ecosystems is associated with introductions of pelagic species to develop fisheries, but with varying long- term success and unclear impact on native inshore (formerly riverine) species. The freshwater sardine, Limnothrissa miodon was introduced into the Lake Kariba reservoir in the late 1960s to develop a pelagic fishery in this reservoir. Fishing started in 1974 and catches peaked in 1990 but declined steadily thereafter, and with that also the socio-economic benefits from the fishery in this lake. This paper examines the trends in catch and temperature data in order to understand these trends and better manage this fishery. A multiple regression analysis shows that fishing effort and lake level explains the variation in the total catch whilst fishing effort and maximum temperature explained variation in catch per unit effort (as indicator of biomass). As the catches were at and above estimated Maximum Sustainable Yield (MSY), the collapse of the sardine fishery may be explained by overfishing. A combination of high fishing pressure with less suitable ecological conditions due to warmer water may together explain the declining harvest. Understanding the life-history characteristics of L. miodon highlights aspects of its vulnerability to fishing. Limnothrissa miodon inhabits both the pelagic and inshore areas. Juveniles are found in the marginal area only, while adults use both the pelagic and shallow areas. Large sardines caught in waters less than 30 m deep may stay close to the margin to feed on their own juveniles, perhaps also to breed. Overexploitation of young adult fish by the fishery might be compromising recruitment by reducing the number of breeding fish, and the number of large fish that carry relatively larger numbers of eggs compared to the young adults. The temporal rise and fall of the L. miodon catch was synchronous with that of its predator, the tigerfish Hydrocynus vittatus. Other inshore fish in addition to tigerfish are caught in the sardine fishery, indicating their tendency to venture into deep water to feed. Bycatch of juveniles of inshore species of commercially importance was low (0.85% in weight), hence capture was unlike- ly to have a large impact on the gill-net fishery on these species. To ensure sustainability of the sardine fishery both fishing effort and environment need to be considered in the management of the sardine fishery. Fishing effort needs to be reduced to sustainable levels through constant monitoring of the size of the sardine population. It is also critical that a detailed ecosystem production model of Lake Kariba is developed that can inform sustainable fisheries.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 94 Developing a sustainable pelagic fishery in an African reservoir

Introduction The formation of reservoirs for hydroelectricity, irrigation and drinking water in areas where previously rivers were the main freshwater habitat, has major consequences for their fish fauna. In general, riverine species will not be adapted to the deep waters of such reservoirs, leaving open niches (Fernando & Holčík 1991; van Zwieten et al. 2011)), limiting opportunities to also use reservoirs for fishing. In several cases, this has led to species introductions to develop new fisheries in newly formed reservoirs (De Silva & Sirisena 1987; van Zwieten et al. 2011). However, such introductions have varied in success, calling for more insight in the processes underlying changing catches (Fernando & Holčík 1991). The Lake Kariba dam on the Zambezi River was completed in 1958 and the lake filled to capacity in 1963. At 5 820 km2 surface area it was then the largest man- made reservoir (Coche 1974). It is made up of five hydrological basins separated by narrows or chains of islands (Figure 6.1). Uppermost basins, the Mlibizi and Binga 6 basins are small and narrow and the Zambezi River influences the characteristics of these basins. They have riverine conditions for most of the year. The other three basins, Sengwa, Chalala and Sanyati, or Kariba basins, are more lacustrine (Coche 1974).

Figure 6.1 Map of Lake Kariba showing the Lake basins.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 95 PART II – BIOLOGY AND IMPACTS OF LIMNOTHRISSA MIODON

Before the formation of the lake, the Middle Zambezi River was inhabited by at least 31 riverine fish species (Jackson 1961; Bell-Cross 1972). The native riverine species were not able to colonize the deep open water after the lake formed and were restricted to water not more than 15 m depth with the highest catches between 2 and 10 m (Coke 1968; Balon 1972; Sanyanga et al. 1995). To improve fisheries in the lake, Limnothrissa miodon, a freshwater sardine locally known as ‘kapenta’ was introduced into Lake Kariba from Lake Tanganyika between 1967 and 1968 to fill the deep water pelagic ecological niche (Bell-Cross & Bell-Cross 1971). Experimental fishing using different gears and an echo sounder revealed that by 1973 the population of L. miodon was large enough to support commercial fishing (Begg 1974). The L. miodon fishery, also referred to as the ‘sardine’ or ‘pelagic’ fishery, started in the middle of 1974. Fishing takes place at night using lights to attract the fish which are then caught using a lift net with a mesh size of 8 mm. In the early 1970s purse seines were also used for fishing. Catches in the L. miodon fishery increased initially. After reaching a peak of nearly 22 000 tonnes, landed in Zimbabwe in 1990, they started to decrease. Marshall (2012a) attributed the decline to overexploitation, whilst Magadza (2011) suggested climate change to be the cause. Both fishing and environment factors are likely important because biomass of L. miodon was high in the beginning of the fishery and decreased as fishing effort increased (Marshall 1988a). The environment is important because low river flow into the lake, and drought, are associated with poor catches of L. miodon (Marshall 1988b). In addition, climatic factors such as rainfall and maximum temperature had the highest correlation with Catch per unit Effort (CPUE) compared to the hydrological factors ‘river inflows’ and ‘lake level’ (Chifamba 2000). Both human and environmental, often together are involved in the collapse of many fisheries worldwide (Dekker 2003; Daskalov et al. 2007; Ruiz et al. 2009). For example, the collapse of Anchovy kilka (Clupeonella engrauliformis) fishery in the Caspian Sea was attributed to recruitment failure caused by the introduced predator ctenophore (Mnemiopsis leidyi) and overfishing (Daskalov et al. 2007). A holistic approach in the analysis of the causes is recommended (Botsford et al. 1997; Starkie 2003; Borja et al. 2008). This approach was adopted by Ruiz et al. (2009) investigating the impact of large-scale climatic patterns, turbulence, upwelling and river flow on the recruitment and catches of anchovy in the Bay of Biscay, Spain. In the same way, a multivariable approach is needed in the assessment of the causes of collapse of the sardine fishery of Lake Kariba, where both the environment and fishing appear to play a part. Fishing can impact the fish population through selection of a particular size group, thereby affecting level of reproduction and recruitment into the fishery. This tendency is linked to the exponential relationship between the fish size and the number of eggs in its ovary as well as differences in the quality of these eggs (Birkeland & Dayton 2005; Arlinghaus et al. 2010). To avoid capture of juveniles

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 96 Developing a sustainable pelagic fishery in an African reservoir

that inhabit the shallow water, the sardine fishery is permitted only in deep water (≥ 20 m). Cochrane (1984) reported that Limnothrissa miodon are found at all depths, with the juveniles and inhabiting the shallow marginal areas and the size increasing with depth. In Lake Kivu, the juveniles (10 – 35 mm) and the large fish (85 – 110 mm) were always found in the littoral area. The largest were caught close to the margin, whilst the medium sized (65 – 90 and 85 – 110 mm) were pelagic (de Iongh et al. 1983). The Lake Kivu study shows that the sardine moved into deeper water as they grew up to a point, and then moved back into shallower water at a large size. This had not been observed in Lake Kariba, when the current study on the distribution of sardines by depth was started to determine how they utilize space and what stages would interact with the inshore fishes. Though the introduction of L. miodon (a pelagic fish) was not expected to have impact on inshore fish, it changed the habit of a top predator, the tigerfish Hydro- cynus vittatus, and it may have indirect impact on other species. The former riverine H. vittatus, which was first confined to the inshore lake, entered the pelagic area to forage on L. miodon and is caught as bycatch of the sardine fishery (Marshall 1987b). 6 From April 1969 to March 1970, only 1.5% of the stomach content of H. vittatus consisted of sardines, whereas from April 1970 to March 1971 the amount had risen to 41.4% and in samples from 1994 – 1997 it was more than 45% (Kenmuir 1973; Mhlanga 2003). A record of the H. vittatus catch was kept since the beginning of the L. miodon fishery in 1974. It is interesting to know how the catch of this species responded to the changes in the catch of L. miodon. The other fish species caught in the sardine fishery are not routinely recorded because of low representation in catch and small size that make them difficult to separate from the rest of the catch. Even so, the cumulative catch in number of these small fish was thought to be large. It was observed that these fish included juveniles of the gill-net fish species. Lack of information on the other components of the bycatch that was not recorded, prompted the collection on the data aimed to evaluate the impact of this fishery on the inshore fished through the capture of undersize fishes. In this paper, we investigate the causes of the decline in the pelagic catches by analysing the relationships between catches of L. miodon, fishing effort and several key environmental variables: air temperature and hydrological factors (rainfall, river flow and lake level). The effect of depth on fish size and fish catches was assessed to understand the distribution of L. miodon particularly with regard to the large fish largely missing from the fishery. Trends in H. vittatus catches in the pelagic area were examined to detect the response of the fish to the decrease in the catches of L. miodon. The impact of the L. miodon fishery on the inshore fish species through the capture of juvenile of inshore fish species, was evaluated by quantifying and determining the species composition of bycatch.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 97 PART II – BIOLOGY AND IMPACTS OF LIMNOTHRISSA MIODON

Materials and methods Relationship between catches, fishing effort and environment variables Catch and effort data for the L. miodon fishery from 1974 to 2011 were obtained from the Lake Kariba Fisheries Research Institute and hydrology and weather data from Zambezi River Authority. One unit of effort is one fishing boat, fishing one night, and catch per boat per night was taken as the catch per unit effort (CPUE). The environmental variables ‘rainfall’, ‘river inflow’, ‘maximum temperature’, ‘minimum temperature’ and ‘fishing effort’ were all entered into a Generalized Linear Model to test which of the variables explained the variation in L. miodon catch and CPUE. The relationship between catch and effort, based on the combined data from Zimbabwe and Zambia for the period 1974 – 2011 (Kinadjian 2012), was used to estimate maximum sustainable yield (MSY) using the Schaefer and Fox Surplus Yield models (Pitcher & Hart 1982):

2 Ye = U∞ F - b F Schaefer model

- F (q / k) Ye = F U∞ e Fox model

where Ye = yield or catches, F = fishing effort, U∞ = CPUE at infinity; q = catchability coeficient; k = rate of increase of biomass, and b = a constant.

Effect of depth on L. miodon Data from two projects carried out in the Sanyati Basin of Lake Kariba were used for this analysis. The first project was done in 2001 – 2002. Monthly samples, taken at depths of 2 to 55 m, were collected in November and December 2001, and in January, March, April, May and June 2002. Fish were caught at night on a fishing rig, using a 4-m diameter lift net of 8 mm stretched mesh size, lined with a 1-mm stretched mesh net. A site was sampled twice, first without light, to measure fish density, then with light. Three 2800 lumens (80W) mercury vapour light bulbs were used, of which two provided surface lighting and one was used for underwater lighting. The net was set for 30 minutes each time. In the second project, done in March 2013, samples were taken at depths of between 20 and 55 m. Sampling was carried out for 10 days, using a commercial fishing rig with a net of 7.5 m diameter. Three light intensities, 2 800 (80 W), 11 500 (250 W) and 20 000 (400 W) lumens, were used for lighting, because the main objective of this study was to measure the effect of light intensity on catches. As in the 2001 - 2002 study, the fish were captured with and without light. When fishing without light, the net was set for 30 minutes as before but for 2 hours when using light. Catch of each species from each setting was recorded in the field. About 200 g of L. miodon was subsampled and all the fish longer than 7 cm were selected for

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 98 Developing a sustainable pelagic fishery in an African reservoir

determination of length, weight, mouth gape, body height, sex, gonad weight, number of eggs, stomach content and stomach content weight. The body height and mouth gape were measured using a vernier caliper. The mouth gape was measured in the dorso-ventral plane with the mouth fully open. Catches taken without light were used to calculate the density of the fish using the swept area method such that:

Fish density = Catch from one setting without light / Area of net opening.

The surface area of the net opening was 12.6 m2 in 2001 – 2002, and 44.2 m2 in 2013. This method of estimating biomass was also used by Cochrane (1978) and Marshall (1985) to estimate the density of L. miodon in Lake Kariba. Data was analysed using regression analysis to test the effect of depth gradient on fish length, catch and density.

Quality and species composition of bycatch Records of L. miodon and H. vittatus catches, kept since the beginning of the pelagic fishery, were used to analyse the relationships between the L. miodon and H. vittatus 6 catches. For the incidental catches that were not routinely recorded, catch data was obtained from fishing companies operating in the Mlibizi and Binga, Sengwa, Chalala and Sanyati basins on the Zimbabwean side of Lake Kariba (Figure 6.1) from April 1993 to June 1994. Additional data was obtained from experimental mid-water trawling in the Sanyati basin in June 1993. The trawl net was operated during the day between 10 and 30 m depth, shot 7 times and towed for 1 hour. Daily landing of bycatch was recorded on specifically designed forms. The bycatch was split into the following groups for easy identification; H. vittatus, cichlids, Synodontis spp, Schilbe intermedius, Brycinus lateralis and others. A subsample was collected from commercial companies fishing in the Sanyati basin in June 1993, to determine species composition of the cichlid group and size of the fish.

Results Relationship between catches, fishing effort and environmental variables The total annual catches of L. miodon rose from the time the fishery started in 1974 to a peak in 1990, and thereafter declined steadily till 2011 (Figure 6.2a). Fishing effort increased throughout the whole period except from 1992 to 2000, when it was more or less constant. Catch per boat per night decreased throughout the whole period except for a more constant period between 1980 and 1990, suggesting a deterioration of conditions for the fish. The MSY estimated from combined Zimbabwean and Zambian catch and effort data from 1974 to 2011, is 23 185 (Schaefer model) and 22 355 (Fox model) tons per annum, and the estimated effort (Fmsy) used to catch the MSY is 10 867 (Schaefer) and 11 226 (Fox) boat nights (Figure 6.2b). The estimates of MSY and effort were

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 99 PART II – BIOLOGY AND IMPACTS OF LIMNOTHRISSA MIODON

a) Catch, effort & CPUE 35000 1.0 Catch 30000 Effort CPUE 0.8 )

25000 ht g 0.6 20000 tons/boat/ni

15000 ( 0.4

10000 CPUE 0.2 Catch(tons/ Effort (boat nights /10)) 5000

0 0.0 1970 1980 1990 2000 2010 2020 Year b) Catch and effort regression 35000

30000 MSY 25000

20000

15000 Catch (tons)

10000

Data 5000 Schaefer Fmsy Fox 0 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 Effort (boat nights/10) Figure 6.2 Changes in the a) catch, fishing effort and CPUE and b) catch and effort regression in the L. miodon fishery of Lake Kariba. The regression lines are fitted with the Schaefer (R2 = 0.79; p < 0.0001) and Fox (R2 = 0.76; p < 0.0001) Surplus Yield models.

used to make inference on the sustainability of the fishery. The maximum fish landing of 30 000 tons of sardines were higher than the value of MSY. To explore the potential cause of this, we related CPUE to temperature. The pattern in the CPUE mirrors that of mean maximum air temperature (Figure 6.3a). Regression

100

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 100 Developing a sustainable pelagic fishery in an African reservoir

a) CPUE and temperature trends 1.0 33

0.8 32 ) ) t 0C h ( g i n / 0.6 31 oat erature /b p

CPUE tons ( 0.4 temperature 30 CPUE Maximum tem 0.2 29

0.0 28 1970 1980 1990 2000 2010 2020 Year 6 b) temperature and CPUE 1.0

0.8

0.6

0.4 CVPUE (tons/boat/night) 0.2

0.0 28 29 30 31 32 33 Mean maximum temperature 0C Figure 6.3 Relationship between CPUE and maximum temperature in the L. miodon fishery of Lake Kariba in a) trends in CPUE and temperature (1974 to 2011) b) regression of temperature and CPUE (fitted with an exponential model CPUE = 33 3380.6 exp (‐0.04496 × temperature); (R2 = 0.80; p < 0.0001).

analysis confirms a strong statistically significant negative exponential relationship with the fitted function (Figure 6.3b). Both fishing effort and temperature explain a significant amount of the variation in CPUE (Table 6.1). In contrast, the total catch is correlated with total fishing effort and lake level but not with mean maximum air temperature (Table 6.1).

101

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 101 PART II – BIOLOGY AND IMPACTS OF LIMNOTHRISSA MIODON

Table 6.1 The χ2 and p values from the Generalized Linear Model testing the relationship between fishing effort, rainfall, lake level, mean maximum and minimum air temperature and both catch and CPUE.

Catch CPUE Variables Wald χ2 p Wald χ2 p Intercept 16.54 0.000 8.34 0.004 Fishing effort 37.11 0.000 15.84 0.000 Rainfall 1.63 0.200 2.00 0.160 Lake level 15.95 0.001 3.83 0.051 Maximum temperature 3.23 0.072 14.87 0.000

Depth distribution of fish Estimates of fish density were 24.8 ± 12.2 and 38.8 ± 6.2 kg/ha for 2001 – 2002 (≥ 20 m; pelagic) and 2013, respectively. The 10-day survey carried out in March 2013 showed a depth gradient in the catches of all species. The catch per set of L. miodon (R2 = 0.191; F = 8.99; p = 0.0048, H. vittatus (R2 = 0.136), and Synodontis spp (R2 = 0.139), and the catch without light of L. miodon (R2 = 0.145) were highest in shallower water (< 30 m) compared to deep water (> 40 m), even though the amount of variation explained by a linear relationship with depth was rather low (p < 0.05, Regression analysis; n = 40) (Figure 6.4). The other species, which included Tilapia rendalli, O. niloticus, a 1.25 kg O. mortimeri, and small cichlid species, were caught only at Site 1 (< 30 m deep). Samples from the period November 2001 to June 2002 showed variations in mean fish length and catch size with depth (Figure 6.5). Figure 6.5a-b shows that the mean length increased significantly with depth, both in samples taken without (R2 = 0.276; F = 10.29; p = 0.003; n = 30; regression analysis) and with light attraction (R2 = 0.313; F =12.78; p = 0.001; n = 30; regression analysis). The mean size of fish taken using light attraction (3.5 ± 1.73 cm) was significantly larger than that of fish taken without light (2.29 ± 1.37 cm) (F = 6.99; p = 0.011; n = 59; ANOVA). Large fish (> 10 cm) were present especially when fishing with light. Figure 6.5a-b shows that the minimal mean length of fish increased stronger with depth than the maximum mean length. A similar increase in the size of the smallest fish was also found in the samples taken in March 2013. This means that larger fish also went deeper, whilst smaller fish remained shallower. Fish of all sizes could be found in the shallow area, though the small fish (< 2 cm) were restricted to the shallow area (< 20 m deep). Monthly variations in the trends were observed. There is no clear pattern of catch with depth, though catches are always low at the shallowest sites and highest at medium depths (Figure 6.5c-d).

102

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 102 Developing a sustainable pelagic fishery in an African reservoir

a) L. miodon catch in the dark b) L. miodon catch with light 400 100

80 300

60 200 40 Catch/set (g) Catch/set (kg) 100 20

0 0 10 20 30 40 50 60 10 20 30 40 50 60 Depth (m) Depth (m) c) H. vittatus d) Synodontis spp 4000 300

250 3000 200

2000 150

Catch/set (g) 100 Catch/set (g) 1000 6 50

0 0 10 20 30 40 50 60 10 20 30 40 50 60 Depth (m) Depth (m) Figure 6.4 Changes in catch per set of a) L. miodon catch in the dark, b) L. miodon catch with light, c) H. vittatus and d) Synodontis, along the depth gradient in March 2013 (fitted with linear regression line).

Figure 6.5 Depth and a) the mean length of fish caught with light b) mean length of fish caught without light c) caught with light d) catch without light. 103

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 103 PART II – BIOLOGY AND IMPACTS OF LIMNOTHRISSA MIODON

140

120 Bulk 100 large y diet fish 80

60 Frequenc

0 24681012 Length group (cm) Figure 6.6 Length frequency distribution of the bulk of the catch, the large specimen and fish found in the stomachs of 5.9 to 10.5 cm long fish (diet fish) from experimental fishing in March 2014.

Catches from Site 1 in March 2013 included some large fish between 8 and 12 cm total length, though the bulk of the catch were between 2.5 and 6.5 cm with a mode at 4 cm (Figure 6.6). Most of the large fish had food in the stomach and all of them had eaten small sardines (1.9 to 3.0 cm total length) (Figures 6.6 and 6.7a). The ingestible size range for large sardines of a given length is determined by the ratio of mouth gape over body height as a function of fish length. The body-height increase with increasing length is larger than that of the mouth gape (Figure 6.7b). Considering the size of their mouth gape, the large L. miodon had eaten L. miodon smaller than they potentially could swallow. The large fish that could be sexed, were all sexually active, judged from their enlarged gonads as well as the size of the eggs in females. Gonad size increased exponentially with body size in females, but there was no significant trend in males (Figure 6.7c). The relationship between female gonad size and fish length is given by the function: Gonad weight (g) = 0.0066 exp (0.3825 × fish length (cm)) (R2 = 0.336; p = 0.0047; n = 22). The number of eggs increased exponentially with female size (Figure 6.8) and can be expressed by the function: Number of eggs = 199.0030 exp (0.0344 × fish length (mm)) (R2 = 0.6755; F = 158.1882; p = 0.0000; n = 77). The size of fish caught in the fishery is much smaller than the fish that had the highest fecundity expressed by the number of eggs.

Relationship between pelagic H. vittatus and L. miodon When the fishery began in 1974, the CPUE of H. vittatus was high and rose to a peak in 1977, but subsequently fell together with the CPUE of L miodon (Figure 6.9a). The relationship between CPUE of H. vittatus and L. miodon can be described by a power relationship in the function: tigerfish CPUE = 0.0862 × sardine CPUE 3.7845 (R2 = 0.65; p < 0.001; df = 37), illustrated in the logarithmic plot (Figure 6.9b).

104

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 104 Developing a sustainable pelagic fishery in an African reservoir

a) stomach content c) Gonad weight 2.0 0.8 Female Male 0.6 1.5 Exponential female Linear male 1.0 0.4

0.5 0.2 Gonad weight (g) weight Gonad

0.0 0.0 Stomach content weight (g) weight content Stomach

-0.2 2 4 6 8 10 12 2 4 6 8 10 12

Total length (cm) Total length (cm)

b) gape & girth d) Length weight 2.5 20 Mouth gape Girth 2.0 Linear gape 15 Linear girth 1.5 10 6

1.0 (g) Weight

5 0.5 Mouth gape/ body height (cm) height body gape/ Mouth

0.0 0 24681012 2 4 6 8 10 12

Total length (cm) Total length (cm) Figure 6.7 The relationships between the length of L. miodon caught in March 2013 and a) weight of L. miodon found in their stomachs b) mouth gape and body height c) gonad weight for females and males and d) weight.

12000

10000

8000

6000

Number of eggs of Number 4000

2000

0 20 40 60 80 100 120 140 Length (mm) Figure 6.8 The relationship between the number of eggs and the size of L. miodon caught in the Sanyati basin of Lake Kariba. Size range and modal length caught in the commercial fishery are represented by the shaded area and the vertical line, respectively.

105

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 105 PART II – BIOLOGY AND IMPACTS OF LIMNOTHRISSA MIODON

a) trends in sardine and tigerfish CPUE

1.0 0.07 ) ht

L. miodon (pelagic) g H. vittatus (pelagic) 0.06 H. vittatus (inshore) 0.8 ) 0.05 night / 0.04 0.6 tons/set or boat/ni or tons/set ( boat / 0.03 tons ( 0.4 0.02 ic & inshore g

L. miodon miodon L. 0.01 ela

0.2 p

0.00

0.0 H. vitattus 1970 1980 1990 2000 2010 2020 Year

b) regression of Log CPUE 3

(kg/boat/night) 0

-1 H.vittatus Log -2

-3 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 Log L. miodon (kg/boat/night) Figure 6.9 The relationship between a) L. miodon and H. vittatus (pelagic and inshore) CPUE (tons/ boat/night) from 1974 to 2011, b) regression of logarithmic scale (base 10): L. miodon and pelagic H. vittatus CPUE (kg/boat/night).

106

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 106 Developing a sustainable pelagic fishery in an African reservoir

Quality and species composition of bycatch The survey covered all basins of Lake Kariba and 26 out of 62 Zimbabwean fishing companies participated. Catches of the sardine fishery contained very small quantities (0.32%) of other species. From May 1993 to June 1994, 69.2 tons of bycatch were landed compared to 21 370 tons of L. miodon. Bycatch was also small compared to catches from the inshore fishery, where 1 281 and 987 tons were landed in 1993 and 1994, respectively. Of the total bycatch, 93.8% comprised of 3 fish species, Schilbe intermedius (42.2%), H. vittatus (37.1%) and Brycinus lateralis (14.5%) by weight. Catches of cichlids (2.0%) and Synodontis spp (4.2%) were low. The cichlid group consisted of Pharyngochromis acuticeps, Pseudocrenilabrus philander, Serranochromis macrocephalus, Tilapia rendalli and Oreochromis mortimeri. Synodontis species caught were Synodontis zambezensis and Synodontis nebulosus. Negligible quantities of Serranochromis robustus, Labeo altivelis and Mormyrops deliciosus were also caught. In the bycatch from the experimental mid-water trawling, only three species were caught: S. zambezensis, S. nebulosus and P. acuticeps. They constituted 2.47% of 6 the total catch. In the March 2013 survey, the bycatch was 0.85% (14.0 kg) of the total fish catch. Most of the bycatch was H. vittatus (86.1%), and the rest was Synodontis spp (5.3%) and cichlids (8.6%). The bycatch consists of small fish of species that make up the bulk of the inshore fishery catch, but also a portion that do not. Two of the important species in the bycatch, S. intermedius and B. lateralis, are not caught in the gill-net fishery because of their small size. The cichlid group also includes smaller species that are not exploited in the inshore gillnet fishery. The inshore fishery species caught in measurable amounts in the bycatch were H. vittatus and the cichlids S. macrocephalus, T. rendalli and O. mortimeri. Of these, only H. vittatus was caught in relatively large amounts. The catches of H. vittatus in the sardine fishery were 7.6% and 8.9% of the gill-net fishery in 1993 and 1994, respectively. The mean sizes of fish caught in the sardine fishery were small compared to those from the inshore gill-nets (Table 6.2). Because of the small size of the cichlids in the bycatch, their quantity in numbers may be of larger significance than the weight may portray. Using the species composition to partition the cichlid bycatch, the estimated numbers of T. rendalli, O. mortimeri and S. macrocephalus caught during the survey in the whole lake were 41 327, 98 129, and 27 677, respectively. The overall picture masks the temporal and spatial differences that can be important to management. There are spatial variations in the quantity and species composition of the bycatch (Table 6.3). The Sengwa basin had the lowest CPUE, Binga and Mlibizi basins the highest.

107

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 107 PART II – BIOLOGY AND IMPACTS OF LIMNOTHRISSA MIODON

Table 6.2 Mean weight (g) of individual specimens from incidental catch in Kariba, the gill‐net fishery, and the maximum size caught in Lake Kariba (Kenmuir 1983).

Species Incidental (g) Gill net (g) Maximum (g) H. vittatus 92.2 1 970.8 15 500 O. mortimeri 4.4 664.1 3 500 S. macrocephalus 10.2 370.9 1 500 T. rendalli 7.3 658.8 2 100 P. acuticeps 3.1 Nil 23 P. philander 2.1 Nil 8 B. lateralis 2.6 Nil 11.9 S. zambezensis 28.0 no record 900 S. nebulosus 3.4 no record 90 S. intermedius no record no record 770

Table 6.3 The estimated catches (metric tons) from the sardine fishery by species and basins from May 1993 to June 1994.

Binga & Total Corrected % in by‐ Species Sengwa Chalala Sanyati Mlibizi (tons) total catch H. vittatus 5.08 1.77 2.40 16.42 25.67 22.1 31.1 S. intermedius 0.59 0.01 13.79 14.79 29.18 25.1 42.2 B. lateralis 2.21 0.00 2.40 5.46 10.07 8.7 14.5 Cichlids 0.26 0.04 0.86 0.23 1.38 1.2 2.0 Synodontis spp 0.40 0.02 1.00 1.49 2.91 2.5 4.2 L. miodon 2377.03 2048.14 7579.33 12838.74 24843.24 21369.9

Total catch 2379.89 2048.20 7583.59 12845.92 24857.60 21429.4 Bycatch 8.53 1.84 20.45 38.39 69.21 59.5 % in bycatch 0.36 0.09 0.27 0.30 0.28 0.28

Discussion We found that the initial increase in catch of L. miodon in Lake Kariba, could not be sustained after 1990 and the CPUE decreased throughout the whole period, reducing profitability. Increased temperatures were associated with the decline of CPUE (as an indicator of stocks), indicating declining ecologically favourable conditions for L. miodon, especially in the first two decades of the study (1970 – 1980; Figure 6.2). Fishing effort that subsequently increased strongly (1980 – 1990), was another factor

108

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 108 Developing a sustainable pelagic fishery in an African reservoir

strongly affecting CPUE, and fishing effort beyond Fmsy may have caused the collapse of the sardine fishery. Hydrocynus vittatus responded to the decrease in sardine CPUE. Catches and CPUE of H. vittatus were high when the sardine fishery started, then decreased simultaneously with sardine CPUE. Hydrocynus vitttatus and other inshore fish species also utilize the deep pelagic water and were caught in the pelagic fishery as incidental or bycatch. Limnothrissa miodon occurred in high numbers and biomass in all depth zones and fish size increased with depth, with juveniles confined to inshore areas. However, the largest specimens of L. miodon were caught below 30 m depth, overlapping with the conspecific juveniles they fed upon. These large fish have a higher fecundity compared to the size caught in the fishery.

Spatial distribution of L. miodon Different stages of fish are often spatially segregated to limit intraspecific completion and predation (Persson et al. 2000). For example, in Lake Kivu juvenile L. miodon (1.0 – 3.5 cm in length) were always caught in the margins, fish of 3.0 – 7.0 cm) in the littoral and inshore areas, and the next larger size class (6.5 – 9.0 cm) 6 in pelagic waters. Sardines of 8.5 – 11.0 cm were caught in all zones except the lake margins, whilst large fish (10.0 – 15.0 cm) were mainly caught in the margins (de Iongh et al. 1983; Spliethoff 1983). The sardines from the pelagic waters fed ex- clusively on plankton, whereas those in the inshore area fed on a mixture of plankton, a relatively large fraction of chironomid pupae, and juvenile clupeids. The largest sardines (> 100 mm, total length) were cannibals, just like in our study, and lived near the margins (de Iongh & Spliethoff 1983). In Lake Tanganyika, the larvae of L. miodon were predominantly present near the shore, immigrating towards sandy shores as they became larger (Tshibangu & Kinoshita 1995). The small pelagic cyprinid Rastrineobola argentea from Lake Victoria displays a similar behaviour of using the inshore areas for spawning and as a nursery area (Wanink 1999). This study confirms earlier studies showing that juvenile Limnothrissa miodon are always confined to the lake margins (Cochrane 1978), and reports that the large (> 8 cm) adults in Lake Kariba seem to utilize both the pelagic and the marginal shallow areas. This size distribution might partly explain the general scarcity of the largest size group in the deep waters. Comparable habitat segregation was found in 1+, 2+, and 3+ year old Eurasian Perch, Perca fluviatilis in Lake Abborrtjärn 3 in central Sweden. Year classes are segregated according to diet, with the planktivorous youngest in offshore and the oldest carnivores predominantly in the inshore area, where they feed entirely on macroinvertebrates (Persson et al. 2000). Absence of the large fish from the pelagic may be a consequence of their lower foraging efficiency compared to the smaller fish. Persson (1987) found that one year old and 1+ perch have a higher capture rate and a lower handling time than the two years old and 2+ perch. This means that the smaller fish have competitive advantage when feeding on pelagic zooplankton.

109

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 109 PART II – BIOLOGY AND IMPACTS OF LIMNOTHRISSA MIODON

These mechanisms observed in other species, may be driving the segregation of the size classes in Lake Kariba and other lakes in which the sardine occurs. Cannibalism of large individual fish on conspecific juveniles can be of ecological importance and can affect the dynamics of fish populations (van den Bosch & Gabriel 1997; Claessen et al. 2000; de Roos et al. 2003). Cannibalism of juveniles by large (> 8 cm) L. miodon was observed in Lake Kariba by Begg (1974). Other fish species, like Haplochromis philander and Barbus sp, have also been ingested. In Lake Kivu, large (> 10 cm) L. miodon were cannibalistic (de Iongh 1983). Cannibalism observed in large fish that were caught in deep water (> 30 m), is a normal feeding strategy of L. miodon. This suggests that the large fish migrate to the shallow area to feed on juveniles, which may not be necessarily their own species. This is probably a necessary diet shift to compensate for changes in energetic demands, due to somatic growth (Werner & Gilliam 1984). Cochrane (1984) suggests that fish in the pelagic area would be too large to be eaten by these large sardines. This idea is supported by our study on mouth gape and body height. The inshore movement of cannibalistic L. miodon is necessary to find suitably sized prey to optimize the energetic needs. Cannibalism may have consequences on the population dynamics by controlling recruitment pulses. For example, the densities of perch ≥ 2 years old in the pelagic area determined the intensity of cannibalism on the one-year-old fish that inhabit the pelagic area (Persson et al. 2000). In lakes where the cannibal sized perch were reduced by a top predator (pike, Esox lucius), the abundance of the small perch (young-of-the-year and 1-year-old) were more abundant than in lakes were perch occurred alone. This suggests that cannibalism controlled the abundance of the small perch (Wahlström et al. 2000), which can have a stabilizing effect on the fish population (van den Bosch & Gabriel 1997; Claessen et al. 2000). Evolution might favour cannibalism because of the survival advantages in terms of provision of energy that can be directed to population growth through reproduction. Persson et al. (2000) observed that after a die-off that affected mostly the large fish, the surviving cannibalistic individuals gained substantial energy from cannibalism in years with strong recruitments, which increased both growth rate and per capita fecundity. Hence, the energy gained by the cannibals may be essential for the recovery of the population, as this energy is allocated into new recruits (van den Bosch et al. 1988; Persson et al. 2000). In the case of the sardines in Lake Kariba, the larger individuals that survive by cannibalism would benefit the population by being more fecund, which might increase recruitment.

Interaction between Hydrocynus vittatus and its prey L. miodon The interactions between the sardines and tigerfish in Lake Kariba confirm that variation in the prey population affects the ecosystem (Marshall 1987b; Bakan et al. 2000; Smith et al. 2011; Essington et al. 2015). Removal of the prey through overfishing alters the abundance and composition of upwelling pelagic communities

110

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 110 Developing a sustainable pelagic fishery in an African reservoir

(Curry at al. 2000). The initial rise in H. vittatus catches in Lake Kariba, when the sardine fishery developed, seems to be typical of new reservoirs. For example, the increase of Nile perch (Lates niloticus) and H. vittatus in Lake Kainji (Nigeria) and Lake Volta (Ghana), and that of pike (Esox lucius) in Cimljansk Lake due to the abundance of prey species (Lelek & El Zarka 1973). Fishing removes sardines from the ecosystem, denying tigerfish this resource and causing a reduction in the number of predators in deep waters. Generally, the inshore tigerfish seem to be unaffected by the reduction in pelagic sardines because they utilize alternative prey, even though L. miodon remains an important prey item (Mhlanga 2003). Hence the decline in the sardine caused the decline of the pelagic tigerfish, which became more restricted to the inshore area where they supplement their sardine diet with alternative prey.

Quality and species composition of incidental catch Our results provide several additional insights on the importance of bycatch of the L. miodon fishery. The species caught as bycatch are mostly those species that favour deep water, they are either planktivores or piscivores. Hydrocynus vitattus and 6 Brycinus lateraris are known to inhabit pelagic waters, and both species were caught in the initial experimental fishing for L. miodon (Balon 1971; Woodward 1974). One of the species in the bycatch was Schilbe intermedius. This small catfish is common in the inshore areas of Lake Victoria, but it was also found in the pelagic zone (Witte & van Densen 1995). The species is also common in the inshore areas of Lake Kariba but there has been no record of its occurrence in the open water (Kenmuir 1984; Sanyanga et al. 1995). The abundance of Synodontis zambezensis in the catches increased with depth, with the highest catch per unit effort at the maximum depth range sampled (12 – 20 m) between 1990 and 1993 (Sanyanga 1996). In Lake Kariba, Pharyngochromis acuticeps occurs in deep water and on both shelving and steep, eroding shores with no vegetation (Hustler & Marshall 1990; Mudenda 1992). Oreochromis mortimeri, B. lateralis, P. acuticeps, Pseudocrenilabrus philander and juvenile Serranochromis macrocephalus are all plankton feeders (Mhlanga 2000; Zengeya & Marshall 2007). Similarly, planktivorous juvenile European perch, Perca fluviatilis (8 – 19 mm) are predominant in the pelagic zone of an Australian reservoir, Lake Humme, where also planktivorous Australian smelt (Retropinna semoni) and climbing galaxid (Galaxias brevipinnis) occur (Matveev et al. 2002). Clearly, planktivorous and carnivorous fish species in Lake Kariba venture into deep water for feeding, where they are caught in the sardine fishery. Capture of non-target fish species, or bycatch, can have impacts on the ecosystem (Crowder & Murawski 1998). In Lake Kariba, the harvest of juveniles of other species in the sardine fishery is a potential problem. Though the sizes of fish caught in the sardine fishery are smaller than those caught in the inshore area, and the yield was also small, comprising 0.32% of the sardine catch by weight. Only in the trial trawl catches was it higher (2.47%). A large proportion (60.9%) of the bycatch consisted of S. intermedius (42.2%), B.

111

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 111 PART II – BIOLOGY AND IMPACTS OF LIMNOTHRISSA MIODON

lateralis (14.5%). and Synodontis spp (4.2%), species that are not important at all in the gillnet fishery. Hence, their capture in the sardine fishery can only be beneficial. Only a small proportion of the bycatch are cichlids, most of which are small species that are too small to be exploited in the gill net fishery. Hence, the bycatch had little impact on the inshore fishery through the capture of juvenile fish. The quantity of the bycatch is small, demonstrating that the current management of the Lake Kariba sardine fishery, that restricts fishing depth to water deeper than 20 m, limits the catch of non-target species.

Effect of fishing effort and environmental variables on catches Various environmental changes can be responsible for the observed CPUE (as an indicator of the ecologically suitability of the lake) for L. miodon. Previous studies found that air temperature, rainfall and river flow were associated with catches of L. miodon in Lake Kariba (Marshall 1982; Chifamba 2000; Magadza 2010). Variation of L. miodon catches in Lake Kariba was found to be correlated to river inflow, with years of drought associated with low catches (Marshall 1982, 1988b). Maximum air temperature had the highest correlation with CPUE of L. miodon compared to other hydrological factors namely, rainfall, river flow and lake level (Chifamba 2000). The current analysis, using a much longer time series, shows that air temperature still accounts for the highest variation of L. miodon catches, with low temperatures associated with high catches, especially early in the study. In addition, rivers bring nutrients into the lake and temperature is important for internal cycling of these nutrients. Lake Kariba is monomictic and is stratified most of the year, during which time the epilimnion becomes depleted of nutrients and of oxygen. Turn-over occurs during winter, often in July. Then, nutrients from the hypolimnion become available in the photic zone. Hence, turn-over, rainfall and river mouths are associated with increased zooplankton production (Magadza 1980 Masundire 1989). Production cycles in Lake Kariba are thus driven mostly by temperature, causing a high correlation with sardine CPUE. Temperature change can affect fish populations by modifying their environment and food resource (Beaugrand et al. 2003; Cohen et al. 2016). Rising temperature in Lake Kariba is thought to have caused changes in the plankton community, timing of stratification and the depth of the epilimnion (Magadza 2010; Mahere et al. 2014). Magadza (2011) observed that the phytoplankton in the lake is now dominated by Cyanophyceae, particularly Cylindrospermum raciborskii. This author also noticed that the breakpoints in the relationships of the sardine catch with air temperature and lake temperature were at 34.8 and 28.7 0C, respectively, occurring in 1987 – 1988. This was just before the peak catch in Zimbabwe in 1990. A similar consequence of warming was found in Lake Tanganyika, where warming has reduced water circulation and hence nutrient concentrations in the hypolimnion (Verburg 2003; O’Reilly et al. 2003). This is also found over longer time spans: an analysis of fish

112

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 112 Developing a sustainable pelagic fishery in an African reservoir

fossils and temperature in Lake Tanganyika revealed a negative relationship between temperature and both fish and mollusc production in the last 150 years, which was associated with a decrease in diatom abundance, resulting from sustained warming during that period (Cohen et al. 2016). ∼ Environmental factors are expected to affect fish populations through their effect on life history parameters (Gutiérrez et al. 2007; Takasuka et al. 2007; Itoh et al. 2009; Pörtner & Peck 2010). Beaugrand et al. (2003) reported that rising temperature induced changes in the occurrence of plankton species, and led to higher mortality rates in cod larvae, thus lowering recruitment. Effects of temperature could arise from a shift away from the optimum temperature for growth, as shown in the Japanese anchovy (Engraulis japonicus) and Japanese sardine (Sardinops melanostictus) larvae in the western North Pacific (Takasuka et al. 2007). A difference in the optimum temperature for growth of these two species is behind the shifts in the fishery between a warm anchovy regime and a cool sardine one. Small pelagic ﬁsh species (such as L. miodon with a life span of < 3 years; Chapter 7) are particularly vulnerable and react rapidly to environmental fluctuations and global change because 6 of their short life span (Cury & Roy 1989). Hence, environmental change may explain part of the change in CPUE of the sardines in Lake Kariba. An understanding of how this factor affects CPUE is crucial in making decisions on how to manage the fishery to avoid overfishing.

Implication to management of the fishery Overfishing has caused the collapse of fisheries worldwide (Walters & Maguire 1996; Botsford 1997; Marshall 2012a; Watson 2013) At the time of its collapse, the Lake Kariba sardine harvest was within the range of the estimated potential yield and above the classic model estimates of MSY. Using a baseline lake where the species forms a successful fishery as a ‘predictor of ecology’, Pitcher (1995) estimated a potential yield of about 5.5 t km-2 (range: 5 – 6.75 t km-2), amounting to an annual yield of 27 000 – 36 000 t of sardines for Lake Kariba. Our MSY estimates from the Schaefer and the Fox models were 23 185 and 22 355 tons per annum, with an optimum effort Fmsy of 10 867 and 11 226 boat nights, respectively. One major shortcoming of using the classic models is that the MSY is estimated in retrospect. Therefore, other methods of estimating sustainable catches are strongly recommended. A management system that recognizes a catch limit based on the lake productivity, and that incorporates the variations in the fish abundance caused by environmental changes, would improve catches and profitability. Changes in CPUE do not always reflect fish density, due to changes in fish catchability arising from environmental change, fish behaviour and technology creep, or changes in the efficiency of fishing vessels (Chifamba 1995; Rose & Kulka 2011; Marriott 2011). Shoaling fish are particularly vulnerable to capture even at

113

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 113 PART II – BIOLOGY AND IMPACTS OF LIMNOTHRISSA MIODON

low fish density, because fishers can search for, and locate the fish shoals (MacCall 1975; Salthaug & Aanes 2003). A stable CPUE may also occur when fishing vessels concentrate on fish aggregates that have become confined to a smaller area (Rose & Kulka 2011; Kraus et al. 2015). Changes in fishing power of fishing vessels, or technology creep, can also change CPUE (Chifamba 1995). For example, an increased use of echo sounders for locating fish, radios for communicating and the replacement of mechanical with hydraulic winches for lifting sardine fishing nets, has increased the fishing efficiency of boats in Lake Kariba (Chifamba 1995). Another example of technology improvement that affected CPUE is the introduction of skewed hooks and swivel line in the haddock (Melanogrammus aeglefinus) and Atlantic cod (Gadus morhua) longline fisheries off the Faroe Islands resulted in an increase in catch per unit effort by about 51% and 26%, respectively (Eigaard et al. 2011). Hence, reliance on CPUE as a measure of fish density will introduce errors in judging the impact of fishing. Better methods of measuring fish abundance should be used in the long-term monitoring of sardine abundance. Estimation of biomass of pelagic fish, using hydroacoustic surveys and other methods such as the swept area method, provides information on the fish stock (Jurvelius 1996; Nøttestad et al. 2014). Our study provides some estimates of L. miodon density in the Sanyati basin of Lake Kariba (24.8 and 38.8 kg/ha in 2001 – 2002 and 2013, respectively) that can be compared to those from earlier studies. The earliest estimates in the Sanyati basin were made in January (8.9 kg/ha) and April (23.1 kg/ha) 1976 (Cochrane 1978). These are lower than the estimates in our study. Estimates from a more comprehensive survey of the Sanyati basin by Marshall (1988a) were 90.5, 48.1 and 38.7 kg/ha in 1981, 1982 and 1983, respectively. Except for 1983, these values are higher than those in our study. Sanyati basin biomass estimates from a lake-wide hydroacoustic survey in September 1988 were between 16 and 44 kg/ha (Lindem 1988). Lake-wide surveys by Ngalande (1995) yielded 25.3, 35.4, 7.6 and 7.6 kg/ha in January 1992, November 1993, January 1994 and July 1994, respectively, for the Sanyati basin. Another lake-wide survey in August 2014 found approximately 60 kg/ha in the Sanyati basin (Mafuca 2014). These surveys showed temporal as well as within-and-among lake-basin variation in fish densities in Lake Kariba. Most density estimates for L. miodon in Lake Kariba are higher than the value of 23 kg/ha that was reported for Lake Kivu, a natural lake into which the species has been introduced (Guillard et al. 2012). Different methods used in making the estimates, and lack of continuous records of sardine densities make it impossible to determine long-term trends. Therefore, there is a need for extensive continuous surveys in order to understand the contribution of changes in sardine abundance on the fluctuations of the catches in Lake Kariba. A reduced stock has implications for the sustainability of the fishery, resulting from reduced spawning stock biomass and recruitment. Daskalov (1999), reported a correlation between the recruitment of whiting (Merlangius merlangus) and anchovy

114

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 114 Developing a sustainable pelagic fishery in an African reservoir

(Engraulis encrasicolus) in the Black Sea, with stock biomass and the physical environment. Removal of large fish, which produce more eggs of better quality, may reduce recruitment (Birkeland & Dayton 2005; Arlinghaus et al. 2010). Fishing can also cause the evolution of traits, when the vulnerability to capture is positively correlated with fitness traits such as intensity of parental care, as reported for largemouth bass (Micropterus salmoides) by Sutter et al. (2012). In Lake Kariba, the heavy exploitation of small size sardines (< 7 cm) therefore may reduce the reproductive potential through having a small spawning stock biomass and a small number of large fish. This possible connection between heavy exploitation of the small fish and fecundity needs to be investigated further, for us to understand the mechanisms operating in the Lake Kariba sardine fish stock and its interactions with fishing. Fisheries often collapse as a result of the combined pressures from fishing and the environment, which are difficult to separate (Ruiz et al. 2009). Our study is another example of this phenomenon. Therefore, the effects of both fishing and the environment on the catches should be incorporated in the management of a fishery, to avoid overfishing. There are many aspects of the sardine biology and ecology that 6 we need to understand better before we will be able to soundly predict the environmental effects, particularly those of temperature, on the sardine abundance in Lake Kariba. Abundance and recruitment should be monitored and used to estimate total allowable catches, as is done in closely managed fisheries such as the anchovy fishery in South Africa (Cochrane & Hutchings 1995). A better understanding of the ecosystem production of Lake Kariba is essential, as it will inform us of the limits on fish catch to ensure sustainability and long-term profitability.

Acknowledgements This work was carried out at the University Lake Kariba Research Station. We are grateful for fisheries data provided by Lake Kariba Fisheries Research Station and hydrology data by Zambezi River Authority. A grant provided by the Tonolli Memorial Fund was used in the 2001/2 field work.

115

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 115

Limnothrissa miodon: a large specimen was caught eating a small conspecific. The small fish was removed to compare the sizes.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 116 Growth of the freshwater sardine, Limnothrissa miodon (Boulenger 1906) estimated from diurnal increments in otoliths

Portia C. Chifamba Jan H. Wanink Britas K. Eriksson John J. Videler

Chapter 7

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 117 Abstract A commercial fishery in Lake Kariba, on the introduced freshwater sardine Lim- nothrissa miodon (Boulenger 1906) started in 1974, rose to a peak in 1990, and crushed thereafter. Overfishing and population changes in the plankton community, driven by a rise in temperature, might have reduced the sardine population and have changed its life history parameters such as growth and age at maturity. We used otolith reading to test if the population collapse was due to: 1) stunted growth, as a result of impoverished food sources, by estimating changes in growth rate from 1993 to 2013, or to 2) fishing pressure, that in time affected the size and age at maturity. Sagittae, the largest otoliths of L. miodon were examined under the Transmission Electron Microscope (TEM) to determine periodicity of growth increments. Diurnal increments were validated using edge-increment analysis and used to estimate age and growth parameters. Fish for increment validation were sampled over a 24-h cycle, at 2-h intervals. The increments at the margins of all otoliths from fish caught at night were light coloured under TEM, while 80% of those from fish caught during daytime were dark. Each pair of dark and light increments were taken as a diurnal increment and used for ageing. The growth trajectories of L. miodon vary among years and display differences in the growth of juveniles/small adults (< 10 months) and large adults (> 10 months). Juveniles and small adults grew fastest in 1996 and slowest in 2013, which may indicate particularly favourable conditions in1996. This supports the hypothesis of poorer plankton diets in recent years. The Gompertz model fitted the data better than the von Bertalanffy, logistic and power growth models. Esti-

mates of asymptotic length (L∞) from the Gompertz model were 18.0, 9.6 and 15.2 cm total length in 1993 – 1994, 1996 and 2012 – 2013, respectively. These values are comparable to the asymptotic length of L. miodon reported for its lake of origin, Lake Tanganyika. Age at first maturity was 8.02 and 7.90 months for females and males, respectively. First maturity occurred at a much smaller size in this study (females: 3.43 cm; males: 3.63 cm), compared to reported values for 1970 – 1972 (females: 5.2 – 5.6 cm; males: 7.1 – 7.3 cm). Temporal differences in growth rate and length at first maturity reflect a flexible life-history strategy of L. miodon under the influence of fishing pressure and possibly changes in food availability.

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 118 Growth of L. miodon estimated from diurnal increments in otoliths

Introduction The freshwater sardine, Limnothrissa miodon (Boulenger 1906) is a pelagic fish, introduced from Lake Tanganyika into Lake Kariba in the late 1960s to increase fish production (Balon 1974; Bell-Cross & Bell-Cross 1971). Commercial fishing began in 1973 and rose to a maximum yearly catch of 20 112 metric tons landed in Zimbab- we in 1990 (Karenge & Mugwagwa unpublished report). The collapse of the fishery, thereafter, has been attributed to overfishing and increased water temperature (Magadza 2011; Marshall 2012a). Catches of L. miodon are negatively correlated to mean maximum air temperature (Chifamba 2000). Increase in temperature caused a change in phytoplankton species composition, favouring unpalatable cyanobacteria and resulting in a reduction of the zooplankton upon which L. miodon feeds (Magadza 2011). Inadequate food may have reduced growth of L. miodon, which we investigate here.

Growth and reproduction Growth of L. miodon in Lake Kariba has been studied using body length and otolith- based methods (Cochrane 1984; Marshall 1987a; Chifamba 1992). Using length-based 7 methods, Cochrane (1984) and Marshall (1987a) estimated an asymptotic length

(L∞) of 8.1 and 7.4 cm total length (TL), respectively, and concluded that the population of Lake Kariba was stunted. Based on reading otoliths, however, Chifamba

(1992) found an asymptotic length (L∞ = 13.8 cm) comparable to the values reported for Lake Tanganyika and Lake Kivu, a natural lake where L. miodon was introduced (Spliethoff et al. 1983; Moreau et al. 1991; Mulimbwa & Shirakihara 1994). In Lake Kariba, L. miodon up to 15.5 cm length are rarely caught, hence the low asymptotic length obtained in the length frequency analysis by Marshall (1987a). The mean size in the commercial fishery was 5.7 cm in 1993 and 5.1 cm in 1996 (Chifamba, unpublished data). Previous studies in Lake Kariba show that L. miodon reached sexual maturity at a small length compared to populations in Lakes Tanganyika and Kivu (Woodward 1974; Spliethoff et al. 1983; Moreau et al. 1991). The smallest size at first maturity in 1970 was 5.2 – 5.6 and 7.1 – 7.3 cm fork length (FL) for females and males respectively (Woodward 1974). More recent estimates of growth rate as well as length and age at first maturity are needed, considering the ecological changes that occurred as the lake matured, and the L. miodon fishery that has evolved. Here we present such data, based on studying diurnal increments in the otoliths.

Reading otoliths Otoliths (lapilli, sagittae and asterisci; located in the inner ear and used for hearing and acceleration detection) can be used to age fish. Between 90 – 99% of the total

119

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 119 PART II – BIOLOGY AND IMPACTS OF LIMNOTHRISSA MIODON

otolith mass is calcium carbonate. The remaining 1 – 10% is an organic matrix composed of proteins (Degens et al. 1969; Payan et al. 2004; Borelli et al. 2001). During otolith formation and growth, the protein matrix plays a major role in cyclic calcium deposition. Otolith matrix proteins are necessary for the deposition of otolin, a protein that anchors otoliths in the sensory maculae and stabilises the otolith matrix (Murayama et al. 2000, 2002, 2005). Daily increments can be seen in the otoliths as alternating layers of organic materials and minerals, mostly calcium carbonate (CaCO3) (Watabe et al. 1982; Mugiya 1987). Borelli et al. (2003) proposed that organic matrix and calcium carbonate deposition in otoliths varies diurnally, with the organic matrix deposited during the day and CaCO3 at night. The alternate deposition of CaCO3-rich and protein-rich layers results in the formation of daily increments in the otoliths. Fish kept under constant dark or constant light conditions maintain the cyclic rhythm, indicating that the rhythm may be intrinsic or due to other factors (Roberts et al. 2004). Daily increments of sagittae, the largest otoliths in most teleosts (Harder 1975), have traditionally been used for age determination (Pannella 1971; Zekeria et al. 2006; Nyamweya et al. 2010). Validation of increments can be done by reading otoliths of fish of known age and counting the number of rearing days. It can also be done by counting the number of increments during a known period following marking (Campana & Neilson 1985; Geffen 1992). The latter method is recommended for larvae and fish species with a short life span (Brothers 1976). Counting daily increments for age determination is appropriate for L. miodon because the majority of the fish caught are less than one year old, and the life span does not exceed two years (Cochrane 1984). This technique has been applied to age adult and juvenile L. miodon in Lake Kariba by Chifamba (1992) and Mtsambiwa (1993). At the time of these studies, no validation of periodicity of increment deposition was done. Meisfjord et al. (2006) confirmed a daily deposition of increments in otoliths using chemical markers on reared juvenile L. miodon. Marginal increments show the last deposition on the outer rim of the otolith. Alternating organic and mineral layers are deposited discontinuously during the daily cycle. To visualise the separate margins with sufficient resolution, Tanaka et al. (1981) and Zhang & Runhau (1992) studied the formation of marginal increments in Oreochromis niloticus otoliths using scanning (SEM) and transmission electron microscopy TEM). We used TEM to do this for L. miodon.

Objectives The objectives of this study were to: 1) validate the periodic increments of the sagittae of L. miodon from Lake Kariba; 2) use these increments to age L. miodon caught in 1993 – 1994, 1996 and 2012 – 2013 in order to determine any temporal changes in the growth parameters of the fish; 3) establish the age and length at first maturity; 120

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 120 Growth of L. miodon estimated from diurnal increments in otoliths

4) use this information to assess whether the fishery catches mature fish, and if fishing affects size and age at maturity. This assessment is important because catching fish before they reproduce, undermines the capacity of the population to sustain itself. Furthermore, evolutionary changes of life-history parameters induced by fishing can affect the sustainability of the L. miodon fishery.

Materials and methods Fish samples for the validation of otolith increments and those for the estimation of growth parameters were collected and analysed separately.

Collection of samples for validation of increments Fish were collected from the Sanyati basin of Lake Kariba (Figure 1.1) over one 24-h period, using a trawl net by day and a lift net at night. During the night, samples were taken at 2-h intervals, starting at 18:00 h and ending at 04:00 h. Daylight samples were taken at 08:00, 10:00, 12:00 and 14:00 h. Nets were set for one hour in both cases. The fish were preserved on ice in the field, for transport to the laboratory. Capture time was recorded for each sample. In the laboratory, the two sagittae otoliths were removed and stored together in the primary fixative, cacodylate buffered glutaraldehyde. 7

Preparation of otoliths for Electron Microscope examination The sagittae were fixed, stained, dehydrated and embedded following the procedure described by Glamart (1984) and used by Zhang & Runhau (1992), with the following modifications. Cacodylate buffer and dry acetone were used in place of saline buffer and propylene oxide. For primary fixation, otoliths were kept in cacodylate buffered glutaraldehyde, prepared using 50 ml 0.2-M cacodylate buffer, 12 ml 25% glutaraldehyde and 88 ml of distilled water.

Sectioning and Staining Initially, 1-mm thick sections were cut on an OM-U2 Reichert Ultra microtome, using glass knives. The sections were stained with 1% Toluidine blue and examined under an Olympus EHS light microscope for opaque and translucent bands, which appeared dark and light respectively. When increments were found, ultra-thin sections were then cut using the same microtome, placed on copper grids and stained with 5% Uranyl acetate in 30% methanol for one hour, followed by Reynolds' lead citrate for 5 – 7 minutes (Reynolds 1963). Stained sections were examined using a Carl Zeiss EMIOC Transmission Electron Microscope at magnifications between 750 and 16 000 times.

Interpretation of increments For each sample that showed a marginal increment, the type of the increment was noted, measured and matched with the time of capture. At each time period, the numbers of fish

121

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 121 PART II – BIOLOGY AND IMPACTS OF LIMNOTHRISSA MIODON

with either a light or a dark increment were added up and used to assess the pattern of increment formation.

Collection of samples for age and growth determination Fish samples were obtained from commercial catches and by experimental fishing in the Sanyati basin of Lake Kariba in 1993, 1994, 1996, 2012 and 2013, and in the Mlibizi basin in 1992 (Figure 1.1). Specimens from 1992 and 1996 were used to study the width of daily increments, and the relationship between the radial of an otolith and fish length. The total number of increments was counted in otoliths from all samples and used to establish the ages of fish. Additional data from 1988 were taken from Chifamba (1992). Larval and juvenile L. miodon are found in shallow, inshore areas, and the adults in deep, open water (Cochrane 1984; Mtsambiwa 1989). Adult fish were collected from commercial landings, while larvae and juveniles (TL < 2.5 cm) were captured in shallow areas (depth < 10 m) using a lift net lined with a 1-mm mesh net. Total length of each fish was measured to the nearest mm. Fish collected in 2013, and those larger than 8 cm in other years (from samples preserved in ethanol), were sexed. The fish were examined under a dissecting microscope to determine the sex and the stage of gonadal development. The sex of immature fish could not be determined with certainty. In all analyses, this group is assumed to contain both sexes, a procedure adopted by Kimura (1995). Sagittae were removed and cleaned in water, then air dried. The pair of sagittae from an individual fish were wrapped in plastic (stretch wrap) and stored in labelled paper envelops until needed for further analysis.

Ageing using increments The procedures for storage and mounting of otoliths used in this study were those recommended by Morales-Nin (1992). Sagittae were mounted on glass microscope slides using transparent nail varnish, except before 1996, when those larger than 3 cm were mounted in blocks of epoxy resin. To reveal increments, the otoliths were grounded and polished sequentially on silicon carbide paper of decreasing grain size: 40.5 µm (320), 15.3 µm (1200), 6.5 µm (2400) and 2.5 µm (4000), and on 3-µm Imperial lapping film. Grinding and polishing was done with the otolith surface submerged in a film of soapy water. Throughout the grinding process, each otolith was checked periodically to avoid over grinding and to achieve adequate polishing. After drying the slide, increments were read under oil emersion at 1000× and 2000× magnification, using a stereo compound light microscope. The increments were read using one of the following three methods, depending on availability of equipment and size of the fish: 1) direct reading on the microscope, using a calibrated eyepiece graticule and measuring the increments in groups of ten (suitable fish < 3 cm);

122

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 122 Growth of L. miodon estimated from diurnal increments in otoliths

Figure 7.1 Sagittae otolith of a) a juvenile (1.6 cm TL) showing 35 daily increments and b) an adult sardine (11.7 cm TL) with 552 increments, showing the focus and the shaded area within which increments were counted and the radial that was measured.

2) taking digital photographs from the microscope and read on a computer; 3) using a digitizer, a computer and a monitor connected to the microscope, to count and measure increment width using a computer programme described by Andersen & Moksness (1988).

Increments were counted in the cauda of the otoliths, along a radial from the focus 7 posteriorly to the margin in the region marked in Figure 7.1. The radial was selected on the basis of clarity of increments. A set of one dark and one light increment was considered to be a daily increment and used to age the fish. One sagitta per fish was counted and each otolith was counted at least twice, depending on the clarity of increments, and the counts averaged.

Estimation of growth parameters Even though the von Bertalanffy growth model (VBGM) is a prime choice in fish studies including L. miodon, models such as the Gompertz and logistic models better fit data of some species (Cochrane 1984; Marshall 1987a; Kimura 1995; Katsanevakis 2006; Chapter 3 – Chifamba & Videler 2014). Therefore, to find the model that best describes our length at age data, we fitted the following four commonly-used growth models to all data sets and estimated the growth parameters using Sigma Plot 12:

1) VBGM L(t) = L∞ (1 – exp (– k1 (t – t0)))

where L(t) = total length at age t (cm), L∞ = asymptotic length (cm), t = age (months)

at capture, k1 = relative growth parameter (monthly) and t0 = age at which individuals would have been at zero length;

2) Gompertz L(t) = L∞ exp (– exp (– (t – t1) / k2))

123

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 123 PART II – BIOLOGY AND IMPACTS OF LIMNOTHRISSA MIODON

where L(t) = total length at age t (cm), L∞ = asymptotic length (cm), t = age (months)

at capture, k2 = rate of exponential decrease in relative growth rate (monthly) and t1

= 1 / k2 ln λ, where λ = theoretical initial growth rate at zero age;

– k 3) Logistic L(t) = L∞ / (1 + (t / t2) 3)

where L(t) = total length at age t (cm), L∞ = asymptotic length (cm), t = age (months)

at capture, k3 = relative growth rate (monthly) and t2 = inflection point of the sigmoid curve;

b 4) Power L(t) = a1 t

where L(t) = total length at age t (cm), L∞ = asymptotic length (cm), t = age (months)

at capture, a1 = scaling factor or rate at which length increases with time, and b = exponent describing the rate of change in the relationship between length and age. An evidence-based approach was used to select the model that best explains the data in order to avoid making prior assumption of a model. Models were compared using the Akaike Information Criterion (AICc) and the R2 value (the coefficient of determination) (Anderson et al. 2000; Burnham & Anderson 2002; Burnham et al. 2011). The age in days was divided by the average number of days in a month (365.25 days and 12 months in a year) to convert days to months. The mean length at age was calculated and used to compare growth trajectories among years.

Age and length data from 2013 were used to estimate the age (A50) and length

(L50) at first maturity, as the value at which 50% of the fish in the population (estimated from the sample) are mature (Trippel & Harvey 1991; Chen & Paleheimo

1994; Piñeiro & Saínza 2003; Karna & Panda 2011). In estimating A50, the fish in the sample were allocated to two-month age classes, such that Class 0 contained fish from 0.0 to 1.9 months old, Class 1 fish from 2.0 to 3.9 months, and so on.

Length classes for estimating L50 were made at 1-cm intervals, such that they contained fish lengths from 1.0 to 1.9 cm, 2.0 to 2.9 cm, and so on. The age and length at which 50% of the fish in the sample reached maturity, were estimated from sigmoidal curves fitted to plots of the percentage of mature fish in a class against the age and length class. After comparing several methods to estimate maturity in fish, Trippel & Harvey (1981) recommended the use of maximum likelihood fit of a sigmoidal function on data that display successive increase of mature fish with length or age. Similarly, Chen & Palohemo (1994) recommended the use of a two-parameter logistic model, fitted using non-linear least squares method to the estimated 50% maturity. Since L. miodon data showed successive increase of maturity with

age and length, L50 and A50 could be estimated from the sigmoid function given below, using Sigma Plot 12:

124

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 124 Growth of L. miodon estimated from diurnal increments in otoliths

M = 100 / (1 + exp (- (a1 - a2) / b))

where M is the percentage of mature fish, at the age or length a1, a2 is the age or length at the inflection point or mid-point of the sigmoid curve, and b is the slope or steepness of the curve. The curve’s maximum value was set at 100 representing 100% maturity. The length at first maturity is where M equals 50% mature fish,

which in this model is a2.

Results Validation of increments Sagittae of 32 fish, varying in size between 43 and 71 mm (TL), were examined. Twenty- seven showed a marginal increment and were used in the analysis. Magnifications of 4 000 – 16 000× revealed successions of increments consisting of dark and light material (Figure 7.2). Figure 7.3 shows the typical dark and light margins observed at various times of the day. During daytime, eight out of ten sagittae had a dark increment at the margin (Figure 7.4). Of the fish caught at night, all sagittae investigated had a light increment at the margin. The dark increment starts to increase in the morning and peaks around noon. The deposition pattern of the light increment is not 7 clear but appears to start forming in the afternoon (Figure 7.5).

Figure 7.2 Diurnal successions of light and dark increments showing the differences in the density of the protein matrix in a decalcified section of an otolith, viewed at a magnification of 16 000×. The scale bar is 0.5 µm.

125

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 125 PART II – BIOLOGY AND IMPACTS OF LIMNOTHRISSA MIODON

Figure 7.3 Dark and light increments (marked with bracket) at the edges of otoliths taken from L. miodon caught in daytime at a) 10.00 h and b) 14.00 h, and at night at c) 22.00 h and d) 04.00 h (Scale bar = 1 µm).

Increment width and ageing A total of 349 fish were aged and used to study growth of L. miodon. The largest fish analysed was 16.5 cm and the oldest had 815 daily increments. Ontogenetic change in the shape of the otolith has implications to the readability of the sagittae. Distortion of the increments causes difficulties in resolving the increments, especially in specimens > 8 cm. The shape of the sagittae changes with age from almost circular in larvae, to an elongate shape due to an enlarged anterior rostrum in adult fish (Figure 7.1). The antirostrum is another projection in the anterior of the otolith that is separated from the rostrum by a notch. In this study, increments along the radial from the focus to the anterior margin were most consistent among specimens. Therefore, this radial was used to study increments and age. The focus is the centre of the area known as the nucleus which is formed early in the development of the otolith, around which the increments form.

126

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 126 Growth of L. miodon estimated from diurnal increments in otoliths

sunrise sunset

2 Number of fish (n) fish of Number

0 08:00 12:00 16:00 20:00 00:00 04:00 Time of day (h) Figure 7.4 Frequency distribution of L. miodon specimens with dark (black bars) or light (white bars) increments at the edge of the otolith, at a given time of the day (sunrise at 06:26 h and 7 sunset at 17:38 h on the day of sampling). There is no data for 16:00 h (n = 27).

sunrise sunset

2.0

m) 2

 R = 0.81 p = 0.0006 1.5

1.0

0.5 Edge increment width (

0.0 08:00 12:00 16:00 20:00 00:00 04:00 Time of day (h) Figure 7.5 The width of dark (black dots) and light (white dots) increments at the edge of the otolith, at a given time of the day (sunrise at 06:26 h and sunset at 17:38 h on the day of sampling) (n = 27).

127

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 127 PART II – BIOLOGY AND IMPACTS OF LIMNOTHRISSA MIODON

a) radial 1996 600

500

400

300 Radial (µm) Radial 200 < 10 m depth > 20 m depth 100 logarithmic all

0 02468 Total length (cm) b) increment width 1996 10

Increment width (µm) width Increment 2 <10 m > 20 m 0 0 50 100 150 200 250 300

Increment number from focus c) increment width 1992 10

4 Increment width (µm) width Increment 2

0 50 100 150 200 250 300

0 Increments number from focus Figure 7.6 a) Relationships between fish length and radial of sagittae, from fish caught in water < 10 m and > 20 m deep in 1996 (n = 154). A logarithmic curve is fitted through all data, b) change in increment width (± standard deviation) of daily increments against increment number starting from the focus to the margin of sagittae from fish caught in 1996 in shallow and deep water, c) change in increment width (± standard deviation) against increment number in 1992 (n = 30). 128

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 128 Growth of L. miodon estimated from diurnal increments in otoliths

The length of the radial has a logarithmic relationship with fish length, signifying the slowing down of growth with age (Figure 7.6a). For the deep-water (> 20 m) fish, the width of the daily increment increases with age until around the 100th increment, after which it decreases (Figure 7.6b, c). Increment width of otoliths from fish caught in shallow water (< 10 m) is smaller than that of fish caught in open water (> 20 m). Shallow water fish also show an earlier peak in increment width than their conspecifics from deeper water (Figure 7.6b). The average increment width was 3.8 ± 2.2 µm, 3.34 ± 0.15 µm and 1.74 ± 0.15 µm in fish caught in deep water in 1992 (n = 30) and 1996 (n = 57), and in shallow water in 1996 (n = 97), respectively.

Table 7.1 Model parameters and their standard error (SE), R2 (computed in Sigma Plot) and AICc values for the VBGM, Gompertz, logistic and power models. Highlighted are statistically significant 2 (p < 0.05) values of L∞, k and t0, and the best R and AICc values (`Did not converge` is when the programme could not make an estimate because data does not fit the model well).

2 Model L∞ ± SE k / a ± SE t0 / b ± SE R AICc VBGM 2013 51.9 ± 43.2 0.01 ± 0.01 ‐0.25 ± 0.45 0.89 33 Females Did not converge 7 Males 37.3 ± 44.0 0.02 ± 0.02 ‐0.50 ± 0.50 0.82 1996 12.4 ± 0.9 0.10 ± 0.01 0.21 ± 0.13 0.90 ‐47 1993 25.2 ± 7.3 0.04 ± 0.02 1.21 ± 0.78 0.87 32 All data 53.9 ± 23.9 0.01 ± 0.01 ‐0.65 ± 0.25 0.89 78 Gompertz 2013 15.2 ± 1.4 8.00 ± 0.99 8.53 ± 0.94 0.89 31 Females 17.5 ± 1.9 9.32 ± 1.11 10.42 ± 1.13 0.93 Males 10.6 ± 1.3 5.96 ± 0.95 6.12 ± 0.88 0.84 1996 9.6 ± 0.2 3.10 ± 0.23 3.47 ± 0.13 0.91 5 1993 18.0 ± 1.8 7.62 ± 1.39 8.94 ± 1.09 0.88 ‐13 All data 18.1 ± 1.1 9.37 ± 0.68 9.50 ± 0.72 0.89 80 Logistic 2013 61.0 ± 67.6 ‐1.02 ± 0.15 78.04 ± 118.44 0.89 33 Females 2.2×103 ± 1.0×105 ‐0.97 ± 0.13 4.3x103 ± 2.0x105 0.92 Males 50.2 ± 90.2 ‐0.96 ± 0.18 78.53 ± 197.80 0.81 1996 12.6 ± 1.1 ‐1.42 ± 0.12 6.70 ± 1.07 0.90 5 1993 25.5 ± 7.8 ‐1.47 ± 0.31 17.40 ± 7.53 0.87 32 All data Did not converge Power asymptotic a b 2013 0.80 ± 0.09 0.91 ± 0.04 0.89 31 Females 0.67 ± 0.07 0.97 ± 0.04 0.92 Males 0.81 ± 0.09 0.87 ± 0.05 0.82 1996 1.26 ± 0.05 0.76 ± 0.02 0.88 ‐1 1993 0.99 ± 0.22 0.88 ± 0.08 0.86 32 All data 1.01 ± 0.05 0.85 ± 0.02 0.89 77

129

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 129 PART II – BIOLOGY AND IMPACTS OF LIMNOTHRISSA MIODON

a) 2013 juveniles, females & males c)1993 20 20

15 15

10 10 Total length (cm)

Total length (cm) Females 5 5 Males Females Males Juv eniles 0 0

b)1996 d)1988 20 20

15 15

10 10 Total length (cm) Total length (cm) 5 5

0 0 0 5 10 15 20 25 30 0 5 10 15 20 25 30 Age (months) Age (months) Figure 7.7 The relationship between the number of increments and total length of L. miodon in a) 2013 juveniles, males & females, b) 1996, c) 1993 and d) 1988, fitted with Gompertz growth models.

Selection of growth models All the four growth models used fit the data well and explain between 86 and 91% of the variation (p < 0.05; ANOVA). The Gompertz model has the highest R2 values. This is also the best model based on the AICc values for the 2013 and 1993 data (Table 7.1). VBGM is the best only for the 1996 data based on the AICc. All parameters estimated with the Gompertz and the power models are significant (t-test, p < 0.05). For VBGM and the logistic model, the estimates for 2013 are not significant, and the estimates of L∞ are high and therefore not reliable for data set comparisons. Results from the AICc criteria vary, but the power and the Gompertz model fit the 2013 data best (Table 7.1). The Gompertz model appears to be the best for making comparisons among years.

Comparison of growth between years using the Gompertz model

The growth trajectories of L. miodon vary among years, with L∞ being highest in 1993 and relatively low in 1988 and 1996 (Figures 7.7 and 7.8). In 1993, 1996 and

2013, L∞ was 18.0, 12.4 and 15.2 cm, respectively (Table 7.1). In 1993, the estimate

for L∞ was larger than the largest fish in the sample, which measured 16.5 cm TL.

There were annual differences in growth of different age groups. Even though L∞ was lowest in 1996, fish below 10 months of age that year grew distinctly faster than in other 130

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 130 Growth of L. miodon estimated from diurnal increments in otoliths

1988

Length (cm) 8 1993 1996 2013 4 1988 1993 1996 2013 0 0 5 10 15 20 25 30 Age (month) Figure 7.8 The relationship between the mean length at age and the age of L. miodon in 1988, 1993, 1996 and 2013 fitted with Gompertz growth model. 7 Table 7.2 Mean length‐at‐age (in cm), standard deviation (SD) and the number of fish (n) analysed for each age group of L. miodon in 1993, 1996 and 2013. Age 1993 1996 2013 (months) Length ± SD n Length ± SD n Length ± SD n 0 1.0 ± 0.1 15 1.4 ± 0.3 3 1 1.6 ± 0.4 30 1.4 ± 0.4 19 2 2.4 ± 0.8 3 2.2 ± 0.8 78 1.5 ± 0.2 13 3 2.3 ± 0.4 7 3.5 ± 0.9 31 2.2 ± 0.5 14 4 2.2 ± 0.2 4 4.0 ± 1.7 2 3.1 ± 1.1 10 5 4.4 1 6.4 1 3.8 ± 1.6 9 6 2.5 ± 0.6 3 7 4.2 ± 0.5 2 8 8.0 1 8.8 1 5.4 ± 2.2 2 9 5.7 1 7.9 1 7.7 ± 3.0 4 10 9.3 ± 4.4 3 6.9 ± 2.2 5 11 5.9 1 8.9 ± 0.2 3 7.5 ± 2.1 6 12 9.1 ± 0.6 5 8.1 ± 0.6 3 13 9.0 ± 4.1 3 9.1 ± 1.6 2 8.7 ± 2.2 5 14 11.8 ± 2.2 4 8.6 ± 0.1 2 15 11.1 ± 3.3 2 10.0 ± 0.9 4 16 11.3 ± 3.8 2 9.6 ± 0.0 2 9.1 ± 1.9 4 17 14.1 ± 1.5 2 11.1 ± 2.3 4 18 15.1 ± 0.1 2 9.0 1 11.9 ± 0.2 2 19 14.0 ± 1.4 2 12.6 ± 1.4 2 20 14.8 1 13.0 ± 0.7 4 23 14.7 ± 1.3 2 24 15.4 ± 1.1 2 26 16.3 ± 0.3 2

131

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 131 PART II – BIOLOGY AND IMPACTS OF LIMNOTHRISSA MIODON

years (Figure 7.8; Table 7.2). In comparison to 1993, L. miodon shows a generally slower

growth to a lower L∞ in 2013.

Sex and individual growth comparison Apart from the annual differences in growth, a difference between sexes was found. All large fish (> 13.6 cm) collected in 1993 and 1994 were females. Furthermore, there were no male fish above 8 cm TL in the samples from 2013. Generally, L. miodon of the same age group may show a large variation in size (Figure 7.7).

Age and length at first maturity

To estimate age at first maturity (A50), we used 104 fish of 0 – 17 months old (Table 7.3). The youngest mature fish observed was a three months old female, and age at first maturity was 8.02 and 7.90 months for females and males, respectively. All fish were mature at 12 months of age (Figure 7.9a). The relation between fish age (a) and the proportion of mature fish (M) is described by the function:

Females M = 100 / (1 + exp (- (a - 8.02) / 1.82)) (R2 = 0.99; p < 0.0001) Males M = 100 / (1 + exp (- (a - 7.90) / 2.64)) (R2 = 0.79; p = 0.0031)

A total of 173 fish of 1.0 – 8.0 cm TL were used to estimate length at first maturity (L50) (Table 7.3). The smallest mature fish was a female of 2.8 cm long, and all specimens of length classes 6 – 6.9 cm and above were mature. The estimated length at first maturity was 3.43 and 3.63 cm for females and males, respectively. Figure 7.9b shows the sigmoid relationship of length (L) and % maturity (M) described by the functions:

Females M = 100 / (1 + exp (- (L - 3.43) / 0.38)) (R2 = 1.0; p < 0.0001) Males M = 100 / (1 + exp (- (L - 3.63) / 0.44)) (R2 = 1.0; p < 0.0001)

Discussion Validation of increments When examined under TEM, decalcified otoliths of L. miodon show alternating dark (opaque) and light (translucent) increments. We assume that the different increments represent the cyclic deposition of organic and mineral material, which agrees with the results of other studies. Demineralised trout otoliths are composed of proteins (48%), collagens (23%), and proteoglycans (29%), and some carbohydrates and lipids which remain after decalcification (Borelli et al. 2001; Payan et al. 2004; Guibbolini et al. 2006). In our study, the dark increments were deposited during the day, and most of the light ones at night. Similar results were obtained by Takagi et al. (2005) who suggest that formation of the dark increments in rainbow trout (Oncorhynchus mykiss) is caused by increased synthesis of the otolith matrix protein otolin-1 (Murayama et al. 2002) in the early hours of daylight.

132

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 132 Growth of L. miodon estimated from diurnal increments in otoliths

Table 7.3 The numbers of immature and mature fish used to estimate the proportion of mature fish in a sample of L. miodon caught in the Sanyati basin. Size group Immature Mature Total % Mature Age class (months) 0 – 1.9 22 0 22 0 2 – 3.9 26 1 27 4 4 – 5.9 12 7 19 37 6 – 7.9 3 2 5 40 8 – 9.9 1 4 5 80 10 – 11.9 4 6 10 60 12 – 13.9 0 6 6 100 14 – 15.9 0 5 5 100 16 – 17.9 0 5 5 100 Totals 68 36 104

Length class (cm) 1.0 – 1.9 39 0 39 0 2.0 – 2.9 26 1 27 4 3.0 – 3.9 17 10 27 37 4.0 – 4.9 4 36 40 90 7 5.0 – 5.9 1 15 16 94 6.0 – 6.9 0 6 6 100 7.0 – 7.9 0 7 7 100 8.0 – 8.9 0 11 11 100 Totals 87 86 173

The lighter increments of decalcified otoliths of L. miodon contain a lower density of materials than the darker ones, representing diurnal fluctuation in the deposition of matrix proteins. Our results suggest that more proteins were deposited during daytime, which agrees with the observed periodicity of protein concentrations found in the endolymph of other fish species (Edeyer et al. 2000); Borelli et al. 2003). Trout (Onchorhynchus mykiss) show increasing protein levels in the endolymph after the dark–light transition, peaking approximately 5ꞏh after the beginning of the light period (Borelli et al. 2003). Edeyer et al. (2000) reported anti-phasic fluctuation of protein in the endolymph of turbot (Psetta maxima) with higher protein concentrations during daytime. Our study shows that the increment at the edge of a decalcified otolith of L. miodon, viewed under a TEM microscope, can be used to validate growth increments of otoliths. The periodicity of protein and calcium deposition follows a diurnal cycle. Hence, a pair of light and dark increments can be taken as a daily band and be used to age the fish.

133

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 133 PART II – BIOLOGY AND IMPACTS OF LIMNOTHRISSA MIODON

Figure 7.9 The percentage mature female and male L. miodon in each a) age class (n = 104) and b) length class (n = 174) in the 2013 data set, with the age and length at first maturity indicated by arrows at the point where 50 % of the fish are mature. 134

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 134 Growth of L. miodon estimated from diurnal increments in otoliths

Increment width and ageing Ontogenetic change in otolith shape is a genetically programmed process that can be influenced by the environment (Gagliano & McCormick 2004; Hüssy 2008). This change in shape makes it difficult to count increments on every radial of the otolith, because of distortion and compression of increments. Studying L. miodon from Lake Tanganyika, Kimura (1995) observed that in some specimens the increments on the rostrum and the antirostrum become indistinct at the margin of the otolith. That diminishes the suitability of these radials for ageing. Increment width tends to decrease towards the margin of the otolith, making the ageing of old fish relatively difficult (Kimura 1995). In Lake Kariba, recruitment of L. miodon into the fishing ground is based on size rather than on age. Fish of different length are spatially segregated, making it easy to control the size caught in the fishery, by restricting the minimum fishing depth for commercial fishing to 20 m (Cochrane 1984; Mtsambiwa 1989). However, some of the small fish caught in shallow water appeared to be up to five months old, the age at which larger larger conspecifics already recruited into the commercial fishery. This means that fast-growing L. miodon are recruited into the offshore fishery at a younger age than slow-growing individuals, that remain in the lake margin for a 7 longer period. Such a process causes a relatively high mortality risk in the fast- growing part of the population. In an experiment using Atlantic silverside (Menidia menidia), Conover & Munch (2002) demonstrated how selective fishing can cause a genetic change from fast- to slow-growing individuals. This process may have contributed to the relatively slow growth of L. miodon in Lake Kariba in 2013, as compared to earlier years, which can be explored in future studies.

Selection of growth models Based on the significance levels reached when fitted to our data, the von Bertalanffy (VBGM), the Gompertz, the logistic and the power growth model, can all satisfac- torily describe growth independently. Without considering other possible models, the VBGM has been used to describe the growth of L. miodon in Lakes Kariba and Tanganyika (Marshall 1987a; Chifamba 1992; Kimura 1995). When compared to other models, the VBGM is not necessarily the best model to describe the growth of multiple species, or intraspecific growth patterns in different time periods (Katsa- nevakis 2006; Chapter 3 – Chifamba & Videler 2014). The power and logistic models were mostly the best to describe the growth of the tilapiine cichlids Oreochromis niloticus and O. mortimeri in Lake Kariba, using the AICc criteria (Chapter 3 – Chifamba & Videler 2014). This finding demonstrates the importance of selecting the model that best fits the available data, to improve confidence in the parameters. Even though all models used describe our data well, some of the parameter estimates

have large standard errors and are statistically insignificant. The estimates of L∞ for 2013 by the VBGM (51.9 ± 43.2 cm) and the logistic model (61.0 ± 67.6) are un- 135

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 135 PART II – BIOLOGY AND IMPACTS OF LIMNOTHRISSA MIODON

acceptable given the maximum size of the fish encountered in this study (16.5 cm). The Gompertz model is the best with respect to R2, AICc and significance of the parameter estimates. Our results show that applying symptotic growth models to un- suitable data (such as our 2013 dataset) that visually suggests an asymptotic model, can give unrealistic results.

Temporal and spatial variation in growth Growth rate of L. miodon varies between years, between lakes (Tanganyika, Kivu and Kariba), and within Lake Tanganyika (Spliethoff et al. 1983; Kimura 1995) and Lake Kariba (this study). In Lake Kariba, VBGM estimates of growth rate for 1976 (k = 0.145) and 1981 (k = 0.254) were higher than in this study (Cochrane 1984; Marshall 1987a). Many studies have shown that temporal and spatial variation in the growth rate of fish is associated with environmental conditions such as temperature and food availability (Kimura 1995; Yasue & Takasuka 2009; Pörtner & Peck 2010; Itoh et al. 2011). For example, a positive relationship between sea temperature and growth of larval Japanese anchovy Engraulis japonicus in the Kii Channel was reported by Yasue & Takasuka (2009). A deviation from the optimum temperature for larval growth in E. japonicus and Japanese sardine (Sardinops melanostictus) caused a shift in their relative abundance. This resulted in an alternation between a warm anchovy and a cool sardine fishery in the western North Pacific (Takasuka et al. 2007). Therefore, the observed temperature rise in Lake Kariba can affect the growth of L. miodon either directly or indirectly, through its effect on plankton. Rising temperature in Lake Kariba is thought to have caused changes in the plankton community, productivity, timing of stratification and the depth of the epilimnion, which may affect the quantity and quality of food for L. miodon (Chifamba 2000; Magadza 2011; Mahere et al. 2014). A similar consequence of warming was found in Lake Tanganyika, where the water circulation is reduced and hence, nutrients get locked up in the hypolimnion (O’Reilly et al. 2003; Verburg et al. 2003). Temperature, lake turnover and nutrient availability determine the composition of the phytoplankton, with cyanobacteria associated with higher temperatures (Magadza 1980; Ramberg 1987; Zhang & Prepas 1996; Cronberg 1997; Murrell & Lores 2004; Dalu et al. 2018). The ability of cyanobacteria to grow at higher temperatures than chlorophytes, was confirmed in the laboratory (Sibanda 2003; Butterwick et al. 2005). Cyanobacteria are a poor food source for zooplankton, compared to other groups of phytoplankton, because of their toxicity and morphology (DeMott et al. 1991; Wilson et al. 2006). An experimental study on the growth of larval North Sea cod showed that diatoms can affect the growth rate significantly, due to the nutrition they provide in the form of essential fatty acids (St. John et al. 2001). Therefore, the decrease in phytoplankton quality and the rise in temperature in Lake Kariba may explain the faster growth of juvenile and subadult L. miodon in 1976, 1981, 1988 and 1996, in comparison to 1993 and 2013. There is a need to 136

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 136 Growth of L. miodon estimated from diurnal increments in otoliths

directly test these linkages in Lake Kariba, in order to draw firm conclusions on the relationships between growth, population size and environmental variables. Fast growth may be accompanied by a low asymptotic length. Such a pattern was found for L. miodon caught in 1996. Fish younger than 10 months grew much faster than those caught in 1993 and 2013, yet to a much lower asymptotic length. A similar change in growth pattern was observed in another small pelagic species, the zooplanktivorous cyprinid Rastrineobola argentea from Lake Victoria (Wanink et al. 1998). In 1988, R. argentea grew at a faster rate than in 1983 but to a smaller final size, showing a flexibility in growth pattern comparable to that observed for L. miodon in this study. Chifamba & Videler (2014 – Chapter 3) showed that different growth models tend to give either high or low parameters, and that length at age can be used to compare growth among years. Our length at age analysis indicates slow growth rates in 1993 and 2013, as compared to 1988 and 1996, which could have contributed to a lower fish production. In agreement with growth patterns reported for Lake Tanganyika (Kimura 1995), we found that female L. miodon from Lake Kariba grow to a larger size than males. In 2013, the largest males were 8.6 cm long, while females reached a maximum size of 13.9 cm. In the other years, samples with large fish (> 7 13.4 cm) did not contain males. Being large is beneficial to females, as fecundity increases with body size, resulting in increased fitness (Parker 1992; Magurran & Garcia 2000).

Size at first maturity The size at first maturity of L. miodon in Lake Kariba decreased substantially, from 5.2 - 5.6 and 7.1 - 7.3 cm fork length in 1970 (Woodward 1974) to 3.43 and 3.63 cm total length in 2013 (this study) for females and males, respectively. The reported decline in the age at first maturity in Lake Kariba is confirmed by changes in the minimum length of mature fish (from years in which size at first maturity was not estimated). This decreased from 4.0 cm in 1975 - 1976 (Cochrane 1984) to 3.5 cm in 1981 - 1983 (Marshall 1993), then to 2.8 cm in this study. Comparable reductions in the smallest size at first maturity were observed in the Lake Victoria minnow (Rastrineobola argentea), Pacific salmon (Oncorhynchus spp), North Sea sole (Solea solea) and Atlantic cod (Gadus morhua). They have been attributed to high adult mortality, caused by fishing and predation (Wanink et al. 1998; Hutchings 2005; Olsen et al. 2005; Mollet et al. 2007; Morita & Fukuwaka 2007). By studying guppies (Poecilia reticulata) in either a high- or a low-predation environment, and manipulating mortality rates in nature, Reznick & Ghalambor (2005) proved that high adult mortality can reduce size and age at maturity. Therefore, the observed decrease in the size at maturity of L. miodon in Lake Kariba is consistent with fisheries-induced evolutionary change, as predicted by life history theory, where

137

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 137 PART II – BIOLOGY AND IMPACTS OF LIMNOTHRISSA MIODON

fitness is optimized by increasing reproductive effort and decreasing size and age at maturity in response to high adult mortality (Stearns & Koella 1986). The observed sizes at maturity in Lake Kariba in 2013, and before the start of the fishery in 1970, are smaller than those in Lake Tanganyika (7.5 and 6.4 cm fork length, for females and males, respectively) and Lake Kivu (6.24 and 6.06 cm fork length for females and males, respectively) (Ellis 1971; Spliethoff et al. 1983). Ellis (1971) found that maturity varied within Lake Tanganyika, not only for L. miodon, but also the lake’s other pelagic clupeid, Stolothrissa tangnicae. Ndebele-Murisa et al. (2010) reported differences in plankton composition and a lower range of phytoplankton production in Lake Kariba (0.10 – 1.76 g C m2 h-1) than in Lakes Tanganyika (0.16 – 4.30 g C m2 h-1) and Kivu (0.85 – 2.20 g C m2 h-1), which may explain the differences in the size at maturity found between these lakes. We conclude that fishing pressure and environmental variation (temperature increase in particular) cause annual variation in the growth rate of Limnothrissa miodon, and a decrease in the size at maturity, reflecting the species’ flexible life- history strategy. Our study shows that growth of L. miodon varies yearly. To capture and understand these variations and their impact on the fishery, we recommend continuous monitoring of life-history and environmental parameters such as temperature, nutrient availability, plankton composition and productivity.

Acknowledgements Fish were sampled using the fishing vessels belonging to Lake Kariba Fisheries Research Institute (LKFRI), University Lake Kariba Research Station and Mash Fishing Enterprise, for which we are thankful. The analysis of otoliths for validation was done at the Electron Microscope Unit of the University of Zimbabwe. We thank Mr Claudius Mutariswa and Mr Patrick Kurangwa for assisting in the preparation, reading and photographing of the otoliths. We are also grateful for the use of the otolith reading equipment at LKFRI. Mr Paul Daley of Zambezi Proteins provided the 1994 sample of large fish used in the study. We are indebted to the late Joel Chisaka, the boat crews, and others who assisted in many aspects of the project. Financial support from Nuffic - NFP-PhD.11/ 858 enabled data collection in 2012 and 2013, and the stay of Portia Chifamba in The Netherlands for data analysis and thesis writing.

138

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 138 Synthesis

Portia C. Chifamba

Chapter 8

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 139

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 140 Synthesis

Introduced fish species contribute more to the fisheries (capture and aquaculture) in Zimbabwe than indigenous fish species. The most important Nile tilapia (Oreochromis niloticus) and the Lake Tanganyika sardine Limnothrissa miodon are the subject of this thesis. Limnothrissa miodon contributes more than 90% to the fish catches in Lake Kariba, and the fisheries output from Lake Kariba contributes about 90% of fishery output in Zimbabwe. Even though the catches of O. niloticus are much less than those of L. miodon, O. niloticus is also of commercial importance in the artisanal fishery of Lake Kariba, and it is the prime aquaculture fish in Kariba. Knowledge on these two species is needed to secure and enhance socio-economic benefits through informed management decisions. As explained in this thesis, both species were introduced (one deliberately and the other incidentally) into the lake after it formed by damming part of the upper course of the Zambezi river in 1958. In addition to applied aspects, the fish introductions in Lake Kariba present an opportunity to learn more about the reaction of an ecosystem in the case of a planned introduction into a vacant/open niche and an introduction into a system already occupied by a congeneric that shares an ecological niche. Therefore, Lake Kariba was the focus of the current study. The creation of Lake Kariba in 1958 transformed a portion of the Zambezi River from a riverine (lotic) to a lake (lacustrine) habitat. The lake environment improved 8 the survival of the cichlids, which prefer still water and reduced the abundance of fish species that favour flowing water, restricting their distribution to the more riverine upper section of the lake (Begg 1974). The lake came with a deep water and pelagic habitat that the native fish species could not utilize. A survey of the lake by Coke (1968) showed that the native fish were restricted to the inshore area (< 15m deep). Hence, about 70% of the lake, being deeper than 17 m, was unutilized (Begg 1970). Introduction of a pelagic fish species was therefore considered necessary to convert the pelagic plankton production into fish that could be harvested to bring socio- economic benefits. A pre-introduction study of the biology of two candidate clupeid planktivorous pelagic species from Lake Tanganyika, Stolothrissa tanganicae and L. miodon, was carried out by Matthes (1965-66) in order to select the best species. Limnothrissa miodon was chosen because of its inshore breeding habits, and it was introduced into Lake Kariba in 1967 and 1968. Fry were captured from near Mpulungu (Zambia) and, after accounting for mortalities, about 362 400 L. miodon fry were released in the Sinazongwe (Zambia) area of Lake Kariba (Bell-Cross & Bell-Cross 1971). Limnothrissa miodon was considered established by 1969 (Kenmuir 1971). The species was also introduced in a natural lake, Lake Kivu (DR Congo / Rwanda; 1958- 1960) and the Itezhi-Tezhi Reservoir (Zambia; 1992) and became established. Like in Lake Kariba, these introductions were deliberate and intended to fill a vacant pelagic

141

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 141 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

niche, and to increase fish production. Escapees of L. miodon from Lake Kariba reached and populated the next Zambezi dam located approximately 250 km downstream of Kariba, Lake Cahora Bassa (Mozambique; 1976-1982). Cahora Bassa was formed in 1975 but L. miodon was not yet found in that year (Jackson & Rogers 1976). High densities were observed in the first hydro-acoustic survey in 1983 (Lindem 1983) but before 1994 no commercial fishing took place (Tweddle 2010). In all these waterbodies there is a fishery for L. miodon. Unlike L. miodon, the introduction of O. niloticus was incidental and unplanned presumed from escapees from fish farms on the shores of the Lake Kariba (Chifamba 1998). The species was first noticed in the Experimental Lakeside Gill-netting catches in 1993. Between 1993 and 2006, O. niloticus spread from areas close to the fish farming unit to the whole of the Sanyati basin and beyond (Chifamba 2006; Zengeya & Marshall 2008). Aquaculture is not the only reason O. niloticus was introduced into some parts of the country. A reservoir in Zimbabwe which was stocked O. niloticus in 1992 by recreational anglers, resulted in its introduction in downstream Lake Chicamba, Mozambique in 1996 (Weyl 2008). This introduction of O. niloticus and that of L. miodon into Lake Cahora Bassa, demonstrate that once introduced in a river system, fish can spread to the rest of the river basin. The suitability of the new environment to O. niloticus and L. miodon was evaluated through the study of their growth. Fish size and trends in the catches of O. niloticus, since its introduction, were compared to those of O. mortimeri. In addition, the study examined various aspects of competitive advantage of the exotic O. niloticus over the displaced native O. mortimeri, a species endemic to the middle Zambezi River. The aspects compared were growth, diet, aggression, size at maturity and reproductive effort.

In summary, in this thesis I explore the following questions (Figure 8.1): 1. How suitable is the new ecosystem for the two species introduced in Lake Kariba, and is their growth similar to that in their native habitat (Chapters 3 and 7)? 2. Did the introductions cause a competitive displacement of native riverine fish species or did the new species fill an open niche (Chapters 2 and 6)? 3. What are the key aspects of the biology of the two introduced species, with consequences for their competitive interaction, in reproduction (Chapter 2), growth (Chapter 3), diet (Chapter 4) and aggression level (Chapter 5)? 4. Is the current fishing intensity on the introduced species in Lake Kariba pro- ducing sustainable economic benefits or is the lake overfished and negatively affected by the changing climate (Chapters 6 and 7)? 5. What are the general lessons that can be drawn from this system for the management of river systems and new lakes and dams in Zimbabwe?

142

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 142 Synthesis

Figure 8.1 Schematic overview of the main questions addressed in this thesis, focussing on the introduction of Limnothrissa miodon and Oreochromis niloticus, the spatial and depth segregation of the inshore and pelagic fishery, and some factors affecting the catches of the pelagic fishery of Lake Kariba.

8.1 How suitable is the new ecosystem for the two introduced species? Trends in the catches of the two introduced fish species O. niloticus (Chapter 2) and L. miodon (Chapter 6) in Lake Kariba indicate that both have established populations that support fisheries. To become established, they adapted to unique environmental conditions in the new ecosystem. Life-history traits of these introduced species are assumed to reflect the suitability of their new environment, with small size and slow growth indicating poor conditions compared to other waterbodies where the fish occur naturally or were introduced. Growth parameters of O. niloticus, estimated for the first time in Lake Kariba (Chapter 3), provide insight into the suitability of the new ecosystem. The growth rate k (year -1) of O. niloticus in Lake Kariba was (mean ± SE): 0.25 ± 0.14, 0.29 ± 0.25 and

143

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 143 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

0.29 ± 0.32 in 1997, 2003 and 2010, respectively. The asymptotic size (L∞) was: 32.4 ± 3.4, 44.6 ± 0.25 and 37.8 ± 0.32 cm standard length (SL), and the maximum age was: 10, 6 and 8 years in 1997, 2003 and 2010, respectively. Compared to other water bodies, the maximum size (converted to approximate (≈) SL for other studies where necessary), growth rate and maximum age from Lake Kariba are typical (Table 8.1). Estimated

maximum size (L∞ = 21.3 to 53.9 cm) and growth rate (k = 0.07 to 0.56) of O. niloticus varies among the water bodies in Table 8.1. Oreochromis niloticus grows to a relatively large size in Lake Kariba, though the

largest asymptotic size (L∞ = 59.3 cm) was found in Lake Victoria (Getabu 1992). The largest specimen, caught by the gill net fishery in Lake Kariba, is smaller (48 cm) than the largest from Lake Turkana (≈ 53.4 cm) (Lowe-McConnell 1958). Compared to other waterbodies, growth is slow in Lake Kariba and similar only to one value from Lake Victoria (k = 0.25) but higher than in the Lagoon Coatetelco in Mexico (k = 0.07) (Getabu 1992; Gómez-Márquez 1998). Therefore, O. niloticus in Lake Kariba grows slowly to a large size, a strategy also reflected in the size at first maturity (Chapter 2). The size at first maturity in Lake Kariba was 17.91 and 18.28 cm in the data sets from 1993 – 2002 and 2003 – 2012, respectively, and the age at maturity was 2 years (Chapter 2). Size at first maturity in Lake Kariba lies between the values from other lakes (9.8 – 29 cm) (Table 8.2). Lake Victoria has the highest size at maturity, ≈ 26 and ≈ 29 cm for females and males, respectively, even though lower values were recorded in more recent years (Njiru et al. 2006b, 2018; Yongo et al. 2018). Com- pared to Lake Kariba, much smaller sizes at maturity were found in two crater lakes, Lake Nyamusingiri and Lake Kyasanduka in western Uganda (9.8 – 12.3 cm) and in eight reservoirs of varying sizes in Côte ďIvoire (9.1 – 14.7 cm) (Bwanika et al. 2004; Duponchelle & Panfili 1998). The smallest mature specimens of O. niloticus were 8.0 cm in Lake Kariba and ≈ 6.6 cm in Mississippi coastal catchments (USA) (Peterson et al. 2004). Furthermore, the age at maturity varies from between 5.6 and 10 months in the reservoirs in Côte ďIvoire, 2 years in Lake Kariba, and 2.6 years in Tabaru River & Yonaguni-jima Island in Japan (Duponchelle & Panfili 1998; Ishikawa 2013). Using studies where the size and age at maturity is given; it appears that early maturity occurs at a small size. Therefore, O. niloticus in Lake Kariba grows slowly to a large size, maturing later than some of the other populations, and at a relatively large size. This indicates that the environmental conditions in Lake Kariba are favourable, though not the best. The level of primary production, type of food available and the size of the waterbody may explain the variation in growth and maturity among waterbodies. Primary production sets limits to energy in an ecosystem hence the amount available to the fish for growth. Primary production in terms of carbon (C) fixation is 0.1 to 1.7 g C 2 h -1 in Lake Kariba, and 3.3 to 3.5 g C 2 h -1 in Lake Victoria (Ndebele-Murisa et al. 2010). This

144

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 144 Synthesis

‐1 Table 8.1 Growth parameters L∞ (cm standard length) and k (year ) of Oreochromis niloticus in African lakes, a lagoon in Mexico and Japanese rivers, obtained from the von Bertalanffy growth model. Maximum age Amax (year), sex, aging method (LF = length frequency analysis) and year of analysis (YoA) are also given. Values marked by * have been converted from total length to standard length using the equation: standard length = 0.838 total length – 0.2408 (R2 = 0.99; n = 25) from Chapter 2.

Country Lake/River L∞ k Amax Sex Method Reference / YoA Zimbabwe Kariba (i) 32.4 0.25 10 ♀♂ scale this thesis / 1997 Zimbabwe Kariba (i) 44.6 0.29 6 ♀♂ scale this thesis / 2003 Zimbabwe Kariba (i) 37.8 0.29 8 ♀♂ scale this thesis / 2010 Egypt High Dam 36.9 0.55 7 ♀ scale Yamaguchi et al. (1990) Egypt High Dam 42.8 0.38 7 ♂ scale Yamaguchi et al. (1990) Ethiopia Tana 36.3 0.5 ♀ otolith Wudneh (1998) Ethiopia Tana 34.7 0.5 ♀ otolith Wudneh (1998) Uganda Nabugabo (i) 32.7 0.52 7 ♀ otolith Bwanika et al. (2007) Uganda Nabugabo (i) 39.9 0.45 8 ♀ otolith Bwanika et al. (2007) Uganda Albert 39.0 0.50 10 ♀♂ LF Moreau et al. (1986) Kenya Victoria (i) 53.9* 0.25 12 ♀♂ LF Getabu (1992) / 1985‐1986 Kenya Victoria (i) 52.6* 0.35 8.6 ♀♂ LF Njiru et al. (2008a) 8 / 1989‐1990 Kenya Victoria (i) 48.8* 0.56 5.5 ♀♂ LF Njiru et al. (2008a) / 1998‐2000 Kenya Victoria (i) 44.9* 0.50 5.4 ♀♂ LF Njiru et al. (2008a) / 2004‐2006 Kenya Victoria (i) 38.5* 0.69 ♀♂ LF Yongo & Outa (2016) / 2014‐2015 Kenya Turkana 21.3* 0.44 ♀♂ LF Moreau et al. (1995) Ethiopia Awassa 21.8 0.57 5.1 ♀♂ otolith Admassu & Casselman (2000) Japan Tabaru R/ 32.6 0.37 8 ♀ otolith Ishikawa et al. (2013) Yonaguni‐ jima Isl (i) Japan Tabaru R/ 28.2 0.42 13 ♀ otolith Ishikawa et al. (2013) Yonaguni‐ jima Isl (i) Mexico Lagoon 29.19 0.07 ♀♂ scale Gómez‐Márquez (1998) Coatetelco (i)

(i) = O. niloticus was introduced

145

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 145 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

Table 8.2 Standard length (SL), age at first maturity (Am) and sex of Oreochromis niloticus in Lake Kariba and some other waterbodies in Africa, USA, Mexico and Japan. Values marked by * were originally in total length and were converted using the equation: standard length = 0.838 total length – 0.2408 (R2 = 0.99; n = 25) from Chapter 2.

Country Lake/River SL (cm) Am (y) Sex Reference Zimbabwe Kariba (i) 17.91 2 ♀♂ This thesis; 1993‐2002 Zimbabwe Kariba (i) 18.28 2 ♀♂ This thesis; 2003 ‐2012 Kenya Turkana 22.4* ♀ Stewart (1988) Kenya Turkana 24.1* ♂ Stewart (1988) Ethiopia Tana 18.1 ♀ Wudneh (1998) Ethopia Tana 20.7 ♂ Wudneh (1998) Kenya Victoria (i) 26* ♀ Njiru et al. (2006b) Kenya Victoria (i) 29* ♂ Njiru et al. (2006b) Kenya Victoria (i) 19.9* ♀ Njiru et al. (2008a) Kenya Victoria (i) 20.7* ♂ Njiru et al. (2008a) Kenya Victoria (i) 21.5* ♀ Yongo et al. (2018) Kenya Victoria (i) 25.7* ♂ Yongo et al. (2018) Uganda Albert 22.2* 2 ♀♂ Lowe‐McConnell (1958) Uganda George 22.2* 2 ♀♂ Lowe‐McConnell (1958) Uganda George 16.9* ♀♂ Gwahaba (1973) Uganda Nyamusingiri 9.8 ‐12.3* ♀♂ Bwanika et al. (2004) & Kyasanduka Sudan Jebel Aulia Dam 23 ♀♂ Babiker (1984) Côte Small 9.1 ‐ 14.7 0.5‐0.8 ♀ Duponchelle & Panfili ď’Ivoire Reservoirs (i) (1998) USA Mississippi 9.23* ♀♂ Peterson et al. (2004) coastal (i) Mexico Emiliano 12.4* 1 ♀ Peña‐Mendoza et al. Zapata (i) (2005) Mexico Emiliano 12.5* 1 ♂ Peña‐Mendoza et al. Zapata (i) (2005) Japan Tabaru R/ 19.45 2.5 ♂ Ishikawa et al. (2013) Yonaguni‐ jima Isl (i) Japan Tabaru R/ 17.44 2.6 ♀ Ishikawa et al. (2013) Yonaguni‐ jima Isl (i)

(i) = O. niloticus was introduced

146

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 146 Synthesis

difference may explain the comparatively lower values of growth, maximum size and maturation of O. niloticus in Lake Kariba compared to Lake Victoria. In addition, differences in the types of food consumed by O. niloticus in locality may also explain some of the differences in growth. The diet of O. niloticus in Kariba consists of benthic pennate diatoms, which contribute 83% to stomach content by volume (Chapter 4). Most stomachs contain detritus (ca 80%) and sand (ca 70%). This diet may be poorer than the omnivorous diet, including a substantial proportion of increasingly available chironomid larvae, to which O. niloticus in Lake Victoria switched during approximately 1985 – 1995 (Getabu 1994; Balirwa 1998; Wanink 1998). For example, O. niloticus in Lake Nabugabo with a relatively energy-rich omnivorous diet, grew faster than in Lake Wamala, where like in Lake Kariba, they fed on phytoplankton (Bwanika et al. 2007). Oreochromis niloticus of over 4 years of age from Lake Nabugabo were, on average, 10 cm larger than those from Lake Wamala. Lastly, the large size of O. niloticus in Lake Kariba supports the observation that small waterbodies tend to have small fish that mature early compared to larger waterbodies (Lowe-McConnell 1958; Duponchelle & Panfili 1998). This may be attributed to more stable conditions in a large lake than a smaller one, in terms of seasonal and annual variation in the amount of water and its physical and chemical properties. Growth and maturation of O. niloticus are therefore constrained by the existing environmental conditions. The general trend is that Oreochromis niloticus in impoverished conditions 8 miniaturizes by maturing early at a small size and reaching a small maximum size. Therefore, the life-history strategy of O. niloticus switches between one of growth and one of maximizing reproduction, depending on existing environmental conditions. Judging from the growth parameters and the size and age at maturity, O. niloticus has adopted a growth strategy in Lake Kariba in response to favorable conditions. This knowledge is important when introducing this fish in a new reservoir, because its growth rate and its maximum size would vary, depending on the new environmental conditions.

Growth rate (k), asymptotic size (L∞) and age at maturity of Limnothrissa miodon (Chapter 7) also provide insights on the suitability of Lake Kariba to this introduced

fish. Estimates of L∞ (± SE) from the von Bertalanffy model were 12.4 ± 0.9 and 25.2 ± 7.3 cm TL in 1993 – 1994 and 1996, respectively. The estimates for the 2013 data

(L∞ = 51 ± 43.2 cm TL; k = 0.01 ± 0.01) were unrealistic, had large standard errors, and could not be used for comparisons. The Gompertz model fitted the data better and provides a more realistic magnitude of the asymptotic length for 2013. Estimates of

L∞ (± SE) from this model were 18.0 ± 1.8, 9.6 ± 0.2 and 15.2 ± 1.4 cm TL in 1993 –

1994, 1996 and 2012 – 2013, respectively. The estimates of L∞ from the Gompertz model are generally lower than those from the von Bertalanffy model. The Gompertz and the von Bertalanffy growth models are dissimilar in such a way that the former

147

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 147 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

provides the exponential decrease in the relative growth rate, and the latter the relative growth parameter k. Since the von Bertalanffy growth model was used in all the other studies in Table 8.1, I shall use it for comparisons. Growth rate (k) of L. miodon varies within a lake, among years, and between lakes. However, the variation within the two reservoirs, both for k and L∞, is much larger than that in the two natural lakes. A large part of the variation in Lake Kariba can probably be attributed to the use of two different methods for growth assessment: length frequency analysis and otolith reading. However, the lack of sufficient growth estimates from otolith reading in the other two lakes prevents the drawing of sound conclusions on the relative suitability of Lake Kariba for L. miodon in terms of growth rate. Except for 1993 – 1994, estimates for Lake Kariba have a lower asymptotic size compared to Lakes Tanganyika and Kivu (Table 8.3). The two reservoirs, Lake Cahora Bassa and Lake Kariba, have the two smallest asymptotic sizes recorded, suggesting that conditions in these reservoirs differ from those in the natural lakes. Ndebele-Murisa et al. (2010) reported differences in plankton composition and a lower range of phytoplankton production in Lake Kariba (0.10 – 1.76 g C m2 h -1) than in Lakes Tanganyika (0.16 – 4.30 g C m2 h -1) and Kivu (0.85 – 2.20 g C m2 h -1), which may explain the differences in the growth and maximum size found between these lakes. Differences in the Lake Kariba estimates of growth parameters from length frequency analysis and otolith reading may reflect some underlying differences in the fish population. Absence of large fish in the samples used for the length frequency analysis has led to low estimates of L∞, and to the conclusion that the L. miodon population in Lake Kariba was stunted compared to the population in Lake Tanganyika (Cochrane 1984; Marshall 1987a). The lack of large fish in the samples may result from most fish dying at a small size, some ecological bottleneck, or a failure to adequately sample the littoral area where the very large fish live. Sampling at a depth between 20 and 25 m, close to the shoreline, yielded some large specimens (Chapter 6). In agreement with my findings, the largest sizes (110 – 150 mm) were caught using gill nets near the lake margins of Lake Kivu (de Iongh et al.1983). There seems to be a gap in our knowledge with respect to the spatial distribution of the fish with age, or the diet switch required at a particular size, which is evident from the cannibalistic tendency reported for large L. miodon in the littoral areas in Lake Kariba (Chapter 6) and Lake Kivu (de Iongh et al.1983). An-other different aspect in Lake Kariba is the intensive fishing pressure on the medium sized fish, which may reduce the numbers that reach a large size. These are potential areas of investigation in Lake Kariba. Overfishing may have caused the reduction in the size at maturity of L. miodon in Lake Kariba (Chapters 6 and 7) hence size at maturity may not be regulated by environmental conditions only. Size at first maturity of L. miodon in Kariba in 1970 –1972 was close to those in Lake Tanganyika and Lake Kivu, regulated

148

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 148 Synthesis

‐1 Table 8.3 Growth parameters L∞ (cm total length) and k (month ) of Limnothrissa miodon in Lakes Kariba, Tanganyika, Kivu and Cahora Bassa, obtained from the von Bertalanffy growth model. Aging method (LF = length frequency analysis) and year of analysis (YoA) are also given. Total length, where necessary, estimated from *standard length or **fork length.

Country Lake YoA L∞ k Meth Reference Zimbabwe Kariba 1993 25.2±7.3 0.04±0.02 otolith This thesis, chapter 7 Zimbabwe Kariba 1996 12.4±0.9 0.10±0.01 otolith This thesis, chapter 7 Zimbabwe Kariba 1988 13.5 0.079 otolith Chifamba (1992) Zimbabwe Kariba 1976‐77 8.1 0.145 LF Cochrane (1984) Zimbabwe Kariba 1981‐83 7.42 0.254 LF Marshall (1987a) Zaire Tanganyika 1987‐89 18.2* 0.096 LF Mulimbwa & Shirakihara (1994) Zambia Tanganyika 1990 20.6* 0.072 otolith Kimura (1995) Burundi Tanganyika 1980 16.1 0.096 LF Moreau et al. (1991) Burundi Tanganyika 1981 17.2 0.109 LF Moreau et al. (1991) Burundi Tanganyika 1982 17.2 0.095 LF Moreau et al. (1991) Rwanda Kivu 1980 16.3 ** 0.1 LF Spliethoff et al. (1983) Rwanda Kivu 1980 16.7 0.108 LF Mannini (1990) Rwanda Kivu 1983 17.3 0.109 LF Mannini (1990) Rwanda Kivu 1986 15.5 0.101 LF Mannini (1990) 8 Rwanda Kivu 1989 16.2 0.107 LF Mannini (1990) Mozambique Cahora Bassa 1993 7 0.45 LF Gliwicz (1984)

environmental conditions only. Size at first maturity of L. miodon in Kariba in 1970 – 1972 was close to those in Lake Tanganyika and Lake Kivu, though in more recent this was not the case (Chapter 7). In 2013, age and size at maturity were 8.02 and 7.90 months, and 3.43 and 3.63 cm for females and males, respectively. In 1970 – 1972, the size at maturity in Lake Kariba was 5.2 to 5.6 cm for females and 7.1 to 7.3 cm for males (Woodward 1974). Over-exploitation may have reduced the size at maturity, as observed in some fisheries (Wanink 1998; Hutchings 2005; Olsen et al. 2005; Mollet et al. 2007; Morita & Fukuwaka 2007). Early estimates of maturity in Lake Kariba were comparable to those in Lake Tanganyika (7.5 and 6.4 cm for females and males, respectively) and Lake Kivu (6.24 and 6.06 cm for females and males, respectively) (Ellis 1971; Spliethoff et al. 1983). The more recent estimate is closer to that obtained from Lake Cahora Bassa in 2003/4 (≈ 4.3 cm for females, reported as 3.9 cm FL) (Mafuca et al. 2011). Similarity between these two reservoirs

in age at maturity, also observed in growth rate and L∞, may reflect similar environmental conditions and heavy fishing pressure compared to the natural lakes, Lake

149

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 149 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

Tanganyika and Lake Kivu. Conditions in Lake Kariba in the early years without the fishery were more favourable and comparable to other lakes. Both introduced species, O. niloticus and L. miodon, have displayed flexible life- history traits in Lake Kariba. The calculated growth parameters indicate favourable, though not optimal conditions for both species. However, their establishment impacted the native species differently, depending on whether the introduction was meant to fill a vacant or an already occupied niche. This is the next question I shall consider.

8.2 Did the introductions cause a competitive displacement of native riverine fish species or did the new species fill an open niche? Catches of Oreochromis mortimeri, a native and endemic fish species of the middle Zambezi River, in the newly created Lake Kariba were initially high (1960s to 1980s) (Chapter 2). Subsequently, catches declined as nutrients from inundated vegetation were lost, and the ecosystem matured and stabilized. In the 1990s, O. niloticus, introduced into the lake through escapees from an aquaculture farm, displaced O. mortimeri in the Sanyati basin (Figure 1.1). Oreochromis mortimeri almost disappeared from the catches in less than a decade after the first appearance of O. niloticus in 1993. I found a significant negative relationship in the catches of these two fish species from 1993 to 2012 (Chapter 2). Between 1993 and 2003 – 2005, the ratio of O. mortimeri to O. niloticus changed from 1:0.1 to 1:27 (Chifamba 2006). Thus, the unplanned introduction of O. niloticus into Lake Kariba resulted in displacement of the native O. mortimeri, and the introduced species became dominant in the catches. Oreochomis niloticus and O. mortimeri share an ecological niche and as such, have similar diets (Chapter 4) and reproductive strategies (Bell-Cross & Minshull 1988; Trewavas 1983). Reproduction involves nest building in a breeding arena by the males, to attract females to lay eggs which the male then fertilizes and the female subsequently broods in the mouth. Competition between the two species could be predicted, and perhaps also the displacement of the native species. There are ex- amples of displacement of native species by O. niloticus in Africa and other parts of the world. Oreochromis niloticus displaced the native planktivores Oreochromis esculentus and Oreochromis variabilis in Lake Victoria (Balirwa 1992) and three native cyprinids in Lake Luhondo, Rwanda (De Vos et al. 1990). In Lake Chivero (Zimbabwe), O. niloticus replaced another introduced planktivorous species, Oreo- chromis macrochir, and accounts for about 95% of the catches from that lake thereafter (Marshall 1999; Tiki 2011). In Lake Chivero, both O. niloticus and O. macrochir fed mostly on blue-green algae (> 50%) and their diets, in all size classes, overlapped almost completely (Zengeya & Marshall 2008). The dietary overlap coefficient (Pianka 1973) for O. niloticus and O. macrochir in Lake Chivero was 150

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 150 Synthesis

0.98, comparable to the value of 0.95 that was found for O. niloticus and O. mortimeri in Lake Kariba (Chapter 4). Outside Africa, the introduction of O. niloticus in Gargalheiras Reservoir in Brazil, caused the decline of another introduced species, Plagioscion squamosissimus and two indigenous fishes, Prochilodus brevis and Hoplias malabaricus (Attayde et al. 2011). At the same time, the proportion of introduced species increased from 74% to 96% of the total fish catch. Like O. niloticus, Prochilodus brevis feeds mostly on sediments and microalgae that include Chlorophyceae, Bacillariophyceae and Cyanophyceae (Da Silva et al. 2010). Generally, O. niloticus tends to displace species with which it shares a niche. Hence, to avoid displacement of native species, introductions should be into a vacant niche as in the case of the introduction of L. miodon into Lakes Kariba and Kivu. Limnothrissa miodon inhabits the once vacant pelagic niche of Lake Kariba (Chapter 6). Juveniles are found in the marginal area only, while adults use both the pelagic and shallow areas where breeding may occur. The presence of L. miodon had a positive impact on tigerfish (Hydrocynus vittatus) by providing an alternative prey to this native predator. Hydrocynus vittatus showed a dietary switch to L. miodon and began to inhabit the open water in pursuit of its prey (Cochrane 1976; Marshall 1987b, 1991). From April 1969 to March 1970, only 1.5% of the stomach contents of H. vittatus consisted of sardines (L. miodon) whereas from April 1970 to March 8 1971, the amount had risen to 41.4% and remained high (Kenmuir 1973; Mhlanga 2003). Hydrocynus vittatus together with small amounts of some inshore fish species, are caught in the deep water as bycatch in the L. miodon fishery (Chapter 6). Limnothrissa miodon is the dominant fish in the pelagic area. Whilst the pelagic area was a vacant niche when L. miodon was introduced into Lake Kariba, the inshore area that became subsequently occupied by the juveniles, was not. The impact of this introduction on the inshore fish species that compete with the juvenile and adult sardines in the inshore area, is unknown. Inshore interactions of the native and the introduced fish species have never been investigated. When considering future fish introductions, all life stages of the exotic species must be included both in the pre- and the post-introduction stage, for impact assessment. A possible negative ecological impact of the introduction of L. miodon was on a zooplanktivore, Brycinus lateralis, a species from the Upper Zambezi River that was already present in Lake Kariba when L. miodon was introduced and occupying the inshore zone where juvenile L. miodon stay. Brycynus lateralis was believed to be capable of utilizing the pelagic area. In support of this view, early catches of L. miodon from the open water contained 20.5% B. lateralis (Woodward 1974). Expansion of B. lateralis into the pelagic area may have been prohibited by competition from L. miodon (Marshall 1991). Since both species are not native to

151

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 151 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

the Middle Zambezi River, and B. lateralis is still found in Lake Kariba, the introduction of L. miodon may not have had serious ecological consequences. The introduction of Limnothrissa miodon and Oreochromis niloticus into lake Kariba demonstrates the positive consequences of a planned introduction into an open niche, and the negative consequences when the introduction is into a filled niche. An introduction into an open niche has little ecological impact and can be beneficial through an increase in fish production. In contrast, the (unintended) introduction of O. niloticus led to the disappearance of O. mortimeri, a congeneric species with equal productivity. The negative impact of O. niloticus on O. mortimeri may have resulted from competitive advantage of the introduced species, that I will consider in my next question.

8.3 What are the key aspects of the biology of the two introduced species, with consequences for their competitive interaction, in growth, diet and aggression level? High reproductive effort (Chapter 2), growth rate (Chapter 3 – Chifamba & Videler 2014), diet overlap (Chapter 4) and aggression (Chapter 5 – Chifamba & Mauru 2017) could give competitive advantage to Oreochromis niloticus over the native O. mortimeri, that enabled the invader to displace O. mortimeri in Lake Kariba. In Chapter 4, the diets of O. niloticus and O. mortimeri are compared and found to be similar. Schoener’s and Pianka’s overlap indices are 0.75 and 0.95, respectively (Pianka 1973; Schoener 1974). The overlap in the diet between the two fish species is biologically significant according to the 0.60 limit suggested by Galat & Vucinich (1983). In addition, there was no significant difference in the protein content of the ingested material in the stomach, as well as in the relative digestion efficiency of the proteins, a further indication of similarity in their utilization of the food resource. As the digestion efficiencies of O. mortimeri and O. niloticus did not differ, this factor may not have given O. niloticus competitive advantage. In contrast to this finding, O. niloticus in Lake Chivero has been reported to have competitive advantage from a higher digestion efficiency compared to O. macrochir, the species it displaced from that lake (Marufu & Chifamba 2013). This suggests that the diet overlap of O. niloticus and O. mortimeri in Lake Kariba was a comparable potential competition parameter, that would favour the most fit. Even though it has been suggested that a diet of detritus may not be limited in the environment, its quality has been shown to vary in space. Bowen (1979) found that juvenile Sarotherodon mossambicus that feed in the shallow part of Lake Sibaya (South Africa) have access to detritus with a higher protein level than the adults, which inhabit and feed in deep water. Such a heterogeneity in the distribution of food resources may lead to competition between O. niloticus and O. mortimeri. Under 152

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 152 Synthesis

conditions of heterogenous distribution and inadequate resources, access may be enhanced by the ability to win a contest and dominate. Laboratory experiments showed that O. niloticus is less aggressive than O. mortimeri and that a larger fish is more aggressive than a smaller one (Chapter 5 – Chifamba & Mauru 2017). Oreochromis mortimeri attacked O. niloticus first in most of the encounters, irrespective of whether it was the bigger or the smaller of the pair. In those encounters where O. niloticus was smaller, O. niloticus made the first bite two times (8.3%). When bigger, O. niloticus made the first bite eight times (40.0%). Over a 30-min encounter, O. mortimeri was dominant and delivered significantly more bites (7.79 2.31 bites) than O. niloticus (4.53 1.53 bites). Oreochromis niloticus tended to be less aggressive, it attacked first only when it was much bigger than the േ േ opponent, while O. mortimeri attacked even when it was smaller. Considering aggression alone, it was unexpected to find higher aggression in the displaced fish. However, considering the association of large body size with higher aggression, O. niloticus would be at an advantage, because it grows faster than O. mortimeri (Chapter 3 – Chifamba & Videler 2014). At any particular age, therefore, O. niloticus would be generally larger than O. mortimeri, hence would have a size advantage in a contest. Mean length at age of O. niloticus was higher than that of O. mortimeri during 1997 – 2000 (Chapter 3 – Chifamba & Videler 2014), the period in which species replacement occurred. Oreochromis niloticus grew to a larger size (1997 – 2000: 32 ± 3.4 cm; 8 2003 – 2005: 44 ± 17.7 cm) compared to O. mortimeri (1997 – 2000: 30.2 ± 4.2 cm; 2003 – 2005: 3.6 ± 14.6 cm). Being larger at any particular size during the displacement years, gave O. niloticus competitive advantage over O. mortimeri. Dominance status has a competitive advantage because it is associated with growth rate, reproductive advantage and metabolic rate. Ten out of twelve dominant steelhead trout (Salmo gairdneri) grew faster than their paired subordinates, given the same amount of food (Abbott & Dill 1989). Tiira et al. (2009) found that dominance status affected growth in brown trout, Salmo trutta populations, with individuals having the lowest ranks growing less compared to those with a higher rank. There is also a significant association between some agonistic profiles and metabolism (Alva- renga & Volpato 1995). Oreochromis mortimeri has competitive advantage through a high aggression level, which is counteracted by the size advantage of O. niloticus in a contest for resources. Competition between O. niloticus and O. mortimeri for other resources, such as nesting and nursery space, is inevitable because of the similarity in their reproductive behaviour. Breeding males in African cichlids, including O. niloticus and O. mortimeri, make and fiercely defend their nest in an arena. The best nests are optimally positioned for visits by females, and they are occupied by the dominant males (Philippart & Ruwet 1982). Compared to their subordinates, dominant male Oreochromis

153

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 153 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

niloticus have the highest gonadosomatic index and higher levels of gonadotropin hormones that trigger spermatogenesis, whereas the subordinates show reduced gene expression of key factors for steroid production (Pfennig et al. 2012). Reproductive success is higher in dominant fish because they occupy the best nesting sites, which attracts potential mates (Philippart & Ruwet 1982; Seppänen et al. 2009). Hence, the

larger O. niloticus males (L∞ = 32.2 and 44.6 cm SL in 1997 – 2000 and 2003 – 2005, respectively) would be at a competitive advantage for mating and reproduction in

comparison with the smaller O. mortimeri (L∞ = 30.2 and 36.8 cm SL in 1997 – 2000 and 2003 – 2005, respectively) (Chapter 3 – Chifamba & Videler 2014). Oreochromis niloticus has a higher reproductive effort than O. mortimeri, lending it a comparatively greater capacity to increase its population (Chapter 2). The proportion of mature fish with ripe gonads in the catch is always higher for O. niloticus compared to O. mortimeri. In addition, O. niloticus has a slightly lower length at maturity (17.2 cm) than O. mortimeri (19.6 cm). Furthermore, the breeding of O. mortimeri seems to be associated with rainfall more than that of O. niloticus, which is mostly dependent on temperature. This implies that the breeding season of O. niloticus is comparably longer than that of O. mortimeri, which seems to depend more on rainfall as a signal for gonad maturation. Higher reproductive effort of O. niloticus could have also given this species competitive advantage. One important factor that was not considered in this study because of prohibitive costs, is hybridization. Congeneric species such as Oreochromis niloticus and O. mortimeri have a high chance of hybridization, and hybrids within this genus have been observed in nature (Gregg et al. 1998; D’Amato et al. 2007). In the Limpopo River, introduced O. niloticus hybridized with the native O. mossambicus (D’Amato et al. 2007). Gregg et al. (1998) reported that O. mortimeri in Lake Kariba has O. macrochir genes, indicating hybridization following the introduction of the latter species into Lake Kariba. The authors argue that there were no recognizable O. macrochir in the lake because their genes may have been swamped by those of the more numerous O. mortimeri. Due to back crosses or introgressive hybridization, hybrids of O. niloticus with O. macrochir in Lake Itasy (Madagascar) and Lac Ihema (Rwanda), and with Oreochromis variabilis and possibly Oreochromis esculentus in Lake Victoria, resemble O. niloticus (Trewavas 1983; Kudhongania & Chitamwebwa 1995; Lowe-McConnell 2000). Progeny of introgressive hybridization will have phenotypic resemblance of one of the parents. It is possible that O. mortimeri has formed hybrids with O. niloticus in Lake Kariba, but these are not visually identi- fiable. Suspected hybrids, with intermediate characteristics between O. niloticus and O. mortimeri were caught during sampling. This suggests that the introduction of O. niloticus may have caused a decline in genetic diversity, through hybridization and

154

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 154 Synthesis

competition. A genetic study to determine hybridization is essential to establish if this has indeed occurred. Each of the factors discussed here could be associated with a small competitive advantage for Oreochromis niloticus. These advantages act simultaneously, and their effect would be cumulative and synergetic over generations. Hence, combined effects of all the many traits of O. niloticus give it a competitive advantage over the native O. mortimeri. The next question is whether the introduction was beneficial to the capture fishery.

8.4 Did it improve the fishery or just replaced native species that could have been fished otherwise? Since Oreochromis niloticus appeared in the Lakeside experimental gill net fishing programme catches in 1993, the catches of this species increased. At the same time, those of O. mortimeri declined until the species disappeared in 2006, except for an occasional catch (Chapter 2). About 50% of the variation in the catches of O. mortimeri could be explained by the variation in O. niloticus catches. The latter rose to a peak in 2000, thereafter remaining stable with a mean catch of 2.8 kg per set. Combined catches of these two species show the replacement of the O. mortimeri by O. niloticus, with the trends in the catches strongly fluctuating between years but without a clear trend in the total catch during the period of displacement from 1992 to 2012 (Chapter 2). 8 The total amount of fish harvested from Lake Kariba did not improve with the introduction of O. niloticus, an outcome that is also found elsewhere. In Gargalheiras Reservoir, Brazil, introduction of O. niloticus did not improve the overall catches. Following the introduction of O. niloticus, the catch per unit effort of introduced Plagioscion squamosissimus (Heckel) and the indigenous species Pro- chilodus brevis, (Steindachner) and Hoplias malabaricus (Bloch) declined, but the overall catches did not (Attayde et al. 2011). In addition, the number of fisherman and their per capita income did not change. Therefore, in terms of increasing fish production the introduction of O. niloticus may have been unnecessary in this reservoir, as well as in Lake Kariba. The introduction of Limnothrissa miodon in Lake Kariba, in contrast to that of O. niloticus, resulted in a highly productive pelagic fishery, landing 24 000 tons at its peak (Chapter 6). In terms of energy flow in the ecosystem, the new species must utilize a niche that is not already occupied by another fish caught in the fishery. Unlike O. niloticus, that feeds on the same food as O. mortimeri, Limnothrissa miodon utilizes pelagic plankton, that inshore fish are unable to use, converting it into fish production.

155

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 155 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

I conclude that the introduction of Oreochromis niloticus into Lake Kariba was ecologically unnecessary, and a mistake in terms of conservation of biodiversity, because it displaced the native endemic Oreochromis mortimeri.

8.5 Is the current fishing in Lake Kariba economically and ecologically sustainable? Fishing offers social and economic benefits in the form of food, revenue, income and livelihood, thus contributes to food security. These benefits are derived from the fishing operation itself and the associated activities such as marketing of fish, and the manufacture of the fishing necessities (fishing nets, vessels, and equipment). Economic and social pressures drive exploitation, because the venture must be worthwhile for fishing to continue. Maximum economic benefits are derived when optimum fishing effort is applied, and the fish population can sustain itself. Environ- mental factors are key to the natural maintenance of the fished population through its effect on growth, reproduction and mortality. Hence, unfavourable conditions can hamper the ability of the population to sustain itself. A balance between factors that diminish the fish population (fishing and natural mortality) and those that increase the population (reproduction and growth) is needed for sustainable fishing. The sustainability of the fisheries of the introduced fish in Lake Kariba will be discussed in this section.

Economic impact of the L. miodon and O. niloticus fishery The impact of introduced fish species, particularly those introduced for the purpose of improving fish production or for sport fishery, should be judged in an economic as well as an ecological sense. Ecological and economic impacts can be profound. A classic example is the disappearance of the native and diverse flock of haplochromine cichlids following the introduction of Nile perch, Lates niloticus in Lake Victoria (Witte et al. 1995). However, in social and economic terms the fishery for L. niloticus had significant positive outcomes, acknowledged and assessed by Reynolds & Greboval (1988) and Njiru et al. (2005). In Lake Victoria, the artisanal fishery was replaced by a multimillion-dollar export industry and the government derived benefit from licensing of fish processing and fishmeal production. Many new jobs were created in the processing and fisheries-related industries. Similarly, the introduction of L. miodon in Lake Kariba brought into existence a semi-industrial fishery for this pelagic species, which strongly increased the total fish production. The fishery for L. miodon is more productive than the artisanal fishery and contributed about 95% of the total landings (16 387 tons) in 1996 in Zimbabwe (unpublished data from Lake Kariba Fisheries Research Institute, LKFRI). The trends in effort and catches in Lake Kariba are described in Chapter 6. Annual 156

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 156 Synthesis

commercial L. miodon catches landed in Zimbabwe reached a maximum of about 22 000 tons in 1990 but decreased to about 2 500 tons in 2011 (unpublished data from LKFRI), representing a huge loss in food production. At its peak, the Zambian and Zimbabwean L. miodon fishery at Lake Kariba generated about USD 37.7 million per year, against USD 20 million in 2013 (unpublished report of the Programme for the Implementation of a Regional Fisheries Strategy for the Eastern and Southern Africa - Indian Ocean region). In contrast to the intended introduction of L. miodon, the introduction of O. niloticus seems to have had no effect on the total landings, as it was almost directly replacing the native O. mortimeri (Chapter 2). However, the fishery for the introduced O. niloticus is important in many lakes. For example, in Lake Chivero (Zimbabwe) the species contributes about 95% to the total catches (Tiki 2011). The yield of introduced O. niloticus by the artisanal fishery on Lake Chicamba (Mozambique) was 5.2 tons per month (Weyl 2008). An increase in mesh size may have occurred in 2003, to enable the fisherman to catch the larger O. niloticus. This suggestion is based on size selectivity of gill nets and the size distribution of the catches landed by the artisanal fishery (Chapter 2). The size of fish caught by the artisanal fishermen and the mesh size in use need to be monitored and managed to ensure the sustainability of the O. niloticus fishery in Lake Kariba. In economic terms, the use of O. niloticus in Kariba for aquaculture was a huge 8 success in terms of boosting fish production and economic activity. Lake Harvest Aquaculture produce about 9 000 tons of fish per year in Kariba (Zimbabwe) and 2 000 tons in Siavonga (Zambia). That production surpasses the gill-net fishery catches (about 5 000 tons per year, all species in Zimbabwe). When assessing the impact of introductions, the economic aspects should be taken into consideration.

Sustainability of the fisheries The estimates of Maximum Sustainable Yield (MSY) and fishing effort were used to make inference on the sustainability of the fishery. Using the Schaefer and Fox models, MSY was estimated from combined catch and effort data from Zimbabwe and Zambia for L. miodon. MSY estimated for the period 1974 to 2011, was 23 185 (Schaefer) and 22 355 (Fox) tons per annum (Chapter 6). The effort (Fmsy) used to catch the MSY was 10 867 and 11 226 boat nights for the Schaefer and Fox models, respectively. These estimates are lower than those made by Pitcher (1995) who used production from a baseline lake, where the species forms a successful fishery as a ‘predictor of ecology’. Pitcher’s estimate of the yield of L. miodon in Lake Kariba is 27 000 to 36 000 tons per year. The maximum fish landing of 30 000 tons of sardines from Lake Kariba is higher than my estimates of MSY, which indicates that overfishing is the main factor that caused the decline in the catches of L. miodon.

157

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 157 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

Overfishing of L. miodon occurred in spite of the recommendation by the fisheries managers within the Zambia/Zimbabwe SADC Fisheries Project, to limit the number of sardine-fishing rigs/vessels to 230 and 270 for Zambia and Zimbabwe, respectively. The Protocol on Economic and Technical Cooperation stating how the two countries were going to manage the fishery, was signed in 1999. But both countries did not adhere to the effort limits. There were 460 rigs in Zimbabwe and 962 in Zambia in 2013, excluding an unknown number of illegal fishing vessels (Paulet 2014). Lake Kariba is a transboundary resource, bringing its own political complexity into the management of the fishery, which has to do with laws, policies and practices. Effort increases in Zimbabwe arose largely from attempts to address the race imbalance in the allocation of fishing permits which, before independence, were reserved for whites. One of the problems in Zambia was the ineffective management of the number of fishing vessels. Rises in fishing effort despite the knowledge of the limits to the L. miodon fishery, highlights the importance of including social and economic factors in the management of a natural resource. The current management objectives are to reduce the fishing effort by limiting the number of fishing vessels and fishing nights, in order to fish within the limits of MSY. These objectives should broaden and include variation in MSY. MSY may not be the most appropriate management objective, because it varies annually according to environmental conditions and other factors such as fishing, which may affect life- history parameters and replacement rate of the population. Growth of the sardines and population size of the juveniles and adults need to be monitored. Fishing effort, measured as boat nights, also changes. Fishing vessel and gear improvement increase the fishing power of the fishing units (Chifamba 1995). More recent developments include mobile phones that allow the fishermen to communicate with each other more intensively, even during fishing. Effective fishing effort rises, hence a need to adjust for vessel improvement or perhaps, the use of catch quotas to regulate the fishery and prevent overfishing. The size at maturity of L. miodon in Lake Kariba decreased from 5.2 – 5.6 cm and 7.1 – 7.3 cm in 1970, to 3.43 and 3.63 cm in 2013 for females and males, respectively (Chapter 7). A high adult mortality (such as in fishing) tends to reduce size and age at maturity. Reznick & Ghalambor (2005) proved this in a study of guppies (Poecilia reticulata) from Trinidad, which live in either a high- or a low-predation environment, and in experiments where mortality rates were manipulated in nature. The decrease in the size at maturity of L. miodon in Lake Kariba is consistent with fishery-induced evolutionary change. According to life-history, fitness in fish is optimized by an increase in reproductive effort and a decrease in the size and age at maturity, in response to high adult mortality (Stearns & Koella 1986). Fishery-induced evolution of fish can make it hard for the fishery to recover, even when the fishing effort is reduced.

158

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 158 Synthesis

Environmental effects on fish manifest on the reported negative relationship between air temperature and catches of L. miodon in Lake Kariba (Chapter 6). The mean maximum air temperature in Kariba rose from 29 to 32 0C between 1974 and 2008; a manifestation of global climate change. In the absence of continuous data on water temperature, it can be assumed that the water temperature rose together with air temperature (Chifamba 2000). One effect of temperature increase is the change in the phytoplankton, from a community dominated by Chlorophytes to one dominated by Cyanophytes (Sibanda 2003; Butterwick et al. 2005). Due to their toxicity and morphology, Cyanophytes are a poorer diet for the zooplankton on which L. miodon feeds, which may have led to the reduction of L. miodon catches (Magadza 2011). Catches of L. miodon are negatively correlated with mean maximum temperature in Kariba (Chifamba 2000). Another effect of high temperature is an increased stability of the thermocline, as is the case in Lake Tanganyika. This reduces the incidence of circulation that brings nutrients to the euphotic zone which would have been lost to the hypolimnion (Verburg et al. 2003). This cycling provides 95% of the phosphates and 97% of the silicon used for algae production in Lake Tanganyika (Verschuren 2003). Consequently, the stability of the thermocline in Lake Tanganyika is estimated to have reduced primary production by about 20% and fish yield by 30% (O'Reilly et al. 2003). Stability of the thermocline may have reduced the primary production in Lake Kariba in a similar 8 way and, in turn, the sardine production. New production estimates are needed to come with more current estimates of population parameters and fishable portion. Changes in productivity caused by environmental variation could be the cause of the variation in growth of L. miodon that was found in Lake Kariba (Chapter 7). Growth in 2013 was slower than in 1993 and 1996, thus increasing the time needed to grow to a fishable size and thereby reducing the rate of biomass regeneration (Figure 8.2). Environmental conditions may determine the growth and size of the sardine population in Lake Kariba. This must be borne in mind when setting catch limits. Multiple regression analysis shows that fishing effort, lake level and temperature are correlated with catches and catch per unit effort of L. miodon (Chifamba 2000). The environment and fishing thus simultaneously affect fish abundance. Therefore, to manage the L. miodon fishery sustainably, there is a need to reduce the fishing effort, assess the productivity of the lake considering the environmental variables, and determine the desired fishing effort to achieve sustainability. Sustainability of the fishery is also challenged in the case of Oreochromis niloticus and O. mortimeri, judging from the reduction in age groups in the catch. In 1997, the oldest O. niloticus and O. mortimeri were 10 and 13 years old, respectively, but in 2003 the maximum ages had reduced to 5 and 6 years, respectively (Chapter 3 – Chifamba & Videler 2014). In Lake Victoria, O. niloticus reduced its size at maturity

159

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 159 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

4 Size at maturity line Total length (cm) length Total

2 2013 1996 1993 1988 0 02468 Age (months) Figure 8.2 Effect of annual variation in the growth rate of Limnothrissa miodon in Lake Kariba on the age at maturity, marked by arrows from the maturity line drawn from the size at maturity (3.43 cm) for females in 2013.

from 35 cm TL (early 90s) to 21 (males) and 22.7 (females) cm in 1998 – 2000, a sign of overexploitation (Njiru et al. 2008a). Fishing reduces the chances of the fish to survive many years, hence many exploited fish species have few age groups of young fish (Trippel 1995). Reduction in the number of age groups in the catch reduces sustainability and resilience of the fished population. Recruitment failure or other calamities are felt more strongly in a population with fewer age groups. Hence, there is a need to monitor the age of Oreochromis niloticus and O. mortimeri, and to manage their fisheries sustainably.

8.6 What are the general lessons that can be drawn from this system for the management of river systems and new lakes and dams in Zimbabwe? The main lessons from this thesis are that: 1) dams create a new ecosystem that can be filled by exotics, and destroy the river ecosystem to which the native species are adapted; 2) reservoirs create opportunity for aquaculture, which can result in the use of exotic species that can escape into the river system, with attached costs and benefits;

160

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 160 Synthesis

3) general conditions that promote the establishment of introduced species, could guide other introductions to minimize negative impacts; and 4) natural resources need to be utilized sustainably, considering environmental variation.

Dams create and destroy ecosystems The construction of Lake Kariba created an ecosystem that allowed the cichlid species to thrive, and a pelagic niche that could be filled by introducing a pelagic fish. New dams can offer the same opportunities as Lake Kariba. In relatively shallow lakes, the open water area may be shallow enough to be utilized by the native fish species. Deep reservoirs such as the planned Batoka dam, upstream of Lake Kariba, will create a habitat in which pelagic species such as L. miodon are likely to thrive. The Kariba dam is a barrier to fish migration that eliminated the connectivity of the Zambezi River almost completely. River-adapted fish were lost when the lake was formed. One species now absent in the catches from Lake Kariba, is the cata- dromous African longfin eel (Anguilla mossambica). This eel is endemic to the Malagasy area in the south western Indian Ocean and migrates into rivers on the eastern coast of Africa, Madagascar and other West Indian Ocean islands. The dam prevents migration of glass eels into the rivers and silver eels back to the ocean. Some fish species in Lake Kariba, such as Hydrocynus vittatus and Labeo altivelis, 8 are potamodromous: they migrate upstream to breed during the rainy season. Con- struction of dams has to be kept to a minimum, and those not in use or heavily silted to be useful decommissioned, to allow free movement of fish and river continuity. Fish migration passages need to be built into every dam and weir to reduce the impact of these structures on fish migration and recruitment. Unfortunately, the loss of fish species does not raise much public concern because they are not as visible to most people as terrestrial animals. Visibility needs to be raised on the plight of fish in the face of human activities.

Dams create opportunity for aquaculture, which can result in exotic species introduction Availability of water limits aquaculture in Zimbabwe, and water storage is critical for electricity generation, agriculture and drinking-water supply for cities. Hence, a strong driver to construct reservoirs of different sizes in this country without natural lakes. Aquaculture development is strongly encouraged and backed by government policy because of its potential role to increase food production and food security. The most favoured aquaculture species are exotics, which include Oreochromis niloticus. Introduction of this species into other dams is likely to have negative biodiversity consequences, similarly to what occurred in Lakes Kariba and Chivero. In economic terms, the dilemma is that O. niloticus farming in Kariba is a huge success

161

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 161 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

with regard to boosting the fish production and the economy, thus fuelling the drive to repeat this success. The use of exotics in aquaculture is a global concern. About 30% of the fish introductions worldwide were for aquaculture, and only 17% for improvement of wild populations (Gozlan et al. 2010). Therefore, limiting fish introductions will be difficult because of their benefits to humans. However, the economic drive needs to be balanced with the environmental impacts and economic consequences of the introductions, that are often ignored. The threatened biodiversity may require costly measures to protect, and costly measures also may be needed to reduce or remove introduced fish species. Mechanical removal of exotics is most effective in small water bodies. This is also done to limit the exotic populations in large areas, with varying success. For example, mechanical removal of Channel catfish (Ictalurus punctatus) and common carp (Cyprinus carpio) from the San Juan River (Utah, USA) has reduced the population, but cannot eradicate the exotic species despite the money and time invested since the 1990s (Pennock et al. 2018). Traps and oral toxicants were developed to reduce C. carpio menace (Stuart et al. 2006; Poole et al. 2018). Bajer et al. (2011) describe how winter aggregates of common carp are detected using radio and acoustic telemetry, and subsequently removed using seine nets with high efficiency. These costs are often ignored. For Lake Kariba, research is needed to investigate possible hybridization of O. niloticus and O. mortimeri. To salvage O. mortimeri from extinction would require a substantial investment of time and money. For example, measures are needed to seek and protect any refugia populations. The Convention on Biological Diversity recognizes that the main cause of biodiversity loss in freshwater is species introduction, and that there is need to take preventative action. Introduction of exotics should be done only after proper assessment of their impact. The need to increase aquaculture in Zimbabwe should be viewed together with the need to prevent further introductions of O. niloticus or other exotics in new areas. This study found that introducing a fish into and occupied niche does not necessary increase fish production. There are tools that can be developed or adapted to screen potential candidates for introduction, such as the protocol described by Copp et al. (2005). Risk identification and risk evaluation assessment is needed prior to all introductions, and the law on introduction of exotics in Zimbabwe has to be adhered to.

Factors that determine the success of introductions The fish introductions into Lake Kariba present an opportunity to evaluate the environmental factors and biological traits of the species, that determine the probability of their establishment. The case of O. niloticus demonstrates various aspects of an introduced fish that may be associated with competitive advantage. Such data

162

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 162 Synthesis

is useful in building case studies and predictive models, to be used in risk assessments of introductions and predictions of invasiveness (Clavero & García-Berthou 2005; Ruesink 2005; Moyle & Marchetti 2006; Romanuk et al. 2009; McKnight et al. 2017). For example, Ruesink (2005) used 1 424 cases of first introduction inter- nationally, to explore how species traits, environmental traits, match between species and environment, and propagule pressure explain establishment. High probability of establishment is linked to small body size, omnivory, and high endemism in the recipient environment. Establishment is also high when humans intend the establishment (76% of the cases) as compared to the cultivation or use of fish without an explicit desire for naturalization (57% of the cases). Successful establishment and invasion can be predicted from historical invasions in other waterbodies, similarity of the native habitat to the new environment, and the tolerance to a wide range in environmental conditions (Moyle & Marchetti 2006). Oreochromis niloticus and Limnothrissa miodon share some of the above traits and conditions surrounding their introduction. Hence, their successful introduction into Lake Kariba can be used in making predictions for other cases. This thesis shows high plasticity in the life-history traits ‘growth’ and ‘size at maturity’ of O. niloticus (Chapter 3 – Chifamba & Videler 2014)) and L. miodon (Chapter 7). These traits may have enabled the two species to adapt to the new environment, which differs from their native habitat, and in the case of L. miodon, 8 to intense fishing pressure. Limnothrissa miodon has a short life span of less than two years, and maturity is reached within a year (Chapter 7). Short generation time and fast growth rate are useful traits for an introduced species, because they enable rapid population growth and high productivity (Chapter 6). The creation of the novel pelagic ecosystem and the initial high productivity of the newly inundated reservoir were favourable for L. miodon. This species fits into Ruesink’s (2005) high establishment category by being a small sized fish that was intentionally introduced into a suitable environment, hence a model invader. Much effort was applied to carefully capture and airlift L. miodon from Lake Tanganyika into Lake Kariba, from where it spread on its own to downstream Lake Cohora Bassa. This demonstrates that introduced species can spread throughout a river system, unless there are absolute barriers to their distribution. That historical success makes future successful introduction more likely, is also demonstrated for both L. miodon and O. niloticus. Before their introduction into Lake Kariba, L. miodon and O. niloticus were introduced in Lakes Kivu and Victoria, respectively. Therefore, the chances of successful establishment of L. miodon and O. niloticus are high, given the intention to introduce these species, their life-history traits and the suitability of their new environment.

163

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 163 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

Natural resources need to be utilized sustainably, considering environmental variation My thesis shows that gains which were made by introducing Limnothrissa miodon into Lake Kariba, are eroded by overfishing and environmental variation. The fishery for the introduced species needs to be managed and monitored to ensure that it remains sustainable. Biomass, fish size, growth, maturity and environmental conditions are some of the aspects that should be monitored. Current lack of time-series data makes it difficult to fully understand cause-and-effect relationships. Monitoring of the gill- net catches at Lakeside by the Lake Kariba Fisheries Research Institute (LKFRI) provided valuable data, is highly commendable and was useful for many studies, including this one. The gill-net fishery for Oreochromis niloticus shows signs of overfishing, and corrective action is required. Many changes have occurred with regard to fishing effort in the inshore fishery, such as the type of nets and the duration of setting. Many fish stocks in Zimbabwe and the rest of the world are overfished, which has reduced the amount of fish landed. Fishing sustainably within optimum limits makes economic sense, because the landings will be higher than in an overfished state. It is critical that the productivity levels of the waterbodies are assessed and monitored, and the correct level of fishing determined and adhered to, considering the environmental variation that limits and varies primary production in Lake Kariba and other waterbodies.

Limnothrissa miodon (top; mirror image from Poll 1952) and Oreochromis niloticus (mirror image from Boulenger 1907).

164

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 164 References

Refs

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 165

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 166 References

A Abbott JC & Dill LM (1989) The relative growth of dominant and subordinate juvenile steelhead trout (Salmo gairdneri) fed equal rations. Behaviour 108: 104-113. Admassu D & Casselman JM (2000) Otolith age determination for adult tilapia, Oreochromis niloticus L. from Lake Awassa (Ethiopian Rift Valley) by interpreting biannuli and differentiating biannual recruitment. Hydrobiologia 418: 15-24. Ahlbeck I, Hansson S & Hjerne O (2012) Evaluating fish diet analysis methods by individual- based modelling. Canadian Journal of Fisheries and Aquatic Science 69: 1184-1201. Almeida D & Grossman GD (2012) Utility of direct observational methods for assessing competitive interactions between non-native and native freshwater fishes. Fisheries Management and Ecology 19: 157-166. Alp A, Kara C, Üçkardeş F¸ Carol J & García-Berthou E (2011) Age and growth of the European catfish (Silurus glanis) in a Turkish reservoir, and comparison with introduced populations. Reviews in Fish Biology and Fisheries 21: 283-294. Alvarenga CMD & Volpato GL (1995) Agonistic profile and metabolism in alevins of the Nile tilapia. Physiology & Behavior 57: 75-80. Andersen T & Moksness E (1988) Manual for reading daily increments by the use of a computer programme. Flødevigen Meldinger 4: 1-37. Anderson DR, Burnham KP & Thompson WL (2000) Null hypothesis testing: problems, prevalence, and an alternative. Journal of Wildlife Management 64: 912-923. Araya M & Cubillos LA (2006) Evidence of two-phase growth in elasmobranchs. Environmental Biology of Fishes 77: 293-300. Arlinghaus R, Matsumura S & Dieckmann U (2010) The conservation and fishery benefits of protecting large pike (Esox lucius L.) by harvest regulations in recreational fishing. Biological Conservation 143: 1444-1459. Refs Attayde JL, Brasil J & Menescal RA (2011) Impacts of introducing Nile tilapia on the fisheries of a tropical reservoir in North‐eastern Brazil. Fisheries Management and Ecology 18: 437-443. B Babiker MM (1984) Seasonal abundance, breeding and sex-structure of populations of Tilapia nilotica (L.) and Labeo niloticus (F.) in the Jebel Aulia Dam area of the White Nile. Hydrobiologia 110: 287-94. Bacheler NM, Neal JW & Noble RL (2004) Diet overlap between native bigmouth sleepers (Gobiomorus dormitory) and introduced predatory fishes in a Puerto Rico reservoir. Ecology of Freshwater Fish 13: 111-118. Baerends GP & Baerends-van Roon JM (1950) An introduction to the study of the ethology of the cichlid fishes. Behaviour Supplement 1: 1-243. Bagenal TB (1978) Aspects of fish fecundity. In Gerking SD (ed) Ecology of freshwater fish production. Blackwell Scientific, Oxford: 75-101. Baglin RE Jr & Hill LG (1977) Fecundity of the White bass, Morone chrysops (Rafinesque), in Lake Texoma. The American Midland Naturalist 98: 233-238. Bajer PG, Chizinski CJ & Sorensen, PW (2011) Using the Judas technique to locate and remove wintertime aggregations of invasive common carp. Fisheries Management and Ecology 18: 497-505.

167

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 167 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

Balirwa JS (1992) The evolution of the fishery of Oreochromis niloticus (Pisces: Cichlidae) in Lake Victoria. Hydrobiologia 232: 85-89. Balirwa JS (1998) Lake Victoria wetlands and the ecology of the Nile tilapia, Oreochromis niloticus Linné. PhD thesis, Wageningen Agricultural University. AA Balkema, Rotterdam. Balon EK (1971) Replacement of Alestes imberi (Peters), 1852, by Alestes lateralis Boulenger, 1900, in Lake Kariba, with ecological notes. Fisheries Research Bulletin Zambia 5: 119 -162. Balon EK (1972) Possible fish stock size assessment and available production survey as developed on Lake Kariba. African Journal of Tropical Hydobiology and Fisheries 2: 45-73. Balon EK (1973) Results of fish population size assessments in Lake Kariba coves (Zambia), a decade after their creation. In Ackermann WC, White GF, Worthington EB & Ivens JL (eds) Man-made lakes: their problems and environmental effects. American Geophysical Union, Washington DC: 149-158. Balon EK (1974) Fish production of a tropical ecosystem. In Balon EK & Coche AG (eds) Lake Kariba: a man-made tropical ecosystem in Central Africa. Monographiae Biologicae 24: 249-573. Dr W Junk Publishers, The Hague. Balon EK & Chadwick EM (1974) Time of annulus inception: a pond experiment. In Balon EK & Coche AG (eds) Lake Kariba: a man-made tropical ecosystem in Central Africa. Monographiae Biologicae 24: 643-646. Dr W Junk Publishers, The Hague. Balon EK & Coche AG (eds) (1974) Lake Kariba: a man-made tropical ecosystem in Central Africa. Monographiae Biologicae 24. Dr W Junk Publishers, The Hague. Baras E & Lucas MC (2010) Individual growth trajectories of sibling Brycon moorei raised in isolation since egg stage, and their relationship with aggressive behaviour. Journal of Fish Biology 77: 985-997. Barneche DR, Robertson DR, White CR & Marshall DJ (2018) Fish reproductive-energy output increases disproportionately with body size. Science 360: 642-645. Bbole I, Katongo C, Deines AM, Shumba O & Lodge DM (2014) Hybridization between nonindigenous Oreochromis niloticus and native Oreochromis species in the lower Kafue River and its potential impacts on fishery. Journal of Ecology and The Natural Environment. 6: 215-25. Beaugrand G, Brander KM, Lindley JA, Souissi S & Reid PC (2003) Plankton effect on cod recruitment in the North Sea. Nature 426: 661-664. Begg GW (1970) Limnological observations on Lake Kariba during 1967 with emphasis on some special features. Limnology and Oceanography 15: 776-788. Begg GW (1974) The distribution of fish of riverine origin in relation to the limnological characteristics of the five basins of Lake Kariba. Hydrobiologia 44: 277-285. Bell-Cross G (1965) Additions and amendments to the checklist of fishes of Zambia, 2. Fisheries Research Bulletin of Zambia 4: 99-101. Bell-Cross G (1972) The fish fauna of the Zambezi River system. Arnoldia, Rhodesia 5: 1-19. Bell-Cross B & Bell-Cross G (1971) Introduction of Limnothrissa miodon and Limno- caridina tanganicae from Lake Tanganyika into Lake Kariba. Fisheries Research Bulletin of Zambia 5: 207-214.

168

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 168 References

Bell-Cross G & Minshull JL 1988 The fishes of Zimbabwe. National Museums and Monuments of Zimbabwe. Harare. Beveridge MCM & Baird DJ (2000) Diet, feeding and digestive physiology. In Beveridge MCM & McAndrew BJ (eds) Tilapias: biology and exploitation. Fish and Fisheries Series 25: 59-87. Kluwer, Dordrecht. Birkeland C & Dayton PK (2005) The importance in fishery management of leaving the big ones. Trends in Ecology & Evolution 20: 356-358. Bittencourt LS, Leite-Silva UR, Silva LMA & Tavares-Dias M (2014) Impact of the invasion from Nile tilapia on natives Cichlidae species in tributary of Amazonas River, Brazil. Biota Amazônia 4: 88-94. Blackburn NA & Southgate DAT (1981) Protein digestibility and absorption: effects of fibre and the extent of individual variation. FAO/WHO/UNU EPR/81/32, FAO, Rome. Blow P & Leonard S (2007) A review of cage aquaculture: sub-Saharan Africa. In Halwart M, Soto D & Arthur JR (eds) Cage aquaculture: regional reviews and global overview. FAO Fisheries Technical Paper 498: 188-207. FAO, Rome. Booth AJ & Merron GS (1996) The age and growth of the Greenhead tilapia Oreochromis macrochir (Pisces: Cichlidae) from the Okavango Delta, Botswana. Hydrobiologia 321: 29-34. Booth AJ, Merron GS & Buxton CD (1995) The growth of Oreochromis andersonii (Pisces: Cichlidae) from the Okavango Delta, Botswana, and a comparison of the scale and otolith methods of ageing. Environmental Biology of Fishes 43: 171-178. Borelli G, Mayer-Gostan N, De Pontual H, Boeuf G & Payan P (2001) Biochemical relationships between endolymph and otolith matrix in the trout (Oncorhynchus mykiss) and the turbot (Psetta maxima). Calcified Tissue International 69: 356-364. Borelli G, Guibbolini ME, Mayer-Gostan N, Priouzeau F, De Pontual H, Allemand D, Refs Puverel S, Tambutté E & Payan P (2003) Daily variations of endolymph composition: relationship with the otolith calcification process in trout. Journal of Experimental Biology 206: 2685-2692. Borja A, Fontan A, Sáenz JON & Valencia V (2008) Climate, oceanography, and recruitment: the case of the Bay of Biscay anchovy (Engraulis encrasicolus). Fisheries Oceanography 17: 477-493. Botsford LW, Castilla JC & Peterson CH (1997) The management of fisheries and marine ecosystems. Science 277: 509-515. Boulenger GA (1907) Zoology of Egypt: the fishes of the Nile. Published for the Egyptian government, Hugh Rees, London. Bowen SH (1979) A nutritional constraint in detritivory by fishes: the stunted populations of Sarotherodon mossambicus in Lake Sibaya, South Africa. Ecological Monographs 49: 17-31. Bowen SH (1980) Detrital nonprotein amino acids are the key to rapid growth of Tilapia in Lake Valencia, Venezuela. Science 207: 1216-1218. Bowen SH (1981) Digestion and assimilation of periphytic detrital aggregate by Sarotherodon mosambicus. Transaction of the American Fisheries Society 110: 241- 247.

169

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 169 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

Bowen SH (1982) Feeding, digestion and growth, qualitative considerations. In Pullin RSV & Lowe-McConnell RH (eds) The biology and culture of tilapias. International Center for Living Aquatic Resources Management (ICLARM), Manila: 141-156. Bowen SH, Lutz EV & Ahlgren MO (1995) Dietary protein and energy as determinants of food quality: trophic strategies compared. Ecology 76: 899-907. Bremner JM & Mulvaney CS (1982) Nitrogen-total. In Page AL, Miller RH & Keeney DR (eds) Methods of soil analysis, part 2, chemical and microbiological properties. American Society of Agronomy, Madison: 595-624. Brothers EB, Mathews CP & Lasker R (1976) Daily growth increments in otoliths from larval and adult fishes. Fishery Bulletin 74: 1-8. Brummett RE (1995) Environmental regulation of sexual maturation and reproduction in tilapia. Reviews in Fisheries Science 3: 231-248. Brummett RE & Williams MJ (2000) The evolution of aquaculture in African rural and economic development. Ecological Economics 33: 193-203. Burnham KP & Anderson DR (2002) Model selection and multimodel inference: a practical information-theoretic approach. Springer, New York. Burnham KP, Anderson DR & Huyvaert KP (2011) AIC model selection and multimodel inference in behavioural ecology: some background, observations, and comparisons. Behavioral Ecology and Sociobiology 65: 23-35. Burnham-Curtis MK & Bronte CR (1996) Otoliths reveal a diverse age structure for Humper Lake Trout in Lake Superior. Transactions of the American Fisheries Society 125: 844-851. Butterwick C, Heaney SI & Talling JF (2005) Diversity in the influence of temperature on the growth rates of freshwater algae, and its ecological relevance. Freshwater Biology 50: 291-300. Bwanika GN, Murie DJ & Chapman LJ (2007) Comparative age and growth of Nile tilapia (Oreochromis niloticus L.) in lakes Nabugabo and Wamala, Uganda. Hydrobiologia 589: 287-301. Bwanika GN, Chapman LJ, Kizito Y & Balirwa J (2006) Cascading effects of introduced Nile perch (Lates niloticus) on the foraging ecology of Nile tilapia (Oreochromis niloticus). Ecology of Freshwater Fish 15: 470-481. Bwanika GN, Makanga B, Kizito Y, Chapman LJ & Balirwa J (2004) Observations on the biology of Nile tilapia, Oreochromis niloticus L., in two Ugandan crater lakes. African Journal of Ecology 42: 93-101. C Campana SE (1984) Comparison of age determination methods for starry flounder. Transactions of the American Fishery Society 113: 365-369. Campana SE (2001) Accuracy, precision and quality control in age determination, including a review of the use and abuse of age validation methods. Journal of Fish Biology 59: 197- 242. Campana SE & Neilson JD (1985) Microstructure of fish otoliths (reviews). Canadian Journal of Fisheries and Aquatic Science 42: 1014-1032.

170

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 170 References

Canonico GC, Arthington A, McCrary JK & Thieme ML (2005) The effects of introduced tilapias on native biodiversity. Aquatic Conservation: Marine and Freshwater Ecosystems 15: 463-483. Carline RF & Hall JD (1973) Evaluation of a method for estimating food consumption rates of fish. Journal of the Fisheries Board of Canada 30: 623-629. Chen Y & Paloheimo JE (1994) Estimating fish length and age at 50% maturity using a logistic type model. Aquatic Sciences 56: 206-219. Chifamba PC (1992) Daily rings of otoliths as a method for ageing the sardine, Limnothrissa miodon in Lake Kariba. Transaction of Zimbabwe Scientific Association 66: 15-17. Chifamba PC (1995) Factors affecting the fishing power of sardine fishing vessels and implications for the management of the fisheries on Lake Kariba, Zimbabwe. Fisheries Management and Ecology 2: 309-319. Chifamba PC (1998) Status of Oreochromis niloticus in Lake Kariba, Zimbabwe, following its escape from fish farms. In Cowx IG (ed) Stocking and introduction of fish. Blackwell Science, Oxford: 267-273. Chifamba PC (2000) The relationship of temperature and hydrological factors to catch per unit effort, condition and size of the freshwater sardine, Limnothrissa miodon (Boulenger), in Lake Kariba. Fisheries Research 45: 271-281. Chifamba PC (2006) Spatial and historical changes of indigenous O. mortimeri following introduction of exotic O. niloticus in Lake Kariba. In Odada EO, Olago DO, Ochala W, Ntiba N, Wandiga S, Gichuki N & Oyieke H (eds) 11th World Lake Conference, Nairobi, Kenya, 31 October - 4 November 2005, Proceedings Vol. II. Ministry of Water and Irrigation / International Lake Environment Committee Foundation (ILEC), Nairobi, Kenya / Shiga, Japan: 500-504. Chifamba PC & Mauru T (2017) Comparative aggression and dominance of Oreochromis Refs niloticus (Linnaeus, 1758) and Oreochromis mortimeri (Trewavas, 1966) from paired contest in aquaria. Hydrobiologia 788: 193-203. Chifamba PC & Videler (2014) Growth of the alien Oreochromis niloticus Linnaeus and indigenous Oreochromis mortimeri Trewavas in Lake Kariba. African Journal of Aquatic Science 39: 167-176. Claessen D, de Roos AM & Persson L (2000) Dwarfs and giants: cannibalism and competition in size‐structured populations. The American Naturalist 155: 219-237. Clark RW, Henderson-Arzapalo A, Sullivan CV (2005) Disparate effects of constant and annually-cycling daylength and water temperature on reproductive maturation of striped bass (Morone saxatilis). Aquaculture 249: 497-513. Clavero M & García-Berthou E (2005) Invasive species are a leading cause of animal extinctions. Trends in Ecology & Evolution 20: 110. Coche AG (1974) Limnological study of a tropical reservoir. In Balon EK & Coche AG (eds) Lake Kariba: a man-made tropical ecosystem in Central Africa. Monographiae Biologicae 24: 1-247. Dr W Junk Publishers, The Hague. Cochrane KL (1976) Catches of Hydrocynus vittatus Castelnau during sardine fishing operations in Lake Kariba. Kariba Studies 7: 98-108. Cochrane KL (1978) Seasonal fluctuations in the catches of Limnothrissa miodon (Boulenger) in Lake Kariba. MPhil Thesis, Rhodesia Univiversty, Salisbury.

171

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 171 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

Cochrane KL (1984) The influence of food availability, breeding seasons and growth rate on commercial catches of Limnothrissa miodon (Boulenger) in Lake Kariba. Journal of Fish Biology 26: 623-635. Cochrane KL & Hutchings L (1995) A structured approach to using biological and environmental parameters to forecast anchovy recruitment. Fisheries Oceanography 4: 102-127. Cohen AS, Gergurich EL, Kraemer BM, McGlue MM, McIntyre PB, Russell JM, Simmons JD & Swarzenski PW (2016) Climate warming reduces fish production and benthic habitat in Lake Tanganyika, one of the most biodiverse freshwater ecosystems. Proceedings of the National Academy of Science 113: 9563-9568. Coke M (1968) Depth distribution of fish on a bush-cleared area of Lake Kariba, Central Africa. Transactions of the American Fisheries Society 97: 460-465. Conover DO & Munch SB (2002) Sustaining fisheries yields over evolutionary time scales. Science 297: 94-96. Copp GH, Garthwaite R & Gozlan RE (2005) Risk identification and assessment of non‐native freshwater fishes: a summary of concepts and perspectives on protocols for the UK. Journal of Applied Ichthyology 21: 371-373. Corrêa SA, Fernandes MO, Iseki KK & Negrão JA (2003) Effect of the establishment of dominance relationships on cortisol and other metabolic parameters in Nile tilapia (Oreochromis niloticus). Brazilian Journal of Medical and Biological Research 36: 1725-1731. Cronberg G (1997) Phytoplankton in Lake Kariba 1986 - 1990. In Moreau J (ed) Advances in the ecology of Lake Kariba. University of Zimbabwe Publishers: 66-101. Crowder LB & Murawski SA (1998) Fisheries bycatch: implications for management. Fisheries 23: 8-17. Cucherousset J, Copp GH, Fox MG, Sterud E, van Kleef HH, Verreycken H, Záhorská E. (2009) Life-history traits and potential invasiveness of introduced pumpkinseed Lepomis gibbosus populations in northwestern Europe. Biological Invasions. 11: 2171-2180. Cury P & Roy C (1989) Optimal environmental window and pelagic ﬁsh recruitment success in upwelling areas. Canadian Journal of Fisheries and Aquatic Science 46: 670-680. Cury P, Bakun A, Crawford RJ, Jarre A, Quinones RA, Shannon LJ & Verheye HM (2000) Small pelagics in upwelling systems: patterns of interaction and structural changes in “wasp-waist” ecosystems. ICES Journal of Marine Science / Journal du Conseil 57: 603-618. Cutts CJ, Metcalfe NB & Taylor AC (1999) Competitive asymmetries in territorial juvenile Atlantic salmon, Salmo salar. Oikos 86: 479-486. Cutts CJ, Brembs B, Metcalfe NB & Taylor AC (1999) Prior residence, territory quality and life-history strategies in juvenile Atlantic salmon (Salmo salar L.). Journal of Fish Biology 55: 784-794. D Dalu T, Magoro ML, Tonkin JD, Human LR, Perissinotto R, Deyzel SH, Adams JB & Whitfield AK (2018) Assessing phytoplankton composition and structure within micro-estuaries and micro-outlets: a community analysis approach. Hydrobiologia 818: 177-191.

172

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 172 References

D’Amato ME, Esterhuyse MM, Van Der Waal BC, Brink D & Volckaert FA (2007) Hybridization and phylogeography of the Mozambique tilapia Oreochromis mossambicus in southern Africa evidenced by mitochondrial and microsatellite DNA genotyping. Conservation Genetics 8: 475-488. Darwall WRT, Smith KG, Tweddle D & Skelton P (eds) (2009) The status and distribution of freshwater biodiversity in southern Africa. IUCN, Gland, Switzerland and SAIAB, Grahamstown, South Africa. Da Silva MJ, Figueiredo BRS, Ramos RTC & Medeiros ESF (2010) Food resources used by three species of fish in the semi-arid region of Brazil. Neotropical Ichthyology 8: 819- 825. Daskalov G (1999) Relating fish recruitment to stock biomass and physical environment in the Black Sea using generalized additive models. Fisheries Research 41: 1-23. Daskalov GM & Mamedov EV (2007) Integrated fisheries assessment and possible causes for the collapse of anchovy kilka in the Caspian Sea. ICES Journal of Marine Science 64: 503-511. Deacon AE, Ramnarine IW & Magurran AE (2011) How reproductive ecology contributes to the spread of a globally invasive fish. PLoS One 6(9): e24416. https://doi.org/10.1371/journal.pone.0024416. Degens ET, Deuser WG & Haedrich RL (1969) Molecular structure and composition of fish otoliths. Marine Biology 2: 105-113. Deines AM, Bbole I, Katongo C, Feder JL & Lodge DM (2014) Hybridisation between native Oreochromis species and introduced Nile tilapia O. niloticus in the Kafue River, Zambia. African Journal of Aquatic Science 39: 23-34. de Iongh HH, Spliethoff PC & Frank VG (1983) Feeding habits of the clupeid Limnothrissa miodon (Boulenger), in Lake Kivu. Hydrobiologia 102: 113-122. Refs de Iongh HH, Spliethoff PC & Roest F (1995) The impact of an introduction of sardine into Lake Kivu. In Pitcher TJ & Hart PJB (eds) The impact of species changes in African lakes. Chapman and Hall Fish and Fisheries Series 18: 277-297. Springer, Dordrecht. Dekker W (2003) Did lack of spawners cause the collapse of the European eel, Anguilla anguilla? Fisheries Management and Ecology 10: 365-376. DeMott WR, Zhang QX & Carmichael WW (1991) Effects of toxic cyanobacteria and purified toxins on the survival and feeding of a copepod and three species of Daphnia. Limnology and Oceanography 36: 1346-1357. De Roos AM, Persson L, McCauley E (2003) The influence of size‐dependent life‐history traits on the structure and dynamics of populations and communities. Ecology Letters 6: 473- 487. De Silva SS (1985) Body condition and nutritional ecology of Oreochromis mossambicus (Pisces, Cichlidae) populations of man-made lakes in Sri Lanka. Journal of Fish Biology 27: 621-633. De Silva SS & Sirisena HK (1987) New fish resources of reservoirs in Sri Lanka: feasibility of introduction of a subsidiary gillnet fishery for minor cyprinids. Fisheries Research 6: 17- 34. De Vos L, Snoeks J & Thys van den Audenaerde D (1990) The effects of Tilapia introductions in Lake Luhondo, Rwanda. Environmental Biology of Fishes 27: 303- 308.

173

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 173 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

Donnelly BG (1969) A preliminary survey of Tilapia nurseries on Lake Kariba during 1967/68. Hydrobiologia 34: 195-206. Doupé RG, Schaffer J, Knott MJ & Burrows DW (2009) How might an exotic fish disrupt spawning success in a sympatric native species? Marine and Freshwater Research 60: 379-383. Duponchelle F & Legendre M (2000) Oreochromis niloticus (Cichlidae) in Lake Ayame, Côte d’Ivoire: life history traits of a strongly diminished population. Cybium 24: 161- 172. Duponchelle F & Panfili J (1998) Variations at age and size at maturity of female Nile tilapia, Oreochromis niloticus, populations from man-made lakes of Côte d’Ivoire. Environmental Biology of Fishes 52: 453-465. Duponchelle F, Cecchi P, Corbin D, Nuñez J & Legendre M (1999) Spawning season variations of female Nile Tilapia, Oreochromis niloticus, from man-made lakes of Côte d'Ivoire. Environmental Biology of Fishes 56: 375-387. Dwyer KS, Walsh SJ & Campana SE (2003) Age determination, validation and growth of Grand Bank Yellowtail flounder (Limanda ferruginea). ICES Journal of Marine Science 60: 1123-1138. E Edenbrow M & Croft DP (2012) Kin and familiarity influence association preferences and aggression in the mangrove killifish Kryptolebias marmoratus. Journal of Fish Biology 80: 503-518. Edeyer A, de Pontual H, Payan P, Troadec H, Severe A & Mayer-Gostan M (2000) Daily variations of the saccular endolymph and plasma compositions in the turbot Psetta maxima: relationship with the diurnal rhythm in otolith formation. Marine Ecology Progress Series 192: 287-294. Eigaard OR, Thomsen B, Hovgaard H, Nielsen A & Rijnsdorp AD (2011) Fishing power increases from technological development in the Faroe Islands longline fishery. Canadian Journal of Fisheries and Aquatic Science 68: 1970-1982. Ejike C & Schreck CB (1980) Stress and social hierarchy rank in coho salmon. Transactions of the American Fisheries Society 109: 423-426. Ellis CMA (1971) The size at maturity and breeding seasons of sardines in southern Lake Tanganyika. African Journal of Tropical Hydrobiology and Fisheries 1: 59-66. Enberg K, Dunlop ES & Jørgensen C (2008) Fish growth. In Jørgensen SE & Fath BD (eds) Encyclopedia of Ecology. Elsevier, Oxford: 1564-1572. Essington TE, Moriarty PE, Froehlich HE, Hodgson EE, Koehn LE, Oken KL, Siple MC & Stawitz CC (2015) Fishing amplifies forage fish population collapses. Proceedings of the National Academy of Sciences 112: 6648-6652. F Fernando CH & Holčík J (1991) Fish in reservoirs. Internationale Revue der gesamten Hydrobiologie und Hydrographie 76: 149-167. G Gagliano M & McCormick MI (2004) Feeding history influences otolith shape in tropical fish. Marine Ecology Progress Series 278: 291-296.

174

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 174 References

Galat DL & Vucinich N (1983) Food partitioning between young of the year of two sympatric tui chub morphs. Transactions of the American Fisheries Society 112: 486-497. Geffen AJ (1992) Validation of otolith increment deposition rate. In Stevenson DK & Campana SE (eds) Otolith microstructure examination and analysis. Canadian Special Publication of Fisheries and Aquatic Sciences 117: 101-113. Getabu A (1992) Growth parameters and total mortality in Oreochromis niloticus (Linnaeus) from Nyanza Gulf, Lake Victoria. Hydrobiologia 232: 91-97. Getabu A (1994) A comparative study on the feeding habits of Oreochromis niloticus (Linnaeus) in Nyanza Gulf, Lake Victoria and sewage fish ponds. In Okemwa E, Wakwabi E & Getabu A (eds) Proceedings of the second EEC Regional Seminar on recent trends of research on Lake Victoria fisheries. ICIPE Science Press, Nairobi. Getachew T & Fernando CH (1989) The food habits of an herbivorous fish (Linn.) in Lake Awasa, Ethiopia. Hydrobiologia 174: 195-200. Glauert AM (1984) Part 1, Fixation, dehydration and embedding of biological specimens. In Glauert AM (ed) Practical methods in electron microscopy. North Holland / American Elsevier, Amsterdam. Gliwicz ZM (1984) Limnological study of Cahora Bassa reservoir with special regard to sardine fishery expansion, Mozambique. FAO/GCP/MOZ/006/SWE, Field document 8. FAO, Rome. Gómez-Márquez JL (1998) Age and growth of Oreochromis niloticus (Perciformes: Cichlidae) in Mexico. Revista de Biología Tropical 46: 929-936. Gómez-Márquez JL, Peña-Mendoza B, Isaías H. Salgado-Ugarte IH & Arredondo-Figueroa JL (2008) Age and growth of the tilapia Oreochromis niloticus (Perciformes: Cichlidae) from a tropical shallow lake in Mexico. Revista de Biología Tropical 56: 875-884. Gompertz B (1825) On the nature of the function expressive of the law of human mortality, and Refs on a new mode of determining the value of life contingencies. Philosophical Transactions of the Royal Society of London, Series A 115: 515-585. Goudswaard PC, Witte F & Katunzi EFB (2002) The tilapiine stock of Lake Victoria before and after the Nile perch upsurge. Journal of Fish Biology 60: 838-856. Gozlan RE, Britton JR, Cowx I & Copp GH (2010) Current knowledge on non‐native freshwater fish introductions. Journal of Fish Biology 76: 751-86. Grammer GL, Slack WT, Peterson MS & Dugo MA (2012) Nile tilapia Oreochromis niloticus (Linnaeus, 1758) establishment in temperate Mississippi, USA: multi-year survival confirmed by otolith ages. Aquatic Invasions 7: 367–376. Gregg RE, Howard JH & Snhonhiwa F (1998) Introgressive hybridization of tilapias in Zimbabwe. Journal of Fish Biology 52: 1-10. Guibbolini M, Borelli G, Nicole Mayer-Gostan N, Priouzeau F, De Pontual H, Denis

Allemand D & Payan P (2006) Characterization and variations of organic parameters in teleost fish endolymph during day-night cycle, starvation and stress condition. Comparative Biochemistry and Physiology 145: 99-107. Gutiérrez M, Swartzman G, Bertrand A & Bertrand S (2007) Anchovy (Engraulis ringens) and sardine (Sardinops sagax) spatial dynamics and aggregation patterns in the Humboldt Current ecosystem, Peru, from 1983–2003. Fisheries Oceanography 16: 155-168. Gwahaba JJ (1973) Effects of fishing on the Tilapia nilotica (Linnaeus 1757) population in Lake George, Uganda over the past 20 years. African Journal of Ecology 11: 317-328.

175

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 175 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

H Hanson JM & Leggett WC (1985) Experimental and field evidence for inter- and intraspecific competition in two freshwater fishes. Canadian Journal of Fisheries and Aquatic Sciences 42: 280-6. Harder W (1975) Anatomy of fishes. E. Schweizerbart’sche Verlagsbuchhandlung, Stuttgart. Harding D (1964) Hydrology and fisheries in Lake Kariba. Verhandlungen der Internationale Vereinigung für Limnologie 15: 139-149. Hillebrand H, Dürselen C, Kirschtel D, Kirschtel D, Pollingher U & Zohary T (1999) Biovolume calculation for pelagic and benthic microalgae 1991. Journal of. Phycology 35: 403-424. Hoenig JM, Morgan MJ & Brown CA (1995) Analysing differences between two age determination methods by testing of symmetry. Canadian Journal of Fisheries and Aquatic Science 52: 368-268. Huchette SM & Beveridge MC (2003) Technical and economical evaluation of periphyton- based cage culture of tilapia (Oreochromis niloticus) in tropical freshwater cages. Aquaculture 218: 219-234. Huchette SM, Beveridge MC, Baird DJ & Ireland M (2000) The impacts of grazing by tilapias (Oreochromis niloticus L.) on periphyton communities growing on artificial substrate in cages. Aquaculture 186: 45-60. Hustler K & Marshall BE (1990) Population dynamics of two small cichlid fish species in a tropical man-made lake (Lake Kariba). Hydrobiologia 190: 253-262. Hutchings JA (1993) Adaptive life histories effected by age‐specific survival and growth rate. Ecology 74: 673-684. Hutchings JA (2005) Life history consequences of overexploitation to population recovery in Northwest Atlantic cod (Gadus morhua). Canadian Journal of Fisheries and Aquatic Sciences 62: 824-832. Hyder M. (1970). Gonadal and reproductive patterns in Tilapia leucosticta (Teleostei: Cichlidae) in an equatorial lake, Lake Naivasha (Kenya). Journal of Zoology 162: 179-195. Hyslop EJ (1980) Stomach content analysis a review of methods and their application. Journal of Fish Biology 17: 411-429. Hüssy K (2008) Otolith shape in juvenile cod (Gadus morhua): Ontogenetic and environmental effects. Journal of Experimental Marine Biology and Ecology 364: 35–41. I Ishikawa T, Shimose T & Tachihara K (2013) Life history of an invasive and unexploited population of Nile tilapia (Oreochromis niloticus) and geographical variation across its native and non-native ranges. Environmental Biology of Fishes 96: 603-616. Itoh S, Yasuda I, Nishikawa H, Sasaki H & Sasai Y (2009) Transport and environmental temperature variability of eggs and larvae of the Japanese anchovy (Engraulis japonicus) and Japanese sardine (Sardinops melanostictus) in the western North Pacific estimated via numerical particle‐tracking experiments. Fisheries Oceanography 18: 118-133. Itoh S, Saruwatari T, Nishikawa H, Yasuda I, Komatsu K, Tsuda A, Setou T & Shimizu M (2011) Environmental variability and growth histories of larval Japanese sardine (Sardinops melanostictus) and Japanese anchovy (Engraulis japonicus) near the frontal area of the Kuroshio. Fisheries Oceanography 20:114-124.

176

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 176 References

J Jackson PBN (1960) The ecological effects of flooding by the Kariba dam upon Middle Zambezi fishes. In Proceedings of the First Federal Science Congress. Salisbury, Southern Rhodesia: 277-284. Jackson PBN (1961) Ichthyology, the fish of the middle Zambezi. Kariba Studies 1: 1-36. National Museums and Monuments of Rhodesia, Salisbury. Jackson PBN (1966) The establishment of fisheries in man-made lakes in the tropics. In Lowe-McConnell RH (ed) Man-made lakes. Academic Press, New York: 53-69. Jackson PBN & Rogers KH (2010) Cabora Bassa fish populations before and during the first filling phase. Zoologica Africana 11: 373-397. Jackson PBN, Marshall BE & Paugy G (1988) Fish communities in man-made lakes. In Lévêque C, Bruton MN & Ssentongo GW (eds) Biology and ecology of African freshwater fishes. ORSTOM, Paris: 325-350. Janse van Vuuren S, Taylor JC, Gerber A & van Ginkel C (2006) Easy identification of the most common freshwater algae: a guide for the identification of microscopic algae in South African freshwaters. North-West University and Department of Water Affairs and Forestry, South Africa. John J (2000) A guide to diatoms as indicators of urban stream Health. Land and Water Resources Research and Development Corporation Occasional Paper 14/99. Canberra, Australia. Johnson JB & Omland KS (2004) Model selection in ecology and evolution. Trends in Ecology and Evolution 19: 101-108. Jubb RA (1974) The distribution of Tilapia mossambica Peters, 1852 and Tilapia mortimeri Trewavas, 1966 in Rhodesian waters. Arnoldia (Rhodesia) 6: 1-14. Jurvelius J, Leinikki J, Mamylov V & Pushkin S (1996) Stock assessment of pelagic three- spined stickleback (Gasterosteus aculeatus): a simultaneous up- and down-looking Refs echo-sounding study. Fisheries Research 27: 227-241. K Kahilainen K, Alajärvi E & Lehtonen H (2005) Planktivory and diet-overlap of densely rakered whitefish (Coregonus lavaaretus (L.) in a subarctic lake. Ecology of Freshwater Fish 14: 50-58. Kakareko T, Kobak J, Grabowska J, Jermacz Ł, Przybylski M, Poznanska M, Pietraszewski D & Copp GH (2013) Competitive interactions for food resources between invasive racer goby Babka gymnotrachelus and native European bullhead Cottus gobio. Biological Invasions 15: 2519-2530. Karenge L & Kolding J (1995) Inshore fish population and species changes in Lake Kariba, Zimbabwe. In Pitcher TJ & Hart PJB (eds) The impact of species changes in African lakes. Chapman and Hall Fish and Fisheries Series 18: 245-275. Springer, Dordrecht. Karna SK & Panda S (2011) Growth estimation and length at maturity of a commercially important fish species i.e., Dayscieaena albida (Boroga) in Chilika Lagoon, India. European Journal of Experimental Biology 1: 84-91. Katsanevakis S (2006) Modelling fish growth: model selection, multi-model inference and model selection uncertainty. Fisheries Research 81: 229–235.

177

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 177 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

Katsanevakis S & Maravelias CD (2008) Modelling fish growth: multi-model inference as a better alternative to a priori using von Bertalanffy equation. Fish and Fisheries 9: 178- 187. Kenmuir DHS (1971) An analysis of data on the tigerfish Hydrocynus vittatus Castlenau obtained from Kariba International Tigerfish Tournaments. Newsletter of the Limnological Society of Southern Africa 17: 13-21. Kenmuir DHS (1973) The ecology of the tigerfish, Hydrocynus vittatus Castlenau, in Lake Kariba. Occasional papers of the National Museums and Monuments of Rhodesia 5: 115-170. Kenmuir DHS (1984) Fish population changes in the Sanyati Basin, Lake Kariba, Zimbabwe. South African Journal of Zoology 19: 194-209. Khallaf EA & Alne-na-ei AA (1987) Feeding ecology of Oreochromis niloticus (Linnaeus) & Tilapia zillii (Gervias) in a Nile canal. Hydrobiologia 146: 57-62. Kimura S (1995) Growth of the clupeid fishes, Stolothrissa tanganicae and Limnothrissa miodon, in the Zambian waters of Lake Tanganyika. Journal of Fish Biology 47: 569- 575. Kinadjian L (2012) Bio-economic analysis of the kapenta fisheries. Mission Report No 1. Report/Rapport:SF-FAO/2012/09. FAO-SmartFish Programme of the Indian Ocean Commission, Ebene, Mauritius. Koebele BP (1985) Growth and the size hierarchy effect: an experimental assessment of three proposed mechanisms; activity differences, disproportional food acquisition, physiological stress. Environmental Biology of Fishes 12: 181-188. Kolding J, Musando B & Songore N (2003) Inshore fisheries and fish population changes in Lake Kariba. In Jul-Larsen E, Kolding J, Overå R, Raakjær Nielsen J & van Zwieten PAM (eds) Management, co-management or no management? Major dilemmas in southern African freshwater fisheries. 2. Case studies. FAO Fisheries Technical Paper. No. 426/2. FAO, Rome: 67-99. Komárková-Legnerová J & Cronberg G (1994) Planktic blue-green algae from lakes in South Scania,Sweden. Part I. Chroococcales. Algological Studies 72: 13-51. Kraus RT, Knight CT, Farmer TM, Gorman AM, Collingsworth PD, Warren GJ, Kocovsky PM & Conroy JD (2015) Dynamic hypoxic zones in Lake Erie compress fish habitat, altering vulnerability to fishing gears 1. Canadian Journal of Fisheries and Aquatic Sciences 72: 797-806. Krupka I (1974) The Kurper bream Sarotherodon mossambicus (Trewavas, 1966). In Balon EK & Coche AG (eds) Lake Kariba: a man-made tropical ecosystem in Central Africa. Monographiae Biologicae 24: 343-349. Dr W Junk Publishers, The Hague. Kudhongania AW & Chitamwebwa DBR (1995) Impact of environmental change, species introductions and ecological interactions on the fish stocks of Lake Victoria. In Pitcher TJ & Hart PJB (eds) The impact of species changes in African lakes. Fish and Fisheries Series 18: 19-32. Chapman & Hall, London. L Lancaster LT, Morrison G & Fitt RN (2017) Life history trade-offs, the intensity of competition, and coexistence in novel and evolving communities under climate

178

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 178 References

change. Philosophical Transactions of the Royal Society B 372: 20160046. http://dx.doi.org/10.1098/rstb.2016.0046 Lelek A & El‐Zarka S (1973) Ecological comparison of the preimpoundment and post- impoundment fish faunas of the River Niger and Kainji Lake, Nigeria. In Ackerman WC, White GF, Worthington EB & Ivens JL (eds) Man-made lakes: their problems and environmental effects. Geophysical Monograph Series 17: 655-660. American Geophysical Union (AGU), Washington, DC. Li HW & Brocksen RW (1977) Approaches to the analysis of energetic costs of intra-specific competition for space by rainbow trout (Salmo gairdneri). Journal of Fish Biology 11: 329-341. Lindem BHT (1983) A preliminary analysis of hydro acoustic data from Maputo Bay, Lake Niassa and Cahora Bassa. FAO/GCP/MOZ/006/SWE. Field Document 6. FAO, Rome. Lindem T (1988) Results from the hydro-acoustic survey of Lake Kariba, September 1988. A report prepared for the Zambia/Zimbabwe SADC Fisheries Project. University of Oslo, Norway. Linton LR, Davis RW & Wrona FJ (1981) Resource utilization indices: an assessment. Journal of Animal Ecology 50: 283-292. LVFO (2005) The state of the fisheries resources of Lake Victoria and their management. Proceedings of the Entebbe Regional Stakeholders’ Conference, 24-25 February 2005. Lake Victoria Fisheries Organization. Jinja, Uganda. Lowe-McConnell RH (1958) Observations on the biology of Tilapia nilotica Linné (Pisces: Cichlidae) in East African waters. Revue de Zoologie et de Botanique Africaines 57: 129-170. Lowe-McConnell RH (1982) Tilapias in fish communities. In Pullin RSV & Lowe- McConnell RH (eds) The biology and culture of tilapias. ICLARM Conference Refs Proceedings 7: 83-113. International Center for Living Aquatic Resources Management, Manila. Lowe-McConnell RH (2000) The role of tilapias in ecosystems. In Beveridge MCM & McAndrew BJ (eds) Tilapias: biology and exploitation. Fish and Fisheries Series 25: 129-157. Kluwer Academic Publishers, Dordrecht. M MacArthur RH & Levins R (1967) The limiting similarity, convergence, and divergence of coexisting species. The American Naturalist 101: 377-385. MacCall AD (1976) Density dependence of catchability coefficient in the California Pacific sardine, Sardinops sagax caerulea, purse seine fishery. California Cooperative Oceanic Fisheries Investigations Report 18: 136-148. Maceina MJ & Sammons SM (2006) An evaluation of different structures to age fresh-water fish from a northeastern US river. Fisheries Management and Ecology 13: 237-242. Mafuca JM (2014) Preliminary results of the hydroacoustic survey conducted on Lake Kariba. Report/Rapport SF-FAO/2014/33, September/Septembre 2014. FAO-SmartFish Pro- gramme of the Indian Ocean Commission. Ebene, Mauritius. Mafuca J, Palha de Sousa L, Abdala S, Brito A, Halare IA, André E & Marguerite M (2011) Some aspects of the reproductive biology of Kapenta, Limnothrissa miodon (Boulenger

179

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 179 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

1906) in the Cahora Bassa Reservoir, Mozambique. Revista de Investigação Pesqueira 29: 18-27. Magadza CHD (1980) The distribution of zooplankton in the Sanyati bay, Lake Kariba; a multivariate analysis. Hydrobiologia 70: 57-67. Magadza CHD (2010) Environmental state of Lake Kariba and Zambezi River valley: lessons learned and not learned. Lakes & Reservoirs: Research and Management 15: 167- 192. Magadza CHD (2011) Indications of the effects of climate change on the pelagic fishery of Lake Kariba, Zambia-Zimbabwe. Lakes & Reservoirs: Research and Management 16: 15-22. Magnan P, Proulx R & Plante M (2005) Integrating the effects of fish exploitation and interspecific competition into current life history theories: an example with lacustrine brook trout (Salvelinus fontinalis) populations. Canadian Journal of Fisheries and Aquatic Sciences 62: 747-757. Magurran AE & Garcia CM (2000) Sex differences in behaviour as an indirect consequence of mating system. Journal of Fish Biology 57: 839-857. Mahere TS, Mtsambiwa MZ, Chifamba PC & Nhiwatiwa T (2014) Climate change impact on the limnology of Lake Kariba, Zambia–Zimbabwe. African Journal of Aquatic Science 39: 215-221. Mannini P (1990) Paramètres de la population de Limnothrissa miodon du lac Kivu (1980- 1989). Projet de Développement de la Pêche au Lac Kivu, Document de Travail No. 32. RWA/87/012/DOC/TR/32 (Fr). FAO, Rome. Marriott RJ, Wise B & St John J (2011) Historical changes in fishing efficiency in the west coast demersal scalefish fishery, Western Australia: implications for assessment and management. ICES Journal of Marine Science 68: 76-86. Marshall BE (1982) The influence of river flow on pelagic sardine catches in Lake Kariba. Journal of Fish Biology 20: 465-469. Marshall, BE (1987a) Growth and mortality of the introduced Lake Tanganyika clupeid, Limnothrissa miodon, in Lake Kariba. Journal of Fish Biology 31: 603-615. Marshall BE (1987b) On the relationship between tigerfish and sardines in Lake Kariba. Journal of the Limnological Society of Southern Africa 13: 78-79. Marshall BE (1988a) A preliminary assessment of the biomass of the pelagic sardine Limnothrissa miodon in Lake Kariba. Journal of Fish Biology 32: 515-524. Marshall BE (1988b) Seasonal and annual variations in the abundance of pelagic sardines in Lake Kariba, with special reference to the effects of drought. Archiv für Hydro- biologie 112: 399-409. Marshall BE (1991) The impact of the introduced sardine Limnothrissa miodon on the ecology of Lake Kariba. Biological Conservation 55: 151-165. Marshall BE (1993) Limnothrissa in man-made lakes: do we understand the implications of their small size? In Marshall BE & Mubamba R (eds) Papers presented at the Symposium on Biology, Stock Assessment and Exploitation of Small Pelagic Fish Species in the African Great Lakes Region. Bujumbura, Burundi, from 25 to 28 November 1992. CIFA Occasional Paper 19. FAO, Rome. Marshall BE (1999) The fishes of Zimbabwe: a century of change. The Zimbabwe Science News 33: 45-52.

180

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 180 References

Marshall BE (2006) A checklist of the fishes of Lake Kariba (Zimbabwe-Zambia). Arnoldia Zimbabwe 10: 257-274. Marshall BE (2011) The fishes of Zimbabwe and their biology. Musée Royal de l'Afrique Centrale, Tervuren, Belgium. Marshall BE (2012a) Climate change does not explain historical changes in the pelagic ecosystem of Lake Kariba (Zambia-Zimbabwe). Lakes & Reservoirs: Research and Management 17: 265-274. Marshall BE (2012b). Does climate change really explain changes in the fisheries productivity of Lake Kariba (Zambia-Zimbabwe)? Transactions of the Royal Society of South Africa 67: 45-51. Marshall BE, Junor FJR & Langerman JD (1982) Fisheries and fish production on the Zimbabwean side of Lake Kariba. Kariba Studies 10: 1-231. Trustees of the National Museums and Monuments of Zimbabwe, Harare. Martin CW, Valentine MM & Valentine JF (2010) Competitive interactions between invasive Nile tilapia and native fish: the potential for altered trophic exchange and modification of food webs. PLoS One 5(12): e14395. https://doi.org/10.1371/ journal.pone.0014395. Marufu LT & Chifamba PC (2013) A comparison of diel feeding pattern, ingestion and digestive efficiency of Oreochromis niloticus and Oreochromis macrochir in Lake Chivero, Zimbabwe. African Journal of Aquatic Science 38: 221-228. Masundire H (1989) Zooplankton population dynamics in the Sanyati basin, Lake Kariba, Zimbabwe. Archiv für Hydrobiologie – Beiheft Ergebisse der Limnologie 33: 469-473. Masundire HM (1992) Population dynamics of Bosmina longirostris (Crustacea: Cladocera) in Lake Kariba, Zimbabwe. Hydrobiologia 243/244: 167-173. Masundire HM (1994) Seasonal trends in zooplankton densities in Sanyati basin, Lake Kariba: multivariate analysis. Hydrobiologia 272: 211-230. Refs Matthes H (1965-66) Peliminary investigations into the biology of the Lake Tanganyika clupeidae. Fisheries Research Bulletin of Zambia 4: 39-46. Matthes H (1968) Preliminary investigations into the biology of the Lake Tanganyika Clupeidae. Fisheries Research Bulletin of Zambia 4: 39-45. Matveev V, Closs G, Lieschke JA & Shirley MJ (2002) Are pelagic fish important in the food webs of Australian reservoirs? Verhandlungen der Internationalen Vereinigung für Theoretische und Angewandte Limnologie 28: 116-119. McDonald JL, Peterson MS & Slack WT (2007) Morphology, density, and spatial pat-terning of reproductive bowers in an established alien population of Nile tilapia, Oreochromis niloticus. Journal of Freshwater Ecology 22: 461-468. McKnight E, García‐Berthou E, Srean P & Rius M (2017) Global meta‐analysis of native and nonindigenous trophic traits in aquatic ecosystems. Global Change Biology 23: 1861- 1870. Meisfjord J, Midtøy F & Folkvord A (2006) Validation of daily increment deposition in otoliths of juvenile Limnothrissa miodon (Clupeidae). Journal of Fish Biology 69: 1845-1848. Mendonça FZ & Gonçalves-de-Freitas E (2008) Nest deprivation and mating success in Nile tilapia (Teleostei: Cichlidae). Revista Brasileira de Zoologia 25: 413-418. Mhlanga L (2000) The diet of five cichlid fish species from Lake Kariba, Zimbabwe. Transactions of the Zimbabwe Scientific Association 74: 16-21.

181

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 181 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

Mhlanga W (1998) Observations on gillnet catches of Kariba tilapia, Oreochromis mortimeri, from Bumi Basin of Lake Kariba, Zimbabwe. Naga 21: 57-60. Mhlanga W (2003) Food and feeding habits of tigerfish, Hydrocynus vittatus, in Lake Kariba, Zimbabwe. In Palomares MLD, Samb B, Diouf T, Vakily JM & Pauly D (eds) Fish biodiversity: local studies as basis for global inferences. ACP-EU Fisheries Research Report 14: 76-81. ACP-EU, Brussels. Mikheev VN, Pasternak AF, Tischler G & Wanzenbӧck J (2005) Contestable shelters provoke aggression among 0+ Perch, Perca fluviatilis. Environmental Biology of Fishes 73: 227-231. Miklósi A, Haller J & Csányi V (1995) The influence of opponent-related and outcome-related memory on repeated aggressive encounters in the Paradise fish (Macropodus opercularis). The Biological Bulletin 188: 83-88. Milinski M & Parker GA (1991) Competition for resources. In Krebs JR & Davies NB (eds) Behavioural Ecology: an evolutionary approach, Vol 3. Blackwell Scientific Publications, Oxford: 137-168. Mollet FM, Kraak SB & Rijnsdorp AD (2007) Fisheries-induced evolutionary changes in maturation reaction norms in North Sea sole Solea solea. Marine Ecology Progress Series 351: 189-199. Morales-Nin B (1992) Determination of growth in bony fishes from otolith microstructure. FAO Fisheries Technical Paper 322. FAO, Rome. Moreau J, Bambino C & Pauly D (1986) Indices of overall growth performance of 100 tilapia (Cichlidae) populations. In Maclean JL, Dizon LB & Hosillos LV (eds) The First Asian Fisheries Forum. Asian Fisheries Society, Manila: 201-206. Moreau J, Munyandorero J & Nyakageni B (1991) Evaluation des paramètres demographi-ques chez Stolothrissa tanganyikae et Limnothrissa miodon du lac Tanganyika. Verhandlungen der Internationalen Vereiningung für Theoretische und Ange-wandte Limnologie 24: 2552-2558. Moreau J, Palomares MLD, Torres FSB Jr & Pauly D (1995) Atlas demographique des populations de poissons d'eau douce d'Afrique. ICLARM Technical Reports 45. ICLARM, Manila. Moriarty DJW (1973) The physiology of digestion of blue-green algae in the cichlid fish, Tilapia nilotica. Journal of Zoology, London 171: 25-39. Morita K & Fukuwaka MA (2007) Why age and size at maturity have changed in Pacific salmon. Marine Ecology Progress Series 335: 289-294. Moyle PB & Marchetti MP (2006) Predicting invasion success: freshwater fishes in California as a model. AIBS Bulletin 56: 515-524. Moyo NAG & Fernando CH (2000) Growth, mortality and reproduction of Sargochromis codringtonii in Lake Kariba, Zimbabwe. Hydrobiologia 438: 237-244. Mtsambiwa MZ (1989) The age structure of pre-recruits of the sardine Limnothrissa miodon in Lake Kariba based on daily increments in otoliths. MSc Thesis. University of Wales, Bangor, UK. Mtsambiwa MZ (1993) Fitting a von Bertalanffy growth model to length at age data for larval Limnothrissa miodon from Lake Kariba. In Marshall BE & Mubamba R (eds) Papers presented at the Symposium on Biology, Stock Assessment and Exploitation of Small

182

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 182 References

Pelagic Fish Species in the African Great Lakes Region. Bujumbura, Burundi, from 25 to 28 November 1992. CIFA Occasional Paper 19. FAO, Rome. Mubamba R (1993) The biology and exploitation of small pelagic fishes in Zambia. In Marshall BE & Mubamba R (eds) Papers presented at the Symposium on Biology, Stock Assessment and Exploitation of Small Pelagic Fish Species in the African Great Lakes Region. Bujumbura, Burundi, from 25 to 28 November 1992. CIFA Occasional Paper 19. FAO, Rome. Mugiya Y (1987) Phase difference between calcification and organic matrix formation in the diurnal growth of otoliths in the rainbow trout, Salmo gairdneri. Fisheries Bulletin 85: 395-401. Mulimbwa N & Shirakihara K (1994) Growth, recruitment and reproduction of sardines (Stolothrissa tanganicae and Limnothrissa miodon) in north-western Lake Tanganyika. Tropics 4: 57-67. Murayama E, Herbomel P, Kawakami A, Takeda H & Nagasawa H (2005) Otolith matrix proteins OMP-1 and Otolin-1 are necessary for normal otolith growth and their correct anchoring onto the sensory maculae. Mechanisms of Development 122: 791-803. Murayama E, Okuno A, Ohira T, Takagi Y &Nagasawa H (2000) Molecular cloning and expression of an otolith matrix protein cDNA from the rainbow trout, Oncorhyn-chus mykiss. Comparative Biochemistry and Physiology 126: 511-520. Murayama E, Takagi Y, Ohira T, Davis JG, Greene MI & Nagasawa H (2002) Fish otolith contains a unique structural protein, otolin-1. European Journal of Biochemistry 269: 688-96. Murua H, Kraus G, Saborido-Rey F, Witthames PR, Thorsen A & Junquera S (2003) Procedures to estimate fecundity of marine fish species in relation to their reproductive strategy. Journal of Northwest Atlantic Fisheries Science 33: 33-54. Refs Murrell MC & Lores EM (2004) Phytoplankton and zooplankton seasonal dynamics in a subtropical estuary: importance of cyanobacteria. Journal of Plankton Research 26: 371-382. N Ndebele MR (2003) Some aspects of the ecology of phytoplankton in Lake Chivero, Harare, Zimbabwe. MSc thesis. University of Zimbabwe, Zimbabwe. Ndebele-Murisa MR, Mashonjowa E & Hill T (2011) The implications of a changing climate on the Kapenta ﬁsh stocks of Lake Kariba, Zimbabwe. Transactions of the Royal Society of South Africa. 66: 105-119. Ndebele-Murisa MR, Musil CF & Raitt L (2010) A review of phytoplankton dynamics in tropical African lakes. South African Journal of Science 106: 13-18. Ndebele-Murisa MR, Musil CF, Magadza CHD & Raitt L (2014) A decline in the depth of the mixed layer and changes in other physical properties of Lake Kariba’s water over the past two decades. Hydrobiologia 721: 185-195. Neat FC, Huntingford FA & Beveridge MM (1998a) Fighting and assessment in male cichlid fish: the effects of asymmetries in gonadal state and body size. Animal Behaviour 55: 883-891. Neat FC, Taylor AC & Huntingford FA (1998b) Proximate costs of fighting in male cichlid fish: the role of injuries and energy metabolism. Animal Behaviour 55: 875-882.

183

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 183 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

Needham JG & Needham PR (1962) A guide to the study of fresh-water biology. Holden-Day Inc San Francisco, Constable & Co Ltd London. Njiru M, Okeyo-Owuor JB, Muchiri M & Cowx IG (2004) Shifts in the food of Nile tilapia, Oreochromis niloticus (L.) in Lake Victoria, Kenya. African Journal of Ecology 42: 163-170. Njiru M, Waithaka E, Muchiri M, van der Knaap M & Cowx IG (2005) Exotic introduc-tions to the fishery of Lake Victoria: what are the management options? Lakes & Reservoirs: Research & Management 10: 147-155. Njiru M, Getabu A, Jembe T, Ngugi C, Owili M & van der Knaap M (2008a) Management of the Nile tilapia (Oreochromis niloticus (L.)) fishery in the Kenyan portion of Lake Victoria, in light of changes in its life history and ecology. Lakes & Reservoirs: Research & Management 13: 117-124. Njiru M, Ojuok J, Getabu A, Jembe T, Owili M & Ngugi C (2008b) Increasing dominance of Nile tilapia, Oreochromis niloticus (L) in Lake Victoria, Kenya: consequences for the Nile perch Lates niloticus (L) fishery. Aquatic Ecosystem Health & Management 11: 42-49. Njiru M, Ojuok JE, Getabu A, Muchiri M, Cowx IG & Okeyo-Owuor JB (2006a) Dominance of introduced Nile tilapia, Oreochromis niloticus (L.) in Lake Victoria: a case of changing biology and ecosystem. In Odada EO, Olago DO, Ochala W, Ntiba N, Wandiga S, Gichuki N & Oyieke H (eds) 11th World Lake Conference, Nairobi, Kenya, 31 October - 4 November 2005, Proceedings Vol. II. Ministry of Water and Irrigation, Nairobi, Kenya, and International Lake Environment Com-mittee Foundation, Shiga, Japan: 255-262. Njiru M, Ojuok JE, Okeyo-Owuor JB, Muchiri M, Ntiba MJ & Cowx IG (2006b) Some biological aspects and life history strategies of Nile tilapia Oreochromis niloticus (L.) in Lake Victoria, Kenya. African Journal of Ecology 44: 30-37. Nøttestad L, Utne KR, Óskarsson GJ, Jónsson SÞ, Jacobsen JA, Tangen Ø, Anthonypillai V, Aanes S, Vølstad JH, Bernasconi M & Debes H (2016) Quantifying changes in abundance, biomass, and spatial distribution of Northeast Atlantic mackerel (Scomber scombrus) in the Nordic seas from 2007 to 2014. ICES Journal of Marine Science: Journal du Conseil 73: 359-373. Nyamweya CS, Mlewa CM, Ngugi CC & Kaunda-Arara B (2010) Daily growth of young- of-the-year of the Baringo tilapia, Oreochromis niloticus baringoensis (Trewavas, 1983). African Zoology 45: 139-143. O Ogutu-Ohwayo R (1990) The decline of the native fishes of lakes Victoria and Kyoga (East Africa) and the impact of introduced species, especially the Nile perch, Lates niloticus, and the Nile tilapia, Oreochromis niloticus. Environmental Biology of Fishes 27: 81-96. Olenina I, Hajdu S, Edler L, Andersson A, Wasmund N, Busch S, Göbel J, Gromisz S, Huseby S, Huttunen M, Jaanus A, Kokkonen P, Ledaine I & Niemkiewicz E (2006) Biovolumes and size-classes of phytoplankton in the Baltic Sea. Baltic Sea Environment Proceedings 106. HELSINKI COMMISSION: Baltic Marine Environment Protection Commission, Helsinki. Oliveira RF & Almada VC (1996a) On the (in)stability of dominance hierarchies in the cichlid fish Oreochromis mossambicus. Aggressive Behaviour 22: 37-45.

184

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 184 References

Oliveira RF & Almada VC (1996b) Dominance hierarchies and social structure in captive groups of the Mozambique tilapia Oreochromis mossambicus (Teleostei Cichlidae). Ethology Ecology and Evolution 8: 39-55. Olsen EM, Lilly GR, Heino M, Morgan MJ, Brattey J & Dieckmann U (2005) Assessing changes in age and size at maturation in collapsing populations of Atlantic cod (Gadus morhua). Canadian Journal of Fisheries and Aquatic Sciences 62: 811-823. O’Reilly CM, Alin SR, Plisnier P-D, Cohen AS & McKee BA (2003) Climate change decreases aquatic ecosystem productivity of Lake Tanganyika, Africa. Nature 424: 766-768. P Pannella G (1971) Fish otoliths daily growth layers and periodical patterns. Science 173: 1124-1127. Parker GA (1992) The evolution of sexual size dimorphism in fish. Journal of Fish Biology 41: 1-20. Paulet G (2014) Kapenta rig survey of the Zambian waters of Lake Kariba. Report No. SF/2014/45. Indian Ocean Commission, Ebene, Mauritius. Payan P, de Pontual H, Boeuf G & Mayer-Gostan N (2004) Endolymph chemistry and otolith growth in fish. Comptes Rendus Palevol 3: 535–547. Peña-Mendoza B, Gómez-Márquez JL, Salgado-Ugarte IH & Ramírez-Noguera D (2005) Reproductive biology of Oreochromis niloticus (Perciformes: Cichlidae) at Emiliano Zapata dam, Morelos, Mexico. Revista de Biologia Tropical 53: 515-522. Pennock CA, Durst SL, Duran BR, Hines BA, Cathcart CN, Davis JE, Schleicher BJ & Franssen NR (2018) Predicted and observed responses of a nonnative Channel catfish population following managed removal to aid the recovery of endangered fishes. North American Journal of Fisheries Management 38: 565-578. Persson L (1987) The effects of resource availability and distribution on size class inter- Refs actions in perch, Perca fluviatilis. Oikos 48: 148-160. Persson L, Byström P and Wahlström E (2000) Cannibalism and competition in Eurasian perch: population dynamics of an ontogenetic omnivore. Ecology 81: 1058-1071. Pesez G & Pesez M (1977) Atlas de microscopie des eaux douces. Lechevalier, Paris. Peterson MS, Slack WT, Brown-Peterson NJ & McDonald JL (2004) Reproduction in non- native environments: establishment of Nile tilapia, Oreochromis niloticus, in coastal Mississippi watersheds. Copeia 2004: 842-849. Peterson MS, Slack WT, Waggy GL, Finley J, Woodley CM & Partyka ML (2006) Foraging in non-native environments: comparison of Nile tilapia and three co-occurring native centrarchids in invaded coastal Mississippi watersheds. Environ-mental Biology of Fishes 76: 283-301. Pfennig F, Kurth T, Meißner S, Standke A, Hoppe M, Zieschang F, Reitmayer C, Göbel A, Kretzschmar G & Gutzeit HO (2012) The social status of the male Nile tilapia (Oreochromis niloticus) influences testis structure and gene expression. Repro- duction 143: 71-84. Philippart J-C & Ruwet J-C (1982) Ecology and distribution of tilapias. In Pullin RSV & Lowe- McConnell RH (eds) The biology and culture of tilapias. ICLARM Conference Proceedings 7: 15-60. International Center for Living Aquatic Resources Manage-ment. Manila, Philippines.

185

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 185 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

Phiri C, Day J, Chimbari M & Dhlomo E (2007) Epiphytic diatoms associated with a submerged macrophyte, Valisneria aethiopioca, in the shallow marginal areas of Sanyati Basin (Lake Kariba): a preliminary assessment of their use as biomonitoring tool. Aquatic Ecology 41: 169-181. Pianka ER (1973) The structure of lizard communities. Annual Review of Ecology and Systematics 4: 53-74. Piñeiro C & Saínza M (2003) Age estimation, growth and maturity of the European hake (Merluccius merluccius (Linnaeus, 1758)) from Iberian Atlantic waters. ICES Journal of Marine Science / Journal du Conseil 60: 1086-1102. Pinkas L, Oliphant MS & Iverson ILK (1971) Food habits of albacore, bluefin tuna and bonita in Californian waters. California Fish and Game 152: 1-105. Pitcher TJ (1995) Thinking the unthinkable: a candidate model for predicting sustainable yields in introduced fish species in African lakes. In Pitcher TJ & Hart PJB (eds) The impact of species changes in African lakes. Fish and Fisheries Series 18: 495-525. Chapman & Hall, London. Poll M (1952) Les vertébrés. In Exploration hydrobiologique du Lac Tanganica, 1946-1947: résultats scientifiques. Institut Royal des Sciences Naturelles de Belgique, Bruxelles: 103-165. Poll M (1953) Poissons non Cichlidae. In Poll M Exploration hydrobiologique du Lac Tanganika (1946-1947): résultats scientifiques. Volume III, Fascicule 5A. Institut Royal des Sciences Naturelles de Belgique, Bruxelles. Poole JR, Sauey B, Amberg JJ & Bajer PG (2018) Assessing the efficacy of corn-based bait containing antimycin-a to control common carp populations using laboratory and pond experiments. Biological Invasions 20: 1809-1820. Pörtner HO, Peck MA (2010) Climate change effects on fishes and fisheries: towards a cause‐ and‐effect understanding. Journal of Fish Biology 77: 1745-1779. Q Quintana L, Silva A, Berois N, Macadar O (2004) Temperature induces gonadal maturation and affects electrophysiological sexual maturity indicators in Brachyhypopomus pinnicaudatus from a temperate climate. Journal of Experimental Biology 207: 1843- 1853. R Ramberg L (1987) Phytoplankton succession in the Sanyati basin, Lake Kariba. Hydro- biologia 153: 193-202. Reynolds ES (1963) The use of lead citrate at high pH as an electron-opaque stain in electron microscopy. Journal of Cell Biology 17: 208-212. Reynolds JE & Greboval DF (1988) Socio-economic effects of the evolution of Nile perch fisheries in Lake Victoria: a review. CIFA Technical Paper 17. FAO, Rome. Reznick DN & Ghalambor CK (2005) Can commercial fishing cause evolution? Answers from guppies (Poecilia reticulata). Canadian Journal of Fisheries and Aquatic Sciences 62: 791-801. Ricker WE (1968) Methods for assessment of fish production in fresh waters. IBP Handbook No 3. Blackwell Scientific Publications, Oxford.

186

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 186 References

Roberts ME, Wetzel II JE, Brooks RC & Garvey JE (2004) Daily increment formation in otoliths of the redspotted sunfish. North American Journal of Fisheries Manage-ment 24: 270-274. Romanuk TN, Zhou Y, Brose U, Berlow EL, Williams RJ & Martinez ND (2009) Predicting invasion success in complex ecological networks. Philosophical Transactions of the Royal Society B: Biological Sciences 364: 1743-1754. Rose GA & Kulka DW (1999) Hyperaggregation of fish and fisheries: how catch-per-unit- effort increased as the northern cod (Gadus morhua) declined. Canadian Journal of Fisheries and Aquatic Sciences 56: 118-127. Rowland WJ (1999) Studying visual cues in fish behaviour: a review of ethological techniques. Environmental Biology of Fishes 56: 285-305. Ruesink JL (2005) Global analysis of factors affecting the outcome of freshwater fish introductions. Conservation Biology 19: 1883-1893. Ruiz J, González‐Quirós RA, Prieto L & Navarro G (2009) A Bayesian model for anchovy (Engraulis encrasicolus): the combined forcing of man and environment. Fisheries Oceanography 18: 62-76. S Salthaug A & Aanes S (2003) Catchability and the spatial distribution of fishing vessels. Canadian Journal of Fisheries and Aquatic Sciences 60: 259-268. Sanches FHC, Miyai CA, Costa TM, Christofoletti RA, Volpato GL & Barreto RE (2012) Aggressiveness overcomes body-size effects in fights staged between invasive and native fish species with overlapping niches. PLoS One 7(1): e29746. https://doi.org/ 10.1371/journal.pone.0029746. Sanyanga RA (1986) Variations in abundance of Synodontis zambezensis (Pisces: Mo- chokidae) Peters 1852, in the inshore fishery of Lake Kariba. Fisheries Research 26: Refs 171-186. Sanyanga RA, Machena C & Kautsky N (1995) Abundance and distribution of inshore fish in fished and protected areas in Lake Kariba, Zimbabwe. Hydrobiologia 306: 67-78. Schemske DW (1974) Age, length, and fecundity of the Creek chub, Semotilius atromacu-latus (Mitchill), in Central Illinois. The American Midland Naturalist 92: 505-509. Schenck JR & Whiteside BG (1977) Reproduction, fecundity, sexual dimorphism and sex ratio of Etheostoma fonticola (Osteichthyes: Percidae). The American Midland Naturalist 98: 365-375. Schoener TW (1974) Nonsynchronous spatial overlap of lizards in patchy habitats. Ecology 51: 408-418. Schwanck EJ (1995). The introduced Oreochromis niloticus is spreading on the Kafue floodplain, Zambia. Hydrobiologia 315:143-147. Seppänen E, Tiira K, Huuskonen H & Piironen J (2009) Metabolic rate, growth and aggressiveness in three Atlantic salmon Salmo salar populations. Journal of Fish Biology 74: 562-575. Sibanda P (2003) The possible effects of global warming on the growth of algae. BSc Thesis, Biological Sciences, University of Zimbabwe, Harare. Skelton P (2001) A complete guide to the freshwater fishes of southern Africa. Struik, Cape Town.

187

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 187 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

Smith AD, Brown CJ, Bulman CM, Fulton EA, Johnson P, Kaplan IC, Lozano-Montes H, Mackinson S, Marzloff M, Shannon, LJ & Shin YJ (2011) Impacts of fishing low– trophic level species on marine ecosystems. Science 333: 1147-1150. Smith RJ (2009) Use and misuse of the reduced major axis for line‐fitting. American Journal of Physical Anthropology 140: 476-486. Songore N & Kolding J (2003) Diversity, disturbance, and ecological succession of the inshore fish populations in Lake Kariba, Zimbabwe from 1960 to 2001. Paper presented for The Linnean Society of London Symposium on the use of long-term databases for the prediction of ecological change. London. October 23rd-24th, 2003. https://www.academia.edu/14642691/Diversity_disturbance_and_ ecological_succession_of_the_inshore_fish_populations_in_Lake_Kariba_Zimbabw e_from_1960_to_2001. Google Scholar Spliethoff PC, de Iongh HH & Frank VG (1983) Success of the introduction of the fresh water clupeid Limnothrissa miodon (Boulenger) in Lake Kivu. Fisheries Mana-gement 14: 17-31. Starkie A (2003) Management issues relating to the European eel, Anguilla anguilla. Fisheries Management and Ecology 10: 361-364. Stearns SC & Koella JC (1986) The evolution of phenotypic plasticity in life‐history traits: predictions of reaction norms for age and size at maturity. Evolution 40: 893-913. Stewart J & Hughes JM (2005) Longevity, growth, reproduction and a description of the fishery for Silver sweep Scorpis lineolatus off New South Wales, Australia. New Zealand Journal of Marine and Freshwater Research 39: 827-838. Stewart KM (1988) Changes in condition and maturation of the Oreochromis niloticus L. population of Ferguson's Gulf, Lake Turkana, Kenya. Journal of Fish Biology 33: 181-188. St. John MA, Clemmesen C, Lund T & Köster T (2001) Diatom production in the marine environment: implications for larval fish growth and condition. ICES Journal of Marine Science 58: 1106–1113. Stuart IG, Williams A, McKenzie J & Holt T (2006) Managing a migratory pest species: a selective trap for common carp. North American Journal of Fisheries Manage-ment 26: 888-93. Sutherland WJ (1996) From individual behaviour to population ecology. Oxford Series in Ecology and Evolution. Oxford University Press, Oxford. Sutter DA, Suski CD, Philipp DP, Klefoth T, Wahl DH, Kersten P, Cooke SJ & Arlinghaus R (2012) Recreational fishing selectively captures individuals with the highest fitness potential. Proceedings of the National Academy of Sciences 109: 20960-20965. T Takagi Y, Toshe H, Murayama E, Ohira T & Nagasawa H (2005) Diel changes in endo-lymph aragonite saturation rate and mRNA expression of otolith matrix proteins in the trout otolith organ. Marine Ecology Progress Series 294: 249-256. Takasuka A, Oozeki Y & Aoki I (2007) Optimal growth temperature hypothesis: why do anchovy flourish and sardine collapse or vice versa under the same ocean regime? Canadian Journal of Fisheries and Aquatic Sciences 64: 768-776.

188

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 188 References

Tanaka K, Mugiya YA & Yamada J (1981) Effects of photoperiod and feeding on daily growth patterns in otoliths of juvenile Tilapia nilotica. Fishery Bulletin: 79: 459-456. Taniguchi Y, Fausch KD & Nakano S (2002) Size-structured interactions between native and introduced species: can intraguild predation facilitate invasion by stream salmonids? Biological Invasions 4: 223-233. Taranger GL, Carrillo M, Schulz RW, Fontaine P, Zanuy S, Felip A, Weltzien FA, Dufour S, Karlsen Ø, Norberg B, Andersson E (2010) Control of puberty in farmed fish. General and Comparative Endocrinology 165: 483-515. Taylor GC & Weyl OLF (2012) Otoliths versus scales: evaluating the most suitable structure for ageing Largemouth bass, Micropterus salmoides, in South Africa. African Zoology 47: 358-362. Thys van den Audenaerde DFE (1994) Introduction of aquatic species into Zambian waters, and their importance for aquaculture and fisheries. ALCOM Field Document 24. FAO, Rome. Tiira K, Laurila A, Enberg K & Piironen J (2009) Short-term dominance: stability and consequences for subsequent growth. Journal of Fish Biology 74: 2374-2385. Tiki M (2011) Growth, mortality and exploitation of Nile tilapia (Oreochromis niloticus) in Lake Chivero and implications on the management of the artisanal fishery. MSc Thesis. University of Zimbabwe, Harare. Trewavas E (1983) Tilapiine fishes of the genera Sarotherodon, Oreochromis and Danakilia. British Museum (Natural History), London. Trippel EA (1995) Age at maturity as a stress indicator in fisheries. BioScience 45: 759-771. Trippel EA & Harvey HH (1991) Comparison of methods used to estimate age and length of fishes at sexual maturity using populations of white sucker (Catostomus commersoni). Canadian Journal of Fisheries and Aquatic Sciences 48: 1446-1459. Refs Tshibangu KK & Kinoshita I (1995) Early life histories of two clupeids, Limnothrissa miodon and Stolothrissa tananicae, from Lake Tanganyika. Japanese Journal of Ichthyology 42: 81-87. Tudorancea C, Fernando CH & Paggi JC (1988) Food and feeding ecology of Oreochromis niloticus (Linnaeus, 1758) juveniles in Lake Awassa (Ethiopia). Archiv für Hydrobiologie, Supplementband. Monographische Beiträge 79: 267-289. Tumbare MJ (2000) Commercial activities of Lake Kariba. In Tumbare MJ (ed) Management of river basins and dams: the Zambezi River Basin. AA Balkema, Rotterdam: 199- 200. Tumbare MJ & Sakala WK (2000) Seismic activities around Lake Kariba. In Tumbare MJ (ed) Management of river basins and dams: the Zambezi River Basin. AA Balkema, Rotterdam: 227-237. Turner GF (1986) Territory dynamics and cost of reproduction in a captive population of the colonial nesting mouthbrooder Oreochromis mossambicus (Peters). Journal of Fish Biology 29: 573-587. Turner GF & Huntingford FA (1986) A problem for game theory analysis: assessment and intention in male mouthbrooder contests. Animal Behaviour 34: 961-970. Twongo T (1995) Impact of fish species introductions on the tilapias of Lakes Victoria and Kyoga. In Pitcher TJ & Hart PJB (eds) The impact of species changes in African lakes. Chapman & Hall Fish and Fisheries Series 18: 45-57. Springer, Dordrecht.

189

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 189 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

Tweddle D (2010) Overview of the Zambezi river system: its history, fish fauna, fisheries, and conservation. Aquatic Ecosystem Health & Management 13: 224-240. V van den Bosch F & Gabriel W (1997) Cannibalism in an age-structured predator-prey system. Bulletin of Mathematical Biology 59: 551-567. van den Bosch F, de Roos AM & Gabriel W (1988) Cannibalism as a life boat mechanism. Journal of Mathematical Biology 26: 619-633. van der Lingen MI (1973) Lake Kariba: early history and south shore. In Ackermann WC, White GF & Worthington EB (eds) Man-made lakes: their problems and environ-mental effects. Geophysical monograph Series 17. Washington DC: 132-142. van Zwieten PAM, Béné C, Kolding J, Brummett R, Valbo-Jørgensen J (2011) Review of tropical reservoirs and their fisheries – The cases of Lake Nasser, Lake Volta and Indo-Gangetic Basin reservoirs. FAO Fisheries and Aquaculture Technical Paper 557. FAO, Rome. Verburg P, Hecky RE & Kling H (2003) Ecological consequences of a century of warming in Lake Tanganyika. Science 301: 505-507. Verschuren D (2003) Global change: the heat on Lake Tanganyika. Nature 424: 731. W Wahlström E, Persson L, Diehl S & Byström P (2000) Size-dependent foraging efficiency, cannibalism and zooplankton community structure. Oecologia 123: 138-148. Wallace RK (1981) An assessment of diet overlap indexes. Transactions of the American Fisheries Society 110: 72-76. Walters C & Maguire JJ (1996) Lessons for stock assessment from the northern cod collapse. Reviews in Fish Biology and Fisheries 6: 125-137. Wanink JH (1998) The pelagic cyprinid Rastrineobola argentea as a crucial link in the disrupted ecosystem of Lake Victoria: Dwarfs And Giants–African Adventures. PhD thesis, Leiden University. EburonP&L, Delft. Wanink JH (1999) Prospects for the fishery on the small pelagic Rastrineobola argentea in Lake Victoria. Hydrobiologia 407: 183-189. Wanink JH (2005) Strangers in Lake Victoria: an evolutionary trap for the unsuspecting furu. In Drent RH & Moore SJ (eds) Seeking nature's limits: ecologists in the field. KNNV Publishing, Utrecht: 244-249. Wanink JH & Witte F (2000) The use of perturbation as a natural experiment: effects of predator introduction on the community structure of zooplanktivorous fish in Lake Victoria. Advances in Ecological Research 31: 553-570. Wanink JH, Witte F & Hoogenboezem W (1998) Life history tactics of the Lake Victoria sardine Rastrineobola argentea, a survivor of the introduced Nile perch. In Wanink JH The pelagic cyprinid Rastrineobola argentea as a crucial link in the disrupted ecosystem of Lake Victoria: Dwarfs And Giants–African Adventures. PhD thesis, Leiden University: 183-193. Watabe N, Tanaka K, Yamada J & Dean JM (1982). Scanning electron microscope observations of the organic matrix in the otolith of the teleost fish Fundulus heteroclitus and Tilapia nilotica. Journal of Experimental Marine Biology and Ecology 58: 127-134.

190

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 190 References

Watson RA, Cheung WW, Anticamara JA, Sumaila RU, Zeller D & Pauly D (2013) Global marine yield halved as fishing intensity redoubles. Fish and Fisheries 14: 493-503. Welcomme RL (ed) (1988) International introductions of inland aquatic species. FAO Fisheries Technical Paper 294. FAO, Rome. Werner EE & Gilliam JF (1984) The ontogenetic niche and species interactions in size- structured populations. Annual Review of Ecology and Systematics 15: 393-425. Weyl OLF (2008) Rapid invasion of a subtropical lake fishery in central Mozambique by Nile tilapia, Oreochromis niloticus (Pisces: Cichlidae). Aquatic Conservation: Marine and Freshwater Ecosystems 18: 839-851. Weyl OLF & Hecht T (1998) The biology of Tilapia rendalli and Oreochromis mossambicus (Pisces: Cichlidae) in a subtropical lake in Mozambique. South African Journal of Zoology 33: 178-188. Wilson AE, Sarnelle O & Tillmanns AR (2006) Effects of cyanobacterial toxicity and morphology on the population growth of freshwater zooplankton: meta‐analyses of laboratory experiments. Limnology and Oceanography. 51: 1915-1924. Windell JT (1968) Food analysis and rate of digestion. In Ricker WE (ed) Methods for assessment of fish production in fresh waters. IBP Handbook 3. International Biological Programme. Blackwell Scientific Publications, Oxford. Winemiller KO & Kelso-Winemiller LC (2003) Food habits of tilapiine cichlids of the Upper Zambezi River and floodplain during the descending phase of the hydrologic cycle. Journal of Fish Biology 63: 120-128. Winker H, Weyl OLF, Booth AJ & Ellender BR (2010) Validating and corroborating the deposition of two annual growth zones in asteriscus otoliths of Common carp Cyprinus carpio from South Africa’s largest impoundment. Journal of Fish Biology 77: 2210-2228. Refs Witte F & van Densen WLT (eds) (1995) Fish stocks and fisheries of Lake Victoria: a handbook for field observations. Samara Publishing, Cardigan UK. Witte F, Goldschmidt T & Wanink JH (1995) Dynamics of the haplochromine cichlid fauna and other ecological changes in the Mwanza Gulf of Lake Victoria. In Pitcher TJ & Hart PJB (eds) The impact of species changes in African lakes. Chapman and Hall Fish and Fisheries Series 18: 83-110. Springer, Dordrecht. Witte F, Goldschmidt T, Wanink J, van Oijen M, Goudswaard K, Witte-Maas E & Bouton N (1992) The destruction of an endemic species flock: quantitative data on the decline of the haplochromine cichlids of Lake Victoria. Environmental Biology of Fishes 34: 1-28. Woodward JG (1974) Successful reproduction and distribution in Lake Kariba. In Balon EK & Coche AG (eds) Lake Kariba: a man-made tropical ecosystem in Central Africa. Dr W Junk Publishers, The Hague: 526-535. Wudneh T (1998) Biology and management of fish stocks in Bahar Dar Gulf, Lake Tana, Ethiopia. PhD thesis, Wageningen Agricultural University, Wageningen. Y Yamaguchi Y, Hirayama N, Koike A & Adam HA (1990) Age determination and growth of Oreochromis niloticus and Sarotherodon galilaeus in High Dam Lake, Egypt. Bulletin of the Japanese Society of Scientific Fisheries 56: 437-403.

191

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 191 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

Yamaoka K (1991) Feeding relationships. In Keenleyside MHA (ed) Cichlid fishes: behaviour, ecology and evolution. Fish and Fisheries Series 2: 151-172. Chapman & Hall, London. Yasue N & Takasuka A (2009) Seasonal variability in growth of larval Japanese anchovy Engraulis japonicus driven by fluctuations in sea temperature in the Kii Channel, Japanese Journal of Fish Biology 74: 2250-2268. Yongo E, Outa N, Kito K & Matsushita Y (2018) Studies on the biology of Nile tilapia (Oreochromis niloticus) in Lake Victoria, Kenya: in light of intense fishing pressure. African Journal of Aquatic Science 20: 1-4. Z Zekeria ZA, Weertman S, Samuel B, Kaleab T & Videler JJ (2006) Growth of Chaetodon larvatus (Chaetodontidae: Pisces) in the southern Red Sea. Marine Biology 148: 1113-1122. Zengeya TA & Marshall BE (2007) Trophic interrelationships amongst cichlid fishes in a tropical African reservoir (Lake Chivero, Zimbabwe). Hydrobiologia 592: 175-182. Zengeya TA & Marshall BE (2008) The inshore fish community of Lake Kariba half a century after its creation: what happened to the Upper Zambezi species invasion? African Journal of Aquatic Science 33: 99-102. Zengeya TA, Booth AJ & Chimimba CT (2015) Broad niche overlap between invasive Nile tilapia Oreochromis niloticus and indigenous congenerics in southern Africa: should we be concerned? Entropy 17: 4959-4973. Zengeya TA, Booth AJ, Bastos AD & Chimimba CT (2011). Trophic interrelationships between the exotic Nile tilapia, Oreochromis niloticus and indigenous tilapiine cichlids in a subtropical African river system (Limpopo River, South Africa). Environmental Biology of Fishes 92: 479-489. Zengeya TA, Robertson MP, Booth AJ & Chimimba CT (2013) A qualitative ecological risk assessment of the invasive Nile tilapia, Oreochromis niloticus in a sub-tropical African river system (Limpopo River, South Africa). Aquatic Conser-vation: Marine and Freshwater Ecosystems 23: 51-63. Zhang Y & Prepas EE (1996) Regulation of the dominance of planktonic diatoms and cyanobacteria in four eutrophic hardwater lakes by nutrients, water column stability, and temperature. Canadian Journal of Fisheries and Aquatic Sciences 53: 621-633. Zhang Z & Runhau NW (1992) Temporal deposition of incremental and discontinuous zones in the otoliths of Oreochromis niloticus (L.) Journal of Fish Biology 40: 333-339.

192

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 192 Authors affiliations and addresses

Auth

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 193

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 194 Authors affiliations and addresses

Portia Chiyedza Chifamba  Department of Biological Sciences, University of Zimbabwe, PO Box MP 167, Mount Pleasant, Harare, Zimbabwe  Centre for Ecological and Evolutionary Studies, University of Groningen, PO Box 11103, 9700 CC Groningen, The Netherlands

Britas Klemens Eriksson  Centre for Ecological and Evolutionary Studies, University of Groningen, PO Box 11103, 9700 CC Groningen, The Netherlands

Tendai Mauru  Department of Biological Sciences, University of Zimbabwe, PO Box MP 167, Mount Pleasant, Harare, Zimbabwe

Han Olff  Centre for Ecological and Evolutionary Studies, University of Groningen, PO Box 11103, 9700 CC Groningen, The Netherlands

John J. Videler  Centre for Ecological and Evolutionary Studies, University of Groningen, PO Box 11103, 9700 CC Groningen, The Netherlands  Zuidlaarderweg 57, 9479 TH, Noordlaren, The Netherlands Auth

Jan H. Wanink  Waterschap Noorderzijlvest, Regional Water Authority, PO Box 18, 9700 AA Groningen, The Netherlands  Institute of Biology Leiden, Leiden University, PO Box 9500, 2300 RA Leiden, The Netherlands

195

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 195

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 196

List of publications

Publ

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 197

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 198 List of publications

C Chifamba PC (1990) Preference of Tilapia rendalli (Boulenger) for some species of aquatic plants. Journal of Fish Biology 36: 701-705. Chifamba PC (1991) Evaluation of some components of the Lake Kariba ‘Kapenta Fishing Unit’. MSc thesis. University college of North Wales, Bangor. Chifamba PC (1992) Daily rings of otoliths as a method for ageing the sardine, Limnothrissa miodon in Lake Kariba. Transaction of Zimbabwe Scientific Association 66: 15-17. Chifamba PC (1993) The life history style of Limnothrissa miodon in Lake Kariba. In Marshall BE & Mubamba R (eds) Papers presented at the Symposium on biology, stock assessment and exploitation of small pelagic fish species in the African Great Lakes region. Bujumbura, Burundi, from 25 to 28 November 1992. CIFA Occasional Paper 19. FAO, Rome. http://www.fao.org/docrep/005/v2648e/V2648E07.htm#ch7 Chifamba PC (1995) Factors affecting the fishing power of sardine fishing vessels and implications for the management of the fisheries on Lake Kariba, Zimbabwe. Fisheries Management and Ecology 2: 309-319. Chifamba PC (1998) Status of Oreochromis niloticus in Lake Kariba, Zimbabwe, following its escape from fish farms. In Cowx IG (ed) Stocking and introduction of fish. Blackwell Science, Oxford: 267-273. Chifamba PC (2000) The relationship of temperature and hydrological factors to catch per unit effort, condition and size of the freshwater sardine, Limnothrissa miodon (Boulenger), in Lake Kariba. Fisheries Research 45: 271-281. Chifamba PC (2006) Spatial and historical changes of indigenous O. mortimeri following introduction of exotic O. niloticus in Lake Kariba. In Odada EO, Olago DO, Ochala W, Ntiba N, Wandiga S, Gichuki N & Oyieke H (eds) 11th World Lake Conference, Nairobi, Kenya, 31 October - 4 November 2005, Proceedings Vol. II. Ministry of Water and Irrigation / International Lake Environment Committee Foundation (ILEC), Nairobi, Kenya / Shiga, Japan: 500-504. Chifamba PC (2007) Trace metal contamination of water at a solid waste disposal site at Publ Kariba, Zimbabwe. African Journal of Aquatic Science 32: 71-78. Chifamba PC & Mauru T (2017) Comparative aggression and dominance of Oreochromis niloticus (Linnaeus, 1758) and Oreochromis mortimeri (Trewavas, 1966) from paired contest in aquaria. Hydrobiologia 788: 193-203. Chifamba PC & Videler (2014) Growth of the alien Oreochromis niloticus Linnaeus and indigenous Oreochromis mortimeri Trewavas in Lake Kariba. African Journal of Aquatic Science 39: 167-176. M Mahere T, Mtsambiwa MZ, Chifamba PC & Nhiwatiwa T (2012) An assessment of catch trends of the Kapenta (Limnothrissa miodon) fishery on the Zimbabwean side of Lake Kariba (1974-2011). Journal of Applied Science in Southern Africa 18: 13-29.

199

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 199 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

Mahere TS, Mtsambiwa MZ, Chifamba PC & Nhiwatiwa T (2014) Climate change impact on the limnology of Lake Kariba, Zambia-Zimbabwe. African Journal of Aquatic Science 39: 215-221. Marufu LT & Chifamba PC (2013) A comparison of diel feeding pattern, ingestion and digestive efficiency of Oreochromis niloticus and Oreochromis macrochir in Lake Chivero, Zimbabwe. African Journal of Aquatic Science 38: 221-228. Marufu L, Barson M, Chifamba P, Tiki M & Nhiwatiwa T (2018) The population dynamics of a recently introduced crayfish, Cherax quadricarinatus (von Martens, 1868), in the Sanyati Basin of Lake Kariba, Zimbabwe. African Zoology 53: 17-22. Mhlanga L, Phiri C & Chifamba P (2014) Challenges surrounding water management within the Lake Kariba environs. In Mhlanga L, Nyikahadzoi K & Haller T (eds) Fragmentation of natural resources management: experiences from Lake Kariba. Lit Verlag, Berlin: 38-64. Muisa N, Hoko Z & Chifamba P (2011) Impacts of alum residues from Morton Jaffray Water Works on water quality and fish, Harare, Zimbabwe. Physics and Chemistry of the Earth 36: 853-864. W Wanink JH & Chifamba PC (1999) Interactions between freshwater fisheries and birds. In Farina A (ed) Perspectives in ecology: a glance from the VII International Congress of Ecology (Florence 19-25 July 1998). Backhuys Publishers, Leiden: 219-225.

200

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 200

Summary

Sum

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 201

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 202 Summary

The biology and competitive displacement of a native fish, Oreochromis mortimeri, by a congeneric species, Oreochromis niloticus, and the biology and exploitation of an introduced pelagic sardine, Limnothrissa miodon in Lake Kariba were studied in this thesis. O. niloticus was introduced from escapees of an aquaculture farm in the early 1990s and displaced an endemic and native fish to the Middle Zambezi River, O. mortimeri, from the Sanyati basin of Lake Kariba. To provide insight on the displacement process, the growth, reproductive effort, aggression and diet of the two species were compared. Kariba dam was constructed for Hydro Electricity generation in 1957, creating a relatively large aquatic lacustrine habitat with conditions that in many ways differed from fluvial conditions of the Zambezi River. The indigenous fish species were able to adapt to living in the inshore area, leaving the pelagic area that constitutes about 70% of the lake, devoid of fish. The pelagic area was hence considered to be a ‘vacant niche’ with plankton that could be used to enhance the fish production of the new lake. A pelagic freshwater sardine, Limnothrissa miodon was introduced into Lake Kariba from Lake Tanganyika, to fill this vacant niche, in 1967 and 1968. By 1974, L. miodon was considered established in the lake and a pelagic fishery based on the introduced sardine commenced that year. It became the most productive fishery on Lake Kariba and in Zimbabwe. Since 1990, the fishery has declined. In the thesis the causes of the decline were evaluated. How well the two introduced fish species became established in Lake Kariba was assessed by comparing their growth in Kariba to that in other water bodies. Based on its growth rate (k), maximum length (L∞) and age at maturity, Oreo- chromis niloticus thrives in Lake Kariba. However, the growth rate determined in Lake Kariba in 1997 (k = 0.25), 2003 (k = 0.29) and 2010 (k = 0.29) is in the middle of the range reported for some populations throughout the world (k = 0.11 – 0.75). Similarly, the maximum size (L∞) in Lake Kariba (32.4, 44.6 and 37.8 cm standard length (SL) in 1997, 2003 and 2010, respectively) is in the middle of the range of Sum the same reported populations (10.2 – 57.2 cm SL). The length at first maturity in Kariba is high (17.26 cm SL) compared, for example, to the small size at maturity (11.2 – 13.2 cm SL) of populations in eight reservoirs of varying sizes in Côte ďIvoire, but it is much lower than that published for a population in Lake Victoria (26 and 29 cm SL for females and males, respectively). The differences in size and age at maturity in the populations of O. niloticus in the various water bodies, reflects the plasticity in growth rate and maximum size of this species in relation to environmental conditions. The oldest specimen of O. niloticus in Lake Kariba were 10, 6 and 8 years in 1997, 2003 and 2010, respectively. Length frequency data from the artisanal fishery and experimental fishing show that the population has many age groups. Such a population is less vulnerable to some years of recruitment failure than

203

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 203 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

a population comprising only a few age classes. A population with this age structure, coupled with a high age at maturity, may reflect a moderate fishing pressure. For the sustainability of this fishery, these population attributes need to be monitored. The introduction of Oreochromis niloticus into Lake Kariba, where O. mortimeri was already present, brought together two formerly geographically separated congeners with a similar niche. Following the introduction, the catches of O. niloticus rose whilst those of O. mortimeri declined, resulting in the displacement of the native species from many parts of the lake. The total catches of these two species, however, remained the same since O. niloticus first appeared in the catches. The mechanisms that conferred competitive advantage to O. niloticus over O. mortimeri, were explored in studies comparing growth, aggression, diet and reproductive effort. The mean stomach fullness was 70.9% ± 9.4 for O. niloticus and 73.4% ± 9.8 for O. mortimeri, and the difference between species was not significant. The two species had a similar diet of plankton and detritus, and a crude protein digestion efficiency that indicates potential competition for food. However, the higher growth rate and maximum size of O. niloticus may have conferred this species competitive advantage over O. mortimeri. A large size offers advantages in the form of increased reproductive effort (since the egg number in the ovary is often correlated to female size) and dominance. The largest fish of a pair in an aquarium contest to test aggression, was more aggressive and became dominant most of the time, in both species. Even though in general Oreochromis mortimeri was the more aggressive species, being larger in size at a particular age, O. niloticus negates this species-specific difference in aggression. Oreochromis niloticus may have a reproductive advantage, because the proportion of fish at a high level of gonadal development every month, was always higher than in O. mortimeri. A high number of reproducing females may signal a higher breeding effort of O. niloticus compared to O. mortimeri. Unlike the unplanned introduction of Oreochromis niloticus, the deliberate introduction of Limnothrissa miodon into Lake Kariba had no major negative impact. The major difference between these two introductions is that L. miodon was introduced in order to occupy a vacant niche. Therefore, interactions with other fish species were rare. The availability of a vacant niche should be an important consideration in all fish introductions, to avoid potential biodiversity loss. The objective of the introduction of L. miodon to increase fish production in Lake Kariba was achieved. In order to maximize the social and economic advantages gained, the management of the L. miodon fishery needs to be optimized through the acquisition of scientific knowledge on sardine biology and fishery, such as collected in the studies contained in this thesis. Growth rate and size at maturity were estimated, and what they reveal about the status of the fish population and fishery was discussed. The maximum size (L∞) estimates of Limnothrissa miodon obtained from the Gompertz models were: 18.0, 9.6 and 15.2 cm (TL), in 1993 – 1994, 1996 and 2012 – 2013, respectively. These

204

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 204 Summary

values, including some annual variation, are comparable to those in Lake Tanganyika. Early studies of L. midon concluded that the population in Lake Kariba was stunted. The lack of large fish may be due to the capture of fish before they reach a large size, and to the migration of the large fish into a more inshore area, where they are not captured by the offshore operating fishing fleet. The age at first maturity is lower than in Lake Tanganyika and has declined since the time the species was introduced into Lake Kariba. Maturity occurred at a much smaller size in this study (3.43 and 3.63 cm for females and males, respectively), compared to 5.2 – 5.6 cm for females and 7.1 – 7.3 cm for males in 1970 – 1972, indicating a large temporal decrease in size at maturity. The decline in age at maturity may be an indicator of overfishing, which causes the selective removal of fast-growing fish and favours the survival of those with a shorter life cycle. This change in the life history of the sardine can slow down the recovery of the fish population. The introduced species in Lake Kariba have provided lessons on the importance of fish introductions to enhance fish production in a reservoir, and lessons on how overfishing and an adverse environment reduce the catches, resulting in an erosion of social and economic benefits. I conclude that management of the fishery should incorporate these two factors in the estimation of productivity, management and monitoring of the fishery. The fish introductions also provided an opportunity to witness the effect of the introduction of a species that shares an ecological niche with a native species. I conclude that an introduction into an occupied niche may not result in higher fish catches and profit. Instead, it may inflict costs in terms of reduced biodiversity through the loss of native and endemic species.

Sum

205

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 205 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

206

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 206

Samenvatting (Dutch summary)

Sam

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 207

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 208 Samenvatting (Dutch summary)

Door de bouw van de Kariba stuwdam, met daarin twee waterkrachtcentrales, ontstond in 1957 een groot stuwmeer (6 000 km2) in de Zambezi rivier op de grens van Zambia en Zimbabwe: het Karibameer. De omstandigheden in het nieuwgevormde lotische ecosysteem verschilden in vele opzichten van die in de rivier. De reeds aanwezige riviervissen konden zich alleen aanpassen aan de omstandigheden in de oeverzone van het nieuwe meer. Daarmee bleef het open water, dat ongeveer 70% van de totale oppervlakte beslaat, verstoken van vis. De pelagische zone werd beschouwd als een onbezette niche voor vis, waarin voldoende plankton aanwezig was om de visproductie van het meer te kunnen verhogen. Om deze niche op te vullen, werd in 1967 en 1968 een uit het Tanganyikameer afkomstige pelagische zoetwater-sardine, Limnothrissa miodon, geïntroduceerd. Vanaf 1974 wordt L. miodon beschouwd als een gevestigde soort. In dat jaar ging een nachtelijke lichtvisserij op de sardine van start, die zich heeft ontwikkeld tot de meest productieve visserij van het Karibameer en van heel Zimbabwe. Na 1990 liepen de vangsten echter sterk terug. In tegenstelling tot de introductie van Limnothrissa miodon, was de introductie van een andere exoot, de Nijltilapia (Oreochromis niloticus), niet bewust. In het begin van de jaren negentig van de vorige eeuw, vestigde deze soort zich in het meer via ontsnappingen vanuit aangrenzende viskwekerijen. In dit geval was geen sprake van een onbezette niche die kon worden opgevuld. De inheemse tilapia Oreochromis mortimeri, een endemische soort van de Midden Zambezi, vormde een concurrent voor habitat en voedsel. Beide tilapias verkiezen de oeverzone als leefgebied, hebben een grotendeels overlappend planktivoor dieet, een vergelijkbare voortplantings- strategie, en zijn genetisch nauw verwant. Na enkele jaren bleek de inheemse tilapia in het onderzoeksgebied (Sanyati basin) al te zijn verdrongen door de Nijltilapia. De biologie van, en de visserij op, de twee geïntroduceerde soorten, Oreochromis niloticus en Limnothrissa miodon, vormen samen met de invloed die ze uitoefenen op de visgemeenschap en de visserij, het onderwerp van dit proefschrift. Door de groeiparameters van beide soorten in het Karibameer te vergelijken met die in andere wateren, zijn de habitatgeschiktheid van het meer en de mate van vestiging van de beide exoten in beeld gebracht. Het proces van concurrentie en verdringing binnen het genus Oreochromis is onderzocht door het vergelijken van de groeiparameters, de reproductieve inspanning, de agressiviteit en het dieet van O. niloticus en O. mortimeri. Inzicht in de ontwikkeling van de sardinepopulatie en de effecten hiervan op de visserij is verkregen door onderzoek aan de relatieve groeisnelheid en de gemiddelde lengte bij geslachtsrijpheid. De relatieve groeisnelheid (k), de maximale gemiddelde lengte (L∞) en de gemiddelde lengte bij geslachtsrijpheid van O. niloticus wijzen op het floreren van de soort in het Karibameer, en daarmee op gunstige milieuomstandigheden. De in 209

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 209 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

het Karibameer gemeten relatieve groeisnelheden in 1997 (k = 0.25), 2003 (k = 0.29) en 2010 (k = 0.29) liggen centraal in het bereik dat voor diverse over de wereld verspreide populaties is gerapporteerd (k = 0.11 – 0.75). Hetzelfde geldt voor de maximale gemiddelde lengte in Kariba (L∞ = 32.4, 44.6 en 37.8 cm standaard lengte (SL) in respectievelijk 1997, 2003 en 2010) ten opzichte van dezelfde over de wereld verspreide populaties (L∞ = 10.2 – 57.2 cm SL). De gemiddelde lengte bij geslachtsrijpheid in het Karibameer (17.26 cm SL) is hoog in vergelijking tot de waarden (11.2 – 13.2 cm SL) die in acht reservoirs van verschillende omvang in Ivoorkust zijn gevonden, maar veel lager dan de gepubliceerde waarden (respectievelijk 26 en 29 cm SL voor vrouwtjes en mannetjes) voor een populatie in het Victoriameer. De variatie in de gemiddelde lengte en leeftijd bij geslachtsrijpheid van Oreochromis niloticus in verschillende wateren weerspiegelt de plasticiteit in de relatieve groeisnelheid en de maximale lengte van deze soort in relatie tot omgevingsfactoren. In 1997, 2003 en 2010 waren de oudste gevangen exemplaren van de Nijltilapia in het Karibameer respectievelijk 10, 6 en 8 jaar oud. De lengte-frequentiegegevens van de artisanale visserij en de bemonstering laten zien dat de populatie bestaat uit een groot aantal leeftijdsklassen. Een populatie met deze leeftijdsopbouw is minder kwetsbaar voor enkele jaren zonder aanwas dan een populatie die bestaat uit slechts een paar leeftijdsklassen. In combinatie met een hoge gemiddelde leeftijd bij geslachtsrijpheid weerspiegelt een dergelijke populatie waarschijnlijk een bescheiden visserijdruk. Om de duurzaamheid van de tilapiavisserij te waarborgen, is het monitoren van de leeftijdsopbouw van de populatie noodzakelijk. De introductie van Nijltilapia in het Karibameer, waar Oreochromis mortimeri al aanwezig was, bracht twee nauw verwante soorten met dezelfde ecologische niche samen die voorheen geografisch gescheiden voorkwamen. Hierna begonnen de vangsten van O. niloticus te stijgen en die van O. mortimeri af te nemen totdat de inheemse soort op de meeste plaatsen in het meer verdrongen was. De totale hoeveelheid tilapia in de vangsten is echter gelijk gebleven sinds O. niloticus voor het eerst in de vangsten verscheen. Uit vergelijkend onderzoek aan groei, agressie dieet en reproductieve inspanning is getracht de mechanismen af te leiden die O. niloticus een concurrentievoordeel op kunnen leveren. De gemiddelde maagvulling van O. niloticus was 70.9% ± 9.4 en die van O. mortimeri 73.4% ± 9.8, een niet significant verschil. Het dieet van beide soorten overlapte grotendeels en bestond hoofdzakelijk uit plankton en detritus. Omdat beide tilapias daarnaast een vergelijkbare verteringsefficiëntie bleken te hebben, is de conclusie dat met de introductie van O. niloticus in het Karibameer een grote kans op het voorkomen van voedselconcurrentie is ontstaan. Hierbij hebben de relatief hoge groeisnelheid en maximale grootte van de Nijltilapia deze exoot waarschijnlijk een concurrentievoordeel gegeven ten opzichte van zijn inheemse verwant. Een grote lichaamslengte levert een vis voordelen op in de vorm van 1) een hogere reproductieve inspanning, omdat per broedpoging het aantal eieren in het ovarium

210

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 210 Samenvatting (Dutch summary)

in het algemeen sterk positief gecorreleerd is met de grootte van het vrouwtje, en 2) dominantie bij conflicten. De mate van agressie van beide tilapiasoorten is getest in aquarium-experimenten. Hierbij bleek dat het grootste exemplaar van een paar in een testsituatie, onafhankelijk van de soort, het meest agressief was en meestal in de loop van het experiment dominant werd. Hoewel O. mortimeri in het algemeen als de agressievere soort kan worden beschouwd, wordt dit soortspecifieke verschil in agressie tenietgedaan door het feit dat voor elke leeftijdsklasse geldt dat O. niloticus groter is dan O. mortimeri. Het maandelijkse percentage vissen met een hoog niveau van gonadenontwikkeling was altijd het hoogst voor O. niloticus. Dit vormt een aan- wijzing voor een reproductief concurrentievoordeel voor de exoot, want een continu groter aantal rijpe vrouwtjes wijst op een relatief grote reproductieve inspanning ten opzichte van de inheemse tilapia. In tegenstelling tot de onbewuste introductie van de Nijltilapia, heeft het bewust uitzetten van de sardine Limnothrissa miodon geen negatieve invloed op het ecosysteem gehad. Het grote verschil tussen beide introducties is het feit dat de sardine een onbezette niche kon opvullen, waardoor interacties met andere vissoorten vrijwel niet voorkwamen. De beschikbaarheid van een onbezette niche zou een belangrijke standaardafweging moeten zijn bij het plannen van visintroducties, om een potentieel biodiversiteitsverlies te voorkomen. Het doel van de introductie van L. miodon in het Karibameer, het verhogen van de visproductie, is gehaald. Maar om de verkregen sociaaaleconomische voordelen te maximaliseren, dient de sardinevisserij te worden geoptimaliseerd via het vergaren van wetenschappelijke kennis over de biologie van en de visserij op de sardine, zoals gedaan in het in dit proefschrift beschreven onderzoek. In dit proefschrift zijn de relatieve groeisnelheid en de gemiddelde lengte bij geslachtsrijpheid van Limnothrissa miodon geschat en is besproken wat deze be- tekenen voor de status van de populatie en de visserij. De maximale gemiddelde lengte (L∞) van L. miodon, zoals geschat met behulp van het Gompertz model, bedroeg in 1993 – 1994, 1996 en 2012 – 2013 respectievelijk 18.0, 9.6 en 15.2 cm (TL). Deze waarden en de bijbehorende jaarlijkse variatie, zijn vergelijkbaar met die in het meer van oorsprong van de soort, het Tanganyikameer. Mijn resultaten wijken af van die van eerdere studies, waarin de conclusie werd getrokken dat de sardinepopulatie in het Karibameer dwerggroei vertoont. Het ontbreken van grote exemplaren in de vangsten van de sardinevissers kan echter worden verklaard door de grote hoeveelheid onvolgroeide vis die wordt weggevangen, in combinatie met de voorkeur van grote sardines voor de oeverzone, waar ze onbereikbaar zijn voor de in het open water opererende vissersvloot. De gemiddelde leeftijd bij geslachtsrijpheid in het Kariba- meer is lager dan die in het Tanganyikameer en is sinds de introductie steeds verder afgenomen. Ook de gemiddelde lengte bij geslachtsrijpheid is sterk afgenomen in de loop der jaren. Dit wordt geïllustreerd door een vergelijking van de waarden die in deze studie zijn gevonden (vrouwtjes: 3.43 cm; mannetjes: 3.63 cm) met gerapporteerde waarden voor de periode 1970 – 1972 (vrouwtjes: 5.2 – 5.6 cm; mannetjes: 7.1 – 7.3 cm).

211

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 211 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

De afname in de gemiddelde leeftijd bij geslachtsrijpheid kan worden toegeschreven aan overbevissing, waarbij een selectieve visserijdruk op snelgroeiende vis de relatieve overleving van exemplaren met een kortere generatietijd bevordert. Deze verandering in de ontwikkelingscyclus van de sardine vertraagt het herstel van de populatie. De introducties in het Karibameer leverden ons lessen op met betrekking tot het belang van het uitzetten van vis in een reservoir om de visproductie te verhogen. Het uitzetten van de sardine Limnothrissa miodon leerde ons hoe overbevissing en on- gunstige milieuomstandigheden de vangsten kunnen reduceren en vervolgens leiden tot een erosie van sociaaleconomische baten. Voor een duurzaam visserijbeheer dienen deze twee factoren te worden geïntegreerd in de productiviteitsschatting, het beheer en de monitoring van de visserij. De introductie van de Nijltilapia bood de mogelijk- heid tot het volgen van de effecten van de introductie van een exotische soort met dezelfde ecologische niche als een reeds aanwezige inheemse soort. Ik laat zien dat de introductie van een vissoort in een bezette niche niet tot hogere vangsten (baten) hoeft te leiden maar integendeel, kosten kan veroorzaken met betrekking tot een afname van de biodiversiteit ten gevolge van het verlies van inheemse en endemische soorten.

212

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 212

Curriculum Vitae

CuVi

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 213

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 214 Curriculum Vitae

According to my passport I am Chiyedza Portia Chifamba but somehow, I was registered at school as Portia Chifamba. Hence, my school certificates are by that name and I have used it thereafter. I was born on the 2nd of August 1963 in Zimbabwe. I am married to Jan Wanink, who is from the Netherlands, and we have three children: Marjo, Nyasha and Paula. I obtained a BSc General degree with a major in Biological Sciences at the University of Zimbabwe (UZ) in 1985. I am employed as a Senior Lecturer in the Department of Biological Sciences (DBS), of the UZ, in Harare. In 2006, I joined the DBS as a lecturer on a Fish Biologist post. I teach ichthyology, fisheries biology, tetrapod biology and environmental impact assessment to students enrolled in the BSc Honours in Biological Sciences degree programme. I also teach fish communities, fish stock assessment, aquaculture, conservation and management of biodiversity in the MSc in Tropical Hydrobiology and Fisheries. I have taught ecology and biodiversity in the MSc Integrated Water Resource Management offered in the Department of Civil Engineering. Through this degree programme, I was once invited to teach a module on Environ- mental Flows Legislation at the University of Malawi, Chancellor College. I have supervised eight MSc and at least seven undergraduate research projects and I currently have two undergraduate students. I was the Chair of DBS from 2009 to 2012. I was an external examiner at Bindura University, Zimbabwe. Completing my PhD means that I will finally have more time for this part of my teaching duties. Before going to Harare, I worked at the University Lake Kariba Research Station (ULKRS) as a Research Fellow from 1995 to 2006. My duties were to research, teach and supervise student researchers. I participated in the development and delivery of short courses on water resources, fisheries management, environmental impact assessment and integrated coastal zones management. I was Project Manager of the Water Group of the SANTREN (Southern African Training on the Environment Network) project run by the Institute of Environmental Studies (IES) of UZ. Within this project I developed and delivered a short training course on Integrated Coastal Zones Management in 2000. I also carried out a training need assessment in Zambia, Zimbabwe and Malawi. I participated in the Water Resources in the Lake Kariba Environs (DARMA) project, where I contributed a chapter in a book. I also participated in the Global International Water Assessment (GIWA). From 1986 to 1994, I worked at the Lake Kariba Fisheries Research Institute (LKFRI) of the Department of National Parks and Wildlife Management (Ministry CuVi of Environment and Tourism) as a Research Officer. My duties were to carry out research, extension and to assist in administration. While on this job, I obtained in 1991 an MSc in Fisheries Biology and Management from then the University of Wales, Bangor (now the University of Bangor) in the UK. I won the Jeremy Jones Memorial Prize for being the best student of my class. After that, I did a Diploma in

215

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 215 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

Management for Sustainable Fisheries at the University of Tromsø (Norway) in 1992. I taught on the US Pitzer College programme, that was based in Harare at UZ. In 2008, at the request of the National Parks and Wildlife Management Authority, I reviewed part of the curriculum of the Diploma offered by the Natural Resources College and assisted the students of this college in undertaking research on Lake Chivero. I have published research in the field of fisheries biology and water pollution. For the research I did on the displacement of the Kariba tilapia (Oreochromis mortimeri) by the Nile tilapia (Oreochromis niloticus) I won the Ibaraki Kasumigaura Prize for outstanding content of a paper on Lake Conservation by the International Lake Environment Commission (ILEC) in 2005. I have interest in integrated management and the environment in general. In line with this, I have coordinated the Lake Chivero Integrated Lake Basin Management program that was sponsored by (ILEC). I was a member of the Zimbabwe National Steering Committee for Integrated Water Resource Management and the Pollution Monitoring of the Zambezi River. I am a member of Wildlife and Environment Zimbabwe. Basically, I love to work with fish, how they are harvested, where they live and so on.

216

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 216

Acknowledgements

Ackn

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 217

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 218 Acknowledgements

The thesis celebrates the end of a long journey. It’s an accumulation of ideas originating from many people associated with collection and processing of samples in the field, analysis of samples in the laboratory, and the writing of manuscripts and thesis chapters. I see it as a magnificent quilt made up of these grand patches. The first definite steps on the journey must have been taken in 1994, when I registered for a PhD with the Aberdeen University in the UK. Then I was working at the Lake Kariba Fisheries Research Institute (LKFRI). Prof Christopher Magadza, then Director of the University Lake Kariba Research Station (ULKRS) and Prof George Turner, then at Aberdeen University, shaped my research project into a PhD project, for which I am most grateful. Thank you, George and Rosanna Robinson, for your hospitality when I came to Aberdeen. My research project was on growth and factors affecting exploitation of sardine is contained in Chapters 6 and 7. Unfortunately, my funding for the PhD stopped and I was also stopped from working on the research project, having moved from LKFRI to work at ULKRS. Later, with assistance from Chris and Prof Ngoni Moyo, I came up with a new research project and in 2002 registered for a PhD at the University of Zimbabwe with the two as my supervisors. I am grateful for their help and encouragement. This time my project was on the impact of the Nile tilapia introduced into Lake Kariba. Sadly, for me, Ngoni left for the University of Limpopo in South Africa and a new supervisor had to be found. Prof Brian Marshall took me under his wings and set about redirecting the research to be focused sharply on the biology. Prof Marshall could not tolerate any socio blah blah, insisting that sociology has no place in science. He gave me encouraging remarks on the aggression experiment in Chapter 5. Too soon, an opportunity arose for Brian to work in Uganda on an EU project on Lake Victoria. After he left, there was no suitable supervisor for me. To make matters worse, the adverse socio-economic situation forced me to stop my fieldwork. Fuel, needed for sampling, became scarce around 2002 and was virtually unavailable thereafter. The situation worsened and was at its worst in 2008, when basic survival was difficult and most of my time was spend in queues for whatever became available. Despite these setbacks I still wanted to pursue my PhD and explored other avenues. Jan Wanink and the late Dr Frans Witte (Leiden University) assisted in arranging my study at the University of Groningen under the supervision of Prof John Videler. Thank you, Frans, for inviting me twice to spend my contact leave at Leiden University, for the help in setting up my experiments in the basement and letting me use your fish. Your hospitality and generosity knew no bounds. Thank Ackn you and Els for opening your house to my family during my stay in Leiden. I fondly remember the evening you brought toys down from the attic for Paula to play with

219

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 219 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

and everyone spent the whole evening enjoying them. What fun we had. You will always be missed. I am indebted to Prof John Videler for accepting to supervise my PhD. Thank you, John and Hanneke, for the warm welcome to your home for fruitful discussions, manuscript writing and wonderful meals. I much enjoyed sampling some of Hanneke’s wild vegetables. John placed enormous emphasis on precision and, thus, had to be very patient with me. Sadly, Prof Videler retired before I completed my PhD, but after we finished the manuscript for the paper that is now Chapter 3. My journey would have ended without the help of Dr Corine Eising, who assisted in the process of getting and maintaining my Ph D supervision on track. Thank you, Corine, for perseverance when arranging papers for my MVV and visas, and for keeping an eye on me. Ingeborg Jansen, thank you for resolving my many problems, in particular that library access. Thank you also for finding me the nicest work space, ensuring I had access to beverages and for offering me your choicest tea. Under the expert supervision and mentorship of Prof Han Olff and Prof Britas Klemens Eriksson, my PhD thesis was remodelled and guided to completion. Han and Klemens, I appreciated your taking me around to meet other students and see fascinating projects being carried out. And thank you for inviting me to the group meetings. I learnt a lot there. With your help, I came to personally experience that every project can be interpreted in many ways; ecological, biological, mathematical, economic, social, engineering and so on. Thank you for your time and dedication to ensuring that the thesis chapters reached an acceptable standard and were finalized. Linnaeusborg is indeed a pleasant working environment and thanks to everyone who is ever so considerate, I had very productive months in The Netherlands. I much appreciated the talks, discussion group meetings and other informal interaction with members of the Groningen Institute of Evolutionary Life Sciences (GELIFES). These have enriched my experience and provided me with ideas I will forever cherish and will try to use to the fullest. The simple questions I was asked by many of you and my attempt to explain, was what I needed to clarify or expand my own ideas. Special thanks to Dr Ruth Howison for assisting me to settle in the first time I came to GELIFES. For the writeup of my thesis I relied on many people beside my supervisors. I acknowledge the inputs from Jan, Prof Tony Pitcher, Chris (to eliminate ‘shonglish’ among other errors), my sister Audrey (English language expert and general sounding board) and others I have not named. Jan spent many sleepless nights giving the thesis a final check and typesetting the book for printing. My brother Mr Mukundi Chifamba resolved all my computer hard- and software issues promptly. Without Mukundi’s support, my work would have been difficult if not impossible at times. The field and laboratory work would have been impossible without the assistance of the late Joel Chisaka, Margaret Gariromo, Zenzo Khali, the late Joseph Chironga, Siagotami Siyasayi, Muleya of ULKRS and others not mentioned by name. With a

220

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 220 Acknowledgements

heavy heart I remember 16 November 2003, when the Volvox capsized during one of my field trips. Thanks to Sanyati Lodge personnel for searching and rescuing my colleagues who had managed to sit on top of the capsized boat. Other staff at ULKRS in administration, especially Elmon Dhlomo, Gero Ebbefeld, Tendo and the workshop team. Elizabeth Munyoro, Godwin Mupandawana and the laboratory team made my research possible in many ways. I depended on Joel for catching fish and keeping them alive. In addition, he could fix equipment that others thought hopeless, without complaining. He was a huge pleasure to work with. His passing left a huge gap. I thank my fellow researchers at ULKRS, the time I was there, who include Moses Chimbari (then Director of ULKRS), Jackie Mangoma, Lindah Mhlanga, Alexio Mbereko, Rudo Sanyanga, Idzai, Crispen Phiri and Nqobizitha Siziba, for valuable comments on my research, especially during our annual planning meetings, and for making working in Kariba fun. Crispen, thanks for noticing that my alcohol did not smell right preventing further loss of my tilapia stomach samples. Technical staff in the Department of Biological Sciences, in particular Gerard Ashley, assisted me to do experiments on aggression and some other laboratory analysis. I also thank my fellow lecturers in the Department of Biological Sciences of the University of Zimbabwe for attending my presentations and making valuable comments on my talks. I wish to also express my appreciation for letting me have the time to work on my PhD. Thanks for covering up for me, especially Lindah for teaching the Vertebrate Biology and Chris the Environmental Impact Assessment courses, when I had to be away to work on my PhD. Also, I thank everyone in administration, in particular Prof Maude Muchuweti, the Dean of Science, Prof Tamuka Nhiwatiwa and Miriro Tarusikirwa, for organizing my leave and travelling. Data from the fishing villages was collected by Mr Manzungu and other three people who lived in the selected villages. Thank you for your dedication and for collecting good quality data. Mr Mufolo and Gamatox assisted in collecting the data and fish samples from Nyaodza fishing village that were used for the digestion experiment. Some of my research I did at Lake Kariba Fisheries Research Institute (LKFRI), where I worked from 1986 to 1994 but afterwards continued to be treated not as a visitor but a colleague. Morris Mutsambiwa, Newman Songore, Itayi Tendaupenyu, Wilson Mhlanga, Nobuhle Dhlovu and Ms Moyo assisted in many ways. Thank you, guys. The late Mr Mushaike was an excellent coxswain and the best at capturing and processing fish. Thank you for allowing him to work on my projects. Some of the data used was kindly provided by a number of organizations. The Zambezi River Authority (ZRA) provided the data for Lake Kariba inflow and lake level, air temperature and rainfall used in this paper. I am grateful for fisheries data Ackn provided by LKFRI. I thank both organizations profusely. Without the data the analysis in Chapter 2 and 6 would not have been possible.

221

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 221 BIOLOGY AND IMPACTS OF OREOCHROMIS NILOTICUS AND LIMNOTHRISSA MIODON

Mr Ronald Bijkerk kindly provided tables of plankton biovolumes used in estimating biovolume. Thank you, Ronald, for also sharing your impressive knowledge on plankton, which helped me to have a better feel for my data. For plankton analysis I was also assisted by Edwin Tambara whose hard work is exemplary. I still marvel at the accuracy of the sketches of plankton he made. Approval to carry out experiments on fish was granted by the Animal Research Ethics Committee in the Department of Livestock and Veterinary Services, Ministry of Agriculture, Mechanization and Irrigation Development, for which I am grateful. I thank Dr Makaya and Dr Swiswa for helping me with the procedure. Field work was done using fishing vessels belonging to LKFRI, ULKRS, Mash Fishing Enterprise (Mapfumo family), Municipality of Kariba People’s Project and Mr Paul Mwera. Mapfumo family provided a boat with fuel and crew for 10 days of sardine fishing in 2013. Mr Nesbert Mapfumo understands the role of research in the management of a fishery and is always willing to offer a hand. The fishing crew of the commercial sardine fishing boat, whose help I am most grateful for, tolerated the nonsense of fishing without light and catching few fish. Peoples Project provided the boat used to sample sardines in 2002. Mr Clive Radcliffe from the University of Humberside operated a mid-water trawl used to catch samples that I used for validation of daily increments on sardine otoliths. Erlend Moksness taught me how to use the computer otolith reading programme they had developed, when he hosted me at Flødevigen Research Station in Norway. I also thank him and his family for the lovely dinner at their home during my stay. The analysis of otoliths for validation was done at the Electron Microscope Unit of the University of Zimbabwe. Thank you, Mr Claudius Mutariswa and Mr Patrick Kurangwa for assisting in the preparation, reading and photographing of the otoliths. I am very grateful for the permission to use the otolith and scale reading equipment at LKFRI. Mr Paul Daley of Zambezi Proteins provided the 1994 sample of large fish used in the study. I appreciate the sponsors of my research projects and my PhD. The International Foundation for Science (IFS) Grant A/3159-1 was used to purchase gillnets and other equipment and for field work. The Tonolli Memorial Fund Fellowship of the International Society of Limnology (SIL) was used in the 2001/2 field work on sardine. Financial support from Nuffic - NFP-PhD.11/ 858 for my PhD studies enabled the data collection in 2012 and 2013 and my stay in The Netherlands. I am grateful to Mr Wiebe Zijlstra for helping with all matters pertaining to the Nuffic grant. The University of Zimbabwe (UZ) financed some of the research work. Electrofishing samples were obtained during a scientific expedition under the Pro- gramme for Institutional University Cooperation of the Vlaamse Interuniversitaire Raad (VLIR-UOS). The VLIR-UOS programme also paid a month salary of a Research Assistant.

222

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 222 Acknowledgements

Thank you, Jan and Kees Goudswaard, my paranymphs, for persistently checking on my progress and encouragement throughout, and for assisting me at my defence. You are right Kees, it’s a relief to get it done. My family, immediate and extended, were very supportive. My cousin Petros kept my car, that I used to travel to Kariba for fieldwork, on the road. My daughters Paula and Nyasha as well as my niece Tadiwa, accompanied me sometimes when I went to Kariba for fieldwork, assisting me where they could. Paula entered some of the diet analysis data whilst Nyasha entered the catch data from the fishing villages and Fadzi assisted with laboratory work. Thank you Nyasha, Fadzi, Chiwoniso, Paula, Tadiwa, Akudzwe and Aiden for tolerating my long periods of absence. Mukundi assisted by taking care of the children in my absence for doing fieldwork and when I was in The Netherlands. Jan you will be rewarded in the usual way for your love and support throughout. Thank you for tolerating my long periods of absence. That should be addressed. Our next major teamwork should be that I learn Dutch. It is awful that I cannot read my own Dutch summaries. That will not do. Those who know me well understand that my memory of names is very poor indeed. That is a lame excuse, I know. I apologise profusely to those I failed to name but have contributed to the process. I am indebted to you all, for without your help the thesis and PhD would not have been attained. Lastly, I thank my loving parents who gave me a solid foundation and instilled in me the importance of perseverance. They would have been most proud of my achievement.

Ackn

223

528794-L-bw-Chifamba Processed on: 6-2-2019 PDF page: 223