Invasion–food security trade–offs in African inland fisheries: multi–year evidence from native and non–native fish assemblages in a subtropical reservoir

Small–scale inland fisheries are critical for food and livelihood security across Africa, yet many freshwater systems now contain native and non–native fish species whose relative contributions to fisheries and the trade-offs for biodiversity and household economies remain poorly quantified. This study assessed the contribution of native and non–native fishes in a subtropical reservoir using a 6–year dataset (2020–2025), complemented by a market and preference survey of 52 fishers and traders. Fish species were classified as native or non–native, and the data were further analysed using multivariate methods. The index of relative importance (IRI) integrated abundance, weight and frequency of occurrence to determine species dominance. Temporal coupling between native and non-native CPUE was tested using Spearman rank correlation to evaluate whether non-native catches compensate for declines in native catches. Twelve species were recorded: 8 native and 4 non–natives (Oreochromis niloticus, Micropterus nigricans, Coptodon rendalli and a crustacean Cherax quadricarinatus). Native species contributed a greater number of individuals (45–207 vs 33 654), but non-native species contributed disproportionately to more harvested biomass and market value. Native species exhibited significantly higher overall CPUE than non–native species, with pronounced differences in the hot–wet and cool–dry seasons. However, the IRI ranked M. nigricans (20.1%) and Coptodon rendalli (20.0%) as the most important species. Non–native species dominated 4–5–inch mesh catches while native species peaked in 3–inch mesh. Market survey results showed that non-native species commanded approximately twice the aggregate market value of native species (USD 45–552 vs USD 22–537 over the study period), with C. rendalli, M. nigricans and O. niloticus preferred by fishers and traders for their larger body size, higher unit price and stronger consumer demand. However, the higher numerical catch of native fish indicates that biodiversity costs of non-native dominance are real and that food-security benefits are mediated primarily through biomass and price, not abundance. Despite comparable annual yields, non–natives functioned as seasonal food–security buffers. The establishment of O. niloticus (16.0% of biomass) raises conservation concerns regarding native cichlid protection and conservation. These findings demonstrate that non–native species generate measurable but uneven food-security benefits while imposing biodiversity costs, necessitating adaptive management strategies that explicitly balance both dimensions.

Human–mediated species translocations are reshaping freshwater ecosystems globally, with fishes among the most frequently moved vertebrates because of their perceived value for aquaculture, recreational angling, biological control, ornamental use and food production (Ellender et al., 2014; Weyl et al., 2015; IPBES, 2023; Chaturvedi et al., 2025), and these introductions are rarely ecologically neutral. In many inland waters, introduced fishes have become embedded within local economies and diets, while simultaneously altering native assemblages, trophic interactions, genetic integrity and ecosystem functioning (Ellender and Weyl, 2014; Lynch et al., 2016; Zengeya et al., 2017; Tickner et al., 2020; IPBES, 2023). This creates a persistent management paradox, where species introduced to enhance production and livelihoods may also undermine the biodiversity, ecological resilience and long–term sustainability upon which inland fisheries depend.

Across Africa, small–scale inland fisheries are central to food and nutrition security, household income, informal trade and local economic resilience (Fluet–Chouinard et al., 2018; Funge–Smith and Bennett, 2019; FAO, 2022). Reservoirs, lakes and river systems supply affordable animal protein and micronutrients to communities that often have limited access to formal employment, livestock production and/or diversified food systems. Zimbabwe’s inland fisheries are no exception, with waterbodies such as Lake Sebakwe supporting domestic consumption, local marketing and livelihood activities, while also functioning as novel ecological systems shaped by dam construction, stocking histories, fishing pressure and environmental variability (Marshall, 2010; Utete and Tsotsonga, 2013; Kupaza et al., 2015).

Over several decades, intentional and accidental fish introductions have produced assemblages in which native and non–native species jointly sustain fishery yields (Ellender and Weyl, 2014; Weyl et al., 2017; Zengeya et al., 2017). This co–production of food by native and non–native taxa is especially important in artificial or heavily modified systems, where introduced species may become numerically dominant, commercially valuable or seasonally critical to household food supply. However, the benefits of non–native fishes are unevenly distributed and must be weighed against ecological risks, including predation, competition, hybridisation, disease transmission and biotic homogenisation (Dudgeon et al., 2006; Cucherousset and Olden, 2011; Weyl et al., 2020; IPBES, 2023). Thus, assessing non–native fish contributions requires more than documenting presence or abundance; it also requires quantifying their temporal, biomass, and size–structured role within functioning fisheries.

Understanding the socio-economic dimensions of fish invasions is equally critical as ecological assessment, as fisher preferences and market dynamics often determine the perceived value and management acceptability of non-native species (de Carvalho-Souza et al., 2024; Jarić et al., 2025). Market surveys from invaded systems have shown that introduced species can achieve high consumer acceptance when they meet taste preferences, affordability criteria, and availability expectations and overcome neophobia (Chakandinakira et al., 2023). In Zimbabwean reservoirs, the integration of non-native species into local food systems has been documented, with species such as O. niloticus and M. nigricans becoming staple components of household diets and local trade networks (Utete and Mabika, 2025). However, the extent to which non-native species are preferred over native taxa, and whether their market value reflects their ecological contributions, remains poorly quantified for most Southern African reservoirs (Weyl et al., 2021). Such information is essential for developing management strategies that acknowledge both the ecological risks and socio-economic benefits associated with non-native fishes (Weyl et al., 2021).

One of the most widely introduced fish species of aquaculture importance in Southern African reservoirs is the Nile tilapia, Oreochromis niloticus (Linnaeus, 1758), with other species such as common carp Cyprinus carpio Linnaeus, 1758, trout (Oncorhynchus mykiss (Walbaum, 1792), Salmo trutta Linnaeus, 1758) and bass species, Micropterus species being used for fisheries enhancements (Ellender and Weyl, 2014; Utete and Mabika, 2025). Oreochromis niloticus has been widely disseminated for aquaculture and capture–fishery enhancement because of its rapid growth, environmental tolerance, high reproductive capacity and strong market acceptance (Canonico et al., 2005; Lorenzen et al., 2012; Mbiru et al., 2021). Micropterus nigricans (Lacépède, 1802) was mostly introduced primarily for sport fisheries, and has established in many Southern African waterbodies and can increasingly occur in artisanal and subsistence catches (Ellender and Weyl, 2014; Weyl et al., 2020). In Lake Sebakwe, earlier survey evidence recorded M. nigricans as a dominant component of the fish assemblage, indicating that introduced predators may already be structurally important in this reservoir fishery (Utete and Tsotsonga, 2013).

The ecological risks associated with these species are substantial, as biological invasions impose major global costs on biodiversity, ecosystem services and human well–being, with recent assessments estimating annual impacts in the hundreds of billions of United States dollars (Cuthbert et al., 2021; IPBES, 2023; Bacher et al., 2023). In Southern Africa, O. niloticus poses a serious conservation concern for native congeneric tilapias because habitat and trophic overlap can facilitate competition, displacement and hybridisation (Zengeya et al., 2015). Hybridisation and introgression between introduced O. niloticus and native Oreochromis spp. have been documented in the Limpopo and Kafue River systems and are implicated in the decline or replacement of native tilapias such as Oreochromis mossambicus (Peters, 1852), Oreochromis mortimeri (Trewavas, 1966) and Oreochromis andersonii (Castelnau, 1861) (Firmat et al., 2013; Deines et al., 2014; Zengeya et al., 2015; Munyai et al., 2023). Similarly, Micropterus spp. can exert strong predatory pressure on native fishes, particularly in simplified lentic systems where prey refugia are limited and native species have not evolved with comparable centrarchid predators (Woodford et al., 2016; Khosa et al., 2020).

Despite these risks, eradication and strict suppression of established non–native fishes may be socially unrealistic where local households depend on them for food, income or seasonal livelihood buffering. This is particularly true in reservoirs where introduced species may now constitute a large proportion of fishable biomass. The key policy challenge is therefore not simply whether non–native species are harmful or beneficial, but under what ecological, seasonal and socio–economic conditions they generate net food–security benefits while imposing biodiversity costs. Addressing this question requires long–term, fishery–dependent datasets capable of capturing changes in catch composition, biomass, size structure and seasonal production. Such monitoring can reveal whether non–native fishes are supplementary, dominant or indispensable components of the fishery, and whether their contribution varies through time, across size classes or in relation to native species availability (Welcomme et al., 2010; Lorenzen et al., 2016; Funge–Smith and Bennett, 2019).

The policy challenge is further complicated by market dynamics and consumer preferences, which may not align with biodiversity conservation objectives. Where non-native species are preferred by fishers and consumers, management interventions to reduce their abundance may face significant social resistance and economic consequences (Gozlan, 2008; Schlaepfer et al., 2011). Understanding the relative market value of native versus non-native species, fisher preferences, and the reasons underlying these preferences is therefore essential for designing socially acceptable and ecologically effective management strategies. This study addresses this knowledge gap by integrating ecological monitoring with socio-economic surveys to assess both the market acceptance and ecological roles of native and non-native fishes in Lake Sebakwe.

The study uses six years of fishery–dependent monitoring data from Lake Sebakwe, Zimbabwe, to evaluate the contribution of native and non–native fishes to inland fishery production and local food supply. Specifically, we examine temporal trends in catch composition, size structure and biomass contribution of non–native fishes relative to native species. We further assess whether non–native species make disproportionate contributions during particular seasons or within specific size classes, and what this implies for balancing biodiversity conservation with household food and livelihood security. We hypothesise that non–native fishes contribute disproportionately to harvested biomass and market value (though not necessarily numerical abundance) in Lake Sebakwe, particularly during specific seasons and size classes, and that their catch rates are temporally decoupled from or negatively associated with native catch rates, indicating a compensatory food-security function, thereby producing a measurable food–security benefit while intensifying conservation trade–offs for native fish assemblages.

Study area

Lake Sebakwe, a multipurpose reservoir constructed on the Sebakwe River in 1957 and subsequently raised by 7 m in 1986, is located in Kwekwe, Midlands Province, Zimbabwe, and supports both artisanal and semi–formal gillnet fisheries (Figure 1) (Utete and Tsotsonga, 2013). The reservoir covers an area of 2 320ha and has a maximum volume of 105–000 m 3 (Utete and Tsotsonga, 2013). Sebakwe region receives 650–700 mm of rainfall annually and temperature ranges between 25 °C and 28 °C (Mugaviri et al., 2023). The catch includes non–native fishes (i.e., O. niloticus, M. nigricans, redbreast tilapia Coptodon rendalli (Boulenger, 1897) and a suite of native species such as Clarias gariepinus (Burchell, 1822), redeye labeo Labeo cylindricus Peters, 1852, yellow–belly bream Serranochromis robustus Gunther, 1864, bulldog fish Marcusenius macrolepidotus Peters, 1852, eastern bottlenose Mormyrus longirostris Peters, 1852, dwarf bream Pharyngochromis acuticeps Steindachner, 1866, Zambezi happy Pharyngochromis darlingi Boulenger, 1911 and greenhead tilapia Oreochromis macrochir Boulenger, 1912 (Utete and Tsotsonga, 2013). A non–native decapod, the Australian redclaw crayfish (Cherax quadricarinatus Von Martens, 1868), is also established in the reservoir as a result of intentional spread from the invaded Lake Kariba (Chakandinakira et al., 2023; Utete and Mabika, 2025).

Data source and structure

Fishery-dependent gillnet records from routine monitoring of the Lake Sebakwe small–scale fishery were analysed for the period January 2020 to December 2025. These monitoring efforts were conducted in collaboration with Zimbabwe Parks and Wildlife Management Authority (Zimparks) officials and fishers across multiple operations. For each monthly fishing event across five fisheries cooperatives (A-E), the dataset recorded species identity, mesh size (3–7–inches), effort, total number caught, total weight, CPUE by mass (kg/100 m), CPUE by number (n /100m), and mean weight. Gillnets were set in the mid–afternoon (approximately 14:00) and retrieved the following morning (approximately 06:00), allowing a soak time of approximately 16 hours. Upon collection, the fish were sorted according to mesh, identified using guides by Skelton (2001) and Marshall (2010) and weighed (to the nearest gram) using a scale.

To assess economic value and consumer acceptance of native versus non-native fishes, asemi-structured questionnaire was administered to 52 respondents drawn from gillnet fishersoperating on Lake Sebakwe and fish traders based in the Sebakwe area (Appendix 1). The questionnaire captured (i) species preference, (ii) reasons underlying the stated preference (e.g. price, taste, size, market demand, and ease of capture), (iii) prevailing farm-gate price per kilogram for each species, and (iv) respondent demographics (gender, age, fishing experience, role in the value chain). Stated species-level prices were applied to the six-year total catch weights to estimate aggregate market value by species and by origin group (native vs non-native). Preference frequencies were summarised as percentages of respondents.

Data analyses

All data were collated and initially processed in Microsoft Excel before statistical analyses were conducted in R version 4.5.2 (R Core Team, 2025). To account for seasonal variability in fisheries production and environmental conditions, sampling records were grouped into three biologically and climatically relevant seasons: hot–wet (November–April), cool–dry (May–August), and hot–dry (September–October). Fish species were further classified as either native or non-native according to their known biogeographical origin within the Lake Sebakwe system. Prior to analysis, data distributions were assessed using the Shapiro–Wilk test, while homogeneity of variances was evaluated using Levene’s test. Fish biomass and abundance data deviated significantly from normality (Shapiro–Wilk, p < 0.05) and exhibited heteroscedasticity among groups (Levene’s test, F = 11.861, p < 0.05). Thus, non-parametric statistical approaches were employed where assumptions of parametric analyses were not met (Zar, 2010).

To evaluate the effects of species origin (i.e., native vs non-native), year, season, and fishery type on catch biomass and abundance, Generalised Linear Models (GLMs) were fitted using a Gamma error distribution with a logarithmic link function, an approach widely recommended for continuous, strictly positive ecological response variables characterised by right-skewed distributions (Zuur et al., 2009; Bolker et al., 2009). In the GLM, the reference categories were set as native species, the year 2020, the cool-dry season and fishery A. Competing models were compared using Akaike’s Information Criterion (AIC), with the model exhibiting the lowest AIC value considered the most parsimonious representation of the data (Guthery et al., 2003b). Differences in biomass, abundance, and catch-per-unit-effort (CPUE) between native and non-native fish groups were assessed using Mann–Whitney U tests. Relative species dominance within the fishery was quantified as the proportional contribution of each species to total catch biomass and abundance. To provide a more integrative assessment of species importance, the Index of Relative Importance (IRI) was calculated following Pinkas et al. (1971). The IRI combines numerical abundance, biomass contribution, and frequency of occurrence, thereby providing a more comprehensive measure of species prominence within fisheries than any single metric alone (Pinkas et al., 1971; Sanyanga, 1996). The index was calculated as:

Where, % W i and % N i are percentage weight and percentage number of each species relative to the total catch and % F O i is percentage of frequency of occurrence of each species.

Temporal dynamics of native and non-native fish biomass between years 2020 and 2025 were examined using non-parametric time-series approaches. Long-term monotonic trends were evaluated using the Mann–Kendall test, while seasonal patterns were assessed using the Seasonal Mann–Kendall test. Trend magnitude and direction were estimated using Sen’s slope estimator, which provides a robust measure of change that is less sensitive to outliers and non-normal data distributions (Kendall, 1975; Hirsch et al., 1982). To assess invasion dynamics through time, an invasion progression index was calculated as the monthly biomass ratio of non-native to native fishes, with increasing values indicating a growing contribution of non-native biomass relative to native biomass. Seasonal differences in fishery productivity were evaluated by comparing CPUE biomass (kg 100 m -1 net) among seasons separately for native and non-native fish groups using Kruskal–Wallis tests. Where significant differences were detected, pairwise comparisons were conducted using Dunn’s post-hoc tests with Bonferroni-adjusted p-values to control for Type I error associated with multiple comparisons (Dunn, 1964). Statistical significance was assessed at p < 0.05.

To directly test whether non-native species compensate for declines in native species, monthly CPUE (kg 100 m - ¹) series for native and non-native groups were aligned and tested for temporal association using Spearman rank correlation. A significant negative correlation would indicate compensation (non-native catches rise when native catches fall), a positive correlation would indicate co-variation driven by shared environmental drivers, and a non-significant correlation would indicate independent dynamics.

Demographic characteristics

Gender composition

Out of the fifty 52 respondents, 79% (n = 41) were male, while 21% (n = 11) were female (Figure 2).

Age distribution

The age distribution indicates that most respondents (40%, n = 21) were between 21 and 30 years old, followed by those aged 31–40 years (38%, n = 20), 41–50 years (14%, n = 7), 51–60 years (6%, n = 3), and over 61 years (2%, n = 1) (Figure 3).

Role in the value chain

The study sample consisted of 41 fishers (79%), 9 fish vendors (17%), and 2 respondents (4%) who performed both roles as fishers and vendors (Figure 4).

Years of experience in fishery sector

The experience distribution indicates that the majority of respondents (71%, n = 37) have less than 5 years of experience in the sector, followed by 23% (n = 12) with 5–10 years of experience and 6% (n = 3) with 11–20 years of experience (Figure 5).

Relative abundance and weight of fishery species

The GLM analysis results showed that fish biomass significantly depended on species classification, year, and season and fishery type. Non–native species had significantly high biomass than native species (β = 0.373, p < 0.0001). Significant (p < 0.0001) increases in biomass were noted for years 2021, 2022, 2023 and 2024, while 2025 year did not significantly (p = 0.363) differ from the reference year 2020. The hot–dry season had significantly high biomass than the reference season (cool-dry) (β = 0.240, p = 0.028), implying that it has about 27% more expected biomass than the reference season. There were no significant differences (p = 0.228) between the hot–wet season and the reference season in terms of biomass. Fishery type significantly affected the biomass, where Fisheries B, C, D, and E had significantly lower biomass than the reference fishery A. The most pronounced differences were seen in Fishery D (β = −1.608, p < 0.0001) and Fishery C (β = −1.405, p < 0.0001), which had about 80% and 75% lower expected biomass, respectively. In general, the results show that non–native species provided higher fish biomass during the period under analysis (Table 1).

In Lake Sebakwe, 11 species were sampled, with M. macrolepidotus and C. rendalli being the dominant fishery species over the sampling period in terms of abundance (21–168 individuals) and biomass (5791.2 kg), respectively (Figure 6; Table 2). The mean total catch weights per year for C. rendalli were 965.2 kg year –1, with M. nigricans, C. gariepinus and O. niloticus having 886.7 kg year –1, 751.6 kg year –1 and 678.8 kg year –1, respectively. Pharyngochromis darlingi had low total weights of 0.11 kg year –1 (Figure 6). The mean total catch weights per year of native and non–native fish species was 1 627.9 and 2 577.2 kg year –1, respectively. Similarly, the native species had a high total abundance (45–207 individuals) compared to non–native species (33–654 individuals). Mann–Whitney U test revealed statistically significant differences (p = 0.006) in fish biomass of native and non–native species, with no significant differences in CPUE (p < 0.001). No significant differences (p = 0.602) in abundances were observed for native and non-native species.

The seasonal total catches for fish species showed marked seasonal and interannual variability from 2020–2025 (Figures 7a–k). The dominant fishery species such as C. rendalli (Figure 7b), C. gariepinus (Figure 7a), O. niloticus (Figure 7h) and M. nigricans (Figure 7f) contributed high biomass, with marked peaks in years 2021, 2023 and 2024. For example, M. nigricans recorded high catches in the first half of 2021, while C. rendalli and O. niloticus peaked in late 2023. In 2024, P. acuticeps increased mid–year and moderately abundant species such as M. macrolepidotus showed peaks in abundance for 2022 and early 2023 (Figure 7e), whereas L. cylindricus displayed relatively low but seasonally distinct increases, particularly in 2023 (Figure 7c). In contrast, low–abundance species S. robustus (Figures 7k, m. longirostris (Figure 7d), O. macrochir (Figure 7g) and P. darlingi (Figure 7j) had low seasonal contributions.

Fish preferences and market value

Of the 52 respondents surveyed, 39% (n=20) preferred non-native fish species, 8% (n=4) preferred native fish species, 2% (n=1) preferred both native and non-native, and 52% (n=27) expressed no clear preference (Figure 8). The most frequently cited reasons for preferring non-native species were their larger size (weight), better taste, and stronger market demand (Figure 9). Similarly, respondents who preferred native species primarily cited taste, weight, and marketability by numbers as the main reasons for their preference.

Market price analysis revealed that all species except C. gariepinus were sold at $3.00/kg (Table 3). The total market value of non-native fishes ($45,552.30) substantially exceeded that of native species ($22,537.05), reflecting both higher biomass and equivalent or superior pricing. The total highest individual species values were recorded for C. rendalli ($17,373.60), M. nigricans ($15,960.90), and O. niloticus ($12,217.80), all non-native taxa.

Temporal trends in native and non–native biomass to identify invasion progression

Using the Mann–Kendall trend test, we observed that the biomass of native species had a negative and non-significant (τ = –0.022, p = 0.789) trend, while the biomass of non–native species also showed a non–significant decline (τ = –0.076, p = 0.346). Similarly, Mann–Kendall trend test assessing seasonal variation showed no significant differences for biomass of native (τ = –0.033, p = 0.745) and non–native (τ = –0.167, p = 0.104) species. The Sen’s slope estimator showed that the extent of biomass change through time was relatively small for both native (–0.17 kg/month) and non–native species (–0.64 kg/month). The invasion progression analysis using the ratio of non–native to native biomass showed no significant increase in the dominance of non–native species (τ = 0.091, p = 0.261) (Figure 10).

Important fish species composition

The Index of Relative Importance (IRI) identified a small group of species that collectively dominated the Lake Sebakwe fishery. Micropterus nigricans (20.1%), C. rendalli (20.0%), and M. macrolepidotus (19.1%) emerged as the most important species, accounting for approximately 60% of total fishery importance. A second tier of important species comprised C. gariepinus (14.0%), P. acuticeps (12.5%), and O. niloticus (11.1%), indicating that both native and non-native taxa contributed substantially to fishery production. In contrast, L. cylindricus, C. quadricarinatus, M. longirostris, S. robustus, O. macrochir, and P. darlingi were of minor importance contributing less than 3% to the overall IRI (Table 2).

Catch per unit of effort

Seasonal variation in catch rates was found to be significant (χ² = 8.797, df = 2, p = 0.012) in both native and non–native fish species CPUE. The Dunn post–hoc test identified significant (p = 0.009) differences between the hot–dry and hot–wet seasons, but no significant differences for the cool–dry vs hot–dry (p = 0.108) and hot–wet (p = 0.906) seasons. Non–native fish species showed significant seasonal differences (χ² = 49.568, df = 2, p < 0.0001), with the Dunn’s post–hoc test showing significant differences for the cool–dry vs hot–dry (p < 0.0001) and hot–wet (p < 0.0001) seasons, but no significant differences for the cool–dry vs hot–wet seasons (p = 0.486). Non–native species had high CPUE compared to native species among seasons. Non–native species reached peaks of 0.45 ± 0.46 kg/100 m during the hot–dry season and native species showed less variation in CPUE (0.23–0.27 kg/100 m) (Figure 11).

Coupling between native and non-native CPUE

Monthly CPUE of native and non-native fishes were positively correlated across the full 2020–2025 series (Spearman ρ = 0.267, p = 0.0236) (Figure 12). Seasonal CPUE of native and non-native fishes were positively correlated across the full 2020–2025 series (Spearman ρ = 0.507, p = 0.0337) (Figure 13). This pattern indicates that non-native catches do not systematically compensate for declines in native catches.

Mesh selectivity and size–quality implications

Gillnet mesh size had a strong influence on catch composition and biomass (Figure 14). The 3-inch mesh yielded high overall catches and was dominated by native species, whereas large meshes (4–5 inches) disproportionately captured non-native taxa. Oreochromis niloticus was the most abundant in the 5-inch mesh, while M. nigricans occurred across multiple mesh sizes but peaked in the 3-inch gear. Native fish biomass declined with increasing mesh size, whereas non-native species contributed a greater proportion of catches in larger meshes. These differences were statistically significant (Kruskal–Wallis: H = 38.13, df = 9, p < 0.001), indicating strong mesh-size selectivity in the fishery.

The six–year assessment of Lake Sebakwe reveals a fishery that is neither conventionally native nor fully transformed by biological invasions, but rather represents a functionally integrated socio–ecological system in which native and non–native fishes jointly sustain ecosystem services and local livelihoods. Such systems increasingly characterise reservoirs throughout Southern Africa, where historical introductions, environmental modification, and fishing pressure have created novel fish assemblages that differ substantially from their pre–impoundment states (Ellender and Weyl, 2014; Weyl et al., 2017). Rather than demonstrating progressive displacement of native species, the Sebakwe fishery appears to have reached a relatively stable post–establishment state in which non–native species contribute substantially to fishery production while native species continue to maintain a significant numerical presence. This stability is reinforced by socio-economic integration, with market surveys revealing that non-native species command equivalent or superior prices to native taxa and are preferred by 39% of respondents. The total economic value of non-native fishes ($45,552.30) exceeded that of natives ($22,537.05) by more than twofold, indicating that the invasion has created substantial economic benefits alongside ecological risks. This finding supports growing evidence that the ecological outcomes of invasions are often more nuanced than simple narratives of native decline and invasive dominance suggest (Schlaepfer et al., 2011; Davis et al., 2011).

A key finding of this study is the disproportionate contribution of non–native fishes to fishery biomass despite similar numerical catch rates between native and non–native groups. The dominance of biomass by C. rendalli, M. nigricans and O. niloticus reflects differences in growth trajectories, body size distributions, and life–history strategies rather than simple abundance effects. Similar patterns have been documented in reservoirs throughout Southern Africa, where introduced cichlids and centrarchids frequently dominate reservoir fisheries not only because of their larger body size and greater market value, but also because they possess suites of invasion-promoting traits, including broad environmental tolerances, rapid growth, early maturation, high reproductive output, dietary flexibility, and resilience to anthropogenic disturbance. These characteristics enable them to exploit novel reservoir habitats and altered environmental conditions more effectively than many native fishes, often resulting in high abundances and disproportionate contributions to fishery production (Kolar and Lodge, 2002; Ribeiro et al., 2008; Ellender and Weyl, 2014; Weyl et al., 2017). From a fisheries perspective, biomass rather than abundance determines food availability, market returns, and household nutritional benefits. Therefore, the contribution of non–native fishes to rural livelihoods may be considerably greater than would be inferred from numerical abundance alone.

The temporal analyses provided little evidence that non–native species are continuing to expand their dominance. Neither biomass trends nor invasion progression metrics indicated directional increases during the six–year monitoring period. In contrast, both native and non–native assemblages exhibited relatively stable trajectories, suggesting that the reservoir has entered what may be described as an invasion equilibrium phase. Such equilibria have been reported in some African impoundments where introduced species successfully exploited underutilised trophic niches and became major contributors to fisheries production without causing measurable reservoir-scale declines in native fish biomass. For example, the introduction of the Tanganyika sardine Limnothrissa miodon into Lake Kariba resulted in the establishment of one of Africa’s most productive inland fisheries, with the species occupying the pelagic zone created following impoundment while much of the native assemblage persisted (Marshall, 1991; Karenge and Kolding, 1995; Kolding et al., 2003). Similarly, in Lake Kivu, L. miodon established a productive fishery by exploiting a previously vacant pelagic planktivorous niche, with relatively limited niche overlap with native fishes (Marshall, 1991; Snoeks et al., 2012). The absence of increasing dominance is ecologically significant because it suggests that invasion impacts may have largely manifested during earlier establishment phases and that present–day community dynamics are governed by longer–term coexistence mechanisms rather than active displacement. The importance of M. nigricans and C. rendalli in the IRI highlights how non–native species have become embedded within the trophic architecture of the reservoir. While bass introductions throughout Southern Africa were historically motivated by recreational fisheries objectives (Ellender and Weyl, 2014), the species now contributes directly to subsistence and artisanal fisheries in Lake Sebakwe, highlighting how introduced species can become embedded in local food systems while still posing conservation concerns for native assemblages. Similarly, the prominence of C. rendalli and O. niloticus indicates the establishment of productive herbivorous and omnivorous pathways that support fishery yields. Together with native species such as C. gariepinus, M. macrolepidotus and P. acuticeps, these species create a diversified fishery spanning multiple trophic guilds. Such trophic diversification can enhance ecosystem resilience by reducing dependence on a single stock and spreading production across multiple energy pathways, a mechanism widely recognised as contributing to stability in exploited ecosystems (McCann, 2000; Rooney et al., 2006).

From a livelihood perspective, one of the most consequential findings is the pronounced seasonal contribution of non–native fishes. Non–native CPUE increased significantly during the hot–dry season, whereas native catches remained comparatively stable across seasons. The seasonal importance of non-native fishes in Lake Sebakwe reflects a broader pattern observed across many inland aquatic systems globally, where harvested aquatic resources function as livelihood and food-security safety nets during periods of agricultural stress, seasonal food shortages, or economic vulnerability (Fluet–Chouinard et al., 2018; Funge–Smith and Bennett, 2019; FAO, 2024). Similar dynamics have been documented in African floodplain fisheries, the Mekong Basin, Bangladesh, and Amazonian river systems, where fishing effort and dependence on aquatic resources often increase when agricultural production declines or livelihood options become constrained (Lynch et al., 2016; Fluet-Chouinard et al., 2018; Funge-Smith and Bennett, 2019; FAO, 2024). Comparable patterns have also been reported for freshwater crab harvesting in Zimbabwe, where utilisation peaked during the non-agricultural winter season, further illustrating the importance of aquatic ecosystems as seasonal resilience mechanisms within rural livelihood systems (Dalu et al., 2016). The increase of non–native fish production during this vulnerable period suggests that these species provide an important buffering function against seasonal food shortages. Therefore, management interventions aimed at reducing non–native populations must carefully consider potential trade–offs between biodiversity conservation and food security outcomes. These findings contribute to an increasingly important debate surrounding the management of non–native species in developing–world fisheries. Traditional invasion-management frameworks often assume that reducing non–native abundance will automatically benefit native biodiversity. However, where non–native species are embedded in fisheries, food systems and local economies, management objectives become considerably more complex because conservation interventions must reconcile biodiversity protection with livelihood dependence, economic value, and social acceptance of the invaded resources (Gozlan, 2008; Schlaepfer et al., 2011; de Carvalho-Souza et al., 2024; Jarić et al., 2025). Lake Sebakwe exemplifies this challenge: although non–native species contribute substantially to fishery production and household nutrition, they also pose potential ecological risks through predation, competition, and genetic interactions with native taxa.

The market and preference survey reinforces the biomass-based interpretation by demonstrating that non-native species realise approximately double the aggregate economic value of natives despite being caught in smaller numbers. This decoupling between abundance and value arises because non-native species are larger-bodied, command higher prices per-kilogram, and are explicitly preferred by fishers and traders for market sale. However, the higher numerical catch of native fishes, combined with the dominance of small-bodied native taxa such as M. macrolepidotus and P. acuticeps in the smallest mesh size, indicates that natives remain the backbone of subsistence consumption and local protein supply, even where they are economically undervalued in formal trade. This dual pattern, where natives sustain household subsistence and non-natives sustain cash income, is consistent with observations from other African reservoir fisheries (Funge-Smith and Bennett, 2019; FAO, 2024) and warns against conflating market value with food-security value.

Particularly concerning is the continued expansion of O. niloticus, a species widely recognised as one of the most problematic freshwater fish introductions globally because of its broad environmental tolerance, rapid reproduction, strong invasion capacity, and documented impacts on native fish communities through competition, hybridisation, and ecological restructuring (Canonico et al., 2005; Cuthbert et al., 2022; Kwikiriza et al., 2025). Although biomass trends did not indicate rapid expansion, the established presence of O. niloticus alongside the low relative importance of native O. macrochir warrants caution. Across Southern and Eastern Africa, O. niloticus has repeatedly been implicated in genetic introgression and demographic displacement of indigenous congeners (Firmat et al., 2013; Deines et al., 2014; Zengeya et al., 2015; Munyai et al., 2023). Hybridisation processes often progress slowly and may remain undetectable through conventional catch monitoring programmes because genetic erosion can occur long before population–level declines become apparent. Thus, the apparent stability observed in this study should not be interpreted as evidence that ecological risks are absent, but it highlights the need for complementary genetic surveillance programmes capable of detecting cryptic changes in native populations. Similarly, the dominance of M. nigricans raises concerns regarding long–term impacts on native fish size structures and recruitment dynamics.

The progressive decline in native catches with increasing mesh size suggests a scarcity of larger–bodied native individuals within the reservoir. Although gillnet selectivity precludes definitive causal inference, the pattern is consistent with predator–mediated restructuring of fish communities documented elsewhere following bass introductions (Woodford et al., 2016; Khosa et al., 2020). Such effects may manifest through direct predation on juveniles, altered habitat use, or changes in competitive interactions. Importantly, these mechanisms often operate over extended temporal scales and may therefore remain undetected within relatively short monitoring periods. The observed mesh–size asymmetry nevertheless presents a valuable management opportunity. Native species were caught mostly in smaller mesh sizes whereas non–native species dominated larger meshes, suggesting that fishing regulations could be strategically designed to selectively target non–native biomass while reducing pressure on native populations. Selective harvesting has increasingly been advocated as a pragmatic approach for managing invasive fishes where eradication is unrealistic and where introduced species support economically valuable fisheries (Britton et al., 2011; Weyl et al., 2021). In Lake Sebakwe, adaptive gear regulations could therefore simultaneously support fishery production, protect native recruitment, and reduce the biomass dominance of non–native species.

Thus, the Lake Sebakwe fishery highlights the emergence of what may be termed a socio–ecological compromise ecosystem, in which non-native species generate disproportionate market value and seasonal biomass buffering, while native species sustain numerical abundance, subsistence consumption and biodiversity value. The data do not support simplistic classifications of non–native fishes as either unequivocally harmful or unequivocally beneficial. These findings show that biological invasions can generate simultaneous livelihood benefits and ecological risks, a pattern increasingly reported across freshwater ecosystems worldwide. This has important implications for the near-future persistence of native species, particularly where non-native taxa continue to spread under intensifying environmental change (Mace, 2014; Seebens et al., 2021). Future management should therefore focus on preventing new introductions, limiting secondary spread, monitoring the genetic integrity of native populations, and using targeted fishery regulations to exploit existing asymmetries between native and non–native stocks.

The six–year assessment demonstrates that Lake Sebakwe functions as a mixed–origin fishery in which native and non–native species jointly underpin food production and rural livelihoods. Non–native species contribute disproportionately to harvested biomass and provide an important seasonal food–security buffer during the hot–dry period, yet they also present ecological risks through predation, competition, and potential genetic introgression. Market survey evidence confirms that the food-security benefit of non-native species operates primarily through higher per-unit price and larger body size rather than through greater numerical catch, since native species remained more numerically abundant across the six-year monitoring period. This distinction matters for management: protecting native abundance is compatible with retaining non-native biomass for market supply, provided selective gear regulations are applied. Importantly, temporal analyses suggest that the reservoir has entered a post–establishment invasion equilibrium rather than undergoing continued expansion of non–native dominance. This apparent stability should not, however, be interpreted as ecological security, particularly given the established presence of O. niloticus and the trophic influence of M. nigricans. Management should therefore move beyond binary eradication–versus–acceptance approaches towards adaptive co–management strategies that prevent further introductions, implement long–term genetic and demographic monitoring, protect native recruitment habitats, and use selective gear regulations to direct harvest pressure towards non–native species. Such an approach offers the greatest potential to simultaneously safeguard biodiversity, sustain fisheries productivity, and maintain the food–security benefits upon which many Lake Sebakwe households depend.