Biological invasions by non-native species are well documented to severely alter and threaten biodiversity and ecosystem functioning (IPBES, 2023; Carneiro et al. 2025). While, non-native invasive species can also have profound negative effects on the economy (Diagne et al., 2021) and human wellbeing (Bradshaw et al., 2016; IPBES, 2023; South et al. 2025) these impacts are harder to document despite global synthesis efforts (Diagne et al., 2021). Economic costs attributed to biological invasions have been reported to have reached a minimum US$ 1.288 trillion in damages (Diagne et al., 2021). On the African continent, the cost of invasion is reported to be between US$ 18.2 billion and US$ 78.9 billion for the period 1970 and 2020 yet are recognized as being under reported (Diagne et al., 2021). Despite such huge costs, implementation of measures to prevent and control biological invasions remains limited mostly due to underestimation of negative impacts by decision makers (Courchamp et al., 2017). Determining economic cost associated with biological invasion is therefore essential to communicate negative impacts for motivation of the development of non-nattive invasive species policy and further strengthening of biosecurity (Essl et al., 2020; Seebens et al., 2017; 2021).

Crustaceans, particularly crayfish (Decapoda: Astacidea) are considered one of the most successful aquatic invasive species and have been introduced globally via capture fisheries, aquaculture, and aquarium trade pathways (Haubrock et al., 2021; Madzivanzira et al., 2020; Twardochleb et al., 2013). Due to their omnivorous nature, ability to attain large sizes and high population densities, invasive crayfish species can cause severe ecological impacts through strong trophic interactions and ecosystem engineering (Lodge et al., 2012; Twardochleb et al., 2013). Negative socio-economic impacts attributed to invasive crayfish include loss of ecosystem services such as food provisioning and increased expenses for agriculture and water management (Kouba et al. 2021; Lodge et al., 2012; Madzivanzira et al., 2020).

The redclaw crayfish Cherax quadricarinatus (von Martens 1868), is a species of concern which is now established and spreading through in several African river basins (Douthwaite et al., 2018; Haubrock et al., 2021; Madzivanziraet al., 2021a; Nunes et al., 2017; South et al., 2025). Invasive crayfish in continental Africa are phylogenetically novel as there are no native crayfish species except for Madagascar (Crandall & Buhay, 2007; Lodge et al., 2012). Cherax quadricarinatus was first introduced to the Zambezi Basin in 2001 for aquaculture purposes, and has since established feral populations in the Kafue Floodplains Ecoregion and Upper Zambezi Floodplains Ecoregion and the Middle Zambezi (Douthwaite et al., 2018; Madzivanzira et al., 2021a). Invasion costs at the Lake Kariba invasion core, in the Middle Zambezi, are significant where US$ 512 352.92 loss is attributed to damage by crayfish (Chakandinakira et al., 2023; Madzivanzira et al., 2023). In the Kafue Flats (Kafue Floodplains Ecoregion), specific impacts are still unquantified but fishers have reported catch spoilage and gear damage, with up to a third of catch damaged by crayfish (Weyl et al., 2017). Here we focus on the Barotse floodplain (Upper Zambezi Floodplain Ecoregion) which, despite having a well documented invasion history (e.g. Douthwaite et al., 2018; Madzivanzira et al., 2021a; Nawa et al., 2024) has not yet had an economic impact assessment. Similar to Lake Kariba, there are anecdotal reports from artisanal fishers which suggest that C. quadricarinatus are spoiling their fish catch through consumption of fish caught on gillnets and as well as damage to the nets (N. Namasiku pers com.). Consumption and flesh damage of fish renders the catch unsuitable for trade, resulting in cascading impacts throughout the value chain (Madzivanzira et al., 2022). Madzivanzira et al. (2022) estimated the cost incurred by extrapolating data from laboratory experiments, however these findings require verification as they represent a maximum possible cost. The potential of escalating economic cost is of great concern as the Barotse floodplain fishery is a fundamental source of livelihoods and food provisioning for approximately 70, 000 people across the floodplain and immediate surrounding areas (CGIAR, 2013). Given that 78.6% of the rural population are estimated to be living below the poverty line in the region (ZSA, 2023), the growing invasion may be a direct threat to human wellbeing. Quantifying the effects that C. quadricarinatus is having on the livelihood of fishers on the Barotse floodplain is therefore critical.

Here, a systematic field-based assessment was used to provide a quantification of economic cost incurred by artisanal fishers as a result of C. quadricarinatus. We aimed to 1) assess crayfish encounter and abundance across season (wet/dry) and invasion gradient (core vs upper/lower edges), 2) assess the prevalence of damage to fish catch and determine if fish catch damage varies by season, invasion gradient and habitat type (lentic/lotic) and if it is species specific, and 3) estimate associated economic costs (catch damage, gear damage and loss of fishing time) across season and invasion gradient.

Data was collected in 2021, using a roving creel survey, a method recommended for estimating catch and effort data for geographically dispersed fisheries (Pollock et al., 1994). Two types of creel surveys were conducted: seasonal survey and monthly survey implemented for 30 days and 14 days respectively. The seasonal creel surveys were conducted bi-annually focusing on both wet season (April) and dry season (October). Each survey was conducted across three strata: Lukulu (upstream invasion edge), Mongu (invasion core) and Senanga (downstream invasion edge) (Fig. 1). The monthly creel surveys were conducted every month from June to September in the Mongu stratum. For data collection, an electronic based questionnaire was utilised, accessed through the ODK Collect App (version 1.30.1) installed on Lenovo phone tablets. The questionnaire was semi-structured, consisting of both open-ended and closed-ended questions (Appendix 1). Seven trained field enumerators (2 in Lukulu, 3 in Mongu and 2 in Senanga) moved randomly through their designated strata and interviewed fishers on landing sites of villages/fishing camps returning from fishing as well as those still fishing in various water bodies across the floodplain.

During each interview, fish caught were sorted into categories based on how they are sold by fishers on the Barotse floodplain, and fish within each category were identified using Skelton (2001). The fish catch categories were: Catfish (Clarias gariepinus, Clarias ngamensis, Clarias stappersii, Clarias theodorae), African Pike (Hepsetus cuvieri), Tigerfish (Hydrocynus vittatus), Redbreast Tilapia (Coptodon rendalli), Threespot Tilapia (Oreochromis andersonii), Purpleface Largemouth (Serranochromis macrocephalus), Largemouths (Serranochromis spp.), Sargos (Sargochromis spp.), Other Large Breams, Small Breams (< 10cm – any bream under 10 cm, such as Pharyngochromis acuticeps, Pseudocrenilabrus philander), Silver Catfish (Schilbe depressirostris), Small Mixed Mormyrids (Hippopotamyrus ansorgii, Petrocephalus okavangensis, Pollimyrus marianne), Bulldog (Marcusenius altisambesi), Synodontis (Synodontis spp), Robbers/ Alestidae (Micralestes acutidens, Brycinus lateralis, Rhabdalestes maunensis), Mixed Small Fish (Cyprinidae, Mormyridae and juveniles of Cichlidae usually sold as a unit in hand, bucket or scoop). Within each fish category, undamaged (whole) and (crayfish) damaged individuals were separated and the weight (kg) of each were recorded using a digital scale. Damage due to crayfish was determined by presence of slicing wounds on the fish and/ crayfish in the net. Slicing wounds are synonymous with crayfish and different from other scavenging predators, like crocodile and terrapins, which tend to cause loss of substantial parts of the fish (Chakandinakira et al., 2023). A maximum of 15 randomly selected individuals from each damaged fish category were further inspected and body parts damaged were recorded. The damaged parts were classified into the following categories, fins, mouth, eyes, stomach, skin and gills. Additional information on fishing effort (number of days fished per week), crayfish presence, gear damage and time lost due to crayfish was also recorded during the interview.

To make seasonal comparisons, analysis of crayfish encounters and abundance, prevalence of fish damage and gear damage and time cost were all based on the seasonal creel surveys. Detailed analysis of fish damaged (i.e., daily catch rate and species composition of spoiled fish) by crayfish as well as associated economic cost was limited to gillnets (~ 100 m net/night) as they accounted for most of the damage. This analysis only incorporated fish damages incurred in Mongu stratum during the dry season. Fish damaged in Mongu stratum (wet season) as well as that of Senanga stratum (both wet and dry season) were not available for quantification as it was considered insignificant by the fishers and discarded spoiled fish, before the time of the interview (N. Nawa, pers. obs.). Lukulu stratum on the other hand was excluded due to lack of fish damage from the surveyed catches.

In-depth analyses of fish damaged in Mongu during dry season were based on the monthly creel surveys given the small sample size from the seasonal creel survey (i.e., October). Two monthly periods (June-July and August September) and habitat type (lentic vs lotic) were used to assess the dynamics of fish spoilage in Mongu stratum during dry season. For standardisation purposes, all analyses were based on individual fishers who had completed fishing at the time of interview implying that daily fishing effort was comparable. Monetary losses arising from gear damage were not determined due to insufficient information for cost quantification (i.e., frequency of gear replacement and/or repairs as well as associated monetary cost).

Prevalence of fish damage was expressed as a relative percentage frequency of the total number of fishers’ who encountered fish damaged by crayfish per season within each stratum. To assess the prevalence of fish damage among fishers between season and across strata, a Chi-square goodness of fit test was used. Daily catch rate was expressed as individual catch per unit effort (CPUE) and calculated for both spoiled fish and intact fish. The calculation was adapted from AES & WWF (2017) as follows:

To assess whether there was a difference in CPUE_spoiled across habitat (i.e., lentic, and lotic) and months (i.e., June-July and August-September) overall fish catch was taken into account by calculating the ratio of CPUE_spoiled : total fish catch (CPUE_spoiled + CPUE_intact). The ratio was arcsine square root transformed and then used as response variable in a Generalised Linear Model (GLM) with a log link after checking for residual distribution and overdispersion. To determine whether fish damaged was related to the crayfish caught as by-catch in gillnets, Spearman Rank Correlation test was performed on the arcsine square root transformed ratio and the number of crayfish caught. To assess species specific damage relative composition of fish species spoiled was expressed as a percentage of the total weight (kg) of catch.

Gear damage was expressed as a relative percentage frequency of the total number of fishers whose fishing gear was damaged by crayfish per season within each stratum. Chi-square goodness of fit test was used to test the difference in proportion of fishers with damaged gear across seasons and strata.

Time cost was expressed as a relative percentage frequency of the total number of fishers who lost time due to crayfish presence per season within each stratum. The Chi-square test of independence and chi-square goodness of fit test were used to difference in proportion of fishers who lost time during fishing due to crayfish across season and strata.

A total of 590 catch interviews were conducted across the three strata during the 30-day wet and dry creel surveys. Mongu stratum accounted for most of the interviews followed by Senanga and Lukulu strata (Table 1). The 14-day creel surveys conducted in Mongu stratum from June to September yielded a total of 422 interviews. The month of June had the highest number of interviews followed by July, August, and September (Table 2).

This study provides the first in field quantification of economic cost associated with C. quadricarinatus invasion on the Barotse floodplain of the Upper Zambezi system. Both crayfish encounter and abundance among fishers depends on season and invasion gradient. Fish damage due to crayfish was limited to gillnets at the invasion core during the dry season and equated to the monetary loss of ~ US$ 22,000 per annum. Our estimates for monetary loss is limited to Mongu but this is predicted to cascade across the floodplain as the invasion progresses. Thus, the monetary loss from gillnets is predicted to increase beyond US$ 50,000 per annum. The entire floodplain has high poverty rates and low economic status (WorldFish Center, 2007), Understanding economic cost associated with IAS has become very important as it informs resource management and policy development aimed at preventing and managing the arrival and spread of IAS species (Diagne, Leroy, et al., 2021; Seebens et al., 2021).

The high encounter rate of C. quadricarinatus among fishers at the invasion core compared to the invasion edges matches the invasion gradient throughout the floodplain (Nawa et al., 2024). The high crayfish encounter and abundance among fishers perhaps explain why significant fish damage was only detected at the invasion core. Despite C. quadricarinatus being caught in various fishing gears, fish damage due to crayfish was exclusively incurred in gillnets, similar to prior studies (Lowery & Mendes, 1977; Weyl et al., 2017; Chakandinakira et al., 2023). Gillnets are passive gears which are are set overnight whereby C. quadricarinatus are attracted to the fish caught (Chakandinakira et al., 2023). During the dry season, low water and reduced resources concentrate higher numbers of C. quardicarinatus into smaller areas which again exacerbates the damage extent (Nawa et al., 2024).

Change in fishing methods during the dry season appears to influence the dynamics of fish spoilage on the floodplain. Lentic habitat gillnetting incurs higher crayfish scavenging compared to lotic habitats, this is compounded by the higher abundance of C. quadricarinatus in these habitats. This may be driven by scent odour plumes being diluted in running water (Larson & Olden, 2016). Crayfish were infrequently retained on the gillnets indicating that further assessments need to be performed as a multi-method assessment with standardised sampling for crayfish (Madzivanzira et al., 2021a, b; Chakandinakira et al., 2023).

Fish that are damaged are considered unmarketable and not suitable for sale (Madzivanzira et al., 2020; 2022; Chakandinakira et al., 2023;). Damaged catch is either kept for home consumption, fed to domesticated animals, or discarded. The most damaged fish species/group by C. quadricarinatus on the Barotse floodplain was Clarias spp. which is likely a facet of high relative abundance of this genus across the region (AES & WWF, 2024). Other fish species/group highly impacted by C. quadricarinatus are H. cuvieri, O. andersonii, S. macrocephalus, C. rendalli, Serranochromis spp. Unlike this study, the C. quadricarinatus invasion on Lake Kariba mostly affected Oreochromis niloticus even though Clarias spp. and C. rendalli are also affected in small proportion (Chakandinakira et al., 2023). The large cichlids are of high commercial value on the Barotse floodplain (Tweddle et al., 2010; 2015) and their spoilage directly impacts fishers’ capacity to generate income. The damage caused to low value species such as robbers, mixed small fish, Synodontis spp., Schilbe depressirostris, and small mixed mormyrids is comparatively smaller. Damage caused may have significant implications on the livelihoods of people in the regions as the low value fish species/group now form a more reliable source of protein amidst declining stock of cichlids (Tweddle et al., 2015; AES & WWF, 2024; South et al. 2025). Such smaller fish species and size-classes are also associated with important micronutrients such as vitamin A, Iron and Zinc (Kawarazuka & Béné, 2011; Islam et al., 2023)

The impact of C. quadricarinatus on the Barotse floodplain is a major food security concern, given the high poverty levels in the region (ZSA, 2023). The total economic cost calculated for the fishery is approximately US$ 21,586.16 per year with each fisher losing about US$ 114.82 annually. The damaged fish represents 1.4% of the total annual fish harvest in Mongu stratum during dry season (AES & WWF, 2022). Although overall loss seems small, taking into consideration the annual average household income of 345.28 US$ estimated for the fishing communities on the Barotse floodplain (AES & WWF, 2022), the annual cash loss at individual level is about 33.25% representing a significant detriment to human wellbeing. The economic cost to the fishery is likely to increase in future as C. quadricarinatus spread further and become more abundant across the landscape.

The economic cost calculated in this study, is much lower than that for Lake Kariba which is approximately half a million US dollars (Chakandinakira et al., 2023). This may be due to two reasons, first Lake Kariba’s fishery has a far higher yield than the Barotse floodplain fishery, thus there is more chance of crayfish encounter (DOF, 2017). Secondly, the Lake Kariba invasion core was established 10 years prior to the Barotse floodplain, thus crayfish are more abundant and larger than in the Barotse floodplain (Madzivanzira et al., 2021a). The stable lacustrine environment of Lake Kariba characterised by slow or static water ideal for C. quadricarinatus (Wingfield, 2002; Haubrock et al., 2021) may have facilitated quicker establishment resulting in higher economic costs. Similarly, lacustrine conditions created by the dammed Kafue River may explain why reported crayfish damage to fishers’ catch is estimated as high on the Kafue flats (Weyl et al., 2017) although associated cost are yet to be quantified.

The potential economic cost on the Barotse floodplain fishery may be much higher than estimated cost from fish damaged given prevalence of gear damage and loss of time. Thin monofilament nets are easily broken and entangled (AES & WWF, 2022; N. Nawa pers. obs) thus monetary losses from gear damage may be considerable. The loss of fishing time as a result of gear reparation and loss of gear efficiency due to crayfish entanglement are similarly unaccounted for in this study (Weyl et al., 2017; Madzivanzira et al., 2022; Chakandinakira et al., 2023). Therefore, both economic costs from gear damage as well as loss time may be adversely impacting the livelihood of fishing communities on the Barotse floodplain.

As economic costs associated with C. quadricarinatus increase on the Barotse floodplain, fishers may increase their fishing effort and resort to use of illegal fishing methods to make up for the lost catch (Chakandinakira et al., 2023). Considering that the use of illegal methods such as water beating ‘kutumpula’, monofilament gillnets, mosquito nets and large seine nets with small mesh size is already rampant on the floodplain, sustainability of the fishery is seriously under threat (Peel et al., 2014; Tweddle et al., 2004; AES & WWF, 2022;). Large commercially valuable cichlids which are already under severe fishing pressure (Peel et al., 2014; Tweddle et al., 2015; AES & WWF, 2022), risk disappearing altogether. Thus, apart to impacting the livelihood of people through economic cost, potential behavioural shift to unsustainable fishing practices as a coping strategy to C. quadricarinatus invasion threatens the biodiversity of fish species on the Barotse floodplain.

Biological invasion poses a serious threat to the livelihood of people, especially in developing countries, as they lack the economic capacity to counteract invasion cost (Diagne et al., 2021). Evidence from this study shows that artisanal fishers on the Barotse floodplain incur economic costs associated with the invasion C. quadricarinatus. Although the costs are limited to the invasion core, the impact on the livelihood of local people is quite significant. Expansion of this population from the core outwards could significantly increase these impacts in the future. As crayfish invasions continue to expand in range across Southern Africa, this will exacerbate the dire lack of food security and high poverty levels in the region. To attain sustainable development goals, South African Development Community countries will have to prioritise prevention and management of non-native species invasions. Given complex socio-economic situations in most African countries, estimation of invasion cost will be very critical in aiding decision to allocate limited resource towards management of invasive species.