Small-scale inland aquaculture has considerable potential to generate socioeconomic benefits in low- and middle-income countries. Hundreds of millions of people globally rely on inland aquaculture, contributing substantial economic and employment benefits to national and regional economies (FAO, 2022). Inland aquaculture also has substantial benefits for rural livelihoods and poverty alleviation (Béné et al., 2016), nutrition and food security (Wang et al., 2024) and gender empowerment (Dam Lam et al., 2022), among others.

Carps are the most widely produced inland aquaculture species (particularly in Asia), accounting for 53.8% of global inland aquaculture production in 2022 (FAO, 2024). Carp aquaculture (especially in Asia) has contributed substantially to global inland aquaculture expansion (Zhang et al., 2022), especially through carp polyculture systems (Naylor et al., 2021).

Small-scale aquaculture systems generally rely on relatively small farms and simple (often rudimentary) technologies, which often constrain the actual output, productivity and efficiency of the sector, and as an extension its potential to enhance food security and offer sustainable livelihoods (Wang et al., 2023). However, far from being homogenous, small-scale inland aquaculture systems (incl. carp) have radically different characteristics in low- and middle-income countries in terms of production models and commercialization practices (Short et al., 2021).

Many efforts have sought to improve the performance of small-scale inland aquaculture systems in low- and middle-income countries. Improvement efforts mainly rely on the promotion of improved production practice/technologies, dissemination of knowledge about production/processing/marketing/nutrition, promotion of improved fish feed and species, or combinations of the above Henriksson et al. (2021). However, many factors have constrained such efforts (Morgan et al., 2017), including our incomplete understanding of differentiation in the characteristics and performance of aquaculture systems (particularly small-scale systems), as well as of the differentiation in the needs, preferences and capacities of producers (Short et al., 2021). This is compounded by the general lack of comprehensive and consistent national/regional aquaculture statistics, the dynamic and fast-changing nature of the sector, and the lack of high-quality and fine-grained datasets compared to land-based agricultural systems. Collectively such information gaps contribute to the general failure to appreciate the considerable heterogeneity in the sector (both between and within countries), and design fit-for-purpose policies and interventions to improve the performance of small-scale aquaculture systems (Short et al., 2021).

Bangladesh experiences many of these circumstances. Aquatic food contributes significantly to the growing demand for animal-based protein due to various interconnected environmental, cultural and socioeconomic reasons. Inland aquaculture caters currently for most of the national fish demand, with the country currently ranking as the 5^th largest aquaculture producer globally (FAO, 2022). The aquaculture sector has expanded substantially in the past decades, with an estimated 4.3 million households (20% of the rural population) operating at least one homestead pond that is usually small in size (0.08–0.10 ha on average). Carps are, by a wide margin, the most widely produced aquaculture species in Bangladesh, accounting for 55.4% of the total aquaculture output (DoF, 2022). Carp aquaculture is practiced across the entire country, and is overwhelmingly geared toward household self-consumption and/or the national market (Henriksson et al., 2018).

However, despite this clear importance of carp aquaculture systems in Bangladesh, and the broader region, we lack consistent and comprehensive datasets about their characteristics and performance across the country (and other parts of South and Southeast Asia). Although many public and private organizations have been actively trying to improve the performance of the sector through diverse interventions, these efforts are constrained by the lack of good understanding of producer differentiation, especially among small-scale producers that are one of the most dynamic segments in the sector (Henriksson et al., 2018; Karim et al., 2016). Arguably, the good understanding of the characteristics, performance and needs of (small-scale) producers could help inform the design and targeting of aquaculture interventions in Bangladesh, and other similar contexts.

This Data Report outlines a dataset that contains the production characteristics and socioeconomic performance of small-scale carp aquaculture systems in Bangladesh (Rossignoli et al., 2023a). The dataset was collected through in-person structured surveys with 4,540 carp producers in 54 areas (upazillas)¹ spanning the entire country. In terms of production characteristics, the dataset contains variables related to farm sizes, fish species, aquaculture production technologies, feed/seed use, aquaculture knowledge, and aquaculture incomes and expenditures, among others. In terms of performance, the dataset contains variables that can generate indicators of rural livelihoods (e.g., on-farm/off-farm income), multidimensional poverty, food security and environmental performance, among others.

Sampling and non-sampling errors can compromise data quality in large-scale farm-level surveys such as the one outlined in this Data Report (Gasparatos et al., 2018; Waha et al., 2016). In the context of the dataset presented here, these errors can compromise the robust identification of technology adoption and performance at the farm level, and ultimately the design of development interventions to improve farm performance and sustainability. The research team was conscious of these risks. From the onset of the study we implemented various quality assurance procedures to reduce these errors to the extent possible.

Sampling errors generally arise from the unrepresentativeness of the sample. During the initial stages of research design we identified that the main sources of sampling errors in our study would mostly be associated with the selection of: (a) study areas with unusual, unrepresentative, and/or very homogenous carp production systems, and (b) individual respondents (i.e., aquaculture producers) with unrepresentative carp production systems compared to the prevailing systems in their respective study area.

Risks associated with study area selection could likely occur by (a) selecting study areas with low/no carp production, (b) selecting study areas in few localities within Bangladesh (e.g., select upazilas only in some divisions or districts for convenience reasons), and/or (c) selecting study areas that collectively fail to capture the diversity of carp production systems in Bangladesh. Such study area selection decisions could individually or collectively lead to a skewed understanding of the national landscape of carp production systems.

To reduce the risks stemming from flawed study area selection we followed the multi-stage process outlined in Section 2.2 that consisted of expert consultations, GIS-based aquaculture suitability analysis and triangulation (to the extent possible) with secondary data compiled by the Department of Fisheries of the Bangladesh Ministry of Fisheries and Livestock (DoF, 2022). By superimposing the upazilas identified through the expert consultation (Figure 1a) with the aquaculture suitability map generated by the research team we ensured that none of the selected upazilas falls completely within areas of low aquaculture suitability (Figure 1b). This multi-stage approach allowed us to avoid selecting study areas with unrepresentative carp production, and particularly areas that individually lack significant carp production or areas that collectively fail to capture the diversity of carp production systems in Bangladesh. By relying on multiple data sources and a large number of experts we reduced to the extent possible: (a) biases that could emerge from reliance on the perspectives of just a few experts, and (b) inaccuracies due to the incompleteness of some datasets.

Risks associated with respondent selection could occur through biases in producer identification and/or lack of proper randomization during respondent selection in each study area. We expected that such types of errors are likely to occur due to the lack of comprehensive, consistent and spatially-explicit datasets about carp aquaculture in Bangladesh (e.g., see Section 1 and 2.2).

To reduce risks associated with respondent identification and selection, we followed a multi-stage process. The respondent selection process relied on a randomized selection from carp producer lists developed by the research team (Section 2.3). This aimed at minimizing (to the extent possible) biases and errors introduced by non-randomized sampling methods (e.g., convenience sampling) or by using outdated producers' lists. To ensure a relatively large and balanced pool of respondents, we randomly selected approximately 84 respondents in each study upazila, from more extensive lists of 130–150 producers that were developed by the research team for this project (see more details in Section 2.3). We must note here that we did not identify every carp producer in each study upazilas. Conducting such a complete census of all carp producers in the study areas, was both outside the scope of the study and would have been prohibitively resource intensive. However, we must also acknowledge that despite our best efforts there is some likelihood that randomization was not conducted perfectly in all upazilas, possibly inserting some selection biases. Unfortunately, this cannot be formally tested considering the lack of comprehensive and cohesive carp producers' lists, as explained throughout this paper. However, we believe that the final sample offers a good representation of the carp aquaculture sector in Bangladesh when taking into consideration the preliminary analysis of the patterns of major variables (Section 4), the insights gained from the expert workshops Rossignoli et al. (2023b), and the findings of other aquaculture studies in the country Henriksson et al., (2018); Hossain et al., (2022); Jahan et al., (2015).

Non-sampling errors refer to all other errors that can arise during data collection, which can cause the collected data to differ from their true values (Phung et al., 2015). Such errors typically occur if (a) data collection deviates from the established study protocol, and (b) responses are biased or fail to capture/report properly the required information (e.g., recollection errors) (Phung et al., 2015). During the initial stages of research design, we identified that the main sources of non-sampling errors would mostly be associated with non-response errors and measurement errors (Waha et al., 2016). We expected that such types of non-sampling errors could likely occur for two reasons. First, was the large enumerator team needed for data collection considering the study's extensive geographical scope (entire Bangladesh) and large sample size (>4,500 surveys) that were necessary to avoid the possible sampling errors mentioned above. Second, was the large amount of information captured in the survey and the long recollection periods for some of the key questions.

To reduce non-sampling errors, we developed and adhered to a robust and comprehensive protocol for the design and implementation of the household survey. Specific actions were to:

Finally, we must acknowledge that it is not possible to formally test the representativeness of the selected sample or externally verify the dataset due to the lack of spatially disaggregated national statistics about the main areas of carp production in Bangladesh, as well as the number, output, and characteristics of carp aquaculture producers. However, we see consistency in the findings of the analysis of the dataset analysis and the expert insights elicited during the expert workshops. In particular, a recent characterization study using the dataset to identify differentiation in the production models within the sample (Dompreh et al., 2025) found the same production models qualitatively identified by the experts during the workshop (Rossignoli et al., 2023b).

Figure 2 visualizes patterns for some of the main production and outcome variables in the dataset, which can be used for different types of applications (Section 5). The log-transformed density distributions reveal substantial variability for all variables. On average (indicated by red lines in each panel), households reported an annual income of 6,869 USD/yr, with aquaculture output contributing 3,755 USD/yr and aquaculture inputs incurring average costs of 2,180 USD/yr. These three income and cost variables exhibit right-skewed distributions, with most households concentrated around lower to mid-income levels and the long distribution tails indicating a smaller number of high-income or high-cost producers. Aquaculture production was on average 2,540 kg/year, with a similarly skewed distribution indicating that a few producers have substantially higher fish output than the majority of producers. The average production area was 0.57 ha, with the distribution showing a steep drop-off, underscoring the substantially higher prevalence of small-scale producers in the sample, and the carp aquaculture sector more broadly. Finally, food purchases were relatively uniform across producers, averaging at 318 USD/yr, and displaying a compact density curve. Collectively, these log-normal shaped distributions imply the significant heterogeneity among carp producers in the sample, justifying the relevance of our dataset that can be used to characterize fully the different carp production models in the country (Section 5).

Supplementary Figure 2 (Supplementary material) summarizes the prevalence of different fish species in the sample. Beyond the many different carp species (see below), many producers cultivated tilapia (Oreochromis niloticus: 43.3%) and different catfish species (Clarias batrachus; Pangasius pangasius; Wallago attu; Mystus tengara: 21.8%) (Supplementary Figure 2a, Supplementary material). The most prevalent carp species were rohu (Labeo rohita: 91.8%), silver carp (Hypophthalmichthys molitrix: 67.9%), mrigal (Cirrhinus mrigala: 61.8%) and catla (Catla catla: 60.8%) (Supplementary Figure 2b, Supplementary material). More importantly, practically all producers (~99.8%) adopt polyculture models (Supplementary Figure 2c, Supplementary material). More than 80% of producers cultivate between three and six fish species, with aquaculture systems combining four (32.4%) and five (24.0%) species being the most prevalent (Supplementary Figure 2c, Supplementary material).

Supplementary Figure 3 (Supplementary material) summarizes some of the main demographic characteristics of the surveyed households. Most household heads are middled aged (41–50 years; average age is 48.0 years) and most households contain >5 members (average household size is 5.1 people).

As outlined in Section 1 small-scale carp aquaculture is a particularly important aquatic food system in many low- and middle-income countries (especially of Asia). The dataset described here can have different applications in Bangladesh and beyond.

First, it can be used to explore the heterogeneity and differentiation of carp, and small-scale more broadly small-scale aquaculture production systems in Bangladesh, in terms of their comparative characteristics and performance. The main variables that can be used to describe the aquaculture systems are included in Module C and D, which contain comprehensive information in terms of fish/crops/livestock species and output, agriculture/aquaculture inputs, aquaculture management practices, intensification levels, and market orientation, among others. Variables in Modules C, D, F, G and H can be used to develop different performance indicators related to income (on-farm income in Module C-D; off-farm income in Module H), multi-dimensional poverty (Module G), food security (Module H), and environmental performance (Module C). Regarding the latter, Module C contains comprehensive information about aquaculture inputs and outputs that can be used to develop inventories for Life Cycle Assessments (LCA) to estimate nutrient use efficiency or environmental impacts such as acidification, eutrophication, and ecotoxicity.

Second, and related to the previous point, the dataset can be used to understand in more depth which factors affect the performance of these aquatic food systems. These factors can be very diverse, spanning multiple domains: technical (e.g., adoption of BMPs), demographic (e.g., household structure), capacity (e.g., education, access to extension services), geography/infrastructure (e.g., access to infrastructure), economic (e.g., access to credit, asset base) or institutional (e.g., participation in producer groups). Related variables can be found in different parts of the dataset: technical (mainly Module C), demographic (mainly Module A), capacity (mainly Module A and F), geography/infrastructure (mainly Modules A and C), economic (mainly Modules C, E, F), institutional (mainly Module E). Such studies can focus solely on Bangladesh or be used in multi-country assessments if combined with similar datasets from other countries.

Third, large-scale integrated modeling studies are increasingly exploring food system transformation and food system vulnerability to climate change and livelihood shocks. However, such models do not usually have good quality data for their aquatic food system components compared to land-based agricultural systems and fisheries. Our dataset contains diverse variables that can help populate such modeling components, depending on the focus of the modeling exercise. Of particular relevance could be variables related to the production characteristics (Module C-D), impacts (Module C, D, F, G and H), and vulnerability to climate change and livelihood shocks (Module I–J).

This Data Note describes a unique dataset that can be used to describe the characteristics and assess the performance of carp aquaculture systems in Bangladesh. Considering the large number of samples (n = 4,540) and it wide geographical coverage (54 upazilas across the entire country), the dataset essentially provides a snapshot of the current state of the carp aquaculture sector in Bangladesh. As the sector is very dynamic and experiences fast change, this information can be used to develop a national benchmark of this critical aquatic food system. This can offer a robust evidence base to both design future interventions and/or assess the performance of future interventions in the study areas. Although the actual data are essentially applicable only in studies focusing solely on Bangladesh, the findings of studies using this dataset can have wider implications considering the global importance of small-scale carp aquaculture, especially in Asian countries.