Water Quality Challenges in Aquaculture: An RCT with Nigerian Aquafarmers

Last registered on July 10, 2023
View Trial History

Pre-Trial

Trial Information

General Information

Title
Water Quality Challenges in Aquaculture: An RCT with Nigerian Aquafarmers
RCT ID
AEARCTR-0011718
Initial registration date
July 04, 2023

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
July 10, 2023, 9:18 PM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Primary Investigator

Name
Mans Soderbom
Affiliation
Department of Economics, University of Gothenburg
Contact Primary Investigator

Other Primary Investigator(s)

PI Name
Håkan Eggert
PI Affiliation
Department of Economics, University of Gothenburg
Contact Investigator
PI Name
Kira Lancker
PI Affiliation
University of Copenhagen
Contact Investigator
PI Name
Nnaemeka Chukwuone
PI Affiliation
University of Nigeria Nsukka
Contact Investigator

Additional Trial Information

Status
In development
Start date
2023-07-16
End date
2024-07-15
Keywords
Agriculture, Firms & Productivity
Additional Keywords
Aquaculture, Economics, Training, Water quality
JEL code(s)
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
Yields in aquaculture farming (aquafarming) vary substantially across regions and remain comparatively low in many African countries. Why this is the case is not fully understood. The premise this project is that low yields can, at least to some extent, be attributed to poor water quality. That this is a reasonable idea is supported by a large body of empirical research in various disciplines. The main hypothesis of the project is that small investments in training and technology can result in lasting and non-negligible positive effects on water quality, and consequently higher yields. To test this hypothesis, we carry out a randomized controlled trial (RCT). The setting is aquafarming in Nigeria, and the RCT will enable us to estimate the causal effects of access to training and access to a toolkit for water quality monitoring on water quality and farming outcomes. The design of the RCT will be two-armed, where the control group and the treatment groups are randomly recruited from a pool of approximately 600 (aquafarmers) based in the area Ekpan, Warri Delta State, Nigeria. The treatment group will be offered 2-3 days of training on how to assess and improve water quality, focusing on acidity and oxygen levels in particular. The training is carried out by staff at the Institute of Aquaculture Development Nigeria. The treatment group will also receive information about the quality of the water in their ponds. This information will be provided by field workers who will use a water quality control kit to obtain measures of water quality with respect to pH and oxygen level on a regular basis. All subjects will be interviewed in a baseline survey and revisited twice with 6-7 months intervals (the average rotation period of fish in the ponds). The program is implemented in partnership with University of Nigeria and Environment for Development, University of Gothenburg.
External Link(s)

Registration Citation

Citation
Chukwuone, Nnaemeka et al. 2023. "Water Quality Challenges in Aquaculture: An RCT with Nigerian Aquafarmers." AEA RCT Registry. July 10. https://doi.org/10.1257/rct.11718-1.0
Sponsors & Partners
Experimental Details

Interventions

Intervention(s)
We design an RCT where the treatment group, consisting of 300 randomly selected aquafarmers, get 2-3 days of training focusing on pH level and dissolved oxygen level as input parameters for improving survival and growth of larvae and fish. Training is carried out by the Nigerian Institute for Aquaculture Development. In addition to training, farmers in the treatment group will also receive information about the quality of the water in their ponds. This information will be provided by field workers who will use a water quality control kit to obtain measures of water quality with respect to pH and oxygen level.
Intervention Start Date
2023-07-16
Intervention End Date
2024-07-15

Primary Outcomes

Primary Outcomes (end points)
There are four main outcomes of interest:
• Water quality, measured by standard indicators of pH levels and dissolved oxygen level
• Total aquafarming (fish) output in kilograms per area unit
• Survival rate as harvested quantity ratio in relation to quantity fingerlings initially put into the ponds.
• Market price of harvested output.
Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)
We will study a range of secondary outcomes that shed further light on mechanisms, side effects, and the distribution of effects:

• Does the intervention affect adoption of new species varieties (e.g., hybrid)?
• Does water quality affect pond rental price?
• Heterogeneity in outcomes: What is the distribution of the main effects across households that differ with respect to gender dimensions and food security?
• Does the intervention have side effects, e.g. in terms of how family labor, including child labor, is used for acquafarming?
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
The design is a two-armed RCT, where the control group and the treatment groups are randomly recruited from a pool of approximately 600 aquafarmers based in the area Ekpan, Warri Delta State, Nigeria. The treatment group will be offered training on how to assess and improve water quality and access to information regarding water quality.

To account for the behavior of aquafarmers before and after the intervention, the experimental design includes ex-ante, mid term and ex-post surveys for the treatment group and the control group. In our empirical analysis, observable characteristics will be used to:
• test for balance at the baseline across the treatment and control groups
• estimate the causal effects with precision
• investigate heterogeneity in outcomes across observable characteristics

In the Ekpan area there are approximately 1000 aquafarmers, most of them are organized in cooperatives which have 100-200 members. We have identified a total of 6 such cooperatives to date.
Experimental Design Details
Not available
Randomization Method
Randomization will be done in office by a computer.


The cooperative leaders in have provided us with lists of all their members. We will use the list to draw an equal number of aquafarmers for the non-treated and the treated group, including reserves that will be asked to join if any of the original subjects on the list do not want to participate or are unable to participate for any other reason.

Randomization Unit
We randomize at the aquafarmer level, stratifying by cooperative. Hence each cooperative will contain farmers in the treatment group as well as farmers in the control group. There is a risk that the treatment “spills over” from farmers in the treatment group to farmers in the control group. We expect such spillovers to lead to a downward bias in the estimated average treatment effects. Hence, if we find an effect, we expect it to be a conservative estimate of the actual causal effect. Power calculations confirmed that randomizing at the level of the cooperative would not be feasible, because there are too few cooperatives.
Was the treatment clustered?
No

Experiment Characteristics

Sample size: planned number of clusters
600 aquafarmers.
Sample size: planned number of observations
600 aquafarmers.
Sample size (or number of clusters) by treatment arms
300 aquafarmers control, 300 treatment.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
We have carried out a power analysis based on the following: • Hypothetical estimable treatment effect: Based on a pilot study we find that mean output per acre is 1500 kg, while the standard deviation was 250 kg. If we assume that the true treatment effect is to raise output by 5%, i.e. 75 kg, and that the spillover effect from treatment to control is such that control subjects increase their output by 25 kg, the expected (downward biased) estimate of the treatment effect with our design is 50kg. • Unobservables: We assume that the error term is the sum of an unobservable component that is common within each cooperative and another unobservable component that is farmer specific. The two unobservable terms have the same variance, which implies that the intracluster (within cooperative) correlation in the error term is 0.50. • Estimation and inference: We use OLS with cooperative dummies (fixed effects). Standard errors are robust to heteroskedasticity. • Sample size as indicated above. Under these circumstances, the power to detect an effect at the 5% level of significance is approximately 90%. Results based on a randomized inference procedure are similar.
Supporting Documents and Materials
IRB

Institutional Review Boards (IRBs)

IRB Name
Ethical Advisory Board, Department of Economics, University of Gothenburg
IRB Approval Date
2023-06-26
IRB Approval Number
N/A
Analysis Plan