2025 Monitoring and Evaluation Report of Our Farm Program
- Jennifer-Justine Kirsch
- 2 hours ago
- 5 min read
Summary
At Fish Welfare Initiative, we closely monitor our farm program, the Alliance for Responsible Aquaculture (ARA), to ensure it improves fishes’ lives as intended. This post shares key findings from our first full year review of the Monitoring and Evaluation (M&E) cycle, covering January to December 2025.
Below you will find the key findings from our latest evaluation. You can access the complete report here:
About the Alliance for Responsible Aquaculture
The ARA is FWI’s primary implementation program and accounts for about 20% of our annual budget. Through it, we work directly with fish farmers in Andhra Pradesh, India, to improve water quality and maintain appropriate stocking densities.
Farmers typically enroll in the ARA to reduce disease prevalence and mortality rates. Participating farms engage with our program through:
Regular Farm Visits: We visit each farm monthly to observe farm conditions, measure water quality, and provide corrective actions as needed.
Stocking Density Cap Commitment: Participating farms commit to maintaining stocking densities within predefined limits.
As of January 2026, the ARA program actively engages 179 farms, monitors approximately 9.5 million fishes, and has cumulatively benefited an estimated 4.8 million fishes. To learn more about the ARA program, see this blog post.
Monitoring and Evaluation at Fish Welfare Initiative
One of our core values at FWI is evidence-based decision-making. To ensure our programs work as intended, we've implemented an M&E plan for the ARA.
We recently concluded the first annual M&E review covering January to December 2025. We had previously done a mid-year review in July 2025 and will continue publishing these reports biannually.
Key Findings from our 2025 M&E Report
(Note: These findings are mostly copied from our internal report and contain somewhat more technical language than typically found in our blog. For context, see the full internal report.)
Water Quality Measurements and Fishes Helped
We detected out-of-range (OOR) water quality events in 7% of farm visits. That’s encouraging for fishes (because there seem to be few acute problems when we visit), but it also limits how often our current model can directly “trigger” measurable welfare improvements. This is one reason we are developing remote sensing models to detect water quality problems more continuously.
When water quality issues were found, about 20% led to measurable improvements in fish welfare—particularly during summer months (Fig. 1).
Stocking Density and Life Stages
Stocking density adjustments accounted for less than 1% of the total fishes helped in 2025. Given this low impact, we’re reconsidering how much to prioritize monitoring this metric.
More than half of the fishes helped were in breeding or rearing stages. This reflects higher stocking densities in early life stages, where water quality improvements often affect many individuals.
Cost-Effectiveness
Currently, the ARA program has a cost-effectiveness of approximately 11 fishes per dollar. This is a slight improvement from 10 fishes per dollar at the end of 2024. Our cost-effectiveness fluctuates throughout the quarters due to weather-induced spikes in water quality issues (Fig. 2).
Stocking Density Monitoring
ARA farmers self-report stocking densities when introducing and harvesting fishes. We verify this through harvest data. Because stocking density appears to contribute little to impact right now, we’re weighing whether this monitoring is worth the staff time. Table 1 details the number of harvest events we evaluated during each report period and how many of the reported densities could be accurately verified.
% of total January-June 2025 | % of total July-December 2025 | |
Total harvest events evaluated: | 82 | 77 |
Farms that overreported their SD: | 0% | 8% |
Farms that underreported their SD: | 6% | 10% |
Farms that accurately reported (± 5%): | 94% | 82% |
Table 1: Analysis of stocking density verified through harvest events that occurred in 2025.
Pond Additions and Drop-Outs
59 ponds were added to the program in 2025, most before July.
32 farms exited the program—a substantial increase from the first half of the year.
The most common reason was farmers shifting from fish to shrimp culture, likely due to pressure from shrimp traders. Fish farming tends to include the use of antibiotics and other chemicals, which are prohibited in shrimp exports. This trader preference may be encouraging farmers to separate fish and shrimp cycles or move away from fish altogether.
Reason | Count | % |
Change to shrimp farming | 19 | 63% |
Change to pangasius farming | 3 | 10% |
Farm lease ended | 3 | 10% |
Unreachable | 3 | 10% |
Refused to reduce stocking density | 1 | 3% |
Logistical reason | 1 | 3% |
Table 2: Reasons for farm drop-outs in 2025.
Performance And Key Uncertainties About Our Theory of Change
The following visualization details our performance on each of the steps of our Theory of Change, comparing performance indicators for January-June and July-December 2025.
The biggest uncertainty about our Theory of Change is the counterfactual improvement of water quality. While we observe improvements when we provide farmers with water quality measurements and corrective actions, we lack sufficient evidence to understand what happens at farms where we don’t intervene. Would water quality improve naturally at a comparable rate?
To answer this, we initiated an outcome evaluation in September 2025. Unfortunately, the data from this study were unusable due to limitations in data collection and a protocol error. We are re-initiating the study and expect to have results by June 2026. You can learn more about this evaluation project here.
Additionally, we will continue to develop and operationalize our Welfare Assessment Protocol (WAP) which we drafted throughout 2025. This tool will help us further assess and compare fish welfare across farms and, hopefully, interventions.
Program and M&E Process Reflections
This M&E cycle included the creation of our M&E report, an M&E meeting with Programs staff, and the review of the report by ground staff members.
Launching a formal M&E process has provided us with useful insights into the program’s strengths and limitations. For example, while stocking density monitoring is currently part of the ARA, its contribution to our impact is limited—an insight which prompted us to reconsider the resources we spend on this activity.
M&E is a daily activity at FWI, and we intend to share updates on our key indicators every six months in the form of the M&E report and this type of blog post. In the meantime, you can follow some of our key metrics on the public ARA dashboard.