Improved Slaughter: Our Initial “Chill Kill” Trial

Summary

This post shares an update on our latest efforts to develop improved slaughter methods for farmed fishes in India. Specifically, it discusses our recent small-scale trial of “chill kill”, which involves slaughtering fishes via an ice slurry. We tested this on three fishes in a treatment group, while three fishes in a control group were killed via the usual mode of asphyxiation.

Our initial findings were encouraging: chill kill proved implementable even in India’s hot, tropical conditions, and fishes in the treatment group became unconscious and died significantly faster than those in the control group—an average of 16 minutes* compared to 51 minutes.

We remain excited about this project as an avenue of reducing suffering at slaughter, and we are planning to conduct more live trials to continue assessing and developing the implementability of this intervention.

This post also includes a Note on Ethics towards the end.

*Note: When we shared these findings in our newsletter, we originally published this figure as 23 minutes, as opposed to 16. This was due to an internal error on our part. Our apologies!

Context

For the past year, Fish Welfare Initiative has been exploring how to reduce suffering during slaughter for farmed fishes in India. The standard slaughter method—mass air asphyxiation—causes severe and prolonged distress, as fishes are left to suffocate slowly in the open air (see picture below).

Earlier this year, we funded three foundational research projects to assess the feasibility of introducing pre-slaughter stunning in India. Unfortunately, those studies indicated that stunning is not currently viable in the Indian market, primarily due to limited demand and significant technical and financial barriers.

However, the projects did identify one potentially promising avenue: slaughter via ice slurry, or “chill kill.” While chill kill is not itself a form of stunning—the fishes likely still experience a period of suffering—it may significantly reduce the duration, if not also the severity, of suffering at the time of slaughter.

To evaluate this possibility, we recently conducted an initial small-scale field trial. This post summarizes the results from that trial and outlines our next steps.

The “harvest” from which we obtained the six fishes used in this trial. This image shows the standard slaughter method for Indian major carps: fishes left in the open air to slowly asphyxiate. Most fishes visible here were still alive when the photo was taken.

Methodology

Before introducing live animals, our team first conducted a series of dry trials to ensure that our protocols were sound and that the ice slurry could maintain sufficiently low temperatures under field conditions.

Dry Trials

These preliminary trials focused on refining the ratio of water, ice, and low-concentration salt (used to depress the freezing point) to maintain near-freezing temperatures during harvest conditions. This was a critical step, as ambient temperatures in Andhra Pradesh often range from 25°C to 40°C, and we needed confidence that the mixture could reliably induce death once live fishes were introduced.

Across six dry trials—five at an ice factory and one at a farm site—we confirmed that our mixture could maintain a temperature of approximately 0°C for nearly two hours. These results gave us sufficient assurance to proceed to a live trial.

Our full internal field report from the on-farm dry trial can be found here.

FWI staff Sanjay, Varna, and Subrata measuring temperature during our final dry trial.

Live Trial

On September 19, our team conducted a small-scale live trial at a partner farm during harvest, with prior consent from both the farmer and trader.

Two groups of fishes were observed:

• Treatment group (chill kill): Three insulated styrofoam boxes were prepared, each containing one fish, 50 L of water, and 45 kg of crushed ice (roughly 90% ice by weight). The mixture’s salinity was 2 ppt, and the temperature was maintained between 0–2°C for approximately two hours through periodic stirring.

  
  

• Control group (air asphyxiation): Three fishes were placed in identical boxes without water or ice—replicating the standard slaughter method, though without the additional crushing from other fishes piled on top.

  
  

Each fish was randomly selected from the ongoing harvest. By that point, all individuals were likely extremely stressed, having spent time in crowded dragnet enclosures and exposed to air for several minutes. Unfortunately, such conditions will likely be unavoidable even in the best-case implementation of chill kill.

Throughout the trial, our team recorded detailed welfare indicators for the treatment group, including free-swimming behavior, equilibrium and righting response, opercular (gill) movement, eye movement, and response to gentle tactile stimulation (tail-grab reflex).

For the control group, since the fishes were in air and unable to exhibit most of these behaviors, we tracked only the indicators used to determine time of death.

Time of death was defined as the point when all three of the following criteria were met:

1. No response to tail grab

2. No eye movement

3. No opercular (gill) movement

   
   

We discuss our findings below.

FWI staff Mani and Subrata assessing eye-rolling reflex in fishes during the trial—an established indicator of consciousness. In conscious fishes, the eyes automatically rotate to maintain orientation when the body is gently tilted. An additional video of the trial can be seen here.

Findings

Our findings were as follows:

• Time to death: Fishes in the chill kill group died significantly faster, with an average time to death of 16 minutes, compared to 51 minutes for those killed via air asphyxiation.

  
  

• Behavioral observations: Fishes in the chill kill group did not appear to struggle violently, while those in the air asphyxiation group did. However, because visibility was limited once fishes were submerged in the ice slurry, this observation should be interpreted cautiously.

During chill killing, we observed a consistent sequence of behavioral changes:

1. Loss of free-swimming behavior

2. Loss of equilibrium (lying on the side)

3. Eye fixation

4. Slowed or irregular gill movement

5. Complete cessation of gill movement

Although this was a small trial, we take these results as a slight positive update that:

• Chill kill can be successfully implemented under field conditions in India.

• Chill kill appears to reduce the duration of suffering compared to conventional asphyxiation.

  
  

You can view the raw data here.

Next Steps

Based on these initial findings, we plan to continue conducting field trials from now through December 2025. Provided that results remain promising, we intend to scale up the size of the trials—including at least one involving over 100 fishes. All future trials will likely continue to take place only during existing harvests (see A Note on Ethics below).

In parallel, this quarter, we are also surveying traders about their willingness to adopt chill kill and conducting a cost assessment. Together, these studies aim to answer a central question: Is chill kill logistically implementable at even a moderate scale?

If that question is answered positively, several further questions remain:

• Program design: How can chill kill be cost-effectively implemented at scale? What would motivate farmers or traders to adopt it—financial incentives, product quality improvements, or market demand? If chill kill does lead to fresher fish, how can that benefit be effectively communicated across the value chain?

  
  

• Welfare assessment: Does chill kill truly constitute a sufficient welfare improvement? While our current approach relies on behavioral indicators, a fuller understanding would require neurological data, such as electroencephalogram (EEG) readings, to detect brain activity. Such testing would clarify whether fishes lose consciousness quickly or instead experience paralysis while still conscious. Our next judgments will also be informed by our own intuition about the likely severity and duration of suffering under chill kill compared to air asphyxiation, as well as our recent internal investigation into these differences.

Given our early results, our judgment is that it is best to focus for now on the implementability questions and return to the welfare question once the method’s feasibility is better established.

We also plan to proceed cautiously: if any stage yields strongly negative results, we will halt the project. For example, there would be little value in investing in EEG-based welfare studies if traders ultimately prove unwilling to use chill kill at scale.

A Note on Ethics

This study—and the broader intervention it explores—is ethically complex for several reasons:

1. It was we ourselves who killed the fishes.

2. We used a control group that we expected would experience more suffering than the treatment group, which indeed appears to have been the case.

3. If we continue developing this intervention, we will likely need to kill many more fishes, and additional control groups may again be necessary.

Trials like these make animal advocates uncomfortable. As animal advocates ourselves, we share that discomfort. Conducting this trial brought a mix of conflicting emotions on our team: on the one hand, satisfaction at obtaining encouraging results; on the other, sadness and unease at having participated in the cruel deaths of these animals.

Despite these complications, we believe such trials are warranted for now. Our reasoning is as follows:

1. We will continue to conduct these trials only at existing harvests. This means that the counterfactual fate of these animals, as was the case with our live trial here, will continue to be asphyxiating to death in a way that is likely more painful than chill kill. Note that limiting ourselves to only running these experiments at harvests does slow our progress, as significant time investments are required to gain trader and farmer buy-in. However, at this time, we believe it is not ethically acceptable for our organization to cause the deaths of animals who wouldn’t otherwise immediately have been killed.

   
   

2. The expected welfare improvement is significant. If chill kill proves feasible and can replace air asphyxiation across even a portion of India’s aquaculture sector, the reduction in suffering could be immense. Air asphyxiation is among the most prolonged and severe slaughter methods used for any vertebrate species. Even if chill kill is far from ideal, shortening death from an hour to half that time would represent a substantial improvement for potentially millions of animals each year.

   
   

More fundamentally, we believe that genuine progress in this field requires us to confront difficult realities directly—both physically and ethically. In our view, working closely within fish farming systems, rather than critiquing them from a distance, offers the most plausible path toward improving them in the long term.

We will continue to grapple with these ethical questions, and we will continue to share our activities and our reasoning openly on this blog. Transparency is especially important in this context, as some potential donors may prefer not to support an organization that conducts trials of this nature. Such donors deserve clarity about FWI’s position and approach.

Acknowledgements

In addition to our own staff, we are grateful to the farmer and trader who consented to our undertaking this study at their site. We are also grateful to Sanu Basu, a friend of the organization, who helped on the day of the live trial.