When the Oskil dam in the Kharkiv region was damaged in 2022, the reservoir drained and land downstream flooded. The destruction altered the river system and the ecosystems connected to it. Ukrainian prosecutors have brought ecocide charges against two Russian generals over the destruction. But the case raises a difficult question for investigators and courts about what exactly counts as harm – the flooded land, the polluted water, or the time an ecosystem takes to recover. That question now sits at the centre of Ukraine’s reform of its ecocide offence and of international efforts to define the crime.

The Legal Puzzle of Ecocide

Ukraine is considering Bill No. 15424-1 to revise its domestic ecocide offence, while Vanuatu, Fiji and Samoa proposed recognising a separate international crime of ecocide under the Rome Statute in 2024. In October 2025, Vanuatu provided an update on the proposal to the Assembly of States Parties’ Working Group on Amendments, indicating that it would not formally table the proposal at that year’s Assembly. The two initiatives belong to different legal systems, yet in bothcases, the drafters must define ecological harm precisely enough for a criminal court to decide what constitutes ecocide.

The 10th Kharkiv International Legal Forum, held on 25 September 2026 in Kharkiv, a frontline city affected by Russia’s war, brought these debates together with the experience of Ukrainian prosecutors investigating wartime environmental destruction. The panel, “Ecocide as a Crime against the Future: Global Challenges and Ukraine’s Experience”, discussed how ecocide is recognised internationally and how Article 441 of the Criminal Code of Ukraine (CCU) works in practice. Much of the discussion turned on evidence.

Ecocide is already a crime under Article 441 CCU, which defines it as “mass destruction of flora or fauna, poisoning of air or water resources, and other actions that may cause an environmental disaster”. It carries a prison sentence of eight to fifteen years’ imprisonment. Having the offence in the Code has not solved the practical problem. Prosecutors still have to turn terms such as “mass destruction”, “environmental disaster”, or – borrowing from international standards – “severe”, “widespread” and “long-term” harm into facts that can be established through evidence and attributed to a particular individual. Bridging this gap between abstract legal definitions and courtroom realities has thus become the central test for prosecutors on the ground.

From Statutory Language to a Workable Threshold     

When does environmental destruction stop being an ordinary environmental offence and become ecocide?

Oleg Bondarenko, Chair of the Verkhovna Rada Committee on Environmental Policy and Nature Management, described parliament’s current attempt to answer this. Bill No. 15424-1, registered on 3 August 2026 and under consideration by theparliamentary committee as of 12 October 2026, proposes defining ecocide as an unlawful intentional act resulting in severe and widespread or long-term damage to the environment. The terms “severe”, “widespread” and “long-term” provide a framework for the offence, but they do not tell courts how much damage is enough to constitute ecocide.

The international legal framework raises similar questions, although its requirements differ. Article 8(2)(b)(iv) of the Rome Statute covers certain attacks during international armed conflict launched with knowledge that they will cause widespread, long-term and severe damage to the natural environment that would be clearly excessive in relation to the anticipated military advantage. Unlike the proposed Ukrainian formulation, these thresholds are cumulative, and the damage is weighed against the military advantage. Neither provision says how a court should measure the damage. An ecosystem-based approach offers a way to assess harm under both frameworks: by examining how far ecological effects extend and how long the affected ecosystem takes to recover. The legal requirements, however, remain different.

Article 441 in its current form already shows the problem. Viktor Checherskyi of the Specialised Environmental Prosecutor’s Office listed several unresolved questions about the objective elements of the offence. Practitioners read the phrase “may cause an environmental disaster” in different ways. On one reading, it qualifies only the residual category of “other actions”, so mass destruction of flora or fauna and poisoning of air or water would not need to risk a disaster at all. On that reading, the line between ecocide and other environmental offences blurs.

The larger gap is that Ukrainian law does not clearly specify which environmental consequences make conduct ecocide. It is also unclear whether the offence can be committed by omission, and when it is complete: once the destruction occurs, or as soon as the threat of an environmental disaster arises. Checherskyi proposed replacing the abstract wording “possible negative consequences” with three categories: (i) irreversible consequences, where the natural value of ecosystems is lost forever; (ii) consequences where restoration of degraded ecosystems is possible but very costly and sometimes takes more than one generation; (iii) consequences where another ecosystem emerges in place of the original one, as after the destruction of the Kakhovka Hydroelectric Power Plant on the Dnipro River in June 2023.

The panel left these questions open, and the disagreement reaches beyond Ukrainian law. An international provision can use threshold language that looks precise, and a court will still have to decide what it means for a drained reservoir or a poisoned river.

Whatever threshold the law adopts, courts need criteria for assessing ecological change that keep ordinary environmental offences out of the ecocide category. Checherskyi’s three categories are a good place to start, because they describe what happens to the ecosystem rather than only what the damage costs.

Oskil Dam as an Illustration         

What unit of environmental harm should investigators measure?

Andriy Kalnytskyi of the Kharkiv Regional Prosecutor’s Office used the destruction of the Oskil dam to examine this question in practice. When the dam suffered extensive damage during fighting in 2022, much of the reservoir drained and areas downstream flooded, disrupting the reservoir and the river ecosystems connected to it. The investigation also offers a way to test what “widespread”, “long-term” and “severe” damage would have to mean in practice if the wording of Bill No. 15424-1 were applied to it.

Take “widespread” first. Investigators can count the square kilometres that were directly flooded or drained, or they can follow the ecological effects through the reservoir, the river system, the floodplains and the ecosystems that depend on them. The definition proposed in 2021 by the Independent Expert Panel, convened by the Stop Ecocide Foundation,already points in this direction: it treats damage “suffered by an entire ecosystem” as widespread, and damage that “cannot be redressed through natural recovery within a reasonable period of time” as long-term. “Long-term” raises a similar choice. It can mean how long the physical consequences of the attack last, or how long the disturbed ecosystem needs to recover. For Oskil, we think recovery time is the right measure. The reservoir may have drained and the floodwaters receded, but the ecological damage did not end with them. Measuring duration by the physical event would therefore understate how long the harm lasts, at Oskil and in any case where an ecosystem remains damaged after the event that disturbed it has passed. For “severe”, an economic assessment of losses is one option. The other is a set of indicators: loss of biodiversity, mass mortality of aquatic organisms, changes in the hydrological regime, pollution, disruption of ecosystem functions and consequences for the population. In our view, a figure for economic loss alone cannot show how severely an ecosystem was harmed.

Kalnytskyi argued that Ukraine is supplying international law with expert environmental assessments and measured consequences of attacks on hydraulic structures. With this material, it becomes possible to test whether legal norms shaped largely by twentieth-century conflicts can protect the environment in twenty-first-century warfare.

Proving Ecocide in Wartime

Measuring harm across a whole ecosystem helps a prosecution only if the harm can then be attributed to a person.

Andrii Tkachenko of the Prosecutor General’s Office explained the difficulty using the Kakhovka investigation. When the dam was destroyed in June 2023, areas downstream flooded and much of the reservoir upstream drained. Theconsequences reached far beyond the dam, cutting water supplies and destroying habitats along the river. For a criminal case, proof that a disaster happened is only the starting point. Investigators must establish how the dam was destroyed and who was involved. They must then show which environmental consequences resulted from those actions and how extensive the damage was.

In wartime, each of these steps becomes harder. Many of the crimes under investigation are attributed to the armed forces of the Russian Federation operating in temporarily occupied territories or directly in combat zones. In these places, investigators often cannot gather evidence by the ordinary methods set out in the Criminal Procedure Code of Ukraine.

Individual responsibility creates an additional evidentiary problem. To establish a commander’s involvement, investigators must first identify the military unit and determine who exercised control over it. They then need evidence showing the individual’s role in the decision that led to the environmental damage. If ecocide is measured across the whole ecosystem, the prosecution must link the accused’s conduct to the wider ecological consequences as well as to the damage at the dam. It must also separate those consequences from conditions that existed before the attack and from other wartime causes. The mental element adds a further burden. Bill No. 15424-1 requires an intentional act. It leaves open whether the required intent must extend to the full ecological consequences or only to the act itself. The answer will determine the additional evidentiary burden that an ecosystem-based approach places on the prosecution. Either way, the extent of environmental damage alone cannot establish the accused’s mental state.

Concluding Remarks

For Ukraine, the practical consequence concerns how environmental harm is defined. Bill No. 15424-1 should do more than repeat “severe”, “widespread” and “long-term”. It should tell courts that widespread damage is measured by how far the effects extend through connected ecosystems, and long-term damage by how long those ecosystems need to recover. We accept that this makes prosecution harder, as the previous section shows, but we think the cost is worth bearing. Defining the extent of harm and attributing it to an individual are separate elements of the offence: a broad definition of harm does not by itself lower the bar for proving causation or intent, so the accused remains protected by the full standard of proof. A narrower definition would ease the prosecution’s task only by leaving out damage that actually occurred. These methods could also inform future discussions on defining environmental damage if states revisit the proposal to include ecocide as a separate crime under the Rome Statute.

Investigators also need agreed scientific methods for measuring ecological damage, so that courts can compare findings across cases and assess how reliable the evidence is.

Evidence gathered for an ecosystem-based assessment also serves purposes beyond the criminal trial. Findings on the extent of environmental damage and the time needed for recovery may help support Ukraine’s compensation claims, for example through the Council of Europe’s Register of Damage for Ukraine, which opened its environmental damage category in September 2026. The same findings can inform decisions about what needs to be restored.

This tension cannot be fully resolved here. An ecosystem-based understanding describes environmental harm more accurately. Criminal law still needs rules clear enough to apply and evidence strong enough to convict a particular person. The wider the harm the law recognises, the harder it becomes to tie that harm to one commander’s decision and state of mind.

The panel at the 10th Kharkiv International Legal Forum was organised by the Centre for Sustainable Development and Environmental Protection and the EU-funded RE-GREEN project (101176210), with support from FORMAS (grant 2024-00299).