The area of land that is covered by water in a big flood (like those which only occur once every 20-50 years) is not the right starting point for determining what the "bed" of a river is for the purposes of the RMA. Rather, the Court determined that the correct starting point is the area covered by water when the river is as full as it can get without overtopping its banks. Those are the normal "banks", which contain the river in "usual" conditions, not in big floods widely covering the flood plains.

This is an important legal issue, because whether land is in the "riverbed" dictates whether a wide range of activities involved with farming (such as cultivation and irrigation , as well as any type of disturbance for structures, fences , sheds or pumphouses) are likely to be lawful on land near rivers. It also has a similar effect on nearly all residential and other urban activities. It also affects what is lawful in the bed of a river and when you can be prosecuted.

In contrast, if it is not "riverbed", these types of activities will be perfectly lawful. As a result, it is much more likely that typical farming and residential activities do not need a resource consent if they are outside the "riverbed", but do require one if they are inside it. Whether the land is inside someone's private title boundary makes no difference to whether it is "riverbed". That all depends entirely on whether it meets the RMA's definition to "riverbed", which makes no mention at all of ownership.

The definition includes terms like "fullest flow" and "banks", which are not defined. In the District Court, the Canterbury Regional Council, which had brought the prosecution, had argued that that left uncertainty, particularly in braided rivers, where it is often hard to find a clear "bank". Often, for example there are tiered terracing leading down to the main braids of a river. In order to ensure that regional councils could protect the natural character of land affected by the river's flooding, and to manage the natural hazard of flooding on that land, those terms needed an interpretation that meant this type of land was within the "bed". That would result in resource consents being needed from regional councils for most activities, including those involved with farming and residential activities, within areas on river terraces affected by flooding, without the need for specific rules requiring this. The District Court had agreed with this and had found that the "fullest flow" was a big flood and the "banks" were raised features that stuck out of those floodwaters (generally on the outer edges of the flood plains).

This put most floodplains, including areas on which people had farmed and lived for decades, within the "bed", making most of those activities illegal without a resource consent. That included a large part of the land belonging to the farmer who had done the works for which the regional council had brought the prosecution, including a big area that had been pasture for decades, on which no works had been done. He had successfully appealed the District Court's ruling in the High Court. ECan had appealed that decision to the Court of Appeal, arguing what it had in the District Court, which ECan said had been right. That appeal has been unsuccessful and the Court of Appeal has dismissed ECan's appeal. Unless ECan appeals the Court of Appeal's decision to the Supreme Court, the Court of Appeal's decision stands. Flood lines are not the tool to define the extent of the riverbed. A genuine "bank" now needs to be determined and then use the area that is covered by water without overtopping that bank to determine the extent of the bed. That approach, the Court of Appeal confirmed, is more in line with the ordinary meaning of the RMA definition.

The Regional Council has until 6 November 2019 to seek leave from the Supreme Court if it wants to appeal the decision to the Court of Appeal.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.