Conservation Comment: Stream protection a murky reality

EFFECTIVE BUFFER: Riparian planting areas need to be at least 10m wide on each bank to have a beneficial effect on stream water quality, Mike Dickison writes.

By Mike Dickison

In June I spent all day standing in the Matarawa Stream on No3 Line, working with schoolkids who wanted to measure water flow and pollution levels.

The stream I was in was fenced, or rather "fenced": there was a single strand of electric wire, with the current turned off. The fence was only a foot high, and hoofprints on both sides of it showed that cows were freely wandering down to the water. In places it was only a metre from the bank, so wasn't protecting much of anything.

DairyNZ has proudly declared that 97 per cent of waterways on dairy farms are fenced, as part of a voluntary accord by farmers to counter the effects of intensive dairying. If you read the fine print, it's actually 97 per cent of "significant" waterways.

So what's a significant waterway? The rule of thumb is "Wider than a stride, deeper than a gumboot": that is, on average more than a metre wide and 30 cm deep. That's a sizeable stream -- most waterways on farms wouldn't qualify. For our native fishes, it's actually quite large; they prefer much smaller streams, and there's no requirement to keep stock out of those. By the time all the trickles and side-streams flow into the nearest "significant waterway", they're already full of mud and cow poo.

Another problem with voluntary fencing is it only applies to dairy cows. Beef cattle don't count, and much of the public outrage from seeing cattle standing in rivers is probably caused by herds of steers. But the fencing initiative also doesn't apply to dairy support pasture, so female calves and underage cows, as well as dried-off cows, can wander into streams and trash them. From the stream water quality point of view, cattle are cattle.

So the "97 per cent fenced" only means 97 per cent of some streams on some paddocks. But surely any fencing is good? Well, fences that cattle can't step over and that are a reasonable distance from the stream can stop soil and phosphates getting into the water. But the biggest problem is nitrogen from cow urine. Gallons of urine soak into the soil and gradually leach into the groundwater, which flows into streams -- and no amount of fencing can stop that. High nitrogen levels in water feed algal slime, which gobbles up oxygen at night, suffocates fish and chokes up streams.

Another thing farmers are urged to do is to plant the banks of streams -- riparian planting. Thousands upon thousands of (tax-deductible) trees have been planted, often heavily subsidised by regional councils, but do they do much good?

There have been plenty of studies showing riparian planting, like fencing, mostly keeps sediment and phosphorus out of the stream. Tree planting can do a lot of good by shading the stream, reducing slime growth, keeping down the temperature and sheltering it from wind.

But to work, it needs to be large -- at least a 10m wide strip of shrubs and trees on each bank, ideally with a high canopy that completely shades the stream. To remove nitrates you'd want a forested strip 20-30m deep. One study only found significant benefits with a 50m buffer -- something I've never seen done on a dairy farm. The DairyNZ guidelines recommend a 5m strip, which is just cosmetic: a pretty strip of trees but too narrow to do much good.

When I asked the kids back in June what they thought of the Matarawa Stream, they said it looked pretty clean and healthy. And true, it was running fairly clear and wasn't full of rubbish or shopping carts. But when we used nets to sample the animals living in it, we were pulling up worms and snails, not the stoneflies and mayflies that indicate a healthy stream. Looks can be deceptive. A well-fenced, well-planted waterway looks very nice. The only way to tell if it's actually healthy, though, is to go stand in it with a net and count bugs.

Dr Mike Dickison is Curator of Natural History at the Whanganui Regional Museum.