Feature

Thirsty Plants

Mixing trees and pasture works best when you understand the different needs of the plants.

Blair Miller, Graeme Buchan and Don Mead

"Agroforestry" is the combination of tree-growing with field crops. In New Zealand, this usually means combining tree-growing with pasture for stock, and offers a potentially profitable diversification for farmers. The long-term return from the tree crop is supplemented with shorter-term returns from animal production. Making it work effectively, however, is a complicated business, since the trees and understorey (pasture grasses) compete for the available soil moisture as well as nutrients. So success in agroforestry depends on achieving a balance between these competing requirements.

Our research at Lincoln University focuses on the competition for water, with the aim of developing guidelines or models which can help farmers optimise the design and management of agroforestry plantings. The project is a multi-party collaboration involving Lincoln University's Departments of Soil Science and Plant Science, and the Forest Research Institute, and is a part of larger research programme investigating dryland agroforestry, based at the university's six-hectare agroforestry trial block. Pinus radiata trees were planted in July 1990 and have been grown over seven different pasture species to gauge the comparative performances of the different combinations. For our water-use investigations we used two common pasture species: "Yatsyn" ryegrass (a cultivar of Lolium perenne cv), and "Wana" cocksfoot (Dactylis glomerata). Pasture was harvested for forage in the first three years and is now grazed regularly by sheep.

Blair Miller cautiously leans out to measure branch diameters so that the tree leaf area can be calculated.

Investigating Water Use

Since starting our water-use project, the character of the agroforestry trial has changed dramatically. Initially the site had 400 stems/hectare and only low-level pruning, while by the end of the project stocking had been reduced to 200 stems/hectare and high-level pruning up to about 5 metres had been completed. This type of manipulation of tree characteristics means continual alteration of the balance between tree and pasture water use.

One of the major results of the tree canopy is to redistribute the rainfall. Trees intercept rainfall and the free water trapped on foliage surfaces is then directly evaporated, but in our agroforestry system with widely spaced trees this canopy interception loss is relatively small. There is however a pronounced redistribution at ground level, with an increase in rainfall to the south-west of the tree, which is the dominant direction from which rainfall comes. Simultaneously, a rainshadow is created to the north-east, reducing moisture input and therefore limiting understorey transpiration and growth.

The pattern of solar radiation is also affected by the trees, and is altered by the removal and pruning of the tree canopies. The complicated shadow pattern that results intensifies the effect of the rainshadow on understorey pasture production, as the area to the north-east of the tree receives more radiation allowing the surface soil moisture to be depleted more rapidly than the south-west.

Soil moisture, rainfall interception, and tree and understorey transpiration have been monitored over two growing seasons to provide a substantial data set from which to develop a physically based water balance model for the system. Eventually such a model will be used to form the basis of a computer-based management model for agroforestry systems which takes into account the different water use requirements of radiata pine trees and typical pasture understoreys, such as ryegrass and cocksfoot.

Results to date have shown the marked effect of spatial differences in the receipt of rainfall and radiation on the water balance of the system. Also, marked differences have been found in the competitive abilities of cocksfoot and ryegrass -- tree transpiration rates are markedly less over cocksfoot than over ryegrass.

For example, tree transpiration over cocksfoot peaked at approximately 12 litres/day in February 1996, while trees growing over ryegrass which had very similar levels of available water peaked at 23 litres/day. These results were for a period when the stocking density was 400 stems/ha. During February 1997, when tree density was reduced to 200 stems/ha, a similar difference occurred with tree transpiration rates of 18 litres/day over cocksfoot, and 31 litres/day over ryegrass.

The greater competitiveness of the cocksfoot is probably due to its more vigorous rooting at depth, and also shifted the balance of above-ground dry matter production. Cocksfoot yields were higher than ryegrass, and correspondingly tree growth over the ryegrass plots was visibly (and measurably!) greater than over cocksfoot.

A successful agroforestry system fulfills the grower's initial objectives. This involves a decision on the trade-off between short-term gains from pasture and longer-term gains from trees. The more water the trees consume, the more you shift the yield away from the pasture. But, of course, those trees don't give you a return for many years. The ideal situation is to find complementary species or a complementary system design, so that increased tree production does not result in a significant reduction in pasture production.

Rain gauges below the trees provide an idea of how much water is intercepted and how much makes its way to the pasture below.

Photos: John McCombe, Photosouth Photographers

Dr Graeme Buchan is in the Department of Soil Science at Lincoln University.
Dr Don Mead is in the Department of Plant Science at Lincoln University.
Blair Miller is a PhD student in the Department of Soil Science at Lincoln University.