MEASURING LEAF-AREA INDEX IN A SPARSE EUCALYPT FOREST - A COMPARISON OF ESTIMATES FROM DIRECT MEASUREMENT, HEMISPHERICAL PHOTOGRAPHY, SUNLIGHT TRANSMITTANCE AND ALLOMETRIC REGRESSION

In the dry sclerophyll jarrah (Eucalyptus marginata Donn ex Sm.) fores ts of the south-west of Western Australia, measurements of leaf area i ndex (L) are required to study tree growth, water use and mine rehabil itation on typically sparse stands (L = 1-2). This study compared thre e indirect estimators of L-sunlight transmittance collected with the D EMON measurement system, hemispherical photography and allometric rela tionships-with the results of a direct measurement (L = 1.05). The est imate of leaf area from the allometric relationship was not significan tly different from the direct estimate (P > 0.20), overestimating L by 9%. The DEMON system and the hemispherical photography overestimated L by 58% and 73%, respectively. These overestimates reduced to 20% and 32%, respectively, when the projected area of stems and branches was also considered in the comparison. When logarithmic averaging of the l ight transmittance was used, the DEMON overestimated L by 132% and ove restimated the projected area of stems, branches and leaves by 77%. Wh ere seasonal or other small differences in L are studied, techniques b ased on light transmission may not detect differences in L that are sm aller than one. The results from this single plot are inconclusive, an d further studies of the performance of hemispherical photography and sunlight transmittance should be made under sparse canopies.