Transmission error in the photometric estimation of leaf area.

A numnber of authors have recommended the use of photometric apparatus for determining leaf area, and at least 1 commercially-manufactured instrument is available. In principle the leaf is placed between a light source and a photocell or battery of photocells: the reduction in photocell output due to the presence of the leaf then gives a measure of leaf area. The transmission of light through the leaves is a source of error, and an attempt to minimize this is usually made by fitting a magenta filter between the leaf and the photocell to absorb any green light passisg through the leaf. Such a filtered system, calibrated with opaque standards, was found to give low values for leaf area when compared with planimetered outline drawings. Leaves of freesia, narcissus, balsam and wintergrow-n tomato and lettuce were all underestimated in area by 9, 10, 15. 18 and 30 % respectively. These results prompted a further analysis of this method of area measurement, and in particular, the spectral comlposition of the light received by the photocell. Figure 1 shows a curve for the output of a photocell, using tungsten light (assuming a filament temperature of K = 29000), and a photocell spectral sensitivity as given in the EEL Data Sheet Reference -No. 0002.1. The same figure shows the spectral transmi-iission of light through a letttuce leaf (with tranismission characteristics similar to the more translucenit winter-grown leaves mentioned above) as given by Rabideau et al. (3). This curve is similar to those obtained from leaves of a range of other species [Rabideau et al. (3), Yocum, Allen and Lemon (4), Gates (1) and Loomis (2)]. Also showin is the spectral transmission of a magenta filter, of the type recommended and used for leaf area photometers, as determined spectrol)hotomiietrically. From these curves it is possible to derive the comiiposition of the light activating the photocell in the presence or absence of the leaf (fig 2). It is clear that only light between the wvavelength 0.6 and 0.7 ,u is of importance. and that the leaf is by no miieans opaque to light of these wavelengths. In the example giveni only 84 % of this light is absorbed or reflected by the leaf. If the leaf w-ere assumed to be opaque. its area would be utnderestimated by 16 %. in reasonable agreement with the underestimiiate referred to earlier. It can also be concluded that -100