Pigment-based identification of ozone-damaged pine needles as a basis for spectral segregation of needle conditions.
ABSTRACT Air pollution affects large areas of forest, and field assessment of these effects is a costly, site-specific process. This paper establishes a biochemical basis for identifying ozone-damaged pine trees to facilitate efficient remote sensing assessment of air pollution damage. Several thousand live needles were collected from ponderosa pine (Pinus ponderosa) and Jeffrey pine (P. jeffreyi) trees at three sites in Plumas National Forest and Sequoia-Kings Canyon National Park. These needles were assembled into 504 samples (based on the abaxial surface) and grouped according to five dominant needle conditions (green, winter fleck, sucking insect damage, scale insect damage, and ozone damage) and a random mixture of needles. Pigment concentrations per unit needle area of chlorophyll a, chlorophyll b, and total carotenoids were measured. The following pigment concentration ratios were calculated for all samples: chlorophyll a/total carotenoids, chlorophyll b/total carotenoids, total chlorophyll/carotenoids, chlorophyll a/chlorophyll b. The group of ozone-damaged needles had significantly lower mean pigment concentrations (family-wise p < 0.01) and significantly lower mean chlorophyll a/total carotenoid and total chlorophyll/total carotenoid ratios (family-wise p < 0.01) than all other groups of needles. Ozone-damaged needles had a significantly lower mean chlorophyll a/chlorophyll b ratio than all other groups except one (family-wise p < 0.01). Linear discriminant analysis with three factors (chlorophyll a concentration, the chlorophyll a/carotenoid ratio, and the chlorophyll a/chlorophyll b ratio) and subsequent maximum likelihood classification of damaged and non-damaged needles gave an overall cross-validated accuracy of 96%. These ozone-damaged needles are biochemically unique in relation to other needle conditions in this study, and further research is needed to generalize these results.
- Remote Sensing of Environment 12/2013; 139:138-148. · 4.77 Impact Factor