Measuring the impact of flooding on Amazonian trees: photosynthetic response models for ten species flooded by hydroelectric dams

• Published in: Trees (Berlin, West) West), 2013-02, Vol.27 (1), p.193-210
• Main Authors: dos Santos Junior, U. M, de Carvalho Gonçalves, J. F, Fearnside, Philip Martin
• Format: Article
• Language: English
• Subjects: Agriculture; Biomedical and Life Sciences; Climate change; Dam construction; Dams; dams (hydrology); Downstream; Flood plains; Flooding; Floodplains; Forestry; forests; leaves; Life Sciences; light intensity; Original Paper; Photosynthesis; physiological response; Physiological responses; Plant Anatomy/Development; Plant Pathology; Plant Physiology; Plant Sciences; Plant species; Rivers; Trees; Tropical environments; Upstream
• ISSN: 0931-1890; 1432-2285
• DOI: 10.1007/s00468-012-0788-2

Increasing areas of Amazonian forest are coming under flood stress due to dam construction and greater variability in river flood levels due to climate change. The physiological responses of Amazonian trees subjected to flooding are important to understand the consequences of these changes. Irradiance response curves for photosynthesis obtained from ten tropical tree species growing in flooded areas were used to fit three empirical models. The study was done in floodplains along the Uatumã River, both upstream and downstream of the Balbina Hydroelectric Dam in Brazil’s state of Amazonas (01°55′S; 59°28′W). Ten species were studied. Models compared were: non-rectangular hyperbola, rectangular hyperbola, and exponential. All models were quantitatively adequate for fitting the response of measured data on photosynthesis to irradiance for all ten species in the non-flooding and flooding periods. Considerable variation was found among the model estimates of maximum photosynthesis (P ₙₘₐₓ), dark respiration (R d) and apparent quantum yield of photosynthesis (α). For photosynthesis, the two hyperbolas overestimated P ₙₘₐₓ while EXP presented more realistic values. For estimating R d, RH presented the most realistic values. To avoid unrealistic value estimates of R d, we recommend adding measured R d values to the regressions. The results suggest that the EXP model presented the most realistic P ₙₘₐₓ and α values, and, in spite of less accuracy in fitting photosynthetic irradiance curves than the RH model, it can be recommended for accessing the information used in photosynthetic irradiance curves for the leaves of tropical trees growing in Amazonian floodplains or in areas that are artificially flooded by dams.