Patterns and controls of aboveground net primary production in meadows of Patagonia. A remote sensing approach

IRISARRI, JGN; OESTERHELD, M; PARUELO JM; TEXEIRA MA

JOURNAL OF VEGETATION SCIENCE

WILEY-BLACKWELL PUBLISHING, INC

Año: 2012 vol. 23 p. 114 - 126

1100-9233

Abstract Questions: (1) Can aboveground net primary production (ANPP) of Patagonian meadows be estimated from remote sensing? (2) How does ANPP of Patagonian meadows change in space and time? Location: Northwestern Patagonia, meadows embedded in a steppe matrix. LAT range: -39º to -43º, LON range: -70º to -72º. Cover area: 85000 km2 Methods: For the first question, we contrasted field ANPP measurements with MODIS high spatial resolution (pixel size: 0.0625 km2) data and we developed a model which estimates the radiation use efficiency. For the second question, we applied the model to a six-year MODIS record for fourteen meadows whose physiognomic heterogeneity was known from previous work. Results: Regarding the first question up to 77% of the field based ANPP variation was accounted for by the absorbed photosynthetic radiation based on a linear transformation of the normalized difference vegetation index derived from MODIS data. Mean radiation use efficiency was 0.54 g-dry-matter.MJ-1. Regarding the second question, ANPP ranged between 610 and 1060 g.m-2.year-1, which represents between 3 and 5.3 times the ANPP of the surrounding arid and semiarid steppes. The interannual coefficient of variation of ANPP was 10%, which is higher than other systems of similar productivity, but much lower than the surrounding steppes (33%). At the level of the management units (paddock) ANPP spatial variations were mainly related to the proportion of prairies, a proxy for low topographic position in the landscape, and longitude, a proxy for precipitation. ANPP interannual variation was most related to latitude, a proxy for temperature. Conclusion: The model developed and tested can be used to infer ANPP from remote sensing data at a spatial resolution that allow one to detect the variability within meadows and management units. Variations at both the physiognomic unit and paddock level were associated with geographic patterns and topography. Meadows were 3 to 5 times more productive and less fluctuating than the nearby steppes. When compared with other ecosystems, their productivity was high, but more variable interannually, likely due to an exceptionally high variability of precipitation in Patagonia.