Hydroclimatic variability of the upper Nazas basin: Water management implications for the irrigated area of the Comarca Lagunera, Mexico

Earlywood ring-width chronologies derived primarily from Douglas-fir trees were used to reconstruct winter-spring (November-May) precipitation and fall-spring (September-June) streamflow volumes for the period 1765-1993 in the forested upper Nazas watershed in Durango, Mexico. The tree-ring data were obtained from mixed conifer stands within or adjacent to the upper Nazas watershed. Precipitation data were derived from one of the longest regional records. The streamflow data were obtained from a guage located in the upper Nazas watershed. The Principal Component 1 (PC1) of nine residual earlywood chronologies accounted for 73% of the variance in November-May precipitation 1967-1993, and 64% for the total period with available data 1941-1993. The mean of three residual earlywood chronologies from Douglas-fir explained 51% of the normalized streamflow at Sardinas from 1971-1992. The 20th century was characterized by severe droughts, especially between 1950 and 1963 that also affected other regions of Mexico and the southwestern United States. Additional droughts of similar or greater magnitude occurred in the 1790s, 1810s, 1860-1870s and 1890-1910s. Similar periods of low flow occur in the September-June streamflow reconstruction between 1765 and 1993. These results indicate that tree-ring chronologies from this region document a high percentage of the precipitation and streamflow variance. Spectral analysis detected significant high periodicities in both records at peaks of 4 and 7 years that could be related to the ENSO frequency bands (approximately 4.0 and 6.25 years). Analysis of the reconstructed records show strong influence of ENSO on precipitation and streamflow amounts on an interannual basis. These results can provide significant inputs to decisions regarding management of water resources that are used to irrigate land in the Comarca Lagunera: specifically they indicate that water budgeting should be managed over longer time periods to account for this ENSO-related variability rather than on the year-to-year basis that is presently used.