The dendroclimatic potential of conifers from northern Pakistan

Abstract

A collection of 28 tree-ring chronologies from six different species located in northern Pakistan were evaluated in terms of their potential for dendroclimatic reconstructions. 15 of the sites are new while the remaining 13 (all Juniperus excelsa M. Bieb.) have been reported earlier. Several species had trees attaining ages of around 700 years ( Cedrus deodara (D. Don) G. Don, Pinus gerardiana Wall. ex D. Don., Pinus wallichiana A.B. Jacks and Picea smithiana (Wall.) Boiss.) but the juniper was clearly the oldest with some trees greater than 1000 years. Correlations between the site chronologies declined with increasing separation distance. This was consistently seen both between sites of the same species and between sites composed of different species. This led to a situation where a much stronger correlation occurred between two different species growing at the same site than between sites of the same species but separated by as little as 0.5 km. Such results highlight the obvious strong elevational gradients present in this mountainous region (where some elevations are over 7000 m). They also lend support to the practice of multi-species combinations for better spatial and temporal coverage. The best prospects for this appear to be C. deodara and P. gerardiana and are consistent with studies from neighbouring India. The comparison to 0.5° gridded climate data was strongest from the same two species though P. smithiana at one site was also highly significant. A general climate correlation pattern from all species was evident that starts with a strong negative relationship to temperature in the previous October, then turns towards positive during winter, before again becoming significantly negative by the current May. The previous October signal is thought to be a lag effect where hot temperatures (and low soil-moisture) stress the trees, thereby reducing reserves available for the following spring. Similarly, hot temperatures in late spring (May) lead to greater soil moisture losses and tree transpiration costs. Conversely, there is an extended strong positive precipitation correlation from late winter to spring (January–May). This ends abruptly and there is no evidence of a summer (June–September) monsoon signal seen in the rainfall correlation functions.