Abstract

The authors requested a poster

View All Abstracts

Climate-forcing of tree-growth across the forest-tundra of northeastern Manitoba

Mamet, Steven D. 1 ; Kershaw, G. Peter 2

1 University of Alberta
2 University of Alberta

The effects of climate and annular modes on radial-growth of Larix laricina (Du Roi) K. Koch, Picea glauca (Moench) Voss and Picea mariana (Mill) BSP were investigated in northern Manitoba. The objectives were to characterize spatial and temporal variations in growth of trees, and assess viability of the ring-width record for Churchill temperature reconstruction. Along a 250 km transect 220 trees were sampled in three zones - northern forest (NFOR), central tundra (TUND) and southern forest (SFOR) (Fig. 1). Standardized ring-width indices were compared with climate and annular modes through response-function analyses. Numerous years of suppressed growth at the three locales resulted from cooling following major volcanic eruptions. June-August temperatures were most limiting, particularly for L. laricina and were modeled using linear regression (Fig. 2). NFOR and TUND trees were most temperature-sensitive due to proximity to Hudson Bay, where near-perennial ice-cover results in pronounced cooling. Spring temperatures were detrimental to growth in NFOR and beneficial in TUND, due to snowpack dynamics. Growth decoupled with temperature during the 1950-1980 period due to atmospheric circulation shifts. NFOR and TUND chronologies were inter-correlated suggesting common environmental forcing, and different influences in SFOR (particularly for P. mariana), the source of which have yet to be determined.

Acknowledgements: This study was supported by grants/contracts from the Northern Scientific Training Program (NSTP) administered through Canadian Circumpolar Institute (CCI), Wapusk National Park, and the Churchill Northern Studies Centre (CNSC) - Northern Research Fund. We thank the staff and volunteers of the CNSC and the Earthwatch volunteers for their dedication and contributions to this research, as well as the comments of D. Froese and V. Lieffers on this manuscript.

 

Fig 1. Locations of tree core sampling from three areas: NFOR, TUND and SFOR, abbreviated as N, T and S respectively

 

Fig 2. Actual (black) and the modeled (blue) summer temperatures (June-August) based on standardized ring-width chronologies at each site and climatic data for the Churchill area from 1929-2002. (a, d, g) NFOR, (b, e, h) TUND and (c and f) SFOR. Species included were *P. glauca* (Pg), *P. mariana* (Pm) and *L. laricina* (Ll). Pearson correlation coefficients for the calibration (rc) and verification (rv) periods are shown (*p<0.05 and **p<0.01 significance).