Extensive Studies of Fertilizing and Thinning Coastal Douglas-fir and Western Hemlock (EP 703)


EP703 is the largest experiment in the province, with 940 plots located at 85 installations established between 1971-1975 throughout coastal B.C. The goal of this study, also known as the Forest Productivity Study, was to create an extensive and comprehensive data base from which to direct future intensive forest management practices in coastal British Columbia. Extensive site (e.g. soil and weather) and ancillary (e.g. vegetation) data were also collected.


  1. To examine basal area and volume growth response of immature Douglas-fir and western hemlock to thinning and/or fertilizer applications under a variety of stand, edaphic and climatic conditions.
  2. To construct growth and yield tables for fertilized and thinned stands of immature Douglas-fir and western hemlock.
  3. To construct volume tables for fertilized and thinned stands of immature Douglas-fir and western hemlock.
  4. To determine levels of fertilization and/or thinning that provide optimal yield from immature Douglas-fir and western hemlock.


The core study was initially planned as a randomized complete block, to test three levels of thinning, including controls (T0, T1, and T2), and three levels of fertilization, including controls (F0, F1 and F2) in a complete 3 x 3 factorial arrangement replicated twice over the nine treatment combinations (18 plots). These thinning levels were removal of 0% (T0), 20% (T1) and 35% (T2) of the initial basal area. The fertilization levels were application of 0 (F0), 225 (F1), and 450 (F2) kg N/ha. Plots within an installation were ranked by preliminary basal area of the plot, and assigned to one of two blocks (“large basal area” and “small basal area”). A Thinning x Fertilization treatment combination was then assigned randomly to one plot from each block, thereby providing two plots (replicates) per treatment combination within an installation. Over time as large, homogenous forested areas become more difficult to find, the experimental design was modified such that the actual number of plots per installation ranged from 4 to 32, with a typical installation consisting of 18 plots.


Of the 85 installations: 42 remain active, 36 have been released and 7 are inactive.