Introduction
Last year, while rummaging around the dusty map drawers in old Marsh Hall, a couple of student found some hand drawn maps from 1942 which were created by a masters student, R. Martin. The maps plot out transects in stands of hemlock mixed with hardwoods at Totoket Mt. in the North Brandford Water Supply property. The students presented the maps to Professor Tom Siccama who was delighted by their findings. Tom shared his gift with the spring 98 class and challenged his students to compare the old data in Compartment 87 shown on the maps with a new survey in the same area to be done by the students.John and I were intrigued by the concept to develop a project of comparing this new found data from the faded 1942 maps and data sheets to current surveys with the assistance of computerized GIS, spread sheet programs, and more accurate quadrant sampling techniques as a part of the Methods of Ecosystem Analysis class. What made this project even more interesting is hemlock (Tsuga canadensis) is on the decline due to the Woody adelgid disease.
According to Mr. R. Martin's report, in 1942 Compartment 87 had just been selectively logged of primarily hemlock (239,869 board feet), some oaks (59,089 board feet), and other wood types (7,295 board feet) to maintain an even age distribution of hemlocks for routine harvesting. Prior to 1942, the lot had not been heavily logged since at least 1860, and was lightly logged around 1920. The compartment was surveyed by Mr. Martin following the logging of 42 by cruising the area in 100 ft. interval transects running in an east-west direction. On these lines circular 1/8 acres plots were taken at intervals of 100 ft. The data was summarized in tables by number of trees, which were categorized into wood types (Hemlocks, Good Hardwoods, Other Hardwoods, Sugar Maples, Oaks and Chestnut Oaks), in one inch interval diameter at breast height (dbh) size classes. We attempted to follow approximately the same pattern of transects as Mr. Martin at a lesser frequency interval due to our time limitations. It is assumed that compartment 87 has not undergone any major harvests following Mr. Martin's surveys in 1942.
Our data shows the composition of the forest stand in 1998 compared to 1942 to be relatively similar (1998 hemlock relative density/ha = 0.43; 1992 hemlock relative density/ha = 0.45). However the overall density is much higher now (1998 total tree density/ha = 439.3) compared to after the stand was logged in 1942 (1942 total stem density/ha = 249.4) showing some
Method
For the sake of efficiency, the area was sampled using the plot quadrant method. The recommended minimum number of plots to obtain satisfactory results is 40 (Cottam and Curtis, 1956). The area of Compartment 87 (x hectares) was used to determine the spacing of plots required to obtain approximately 100 sampling points. Starting in the northernmost point of the area plots were located, by pacing, every 50 meters on east-west (magnetic bearings) transects 50 meters apart. From