August 5, 1985
Black Rock Forest Inventory 1985
Kathleen Stearns Friday and James Boyd Friday
PROCEDURE
Original Stand Classification: The University of Massachusetts prepared a stand map of the forest from 1981 aerial photographs without field checking. One hundred and forty-three stands were delineated and classified under 26 different cover types. Timber cover types were differentiated by canopy height, by 20-foot intervals; canopy cover, less than or greater than 80%; and composition (>80% hardwood, 80-50% hardwood, 50-80% softwood, or >80% softwood). Presumably stand composition was defined by canopy cover, so a "mostly softwood" stand could mean understory hemlock provided a great deal of cover, rather than that softwoods actually predominated in density, basal area or volume. Other cover types included "clearcut/open" and four wetland types. (For details, see Table I). Areas by stand and compartment were also given to the nearest acre. Comparisons of this mapping with forest records show that cuttings, treatments and disturbances were seldom reflected in the stand map.
Sampling Unit Reclassification: As the inventory proceeded, it became evident that the photo interpretation had frequently mislabeled the stands. Once it was apparent that this had happened often, all timber (non-wetland) stands were visited and field checked before plotting sample points. This involved walking the length of the stand and stopping to observe heights and cover every chain. Heights were estimated visually and softwood cover was estimated by basal area. The stand was then reclassified into the appropriate category. After walking along several ridgetops, it was decided that any H4 stand (hardwood 60 to 80 feet) marked on a hilltop or ridgetop should be reduced to H3 (hardwood 40 to 60 feet). The trees on hilltops visited were invariable short and scrubby. To avoid confusion between the originally classified stands and the reclassification system, the reclassified areas are called "sampling units" or "units" on the data sheets and in this report. Usually several small stands were grouped together into a sampling unit when reclassified. Because aerial photos were not available for examination and because it was not feasible to do a very thorough ground reconnaissance, stand boundaries were seldom redrawn. Stands were only subdivided into different units when a clear boundary could be seen, e. g. a road or a trail. Thus the map was greatly simplified. No attempt was made to break smaller units out of the overall H3A stand, as to do this without bias would have involved systematically field checking the entire stand and drawing new stand boundaries. Hence the background "H3A" classification (hardwoods, full canopy, height 40-60 feet) is actually very heterogenous, with many stands of tall or scrubby trees. Table II represents a list of what stands were reclassified and why. This reclassification attempt represents a great improvement over the original stratification in that it has eliminated most outright misclassifications. However, it is not sure that this attempt will result in significant differences between categories either, since the original mapped stand boundaries were used and merely recategorized and no smaller units were broken out of the overall H3A stand. Three recent clearcuts were stratified out. These were the cut at the top of Hulse road (C4), the cut on Sutherlaqnd Pond road just south of the pine plantations (C8), and the small deer habitat cut on Continental Road south of the Stone House (C23).
Plotting Points: Sample points within the stand were plotted according the the prescribed Harvard Forest method. One or several axes were drawn in each stand so as to best sample that stand. Thus the axes might resemble a straight line, a "Y", a "T", a "Z", or an "11" to sample irregular shapes or to skip over an inclusion. After dropping off one chain at each end of the transect, the length was totalled and divided by three to find the plot interval. A random number was then chosen from a random number table in Avery and Burkhardt (1983) and multiplied by the plot interval to find the distance to the first plot. The next two plots followed at the plot interval going north to south. The nearest landmark to any one of the plots was then chosen and a bearing and distance to the plot measured. The other two plots might be accessed from the first one visited or from other landmarks. Landmarks most frequently used were road and trail junctions and property boundary corners. Where a starting point was less well defined, for example a given distance along a road, a tree at that starting point was blazed with orange (rarely red) paint. Thirteen degrees magnetic declination was used. Table III gives a list of all the plots and directions to them. In relocating the plots, it would be wise to follow these directions precisely and not try to find a plot from another starting point. Unmapped bends in trails and roads may throw off plot location slightly, and another set of directions may appear correct on the map but miss the plot altogether.
Visiting Plots: Distances from the landmark to the plot were paced. At the beginning of the summer pacing was calibrated by the use of a hip-chain, a distance measuring device. Throughout the summer, the hip-chain was used when the terrain was very steep or rugged or the distances were very long and it was thought that pacing would be less accurate. If a plot was more than 12 chains from a landmark, a tree was flagged with orange (rarely red) paint at the 10-chain (and 20-chain) interval. A 2 foot by 3/4 inch pvc pipe was driven at each plot center. This was marked with the number of the compartment (the original stand number not our reclassified stand), and the plot number on that transect. A few plots fell directly on rock ledges, and on these a circle with a dot inside was painted and plot stake driven nearby. These are noted on the data sheets and on the plot stakes. A good-sized tree near the plot center was blazed with orange (rarely red) paint unless the plot was easily visible.
Tree tally: Trees greater than 2" DBH were tallied using a 10 basal area factor prism. Borderline trees, usually 2 or 3 per plot, were checked using a 100' tape and a table of horizontal limiting distances from Avery and Burkhart (1983). For each tree, species, diameter at breast height, number of eight foot pieces, overall form, crown class, and any special notes were recorded. Species were recorded according to Petrides (1972). The species list published by Raup (1938) was also consulted, but it was thought that Petrides had more modern names. Refer to Table IV for a list of the tree, shrub, and herbaceous species encountered. The abbreviation used was the first two letters of the genus and the first two letters of the species. Thus Quercus rubra became QURU. Diameter at breast height was measured to the nearest 0.1 inch according to the rules described in Avery and Burkhart (1983). The most important of these are that DBH is measured on the uphill side of the tree and that it is measured 3.5 feet above the fork if the tree forks below breast height. The "fork" is defined at the lowest point at a crotch where a crease is visible, not where the bark of the two stems actually separates. Eight foot pieces of the bole were counted to a 4 inch top. Most hardwoods broke up into branches long before they reached this limit. While only bole and not branches were counted, both sides of a fork were considered bole if they rose vertically rather than horizontally. Thus, by forking, one tree in a stand might have 6 eight foot pieces while its companions of the same overall height had only 4. Height estimates were checked in the morning and after lunch using a Haga altimeter. Especially tall trees were also often checked. Tree forms were recorded as "G" (good), "P" (poor), or "C" (cull). Form itself was defined as being good if the tree had the potential to grow one 12 foot sawlog at or near the butt. If more than half of the 12 foot butt log (i. e. the first 6 feet) was defective, the tree was given poor form. Reasons for assigning poor form included sweep, crook, excessive branching, injury, and signs of heartrot. Leaning trees, if they had no other defects, were given good form. Trees which were too rotten to have value even for firewood were classified as culls. Crown classes were defined as follows: trees receiving sidelight were "d" (dominant), trees receiving toplight only (or no toplight but a lot of sidelight) were "I" (intermediate), and trees receiving no direct sunlight were "O" (overtopped). Once a tree was measured, the dbh line and tally number were painted on the side facing the plot center. Lastly, the trees were mapped on a circular grid to aid in relocating. Trees were numbered sequentially, proceeding clockwise and beginning with north. When a plot was close to the stand border, it was "mirrored" (Avery and Burkhart, 1983) so that trees outside the stand were not tallied but trees at the edge of the stand could possibly be tallied twice. This procedure was rarely necessary and no trees were ever tallied twice. At least once a large tree outside of the stand was not tallied although the prism indicated that it would be were it in the same stand.
Regeneration tally: All trees less than 2 inches dbh were tallied within a 2 meter radius plot. For a list of which small woody species were considered trees and which shrubs consult Table IV. When there were more than 100 seedlings of one species in a plot, the plot was divided into quadrats and two quadrats were sampled. This only happened three or four times and only with spruce and hemlock. Regeneration that reached breast height was tallied in a separate column and not included in first tally. It was thought that this would give a more realistic picture of what the state of the regeneration in the forest actually was, since almost all seedlings seemed to be browsed off year after year and never reached the sapling stage. Most softwoods did not seem to get beyond the cotyledon stage. To get a larger sample size, timber trees of greater than breast height but less than 2 inches dbh were also tallied in a plot extending from 2 to 4 meters radius (see Table IV for list of timber trees). The data of the first and second columns can be added to get total regeneration in a 2-meter plot; for timber species, the data of the second and third columns can be added to get the total large regeneration in a 4-meter plot.
Shrub and herb tally: Shrubs and herbs were also tallied in a 2 meter plot. The percent cover due to each species was recorded, or if less than 10% the species was marked present. Identification was done according to Newcomb (1977) and Petrides (1972). Latin names were abbreviated the same as with trees. Plants that were not able to be identified were called UNID. Some plants were identified only to genus. There was some confusion in the genus Vaccinium and between this genus and Gaylusssacia, and it would probably be better to lump the two groups together in analyzing the results. For a list of shrubs and herbaceous plants encountered, see Table IV.
Calculations: Board foot volume was calculated for trees of "Good" form, 9.0 inches DBH and larger (the "10-inch" DBH class included trees of DBH 9.0-10.9). For hardwoods, Table V was used (Forbes, 1955), and for softwoods, Table VI (Forbes, 1955). Volumes for chestnut oak were reduced by 12% to account for Girard form class. These tables were intended to be used with a 6" minimum log diameter, but out heights were seldom actually to a 4" diameter. Cordwood in the tops of sawtimber trees was computed with Table VII (provided by Dr. Gould). Again, the minimum log diameter was larger (8"), but because no other tables were available this one was used. Bole cordwood was calculated for trees of "Poor" form and trees 4.5 inches DBH (5-inch cordwood size class) to 8.9 inches DBH (below 10-inch sawtimber size class), using Table VIII (Beers, 1964).
