| Biomass and nutrients removed from Watershed 5 in the whole-tree harvest: |
What follows is an explanation of how we estimated the total biomass and nutrients removed from W5 in the experimental whole-tree harvest. Removal of biomass from the watershed was not complete, and we needed to account for what was left when the harvest operation was over. These estimates of the biomass remaining were then subtracted from the total estimated biomass on W5 before the harvest began to arrive at an estimation of the total aboveground biomass removed from the watershed.1. Summary of aboveground biomass removed
2. Using biomass equations
3. Pre-cut biomass
4. Biomass left on the watershed
a. Leaves
b. Edge trees
c. Felled trees not removed
d. Slash
e. What happened to dead trees and small trees?
5. Nutrients removed
Aboveground biomass removed from W5: (leaves are excluded in these estimates of aboveground biomass) Dry tons # 25 x 25 m (Mg) grid units # hectares --------------------------------------------------------------------- 1. Total W5 pre- cut biomass 1983* 4397.18 360 22.5 2. Total W5 left as living edge trees - 346.30 73 - 1.75 (38.3% of the grid units, see text below) ------------------------------------------------------------------- total cut/killed 4050.88 20.75 3. Total W5 cut & left (not skidded) - 163.74 19 - 0.84 (71% of the grid units, see text below) -------------------------------------------------------------------- 3887.14 Mg from 19.91 ha 4. Dead branches left as slash - 46.99 -------------------------------------------------------------------- total removed 3840.15 Mg from 19.91 ha * This is "potentially harvestable" biomass, which includes only live trees >10-cm dbh. |
Web page created June 2000
Our methodology for estimating biomass is based on 30 years of work with the Hubbard Brook allometric equations developed by Whittaker et al. (1974). Whittaker et al. published biomass and nutrient estimates for Watershed 6 for the year 1965, as well as selected equations for various species and plant components at different elevations. They left no detailed instructions indicating exactly which equations to use to arrive at their W6 estimates, and it should be pointed out that none of the equations published by Whittaker et. al. were actually used by them in the calculation of the biomass of W6. For instance, they used the equations for elevational "thirds", yet they published pooled "low + mid" thirds equations. For the higher third, they published only the pooled equations for the hardwoods while they used the individual species equations in the calculations. We have access to all the equations which they prepared (unpublished), and extensive study of these in relation to the published biomass equations has allowed us to derive conclusions as to which ones they actually used for W6. Through these efforts, we have been able to reproduce the W6 biomass estimates published by Whittaker et al. to within a few percent of what they reported.
In the estimates that follow, we have used the same combination of equations for W5 that we believe Whittaker et al. used for the W6 biomass estimates. Even so, there were innumerable points along the way in which arbitrary decisions were made in proceeding with the calculations, and thus there are numerous ways in which we could have arrived at a best estimate of how much biomass was removed from W5.
One exception to our strict use of the Whittaker et al. equations is balsam fir bole wood and bark. We completed a new dimension analysis for the wood and bark of balsam fir in 1985 using trees at the top of W5, and calculated equations for these components of this species which Whittaker et al. did not do in 1965. Since fir is more abundant than spruce at these upper elevations, it seemed appropriate to make these further estimates rather than use the spruce equations to estimate fir biomass. We did not make new equations for branches or foliage of fir, but used spruce equations to make these estimates as Whittaker et al. did.
Eventually we hope to make the following available: the original 1982 W5 forest inventory data (species and dbh), the allometric equations used, tree height data and equations, inventory data for edge trees left, the chemistry data used, and the computer programs (BASICA) used.
In order to arrive at an estimate of the total aboveground biomass on W5 just before the harvest, we used the pre-cut inventory taken one and a half years earlier, but had to make an adjustment to account for biomass accrued in the intervening time. The pre-cut inventory was done during the summer of 1982 and is assumed to have been done half-way through the diameter growth growing season (June-July). Cutting and removal of the trees started in October of 1983 and continued into the spring of 1984, before growth started. Therefore, the last season of growth for all trees cut was the summer of 1983. Thus the harvest took place 1.5 growing seasons after the inventory--half of the 1982 season and all of the 1983 season.
To account for the biomass accrued in the 1.5 growing seasons since the W5 inventory, we used longer term data we had for Watershed 6 (the Hubbard Brook reference watershed). W6 is immediately to the west of W5 and was presumably very similar to W5 before the harvest. The estimated W6 biomass was calculated from inventory data for the years 1977 and 1982. From these data it was possible to estimate an average annual biomass increment on W6 in the years before the W5 harvest--2.98 Mg/ha/yr. This average increment was then applied to W5 for 1.5 years of growth and added to the W5 1982 biomass estimate, to arrive at a total aboveground biomass estimate. These calculations for W6 and W5 are outlined in the two tables below.
Note: These biomass estimates are for "potentially harvestable" biomass. They include only live trees >10-cm dbh since dead trees and small trees were expected to be left on the watershed (see What happened to dead trees. . .). Also, leaves have been excluded from the estimates since they were not removed with the hardwoods, nor with most of the conifers (see Leaves).
Watershed 6 forest biomass increment data
for live trees >10-cm dbh:
Plant part Total biomass (Mg/ha) Biomass accrued (Mg/ha) 1977 1982 in 5 yrs per year Bole wood 106.08 114.60 8.52 1.70 Bole bark 11.22 12.08 0.86 0.17 Branches 48.43 53.57 5.14 1.03 Dead branches 4.45 4.95 0.50 0.10 Leaves & twigs* [3.46] [3.65] Twigs only 0.170 0.179 0.009 0.002 Total** 14.849 2.98 * twigs = 4.95% of leaves & twigs
** totals do not add exactly due to rounding error
Watershed 5 total pre-cut aboveground biomass
for live trees >10-cm dbh:
Plant part Biomass
(Mg)Bole wood 22.5 hectares * 119.92 Mg/ha in 1982 = 2698.2 Bole bark 22.5 * 12.93 = 290.93 Branches 22.5 * 53.20 = 1197.0 Dead branches 22.5 * 4.72 = 106.2 Twigs 22.5 * 0.190 = 4.27 Total at time of inventory 4296.60 Added biomass in 1.5 years 2.98 Mg/ha/yr * 1.5 years * 22.5 ha = 100.58 Total aboveground biomass at the time of harvest 4397.18
4. Biomass left on the watershed
For various reasons, some components of the total aboveground biomass were left behind on the watershed. These components were measured and subtracted from the total pre-cut biomass to arrive at a total amount of biomass removed from W5 as outlined in the summary table. The measurements and calculations for each component are explained below.
Since the logging was done in the winter, no leaves were removed with the hardwoods. Most of the conifers were also removed without needles. Due to problems with the contractor, the felled trees on much of the upper third of the watershed were not removed until the spring of 1985. Thus many of the conifers were left lying for a year and had lost their needles by the time they were finally removed. Any needles left after that time probably fell off in the process of being dragged off the watershed. There were so few conifers on the rest of the watershed that it was not worth the effort to modify this part of the calculations to account for the differences. As a result, all of the biomass estimates were done without including leaves.
In order to keep a forest buffer strip between the harvested Watershed 5 and Watersheds 6 and 4, trees on portions of 73 grid units around the edge of W5 were not cut and continue to grow (see map of grid units). The diameters of all of these trees were measured after the harvest and before growth started in the spring of 1984. The biomass of these residual buffer trees was estimated and subtracted from the total aboveground biomass estimates in the summary table. Since nutrients are estimated on a per hectare basis, we also needed to calculate the area these standing trees represented. This was done in the following steps:
1. The total biomass for the edge trees left standing on these grid units after the harvest was calculated based on the inventory of these grid units done in the spring of 1984.These calculations are summarized in the following table:2. The total 1982 live biomass for trees >10-cm dbh on the 73 grid units was calculated from the pre-cut inventory data. This estimate was increased by 1.5 years using the estimated annual biomass increment (as calculated above based on W6 data) to reflect the estimated biomass which existed on these grid units at the time of the harvest.
3. The ratio of biomass left standing to biomass at the time of harvest was then used as an estimate of the proportion of the area of the 73 grid units that was left forested.
Biomass estimates for edge trees left standing on W5
for live trees >10-cm dbh:
Plant part Biomass in 73 edge grid units (Mg) in 1982 post-harvest Bole wood 554.02 216.86 Bole bark 60.82 23.88 Branches 245.37 96.72 Dead branches 21.72 8.49 Twigs 0.91 0.35 Total 882.84 346.30 Biomass added in 1.5 years + 20.32 Total estimated biomass at time of harvest 903.16 Biomass left standing / total biomass = 346.30 / 903.16 = 0.383 = 38.3%
Total area of 73 grid units = 4.56 ha
Area of 73 grid units left forested = 4.56 * 0.383 = 1.75 ha
Trees on portions of 19 grid units in an area of steep slopes were cut and left on the watershed because the logger could not get equipment in to pull them out (see map of grid units). After the harvest, each of these 19 grid units was evaluated by Wayne Martin in terms of the proportion of the grid unit from which the trees were removed. The biomass of these felled, inaccessible portions of the grid units was estimated and subtracted from the total aboveground biomass estimates in the summary table. As with the edge trees, since nutrients are estimated on a per hectare basis, we also needed to calculate the area in hectares these residual trees represented. This was done in the following steps:
1. The total 1982 live biomass for trees >10-cm dbh on the 19 grid units was calculated from the pre-cut inventory data. This estimate was increased by 1.5 years using the estimated annual biomass increment (as calculated above based on W6 data) to reflect the estimated biomass which existed on these grid units at the time of the harvest.These calculations are summarized in the following table:2. When the harvest operation was complete, it was determined that within the 19 grid units an area equivalent to 13.5 grid units (or 71%) was felled but the trees not removed. Therefore, 71% of the estimated total biomass (from the first step) for the 19 grid units was left on the watershed.
3. The area in hectares of felled trees left on the watershed was calculated using the same percentage.
Biomass estimates for felled trees not removed from W5
for live trees >10-cm dbh:
Plant part Biomass in 19 steep grid units in 1982
(Mg)Bole wood 138.96 Bole bark 14.875 Branches 66.628 Dead branches 4.641 Twigs 0.2134 Total 225.317 Biomass added in 1.5 years + 5.30 Total estimated biomass at time of harvest 230.62 Proportion of grid units felled but not cleared = 13.5 / 19 = 0.71 = 71%
Biomass of felled trees left on watershed = 230.62 * 0.71 = 163.74 Mg
Total area of 19 grid units = 1.19 ha
Area of 19 grid units with felled trees not removed = 1.19 * 0.71 = 0.84 ha
Of the trees that were cut and removed from the watershed, not all of the calculated biomass of each tree was removed. In any harvest operation, many of the dead and smaller branches are left behind. In order to estimate how much slash was left on W5, we conducted a survey to determine the weight of woody material in several size classes that was left on the watershed. The survey was conducted in the fall of 1984, while the wood was still fresh enough to distinguish remnants of living trees from older dead material. The methods and results of this slash study are summarized separately, but we estimated an average of 19.2 Mg of slash per hectare.
Of the slash present on the watershed, we must make some assumptions about where it came from. The table below outlines the total biomass in 1982 for live and dead trees in two size classes. (We did not bother to add 1.5 years of growth to the live trees because the extra biomass would be too small to worry about in this rather rough estimate of slash.) We are somewhat arbitrarily attributing the 19.2 Mg/ha of slash to all the dead trees, all the trees 2 to 9-cm dbh, and half of the dead branches on the live trees >10-cm dbh. Undoubtedly some of the live 2 to 9-cm dbh trees were removed from the watershed as well as some dead trees, but many of the live branches from the larger trees were broken off and left. We are assuming these cancel out in terms of a total estimate of slash that was left on the watershed.
Note: The original description of slash collection states that the wood must appear "alive" to be collected. As no one can remember exactly what happened, we are assuming that all material, live and dead, that had obviously just fallen as a result of the harvest was collected. We assume the "alive" criterion only applied in the few cases where it was hard to distinguish between wood that was down before the harvest and wood that came down during the harvest.
Biomass attributable to slash on W5:
Forest component Biomass (Mg/ha) Total aboveground in 1982 Slash in 1984* Live trees >10 cm dbh 190.96 [Dead branches] [4.72] 2.36 Dead trees >10 cm dbh 10.85 10.85 Live trees 2 - 9 cm dbh 5.40 5.40 Dead trees 2 - 9 cm dbh 0.60 0.60 Total 19.21 * The slash collected in sample plots amounted to an average of 19.2 Mg/ha. We are attributing the slash collected to the forest components on the left in the amounts listed in the column on the right, i.e. all dead trees, all small trees, and half of the dead branches of large live trees (4.72 * 0.5 = 2.36).
Since dead trees and small trees were in theory to be left on the watershed as slash, the starting point for total aboveground biomass that was potentially harvestable included only the live trees >10-cm dbh. For this reason we only need to subtract the portion of slash assumed to be from the large live trees--half of the dead branches. The calculation is summarized below.
Biomass estimate for dead branches left as slash on W5
for live trees >10-cm dbh:4.72 Mg dead branches/hectare * 0.5 * 19.91 hectares =
46.99 Mg of biomass left as slash
4e. What happened to dead trees and small trees?
In theory, dead trees and smaller trees, 2 to 9-cm dbh, were to be left on the watershed. The logger was asked to fell these trees, but it was not expected that he would remove them since they were not likely to be merchantable. They were indeed felled, as is obvious in the "clearcut" look of W5 in post-harvest photographs. However, we do not know how many of these trees the logger chose to take and, thus, were removed from the watershed.
Certainly some of the more recently dead trees were deemed salvageable, but we can not make a determination from the 1982 survey as to which trees were more recently dead and might have been taken. In that survey, dead trees made up about 15% of the basal area and density, and were categorized as either a snag (stub) or a standing dead tree. Standing dead trees were standing with limbs and generally more recently dead than snags. A snag was a dead tree which had broken off somewhere up in the stem above breast height and had no remaining branches of any significance. Many of these were standing dead that finally rotted enough to lose their top, but some were recently alive and broken off by wind or ice. Thus, it is impossible to tell from its category how recently each tree died.
It is apparent from photographs that many trees less than 10-cm dbh were removed from the watershed. Some are visible in the piles at the landing in photographs of the chipping operation. However, we have not come up with a method to estimate what percentage of small trees were taken.
For the sake of calculating the amount of biomass removed from the watershed, we are assuming that all the dead and small trees were left. The pre-cut biomass of these trees was never included in the estimate of total aboveground biomass potentially harvested. Fortunately, we do have a good estimate of how much slash was left in total on the watershed. So, for a simple estimate of the total biomass removed, the exact amount of slash left as dead and small trees rather than branches of large trees is not so important. The biomass of dead and small trees was removed from the slash estimate, and the amount of slash remaining was attributed to live trees >10-cm dbh and subtracted from the total pre-cut biomass estimate (see Slash).
We have a standardized computer program for estimating total nutrient amounts based on biomass estimates by plant part and by species. Nutrient concentrations of individual plant parts are those reported by Whittaker et al. (1974) and Likens and Bormann (1970). To estimate the nutrients removed from W5, we modified the program to reduce the dead branch mass by one half to account for the dead branches we believe were left on the watershed as slash (see Slash). Then at the end of the program, we increased the nutrient estimate by 1.16 % , which is the best estimate of the increase in biomass in the 1.5 years from the 1982 survey until the time of harvest. The output of this program gave us nutrients on a per hectare basis, which we multiplied by 19.91 hectares* to get total the total nutrient masses removed. We plan to eventually provide a detailed presentation (computer programs and data) of the nutrient data by plant part, including the methods we used in making the estimations of biomass and nutrient mass.
* An estimated 19.91 hectares was cleared in total after accounting for edge trees and felled trees not removed (see summary table).
Nutrients removed from W5:
Nutrient Total nutrients Total nutrients removed
(from 19.91 ha)kg/ha moles/ha kg moles Calcium 525.68 13115.72 10466 261134 Magnesium 43.29 1780.06 862 35441 Potassium 170.10 4350.51 3387 86619 Phosphorus 43.34 1399.32 863 27860 Sulfur 55.15 1719.90 1098 34243 Nitrogen 434.51 31018.68 8651 617582 Sodium 2.24 102.00 45 2031 Iron 4.74 84.79 94 1688 Zinc 6.92 105.83 138 2107 Copper 0.54 8.52 11 170 Manganese 46.88 853.29 933 16989