Yale University School of Forestry & Environmental Studies FES519B - Methods of Ecosystem Analysis 1998
 ABOVE-GROUND  BIOMASS  and NUTRIENT ESTIMATES at a MIXED DECIDUOUS  and  HEMLOCK-HARDWOOD FOREST TOTOKET MOUNTAIN,  NORTH BRANFORD, CT 1998

Photo credit: http://bluehen.ags.udel.edu/gopher-data2

The data presented below are the result of a term project for Professor Tom Siccama's FES 519b: Methods of Ecosystem Analysis class at the Yale University School of Forestry and Environmental Studies.  Field sampling was done from January-March 1998 at Totoket Mountain near North Branford, CT.  We are indebted to the Regional Water Authority for providing access to the site.

The results are based on field measurements of 101 plots, each 6 m x 30 m, arranged along two transects.  Plots were established to study the composition and structure of the 70-100 year old mixed deciduous and hemlock/hardwood forest at Totoket Mountain (click here to get more information on the forest composition and structure).  The "Uphill" transect, plots 1-51, ran in a north-west direction, and ranged in elevation from roughly 290 ft to 510 ft above sea level.  The "Countour" transect, plots 52-101, ran in a  north-east direction, and stayed as near as possible the 350 ft contour line.  The "Uphill" transect is considered more xeric, the "Contour" transect is more mesic.

Please note: Biomass results are expressed in Mg / ha, or metric tons per hectare.  1 Mg = 1,000 kg = 1,000,000 g

What's Below
On this page, you'll find the following information about the above-ground live tree biomass and above-ground nutrient estimates at Totoket Mountain, CT:
How are biomass estimates computed?

Allometry equations, usually in one of two forms,
biomass = a x dbh b
biomass = a + b log (dbh)
allow a reasonable estimation of the above-ground biomass to be made, based solely on knowledge of tree species and tree diameter (the equation coefficients, a and b, are different for different species).

The following figure illustrates how biomass estimates vary by species of the same diameter at breast height. One reason that different species trees of the same diameter have different amounts of above-ground biomass is that different species have different architectures.  Another reason is that the wood of some species is much more dense than that of other species: for example, hemlock has a wood density that is roughly one-half that of most hardwood species.

Based on data collected in the field at Totoket Mountain, we applied (where possible) the species-specific allometry equations contained in Tritton and Hornbeck's (1982) "Biomass Equations for Major Tree Species of the Northeast" (USDA Forest Service, Northeastern Forest Experiment Station, GTR NE-69).   Generally, Brenneman et al.'s (1978) equations for West Virginia hardwoods were applied.  In the cases where Tritton and Hornbeck did not give an equation for the species in question, then we applied the Monk et al. (1970) "General Hardwoods" equation, or the Monteith (1979) "General Softwoods" (see Tritton and Hornbeck (1982) for Monk and Monteith sources).

The numbers presented below are estimates: the validity of applying equations derived in West Viriginia to a site in Connecticut can be debated until the cows come home.

For comparative purposes, 1997 data from the 80 year old forest at Watershed 6, Hubbard Brook, NH, are provided where appropriate.

Biomass Estimates: 1993 vs. 1998

 Above-ground live tree biomass, Mg/ ha 1993 1998 Net % Change Totoket Mountain Uphill plots (#1-51) (xeric) 142 148 + 5% Contour plots (#51-101) (mesic) 148 162 + 9% Total 145 155 + 7% a 1997 Hubbard Brook 195

Based on the above, aboveground live biomass at Totoket Mountain increased at approximately 2,000 kg / ha annually over the five-year period 1993-1998.

Biomass Estimates: Major Tree Species

 Above-ground live tree biomass, Mg/ ha 1993 1998 Net %  Change % 1993  Total % 1998 Total Totoket Mountain Acer saccharum 34 37 +6% 23% 24% Quercus alba 35 36 +1% 24% 23% Carya sp. 14 15 +4% 10% 10% Quercus rubra 14 13 -6% 10% 8% Fraxinus americana 6 8 +34% 4% 5% a Hubbard Brook 1997 % 1997 Total Acer saccharum 71 36% Fagus grandifolia 62 32% Betula lenta 40 21%

QA / QC of Biomass Estimates

We wondered "does it matter whether we use species-specific allometric equations? Or do we get a pretty good estimate of total biomass by just using a general hardwoods equation?"

We compared our "best estimate" of total biomass (using species-specific equations) with  biomass calculated using

•    general hardwoods equation (> 10 cm dbh) (Harris et al. 1973)
•    general hardwoods equation (> 2 cm dbh) (Monk et al. 1970)
•    Acer saccharum equation (Brenneman et al. 1978)
We found that the Monk equation comes remarkably close to the "best estimate" of total biomass:

 Above-ground live tree biomass, Mg/ ha 1993 1998 Species-specific equations 144 155 Harris et al. 115 122 Monk et al. 147 156 Brenneman et al. 177 188

Nutrient Analysis
We estimated the total calcium and magnesium content of the aboveground biomass by applying plant tissue nutrient concentrations measured by Hunter and Siccama (1996) at the Quinnipiac River and by a previous FES 519b class at Sea Hill.  For each tree species we used an average of the available data.  For an entire tree, the amount of a given nutrient is a function of four factors:
• the nutrient concentration in wood (usually low)
• the nutrient concentration in bark (usually high)
• the total biomass of the tree (derived from allometric equations, as above)
• the bark biomass percentage
The fourth factor is somewhat problematic.  At Hubbard Brook, bole bark accounts for 4.0% of tree biomass for beech, 6.4% for yellow birch, and 7.8% for sugar maple (average 6.1% for the three species).  This does not take into account the amount of bark on the branches, however, so we did our best to adjust for this.  We believe that there is a higher percentage of bark in branches than in the bole, and therefore we thought it reasonable to add an additional four percentage points to the bole bark figures to arrive at a very approximate bark biomass percentage for the entire tree.  Therefore, we used 10% as a rough estimate of the total bark biomass percentage for all tree species.

Based on this estimate, we calculate the following nutrient totals in the aboveground live biomass at Totoket Mountain:

 kg / ha 1993 1998 Net %  Change Totoket Mountain Calcium 558 583 +4% Magnesium 34 37 +8% a Hubbard Brook 1997 Calcium 551 Magnesium 48

Nutrient levels at Hubbard Brook appear to be of the same order of magnitude as at Totoket Mountain.  Magnesium levels are quite a bit higher at Hubbard Brook, whereas Calcium levels are only a small amount higher at Totoket Mountain.

As mentioned above, these estimates are sensitive to the bark biomass percentage assumed.  The following two graphs show how estimated nutrient totals change as the bark biomass percentage is increased or decreased.

Biomass Modeling

As an experiment, we decided to try define a model to estimate the above-ground live tree biomass at Totoket Mountain five years from now (2003).   First, we checked the model using 1993 data to estimate 1998 data: the model performed quite well when the totals were compared with what we estimated based on this year's field measurements.  We then ran the model using the actual 1998 data to forecast 2003 data. We will leave it to a future FES 519b to check the validity of  these results!

In order for the model to work, we had to make a number of assumptions above mortality, tree growth, and regeneration.  These assumptions are as follows:

Basics of the Model

Mortality

• define an annual mortality rate: we assumed an average annual mortality of 1.0%
• randomly determine individual tree mortality
Tree growth
• each tree allowed to have its own growth rate
• growth rate a random variable
• trees not allowed to shrink in size
• average radial growth rate = 1.2 mm / year (click here to get to more tree ring information)
Biomass calculations
• determine which trees are still alive
• apply annual growth
• calculate biomass of still-living trees (allometric equations)
Estimate regeneration
• allow regeneration in the form of sapling ingrowth
• randomly determine the number of new trees
• average number of new trees = 130 / ha over five year period
• assume new trees have dbh = 5 cm
• calculate biomass accordingly

Model Results

Calibration test

• 1993 Total Live tree biomass: 145 Mg / ha

• Model run five times to predict 1998 total above-ground live tree biomass:

 Above-ground live tree biomass, Mg/ ha Established Trees Ingrowth Total Tree Biomass Run 1 158 0.6 159 Run 2 157 0.8 158 Run 3 158 0.7 158 Run 4 159 0.7 160 Run 5 155 0.9 156 Average 157 0.8 158 ± std dev 2 0.1 2

The average 1998 biomass predicted by the model is 158 Mg / ha, which is quite close to the estimated biomass of 155 Mg / ha.

Forecasting 2003 Biomass
• 1998 data used as base

• Model run five times to predict 2003 total above-ground live tree biomass:

 Above-ground live tree biomass, Mg/ ha Established Trees Ingrowth Total Tree Biomass Run 1 165 0.6 166 Run 2 167 0.4 167 Run 3 167 0.5 168 Run 4 170 0.8 171 Run 5 163 0.6 164 Average 166 0.6 167 ± std dev 2 0.1 2

The average 2003 biomass predicted by the model is 167 Mg / ha.  Hopefully someone will re-survey our transects in five years to see how close the model is to being right!

Model Sensitivity

The model is quite sensitive to the mortality rate and annual growth rate assumed.  On the other hand, the amount of ingrowth is so small as to be more or less insignificant.

The following figures compare the effect of changes in the level of annual mortality and annual growth rate on the forecasted biomass.

Species-by-species Biomass and Nutrient Data