The Camp Riley Cruise Experiment

Introduction & Comments

George Deegan sent along this report of some work that he did with the DNR.  He gave us permission to edit it slightly for the newsletter, but is essentially as he sent it to us.  It is a wonderful example of the trade-off you get when you try for more data (tree VBARs) by estimating heights vs. carefully measuring a smaller number of tree VBARs that balance the Basal Area estimate.  I remember seeing this same affect in cruises done in Michigan.

When you measure too many trees and have to estimate everything you have the worst of two processes.  First, you waste your time because the SE% for the VBAR is already too good compared to the SE% for the Basal Area, so the overall SE% hardly changes.  Second, you get a bias in all those measurements because the measurements were not careful.  I would suspect that the tree count with his larger (20 BAF) prism was probably more accurate as well.

It is true that having lots of measured tees is psychologically comforting – but it’s often a waste.  The data can tell you that.  George’s work here is a great example of practical research and analyzing the results of that work.  We were very pleased to receive it.  The relative difference in the cut-outs tells the story. 

                                                                                K.I.

Alternative Cruise Designs – Results of Two Comparisons

By George Deegan

The following discusses the results of two cruises where a traditional 10-BAF cruise was compared to alternative cruises using a 20-BAF factor and elements of “Big BAF“ and “Sub‑sampling” cruising.

Camp Ripley State Sale

An alternative cruise combining elements of "Big Basal Area Factor" and a “Sub‑sampling” 20‑BAF cruise method was compared against a traditional 10-BAF cruise on a 28.3 acre aspen tract in Camp Ripley.

• In the 10-BAF cruise, 46 plots were installed and a total of 520 trees were tallied for basal area.

1.         On every plot the number of 100 inch sticks per tree was ocularly estimated for each of the 322 Aspen, 60 Red Maple, 73 Red Oak, 36 Paper Birch, 8 Sugar Maple, and 16 White Pine trees tallied.  The mean Volume-to-Basal-Area-Ratio (VBAR) was then computed for each species.

• In the alternative cruise, 58 20-BAF plots were installed and a total of 333 trees were tallied for basal area.

1.         On every 5th plot, a 100-BAF sweep was also taken, and the number of sticks per tree was measured by clinometer on the 11 "in" Aspen trees to calculate VBAR.

2.         Also, on every 5th plot, the number of 100 inch sticks per tree was ocularly estimated for each of the 11 Red Maple, 8 Red Oak, 3 Paper Birch and 1 White Pine trees tallied on the 20-BAF sweep, and the mean Volume-to-Basal-Area-Ratio (VBAR) was computed for each species

The cruises were conducted March, 2005 by John Korzeniowski, Walker Wearne and George Deegan. The sale was harvested in September 2006 with Cut-to-Length equipment.

Results for Aspen and Red Maple are tabulated below:

Aspen

Red Maple

In this trial, the Big BAF/Sub-sampling method produced results closer to the actual scale from a smaller number of tree observations.  The disparity between cruise volumes is a product of the difference between VBARs and BA/acre estimates.

For Aspen, the BA/acre estimates were almost identical (70.6 sq.ft. vs. 70.7 sq.ft).  
The VBAR estimates varied by 6.7%.  The difference in the two cruise volumes can be attributed almost entirely to the greater average stick height found on the Big BAF/Sub‑sampling Cruise (measured vs. estimated lengths).

Note that the average stick height for Aspen in the Big BAF/Sub-sampling cruise was based solely on 11 observations.  However, each of these observations was made using a clinometer.  

In contrast, the average stick height of Aspen in the 10-BAF cruise was based on 322 observations.  However, each of these observations was based solely on an ocular estimate.

For Red Maple, the BA/acre estimates differed by 13.5% (13.04 sq.ft. vs. 14.08 sq.ft).  The VBAR estimates differed by an additional 3.4%.  The difference in the two cruise volumes (17.4%) can be attributed to the combined effect of the BA/acre estimates and the VBAR estimates.

Paper Birch was also harvested.  Both cruises estimated 66 cords of birch.  The scale volume was 73 cords.

Becker County Private Sale

Presented below are the results of a cruise comparison for a 26 acre private tract in Becker County.

In the 10-BAF cruise, 21 plots were installed.  The number of 100-inch sticks per tree was ocularly estimated for each of the 136 trees tallied.

In the Big BAF cruise, 41 20-BAF plots were installed and 140 trees tallied for basal area.  On the same plot centers, 41 100-BAF plots were also taken and the number of sticks per tree was measured by clinometer on the 28 “in” trees to calculated VBAR.

The cruises were conducted Jan. 31 2002 by Kent Wolf and George Deegan.  The sale was harvested and scaled a year later.  Results are tabulated below.

ASPEN

In this trial, the Big BAF method produced results closer to the actual scale from a smaller number of tree observations. The disparity between cruise volumes is a product of the difference between VBARs and BA/acre estimates.

Note that the average stick height for Aspen in the Big BAF/Sub-sampling cruise was based solely on 28 observations. However, each of these observations was made using a clinometer.

In contrast, the average stick height of Aspen in the 10-BAF cruise was based on 136 observations. However, each of these observations was based solely on an ocular estimate

The result of the difference in stick heights alone contributed about a 9% difference in volumes.

Note also that the Big BAF cruise returned a BA/acre approximately 5% greater that the 10‑BAF cruise

Conclusions:

1. Ocular estimation of sticks may be a significant source of error.
2. A precise estimate of VBAR can be obtained from a relatively few trees.
3. 20-BAF plots can yield a good estimate of Basal Area/acre in stands with variable stocking.
4. Making more precise stick height estimates on a smaller number of trees drawn from a larger number of sample plots may improve cruise accuracy.