Eagle Lake Ranger District and The "Big/Little BAF" Concept

by Rick Crowther, Marking Crew Foreman, Eagle Lake Ranger District, Lassen National Forest

In the 1980's the timber cut on the Lessen National Forest was much different than today. At that time 1 was working on a small matting crew on the Eagle Lake Ranger District. It seems at that time we were told by our supervisor to mark trees, big ones, and every once in a while measure one. At that time the standard method of cruise sampling was strip cruise and sample tree. This was appropriate because each tree was being visited and there was quite a large variation from free to tree.

The Forest has seen 3 major changes on the district in my short career ranging from Overstory Removal (big trees) to a massive Insect Salvage in the early 1990’s, and today where we primarily thin from below to improve Forest Health. Even though our selection of timber to be harvested had changed our sample methods failed to change with the method of cutting. The Eagle Lake District, like others in the Region was over sampling and measuring too many small trees. This was not evident until the District sent me to a workshop put on by John Bell & Associates, and Kim Iles & Associates. I had beard quite a lot about John Bell, but who was this Kim Iles guy? I learned basic statistics. During the field exercise Kim had us do this weird thing during the variable plot exercise by using 90 Basal Area Factor (BAF) for selecting our measure trees. The class was a huge success!

In April 1996, at another Bell and Iles workshop at Oregon State University, Kim and John discussed this Big/Little BAF method. After two years of cruise design work on a variety of timber sales and venturing into sampling soil compaction and crown density. I had a much better handle on sampling methods, and was hungry for new ideas. It was at this session that the light came on with the Big/Little BAF concept. I got it!

Upon returning to my home unit on the Lassen I informed my cruisers that things were never going to be the same at Eagle Lake. Life was going to get much easier because they did not have to measure every 12" white fir. The cruisers were somewhat confused by this new concept but after a few plots they loved it.

The first attempt using the new Big/Little BAF design was on the Logan Multi-Product Thin. This large area had a leave tree mark with small timber to cruise. The area was not stratified by stand type. With a BAF of 5, we gathered pre-cruise information for Strata 1. The coefficient of variation (CV) was 94 for tree count. The CV was 48 for the tree measurements. I was trying for a standard error (SE) of 14%. This told us we needed 180 count plots and 47 cruise trees! Can it be we only need 0.261 measure trees for every plot? Being a conservative cruiser and not comfortable with the answer, I went with a much safer plan. This way we would capture about 1 measure tree in every plot. The final cruise design was a 5 BAF tally, with a 20 BAF measure selection factor. We came out with a SE of 13% for the unit. Although production was almost doubled with this method, I knew we measured too many trees. Several methods were tried on this sale but they bogged the crew down and production was lost. They began to fall behind our completion target date for this multi-product timber sale and I was searching for the right design to finish on time.

There were 500 acres left to cruise in 3 different units with about a week to finish. These units had a tree count CV of 65 and a volume to basal area ratio (VBAR) CV of 41; we were trying for a SE of around 13%. I needed about 130 plots and in those plots we needed 39 measure trees. From the pre-cruise plots we knew that a 71 BAF gave us around 0.3 trees per plot, so now it was "rock and roll time." The cruise design for stratum six was a 5 BAF tally, 71 BAF measure selection factor with 120 yards between plots. It took the crew 2.5 days to complete the cruise, which ended up with a SE of 19% and 34 measure trees in 131 plots. This was not the best answer, but when combined with the other strata it ended up with a SE of 6.33%. I thought the District could live with that.

Convinced that Big/Little BAF was the only way to go, we began our next sale, Harley Timber Sale, which was also a thin from below leave tree mark sale. I used aerial photos to establish strata, and created a spreadsheet for use during my pre-cruise to sample various basal area distributions. The total area of the sale was 652 acres and had 3 strata.

Region 5 standards required us to get a SE of less than 10% in a tree measurement sale like Harley.


CVtc = CV of tree count

CVvbar = CV of VBAR




Desired SE

















Stratum 1. To capture the necessary number of count trees we used a 5 BAF for the tally plots and needed about 207 plots (I bumped it to 300, some people were nervous). Within these plots we needed about 81 measure trees – around 0.39 trees per plot. I knew from pre-cruise information that a 40 BAF would give us around 0.5 trees per plot, (108 trees too many, but I could live with that)! The crew began the cruise averaging around 50-60 plots per day. The entire stratum took 4.5 days to complete. But what was the answer for this cruise you ask? We took 318 plots, measured 167 trees, and achieved a standard error of 8.76%.

Stratum 2. We used a 10 BAF for tally and 54.4 BAF selection factor for measure trees. The crew took 283 plots, with 134 measure trees, and had a standard error of 9.71%. It took 4 days to complete this stratum.

Stratum 3. We used a 20 BAF for a tally, 90 BAF for a measure trees (only because at the time I couldn’t find anything BIGGER [Editor’s note: See Issue No. 36]). We took 211 plots, measured 157 trees, and had a standard error of 9.73%.

After lumping the three strata together the final answer was a volume 1.6 mmbf and a combined SE of 6.1%. The new system made the crew happy, my boss happy (a good thing) and was about 50-60 % more efficient than the previous method we had been using at Eagle Lake. It gave us a better answer faster because we focused on the thing that was most variable, the tree count. The ironic thing is the cruise could have been done much faster if we could deal with biomass differently. Half of the crew’s time on a plot was spent measuring junk.

This is a problem that I brought up at the Advanced Cruiser session at OSU in November. The question was asked, "If there is one thing you can get out of this course what would it be?" I immediately thought of my biomass problem, and at the end of the course my problem was addressed by Kim Iles. He said that the problem was with the procedure that my district used to expand biomass material. Kim’s answer was to get a new procedure.

Upon returning to the District, and with the arrival of our first major winter storm I decided to test out an idea. We kept the crew in that day because of hazardous driving conditions. To keep them busy they were split up into groups of 3 to test out the idea of Big/Little BAF for Biomass (trees under 10" DBH). Group 1 would use the traditional 1/40th acre fixed plot and measure lots of trees. Group 2 would use a 5 BAF for tally trees and 40 BAF for measure trees. Group 3 would use a 5 BAF for tally trees and 90 BAF for measure trees.

The area they had to work in was 40 acres of east-side pine. Each team had a time limit of 3 hours to cover the area and started in different parts of the property. Group 3 had a data recorder problem that caused them to lose 30 minutes of cruising. The results were:

Type # of Plots # of Sample Trees SE% CVtc CVvbar
Fixed Area 37 89 53.5 188  
5/40 BAF 50 10 42.9 132 33
5/90 BAF 42 9 34.6 112 18

Besides getting better results with the Big/Little BAF scheme in biomass, data entry went from a 3 to 4 hour task to a 10 minute process. We have saved hundreds of pages of paper besides time and money.

In closing I would like to thank John and Kim for all their support and advice, and in helping me understand that there IS a better way.

Originally published April 1999

Return to Home
Back to Guest Index