Methods Testing and Group Development

After developing the transects for the study area I was pretty confident that our first field day was going to run smoothly. Prior to going to our site, I created a map that had the overall layout of the study area as well as the transects drawn in. I also wrote down the coordinates correlated to each transect to insure that we could accurately randomly sample them.

Aerial image of the study area and each transect. The total area is 58,869.42 m^2.

Unfortunately, when we were ready to begin we found that all of the coordinates were shifted to the southeast about 30 meters or so. This meant that the starting point for each transect would be more difficult to find although the total plot area was still accurate. In order to deal with this unfortunate setback, we decided to only pick one of the original transects at random and then measure every other from there on out. In this way, we could insure that we were not violating any statistical assumptions while minimizing the work to correct the coordinates.
After establishing this methodology, we began walking the transects. Some quickly realized that this steep hiking through poison oak was not for them (I was the only one to wear boots and long pants). So, we decided to stop after sampling 5 of the transects. The results we got were very interesting. Each transect count was within an encouragingly tight spread ranging from 36 to 49. This tells us that our method of transect sampling is precise, even though it may not be accurate.

Part of the Bloody Nose Trail running through the study site.

From there, we tried a different method of sampling in order to see if there was a better tactic. The method we tried was a fixed area circular count. Essentially, a spot is chosen at random and then a circle with a radius of 15 meters is drawn around it and every individual tree within that area is counted (15m was chosen for the radius because that would give an area similar to that of the transects, 715m^2). That way we could compare our counts and check for accuracy between the two survey techniques. To our delight, the counts were very similar and in the 36-49 range just like we hoped. If we were to derive a total estimate for the number of eucalyptus trees in this area of Montano de Oro using the data that we have, we would get approximately 3,229 trees.


What We Learned

The transect sampling is difficult for some on this particular site due to the steep topography. Using the fixed area circular sampling is less labor intensive although setting up the plots will be more difficult (an adventure for next time). When we return to the forest, we will be using this new method of sampling and will be ensuring that our GPS coordinates are accurate both on the program I used to make the map and on google.maps.

Post by Eucalyptus Veteran,
Nathan Tallman