When Nature Calls

 When Nature Calls

by

Tomas Jimenez

Figure 1. Arrival at Montana de Oro

Day two of data collection started off much better than the first. Well rested, hydrated, and with plenty of time for breakfast spirits were high for this eucalyptus enthusiast. Upon arrival, we quickly split into two teams: The Alpha team (Jefferson and Nathan) and the B team (Tomas, Nick and Adam). What we lacked in raw talent we made up for with scrappy determination. Although we were the underdogs, we would not be discouraged. We quickly counted two randomly selected circular areas using our three man counting method. Adam and I would walk out fifteen meters from Nick to act as markers. Nick used his excellent tree counting ability to count the trees within the section of the circle that Adam and I marked off. We would then rotate around the circle to count the next section, and so on, until all the trees in the circle were accounted for. 

Figure 2. Adam acting as a marker or a bouncer in the woods

At this point, against all odds, we were beating the A team, until the unthinkable happened. My big, beautiful breakfast that I had been so grateful for was about to turn on me. Nature was calling in a way that it had never called before. Panic overcame me. With no known toilets around, few good wiping leaves, and a wildlife ecology group that found it "funny" to harass me at my most vulnerable. I searched frantically through the woods for the smallest sign of hope, until finally I stumbled upon a small wooden outhouse. All was saved! I'm not sure what good I had done in my life to deserve to find this beautiful outhouse and I didn't care. In that moment everything was right in the world. All my problems blissfully dissolved away into thin air.

Figure 3. Catharsis 

Unfortunately, things weren't all that they seemed. Once I figured out how to close the door everything was going smoothly until one of my companions through a large tree branch at my humble sanctuary. It made a loud noise which startled me, but it didn't startle me nearly as much as what the noise caused. Out of every nook and cranny of the old wooden hut, came every type of creepy, crawly critter that you could imagine. The bugs coated the inside of the walls completely surrounding me. If that weren't enough there was an odd scratching noise from outside the outhouse that sounded like something out of a horror movie. When I tried to open the door I found that my fellow teammates had wrapped the transect around the bathroom a couple times to trap me. Eventually they let me out and we counted trees.

Fin.

Figure 4. The B team (aka team T.A.N) after hard day's work

B Squad

Team TAN

Team 1 set out for Montaña de Oro in the hopes of finally gathering some concrete data. We sat down the previous week and discussed what went wrong with our initial sampling and how to adjust our technique. We made several changes, which have been discussed in previous posts. One of the key changes to our methodology was breaking up into separate groups. Alpha Team, consisting of Jefferson and Nathan, set of on their own to gather a smaller sample of data because they were only a tandem. The team consisting of myself, Adam and Tomas, affectionately named Team TAN gathered a slightly larger data set.

Data Collection Can Get a Little Messy

Testing out our new data collection methodology excited us all. The data came fast and furious as we worked at a fever pitch. We had to be careful though, because when you're sciencing as rapidly as we were, things can get out of control quickly. One of our transect tapes got entangled with a tree, it was an utter and complete disaster. We were uncertain if we could recover from the tragedy. Team TAN buckled down and persevered though, we would not be undone by knots. We undid the knots and moved on.

The rest our day was a wild success. Through the data we collected we were able to calculate an estimated population size of 3,457. To gauge the accuracy of this estimate we will probably have to head back out to Montaña de Oro and take a complete census of the area.

Circles Are Better Than Lines

As Nathan mentioned, a newer and improved method of measurement was employed on the second day of tree tallying. The transect grid utilized was the same as before, however the transect lines were now used to create intersections on the land. 

Measuring the transects at the border of the plot (road=border)

Each intersection would result in a measuring spot that would be 15m in radius, to create a 148m^2 measuring area. By following a transect line to the intersection of interest (versus trying to walk 5m apart in a straight line down a hill), we were able to more accurately find the measuring point of interest.  

Tape from the road was walked in a relatively straight line to each intersection point. #accuracy

Sometimes life gives you lemons.

Nonetheless, the growth covered landscape still provided enough obstacles to interrupt the seemingly flawless tape-walking-to-measuring-point method. Even so, the new method also provided a much simpler and easier means of counting the trees. Walking in a circle with a fixed15m tape, and counting trees inside of it, proved to be much more consistent than the traveling human transect used in the tree counting 1.0 blog post.


One of the issues Nathan and I ran into dealt with the dispersion of trees in our plot. One of the randomly selected intersections provided us coordinates to an area on the hill that contained substantially less trees than any other spots counted. Oh well.

Leonardo DiCaprio has more Oscars than this plot of land has trees.

-Jeff

Results and Discussion

Results:

With the use of the two different sampling techniques we were able to create two individual population estimates. This helped us to generate a range of possible population sizes. Using the transect method we estimated the eucalyptus population to be 3,229 individuals. The average count in our transects was 43 individuals. The area of the transects were 795m^2 and our total study area was 58,869.42m^2. By dividing the total area by the area of one transect and then multiplying by the average transect count we were able to create a population estimate.

The next sampling method we used was the point count method. Using this we estimated the population size at 3,457. To calculate this we divided the total study area by 706.858m^2 (the area of a single point count) and then multiplied by 41 which was our average count. To our delight these two independent population estimates were in a relatively tight range. This leads us to believe that our estimates were at least precise if not accurate.

Discussion:

The reason that we were interested in the population size of the eucalyptus trees was because it was a unique sampling opportunity. We were curious how difficult it would be to estimate the total number of trees within a forest. On top of that, there are good reasons to be aware of this information for the MDO groves. For example, some people find the eucalyptus trees in MDO to be a pest and an invasive species that prevents the growth of native plants through their allelopathic qualities (Chu et al. 2014). Perhaps these people would be curious of the population size because they would want to know how much it would cost to remove them. Others find the trees to be a crucial part of the state park's look and would want to preserve this grove. Additionally there have been studies that suggest eucalyptus plantations are actually hosts to quite substantial species richness (Timo et al. 2014). They may be interested in if the population size is fluctuating through time. Whatever the case may be, it was an interesting study to tackle and it was a good exercise to sample in such a large study area.

If we were to sample again, using the point count method would probably be our best option. It was not labor intensive and it was easier to be accurate with our counts in this manner. To further refine it, we would use an accurate GPS unit to pinpoint location. We also would find a way to better delineate the outlines of our point counts just to insure that we are 100% accurate in our counts. Regardless, the estimates that we came up with are remarkably close to one another.

Thank you so much for joining us on this adventure. It was a blast and we learned a lot. This is the Eucalyptus Team signing out.

Sources:

Chu, Chaojun, et al. "Allelopathic effects of Eucalyptus on native and introduced tree species." 

     Elsevier: Forest Ecology and Management 323 (2014): 79-84. Print.

Timo, Thiago, Maria Carolina Lyra-Jorge, and Carla Gheler-Costa. "Effect of the plantation age on 
     the use of Eucalyptus stands by medium to large-sized wild mammals in south-eastern Brazil." 
     Biogeosciences and Forestry 8 (2014): 108-33. Print.

Second Day of Data Collection: Seasoned Veterans

View from atop the sand dune looking towards our site.

Early on the third of November, we reembarked to our study site in MDO. Our goal of this outing was to finish collecting sample data using our new point count technique and not lose any members along the way. I arrived at the park well before the rest of the team which allowed my to get another perspective on our study site. Across Pecho Valley Road from our study site are some tremendous sand dunes. Before beginning, I climbed to the top to enjoy a little breakfast by the ocean and get more of a bird's eye view of our study site. There is truly a remarkable number of eucalyptus trees in the park!

Once there we divided into two teams each tasked with collecting several point counts. Jeffrey and I decided to use our old transects to randomly sample from. What we did was generate a number between 1 and 43 that would correspond to a specific transect. Then we selected a number from 1 to 140 that would tell us how many meters to travel down the transect before getting to our sampling area. Once at the randomized location, we used a transect tape to create a large circle with a radius of 15 meters.

Transect tape disappearing into the distance.

This sampling technique did turn out to be less arduous than the transect sampling we tried earlier. There was still a good deal of hiking involved, however, we were not trying to count trees while hiking. This allowed a more focused effort on the sampling task and perhaps gave us more accurate results. Interestingly enough, during a part of the sampling I noticed we were walking on a game trail that was littered in feces. I later learned that they were coyote feces which I though was very exciting to see. This is completely off topic though. So, let's get back to it!

With the combined effort of our two groups were were able to complete 13 point counts which was higher than our goal. It seems that we have developed our forest legs. The results of our second day of data collection are very exciting. There is a remarkable correlation between the estimations generated by our two techniques which makes me confident that we have created a precise range for the eucalyptus population size. I'm not going to give away the punch line here though. For that, stay tuned for our Results and Discussion post, coming soon!

    -Nathan

New Methodology: The Future of the Eucalyptus Team

As we have said in our previous posts, transect sampling turned out to be more difficult to implement than we initially thought. Practically speaking, it required us to walk up and down steep transects and we struggled to keep our lines even as we trudged through poison oak. This sampling method also turned out to be very time consuming. So, we have moved on to plan B: point count sampling.

My footprints looking back down the Bloody Nose Trail

When we go out in the field tomorrow we will be randomly sampling from a selection of many points scattered throughout the study area. We hope that this will give the opportunity to sample a representative grouping of the entire population. Similar to our first field day, we will locate the points using GPS while out in the field. At each point, an imaginary circle with a 15 meter radius will be drawn and every individual eucalyptus tree inside will be counted. If our initial methods testing has any predictive power, than we expect most of our points to contain somewhere around 40 individuals. We hope to sample from at least 10 different points tomorrow. The more we are able to sample from, the more confident we can be in our estimation. Unfortunately, our time constraints may force us to do fewer.

After collecting our data, we will be able to calculate an estimation for the total eucalyptus population within the study area. This is possible because we know the total area of the study site as well as the area of our sampling points. Using the average eucalyptus count from the sampling points,we can calculate the eucalyptus density. From the density we can derive the total population by multiplying the average by the number of times the point area goes into the total area.

But, before we get ahead of ourselves, some field work must be done! Join us next time where we will explore our second sampling excursion and discuss the results of our study.

   -Nathan

First Day in the Field

First Day in the Field

During our fist day in the field our eucalyptus team encounter a few problems, but ended with positive feedback for our next survey. We plan our method of attack before stepping into the field by mapping out of population and documenting possible transects. Our original plan was to transect sample at least ten vertical lines through our population with a ten meter long transect. As we soon found out that would be too difficult due to the amounts of branches and obstacles on the ground, so we reduced the transect size by 5 meters.




After Random selecting our first transect based on latitude and longitude within our population area, we measure 10 meters to the next transect location. Due to the under calculated slope of our terrain and the overestimated stamina of our team we were only able to count 5 transects within a 3 our window. We were not able to record as much data as we liked, but we were able to get consist data.

The transect measurements were accurate but very exhausting so we tried spot counting. Spot counting gave us a very similar range of data and required a lot less energy on our part. We were only able to do a few spot counts but we feel confident in that method. Our next field day will focus on spot counting a in order for us to begin compiling data for analysis.

The only other issue we had during our first field pay was some patches of poison oak. All in all we learned a lot more while being out in the field on the method we should use and the amount of data we will need for an accurate population estimate.

-Adam Ibrahim

Tomas' first day of data collection

The day started bright and early. Nick came to pick me up without telling me when he was coming. this lead me to taking my bagel with lox on the go and forgetting to grab water for the trek. The trek was also a surprise because the area that we were studying looked a lot flatter on google maps. The data collection began with no water and salmon mouth. We mapped out a bunch of transects that cut through our study area,  each 10.8 meters apart. using a random number generator we picked our starting point, and would do every other transect from then on. Adam and I stood  five meters apart and walked along our chosen line. Nathan would count all the eucalyptus trees in between Adam and I. He counted the trees on Adam's border but omitted the ones on my border. Nick and Jefferson made sure we were the right distance apart and kept us going in a straight line. This was easier said than done because of all the obstacles and poison oak. After doing that a few times Jeffery and Nick went and did circular areas in random locations in our study area. They came up with about the same area and got numbers that were consistent with our first method of counting. The second method proved to be much easier. On one of the treks up I started to develop a migraine that started with my loss of vision, so I had to be taken home early. I spent the rest of my day in my dark, dark room hating life and missing out on all the fun eucalyptus tree action.

Tree Troubles

Before heading out to Montana de Oro I had some idea of the terrain we would be working in, but not a complete picture. From what I understood, the trees had been uniformly planted decades ago, with the intention of cutting them down to use for railroad ties.  I believed that the area was going to be flat and easy to traverse, it just seemed like it would be logical to plant the trees somewhere that would be easy to harvest them.

The evenly spaced trees

The reality partially matched my expectations. The trees were evenly spaced but the ground was not easy to cover as all of the trees were planted on a steep slope. Some areas of the plot were impossible to cover because of fallen trees and other areas were littered with poison oak. We made it out to Montana de Oro around 8AM and by 10 the weather started to heat up, only making it harder to climb up and down the slope.

We had planned to do transect sampling to get a population estimate, but three samples in and our stamina started to give out. After two more samples we decided that we needed to try another method. We weren’t getting the kind of accuracy we needed out of our transect measurements. Our GPS wasn’t accurate enough to keep us on track for the entire length of a transect and the difficult terrain meant we had to wander off track occasionally.

Jefferson and I decided to break off from the transect group and give spot counting a try. We chose a random set of GPS coordinates in the area and measured out a circle with a 15 meter radius. From there we counted every tree inside the circle. We did this twice and compared our data to the transect counted trees. 

The area of the transects was roughly 715m² while the circles measured 707m².  The transect counted trees averaged between 36-49 per plotted area, the spot counted trees were counted as 49 and 41. The methods provide similar data, but the spot counting was substantially easier to do. In the future we will go out to Montana de Oro again and do a more intensive series of spot counting.

My main takeaway from the whole event was that, before deciding on a population estimate method, you need to strongly familiarize yourself with the area you plan on working in. We could have finished our data collection if we had done spot counts from the beginning.

Team Transect feat, Jefferson

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

Ch. 1 Trouble in the Forest

Group Number 1. Aspiring eucalyptus trackers. 

Tuesday morning began like any other day. The rising sun unveiled a quaint fog blanking the sleepy town of San Luis Obispo. Despite the crisp morning air, an outfit consisting of shorts and a t-shirt were donned in anticipation of the field exertion only wildlife ecology sampling could induce. Fast forward 40 minutes, and the musketeers of group number one were damp with exhaustion, scaling up and down the biased landscape of Montano de Oro. The early reconnaissance mission through Google Maps failed to provide insight on the topography of the chosen plot. The sampled area between the Bloody Nose Trail and Pecho Valley Road proved to be a uniformly laid out and evenly sectioned plot for measurement, but it all resided on a hillside which made surveying wearisome.

Top of the plot area on the Bloody Nose Trail

The plot of land was sectioned into 43 transects. Transect 11 was picked, through a mobile random number generator, to be measured first. Each measurement was decided be taken in a width of 5m of the chosen transect, for every other 10m (After transect #11) going north to transect 43.

Students in the forest

Eucalyptus trees that fell within the borderline of the transect were only documented for the south facing border. The length of each transect from the Bloody Nose Trail to Pecho Valley Road was around 143m, giving the observed area of land for each transect around 715m^2.






Many questionable measuring tactics arose throughout the day, most noticeably the consistency of the transect area. The starting point for each transect was carefully measured out at the top of the valley on the Bloody Nose Trail, and was supposed to be followed down in a straight line to Pecho Valley Road. However, the varying topography and tree obstacles caused our transect counters to shift the plot's boarders as we moved through the forest.

Expert team of eucalyptus counter in action,
note: distance between two counters may not actually be 5m. 

This slight migration occurred during each trip up and down the mountain, despite our best efforts in realigning our travel path mid mountain. Also, the 5m transect area of measurement could not be sustained by a physical tape through the count, as it would get snagged on any of the numerous growth. Our group ended up deciding to estimate the distance of 5m between two people and keep that estimated distance down/up the hill as we counted. Of course, that distance was not always maintained the whole time, and it was difficult to keep track if one side was too far from the other or vice versa.

Trails in the forest.

Despite the possible inconsistency in counting and transect sampling, the first two transects sampled both contained 49 trees. The following transects measuring yielded a similar amount of trees, around 50. Although, this "phenomenon" could simply be attributed to the row-like plot pattern the eucalyptus trees were planted in, various variables suggest otherwise. First off, the rows in which the trees are planted in were not parallel with the transects that were in place, so the count did not simply neatly follow each row.  Moreover, only live trees were recorded in each plot, seeming to only produce more discrepancy among the tally. Through all this however, a consistent population of trees was observed for each transect of land, possibly suggesting a stable density of growth.

A spot count was later experimented with to compare with the transect method. The spot count location was haphazardly chosen to be centered around a large eucalyptus tree (pictured below). (We decided that total randomness wouldn't be too crucial in this particular spot count, because we were simply looking for any drastic change in tree count from the transect method).

Nick standing next to the majestic eucalyptus tree, used as the center for our spot count.

The spot count was measured in a circle fashion, with every tree in a 15m radius being documented. With a radius of 15m, the total area of the count was around 706.5m^2 (not too far below the transect area). For this area, a total of 49 trees were observed, which is very similar to the quantities seen through the transect method. The biggest strength the spot count provided our group was not in its effortless approach in measurement, but that each tree could be distinctly be accounted for in the chosen area by walking the transect tape around in a circle of a concrete center, versus the transect walking method that included inevitable changes in area and direction was we walked down the hill.

Due to these reasons, a spot count will probably be utilized for future eucalyptus quantification. Before this can happen though a new grid must be created to encapsulate the region in a way that works with circular spot counts. Also a GPS unit will tremendously help with the accuracy of counting regions.

Until next time.....
-Jeff (aspiring eucalyptus counter)

Eucalyptus Euphoria

Being assigned group number one was a great honor, and with this large responsibility also came the heightened expectations in the science community. Thus we had to diligently construct a project that would leave no disappointment among our audience. Therefore, it was no question that the majestic blue gum eucalyptus tree, or Eucalyptus globulus, would be selected for observation. This Australian native has made its way all over the world, to places like Spain, South Africa and, of course, California.  Juvenile leaves are covered in a blueish waxy bloom which gives it its name "blue gum". The eucalyptus trees in Montano de Oro were originally planted to be harvested for railroad tie production, but ended up being abandoned. Now, many people feel that these trees are an invasive species that use their allelopathic features to prevent native plants from colonizing the area. Other believe, however, that these trees are an integral part to the overall look and "spirit" of the state park and they also site that these trees are common living places for migrating Monarch Butterflies. No matter what opinion you have, it is still useful information to know how many eucalyptus trees are in the park in order to: establish trends in population growth through time, determine the cost of removal, or set a precedence for conservation.

Study Site

The location that we have selected to survey is located in Montano de Oro State Park along Pecho Valley Road and the Bloody Nose Trail. There are multiple reasons for choosing this site. First of all, its boundries can be clearly defined because there are two hard dividers on its western and eastern sides (Pecho Valley Road and the Bloody Nose Trail respectively) and a clear stream bed to the north. Due to these geographic features, not only can we clearly define the study area on a map but we can also stick to that area while taking measurements on site. The second reason for choosing this particular area is because the eucalyptus trees plated here are relatively evenly spaced and almost take up the entire plot of land. This will make generating a population estimate in such a large area more feasible.

Methods

Due to the large size of the study area, we plan to use two different sampling methods in the hopes of generating a more accurate population estimation. One method we will use is is fixed area transects on the site. This sampling method will give us populatin density from which we can extrapolate the total population. The reason that this is preferred over quadrats (on site) is because  creating a quadrat large enough to encompass multiple individuals would be difficult and generating a fixed are transect and subsequently walking the study area in a west-east fashion simply makes more sense due to the even spacing of the individuals. The second method we will be exploring is fixed area spot counts. This method is similar to the transect sampling except we will not be walking along transects. Instead, we will be standing at a single point and counting trees around us.