Burning to Death in the Santa Ana Mountains

Last semester I took a field methods class and got to go to several places around southern California, experiencing both physical and human geography field studies.  The first half of the class involved the physical side and so I got to trek around in the wilderness in heat wave temperatures.  Yeah, not my idea of fun but the scenery was beautiful, so at least I had that to look at while I tried not to die.

The first trip was to the Santa Ana Mountains, specifically we hiked along the El Cariso Nature Trail in the Cleveland National Forest.  Our assignment was to examine and identify the types of Chaparral vegetation there.

One of the plants along the El Cariso Trail.  Photo by Laylita Day.

One of the plants along the El Cariso Trail. Photo by Laylita Day.

The trip seemed simple enough but because of the predicted high temperatures, the class elected to leave early in the morning and I am NOT a morning person.  The drive took more than an hour and my group got a bit lost along the way.  Once we actually started hiking the heat was already at an unbearable temperature.  The hike itself was not too strenuousness but the heat and lack of shade had me wishing I was anywhere else.  But enough of my complaining and back to the Chaparral.  (Warning: the rest of this post is really academic).

First we learned about the history of Chaparral in California.

Chaparral is part of the Sclerophyll Forest class, which is a subdivision of the forest biochore, a classification of natural land vegetation. Sclerophyll plants are both small trees and shrubs categorized by their hard leaves. The canopy coverage is about 25 to 60 percent and shrubs are usually classified by a height of less than 15 feet. These plants are evergreens and drought resistant, keeping their thick leaves even during severe droughts. Sclerophyll plants such as chaparral can be found in Mediterranean climates where there are hills and low mountain slopes as well as along the coastal ranges. They are usually found above 1500 feet.

In California and specifically the Santa Ana Mountains, there are many different species of chaparral, which corresponds with elevation and land exposure. The Mediterranean climate consists of dry summers with high air temperatures, making fires easier to break out. This combined with the high resin and vegetable oil content of chaparral makes the plants highly flammable and fires prone to consuming large areas quickly and intensely.

Map of study site (El Cariso Nature Trail) and points of data collection.  Basmap from Google maps; design by Laylita Day.

Map of study site (El Cariso Nature Trail) and points of data collection. Basmap from Google maps; design by Laylita Day.

Chaparral though has adapted both to the fire regime of the area and the dry hot weather. The high temperatures allow the seeds from particular species to be released and the plants grow quickly during the mild wet winters. During droughts chaparral can use a number a measures to conserve moisture, such as changing their leaves into small, compact and fleshy or reducing leaf surface area or producing light colored hairs to reflect sunlight. They can also have long roots that can access water from deep levels underground. The same resin and oil that makes the chaparral flammable is also a measure they use to prevent water loss.

Because of the fire regime associated with chaparral and climate factors, chaparral can be a hazard and thus study of it and its environment are important. The chaparral and its locations in California and parts of Mexico, create biodiversity hotspots where rare species of plants can be found. This also makes a case for the importance of studying and maintaining the region where chaparral are found.  [Info from Splansky and Laris]

For this study we had three hypotheses. The first was that a community of mountain chaparral species on an exposed slope exhibits a hierarchical frequency distribution in accordance with the following rank order:

  1. Scrub Oak
  2. Black Sage
  3. Chamise
  4. Hoary Leafed Ceanothus
  5. Sugar Bush

The second was a species/sample (or species/area) curve is an effective technique for determining an acceptable number of samples needed in order to provide a reliable vegetation analysis for the study site.

The third was “Does species diversity and total biomass in a chaparral community increase over time since disturbance indicating stability and seral climax as suggested by Clement?”

A sub-hypotheses was that two chaparral locations, Area #A (burned over) and Area #B (fire free), each at a different seral stage, will demonstrate differences in: (i) their hierarchy of species when ranked by number of individuals per species, (ii) by extent of diversity, and (iii) when ranked by average height (a proxy for biomass) of individuals belonging to the same species.


To help my group answer hypothesis one, we used a linear transect sampling method. A position along the trail was selected at random by our professor Dr. Laris, facing west and overlooking Highway 74. The trail was curved and within an area that had not been burned for about 25 years. We used a compass for direction and a transecting tape, placed on the ground along the edge of the trail. The transect distance totaled 20 meters and we noted the closest species at every meter. In some cases the same plant was present at multiple meter marks, and so we instead noted the next closest one, usually the one behind it.

For hypothesis two, we used the point quartering sampling method. My group walked 10 meters north of the bench that is the dividing line between the burned and unburned sections of the trail. We walked toward the unburned section to establish our base point. From there we rolled a 9-sided die to generate a random direction. The point of the die was used to decide direction. Then a 12-sided die was used for generating a random distance in meters.

For the first site, we went east and upslope 4 meters from the base point. For the second site, we started from our base point again and went northwest downslope for 8 meters. For the third site, we started from site 1 and went south upslope two meters. For the fourth site, we traveled from site 3 downslope for 10 meters.

At each site, we noted the two closest species in four directions and estimated the height by comparing it to the known height of a group member when she stood next to the plants. This method was to be done 3 times if we did not identify any new species, but since we found a new one at site 3, we continued on to site 4 and then stopped. Because we were not consistent in our starting points, returning to the base to do site 2, there may be some bias present in our results.

For hypothesis three, we used the data we had collected for hypothesis two and also data from another group, who had collected data in the burned section of the trail. We then compared the two sets of data and calculated average height and frequency of species using excel.


My group found that our species ranking did not match the one presented for hypothesis one. While the dominant species was the same, the others were ranked differently with ceanothus proving to be much higher in presence than expected. A plant not even listed in the hypothesis was also present, yucca.

Data collected for hypothesis two and three showed a clearly repetitive pattern with the same types of species of chaparral being counted at each site with only two new species, black sage and chamise being noted at site 3 thus the need for sampling from site 4.

Chamise.  Photo by Laylita Day.

Chamise. Photo by Laylita Day.

A dominant species, black sage, is apparent there as well. Though again another group’s type and frequency of plants found differ from ours in the unburned section of the trail. Thus for hypothesis three, there does seem to be different species and growth patterns for the areas that have been disturbed versus areas that have not been disturbed.


For hypothesis one, there is clearly a difference between the ranking given and the one my group found as well as the one found by another group. The ranking given does not state for what type of area, burned or unburned this hierarchy should apply to or whether it should be applicable to either area. Whichever it is, the hierarchy found by my group somewhat correlated because in both instances scrub oak was the dominant species. The biggest difference was ceanothus, which jumped from 4th place to 2nd in the ranking. Black sage was only down by one spot.

For the other group, the differences seemed to be more striking because of the absence of scrub oak and dominance of black sage, yet the other three species correlated with the hypothesis ranking. Using these results, I would conclude that the hypothesis is partially acceptable. It seems that based on the knowledge that black sage can grow well in open spaces after a fire, the hypothesis ranking may have been taken from a place that was burned not too long ago, allowing the black sage to have a strong foothold but long enough ago that scrub oak was now dominant and ceanothus still working its way into beating out the black sage.

Bias could have been introduced in the fact that we were required to do our transect sampling along the trail, where active maintenance of the trail, cutting back of the vegetation, could mean some species have greater advantages than others. There could have been some biases in the fact that we were told to go to a certain spot by our professor since he has some knowledge about the area that my group may not have had. My group’s inexperience in plant identification and our reliance on the inaturalist application may have also introduced some bias in the form of identification error.

For hypothesis two, the species/sample curve seems to be a partially effective technique in combination with the quartering sampling method because it did get a decent sampling size for the area (16 individual plants for the unburned area) and a pattern of plant species did seem to be apparent. These results appeared to correlate with the fact that my group’s sampling area was unburned for a long time and had lots of scrub oak and ceanothus rather than black sage, which is more prominent after a recent fire. This was shown by the other group’s data where the burned section exhibited high amounts of black sage.

Limitations and bias though were also present. In some cases there is only so far one can go into the chaparral because of the denseness of the vegetation so the sampling area is thus confined. The number generated by our dice is also limited by how many side each has. There is also the uncertainty of where exactly the boundary between burned and unburned is; it may not be quite as clear cut as the bench that we used.

Also since my group did the quartering process differently from the stated instructions there is possible bias in our data. Perhaps we did not get a wide enough range of distance, which could have limited our sampling area. Overall, I do not think that this caused a strong bias since the sampling method was for essentially random locations and each location was still found randomly if not at the correct starting location.

For hypothesis three, using just the data from two groups makes it difficult to really establish whether Clement’s theory is correct or not. As far as biomass goes, the plants in the unburned section were mostly taller than or about the same height as their burned counterparts. In this way, Clement’s theory seems to be correct.

As for diversity, the types of species did not seem to really change. The main five were present (except for scrub oak in the burned area) and there was only two different species, yucca and laurel sumac to add to the diversity. The amount of certain species varied quite a bit between the two datasets. Taking in both datasets as a whole, certain species clearly dominant at different times.

Going from burned to unburned, one species dominates while the others are somewhat similar while the unburned section does seem to show that there is a little more balance: the two highest are close to each other in number while the other three are similar to each other as well.

I believe there is stability in that certain species will dominate an area at a certain time after disturbance and that it will become somewhat more balanced as time goes on. There also could be a lot of variability even within the same area, such as one spot in the unburned compared to another spot in the unburned. Because of that I conclude that Clement’s theory is partially to mostly correct though more data would be needed to make a more accurate conclusion.

Bias was most likely present because of the limited amount of information done when using only two sets of sampling data. To better establish a clear pattern of species development a larger range of area and areas with different growth times would be more appropriate. Another area of bias could be because of the high amount of black sage and absence of scrub oak from the other group’s data, that area might be an anomaly if other burn locations did not exhibit such a wide difference. Perhaps there were more equal amounts of scrub oak and black sage in other burn spots. Thus getting data from more than two locations would give a more accurate picture of the area’s ecology.


The presence of not only the main five chaparral species mentioned in the original hypotheses but of also other species, such as yucca and laurel sumac, show that this area does have biodiversity. The types and amounts of species found in particular areas show how disturbances, such as fire can have a strong effect on what types of species grow where, when and how much.  The type of sampling technique used needs to be considered carefully in relation to the area’s history and what is and is not possible in the area’s terrain.

Ways to improve this study would be to try and pick a location that is not affected by human alteration (maintaining of trail) though this would mean needing equipment to get one through the dense chaparral to a more natural spot to acquire data from. Expanding the distance range used in the quartering technique might also provide a better sample rather than recording ones that are somewhat clustered together in the same general area. Overall this study did show that mostly accurate results can be obtained and a general pattern of species types and frequency can be obtained.

Note: This post was adapted from a report I wrote for the field methods class.


Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s