Species Distribution and the Buddy Factor at Hopkins

Title: 0 pts

Next time, give me a brief Abstract.

John Harriman

Introduction: [0, 0, 0, 0, 2, 2, 4, 0, 2, 2, 2] 14 pts

Species distribution within the kelp forest depends on varying physical features (Roff and Taylor 2000, Méléder 2010). These features make an ideal environment for a wealth of species to exist in. Different species inhabit various niches within the kelp forest, and differing hydrodynamic effects, light penetration, temperature differentials, and substrate types on which species live are creating these niches (Stevens and Connolly 2004, Méléder 2010). We are interested in studying these different variables because we want to observe patterns in species distributions based on the different abiotic factors that the kelp forest consists of. [why??? what does this tell us?]

We used qualitative field sampling methods [why?] to determine the distribution of multiple species in the kelp forest.[why?] These species are part of the macro-fauna and flora of the kelp forest including fishes, invertebrates, and algae. We did our sampling at different depths, [why?] which allowed for hydrodynamic, light, and temperature variables to influence the data. We also surveyed different habitats where some areas were generally sandier while the others were composed of rocky reef to account for substrate differences (Stevens and Connolly 2004). Buddy pairs were sent out to do qualitative analysissampling, but between-buddy differences in the data, from here out the “buddy factor,” may get in the way of seeing true variability given the environmental influences (Toma 2000). [this sentence is not clear]

We are interested in finding trends that define species distribution in the kelp forest based on environmental differences. [really? Or are we interested in assessing the precision of qualitative surveys and what factors influence that?] We would ideally like to have unbiased data to assess these differences in distribution. This gives rise to the questions of whether the buddy factor, as well as the overall qualitative nature of the survey itself, are interfering with the results we want to obtain.

Methods: [2, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 4, 2, 2, 0, 2] 16 pts

[look at writing guidelines… I need a general approach, then a detailed study system description, subheadings of questions, the design and analysis used for for each question, including predictions. ]

In order to find a distribution pattern of species, we did a qualitative survey of the Hopkins kelp forest using SCUBA. Buddy pairs were sent out to swim four transects at each five meter interval along the main transect line running through the middle of the kelp forest (Fig.1). These buddy pairs were told to give a qualitative numerical description (1-5) for rare to commonly seen species, respectively (Fig. 2).

We were not prescribed any exact methods of how to justify our qualitative analysis; in fact, we were told not to discuss it between our own buddy pair or among the group as a whole.

Starting at our meter point on the permanent transect line, we laid out a thirty meter transect and, along the way, observed the scenery as far as visibility allowed to get an overall idea of the abundance of each species. At the end of the transect, we turned around to reel the tape in and repeated the same observation in the other direction, but counted it as a new transect. These two transects were considered the deep transects because they went offshore into deeper water. These transects were followed by two similar transects inshore to shallower water where qualitative observation continued in the same manner.

A statistical analysis of the data was derived to see if there was variance between the different environments surveyed. (???) Variance graphs were made between different depths, distances along the permanent transect, and between buddies. [why?] The buddy factor was also statistically compared between individual species and among the different qualitative numerical values to see if the buddy factor had any influence on the data. (really??)

Results:

We found that there was significant variance between all aspects of the data (i.e. distance, depth, and buddy) (Figure 3). The variance in distance along the permanent transect (meter) was the most notable, and variance was also accounted for by depth. We also found that there was significant variance in the data between buddy pairs. This buddy factor that accounts for almost 40% variance in the data is very notable throughout the study.

The correlation between the mean abundance of any particular species and the difference between buddies is very strong (Fig. 4 and 5). Where there is a high abundance of species, such as the bat star, Patiria miniata, there is a relatively low difference between the qualitative descriptions that any given buddy pair gave. Buddy pairs were in close agreement for species that were very rarely seen, such as the red abalone, Haliotis rufescens. As for species with a medium abundance (qualitatively 2-3), there seemed to be a large discrepancy between buddy pairs. The orange cup coral, Balanophyllia elegans, represents this well where its relative abundance is around 2.5 but the difference between buddies is almost 60% disagreement. These examples are only from the invertebrates, but they can be seen across the board in algae and fish as well (Fig. 6).

Disscussion:

The variance that exists in these data for the distribution of species can be attributed to the abiotic differences in the environment; however, these differences are offset by the buddy factor. The different distances along the permanent transect changed the overall environment from mostly sandy areas to rocky reef. Different organisms inhabit these different environments, so a shift of different inhabitants occupying these spaces was seen. At different depths within the reef, varying abiotic features such as light, surge, and temperature shaped the distribution of species. To test the distribution of species, a good sapling technique is needed to diminish bias and fully prove that these environmental factors are the explanation for the distribution of species. The buddy factor is what is getting in the way of truly explaining the distribution of species.

The buddy factor can be attributed to many different variables. Diver experience is one such variable that can greatly influence differences in data because one buddy may have 100 dives and knows what to look for, while the other only has 10 and may not be familiar with all the species being sampled. The way this survey was conducted also led to their being buddy differences. Each buddy was surveying on different sides of the tape and most likely looking at different things along the tape, which brought about bias in the data. The bias that the buddy factor brought out in the data is very strong and makes this data set unattractive for noting the true distribution of species in the kelp forest. This is not to say that buddy pairs should not be used for conducting studies, but rather that rules and guidelines should be set on the methods of how to collect the data.

The data should have been collected in a much more standardized way. It seemed that we were all given tapes and data sheets and told to just go for it. If we were trained on how to collect the data, the results may have seen much less variance between buddies and, therefore, much less bias in the data. We should have been trained on how much area to cover, how much time to take on each transect, how to look for more cryptic species, and especially how to give our qualitative descriptions and what each number actually means. After all of this training we could then go out into the field to collect unbiased data in a qualitative fashion.

Even after all this training, there may still be a problem with the collection of qualitative data. Perhaps a quantitative approach would be more appropriate. Quantitative data is much more concrete and has a level of accuracy that cannot be matched by qualitative approaches (Rajendran 2008). Actual counts of different species can be seen instead of biased qualitative relative abundance. Quantitative is more accurate means of collecting data and no personal bias is put into it. Quantitative surveying is used in the field by many researchers and survey groups like PISCO, to monitor the biotic and abioic features of the subtidal environment inside and outside of marine protected areas, or MPAs. The data sets that come out of these surveys are unbiased and are used by the Fish and Game to determine the significance of MPAs (PISCO website). Qualitative processes are still used in many groups such as REEF, which sponsors The Great American Fish Count, and Reef Check. Perhaps a blend of qualitative and quantitative is appropriate based on the means of what a scientist is trying to determine (Rajendran 2008).

A quantitative approach seems the best method if accuracy is desired. While qualitative surveys are still used regularly, it is quantitative surveys that produce the most reliable data without personal bias.

References: [3, 2] = 5 pts give full journal names.

Admin. May 2, 2011. PISCO website. October 10, 2011. http://www.piscoweb.org/.

Méléder, V. 2010. Predictive modeling of seabed habitats: case study of subtidal kelp forests on the coast of Brittany, France. Marine Bbiology, 157(7), :1525-.

Rajendran C.P. 2008. Quantitative vs qualitative: Which side of the fence?. Current Science, 94(1), :20-XX?.

Roff J.C., Taylor M.E. 2000. National frameworks for marine conservation—a hierarchical geophysical approach. Aquat Con- serv 10: 209–223.

Stevens T., Connolly R.M. 2004. Testing the utility of abiotic surrogates for marine habitat mapping at scales relevant to management. Biol Conserv 119: 351–362.

Toma J.D. 2000. How Getting Close to Your Subjects Makes Qualitative Data Better, Theory into Practice, Vol. 39, No. 3: 177-184.