As is typical, for this blog, I will meander about; even though the focus for my reading has been rocky shore methods, I wanted to highlight some of the methods I have read on, rocky shore or not.
One thing of critical importance is to make sure your method helps to answer your question(s). William Sutherland, who we will hear from again, uses a method he calls reverse planning by thinking backward, for example, by producing potential graphs that your data would provide, then asking if they help to answer the research question. For this blog, I will focus on methods to get after questions of population numbers, species abundances, and species diversity.
Let me start with the aforementioned African grasslands. I stole these methods from a book by Larkin Powell and George Gale. Larkin is sitting about 10 feet from me as I type this, but I’m not going to tell him. One of the methods they outline under Distance Sampling is the double-observer method. This method accounts for imperfect sampling; each observer does not see all the individuals in the sample area. By driving through it, you could do this in a large grassland; each observer would record their findings without communication with the other observer. Observers note specifics (e.g., distance away, direction, a large clump of trees) to reconcile the data later. This gives an idea of the number of individuals not detected. A downside is that you need two people.
I tried this method in a forest after a rainstorm. Instead of driving through a grassland counting large mammals, I walked and counted Limacus flavus — the Tawny Garden Slug. I walked a pathway from the road to the seashore through this woodland, noting each slug 1/2 a meter on either side of the path center. Since I had no second observer, I walked back on the same path, again noting each slug. I missed slugs in both directions, which surprised me. This orange slug stands out among the pathways substrate, or so I thought. My quick survey gave me an estimate of the slugs along the pathway, 41, and slugs in the entire woodland if the distribution is random—about 11,000 slugs in total.
The math involved is something like this:
x11 = Animals detected by both observers 1 and 2
x10 = Animals detected by observer 1 but not by observer 2
x01 = Animals detected by observer 2 but not by observer 1
N = true abundance (# of animals in the area sampled)
pi = detection probability for observer i
Detection probability for observer #1: p1=x11 /(x11 +x01)
Detection probability for observer #2: p2=x11 /(x11 +x10)
P(both miss animal) = (1-p1)(1-p2)
So: p1+2 = 1- (1-p1)(1-p2)
N then is = (x11 + x10 +x01)/p1+2
I’m trying to solidify a good survey method for fouling community organisms; the double-observer method is inappropriate. Fouling organism populations are often estimated by starting with a blank slate (literally). Researchers immerse plates of some material (ceramic, plastic) in the water and collect them at later dates. The percent cover of each organism that had adhered can be estimated.
One set of researchers (JP Sutherland and Karlson) drew an image of the coverage on each plate by first drawing onto a glass plate placed over the immersed plate and then tracing that onto paper. They mentioned that this method was particularly time-consuming, and they altered their methods within a few months. As a replacement method, they opted to use computer-generated random points (mapped within the area of the plate). After some trials, they settled on using 75 points. They recorded the species at each point; they did not record the canopy, but they did record both species when one was on top of the other. It is telling that this set of researchers altered their method; planning a research method may be the most important aspect of the study. Below, Below I quote ecologist and master research planner William Sutherland from the book Ecological Census Techniques, British spelling and all. FYI, there are two different authors named Sutherland mentioned herein:
This chapter was written in response to my frustration when talking to people who had worked hard
collecting data but who had missed opportunities and largely wasted their time as a consequence of
poor planning. Planning relates to any research programme, not just those involving carrying out a census.
Planning is important in deciding the method for a number of reason, but mostly because it leads to the correct type of analysis (statistical analysis). I do a fair amount of consulting in statistics. The biggest frustration I have is trying to analyze a dataset that was collected without an actual plan. The reverse planning method advocated by Sutherland will help us here; for example, the type of graphs created typically indicate the statistical method that will point you toward answering your research question. FYI, I had no idea how important planning was for a research project when I started doing them. Yes, that made for some frustrating times during my Master’s program.
The point to drill home is that the method is important so you can collect data that align to math/a statistical method that will answer your question. As JP Sutherland and Karlson discovered, the method should also be doable and not take all the time and/or money. This is, of course, why we are estimating, for example, a population. The time to count every member of the population is typically beyond our abilities, our time, and the available funds.
I came across a method for surveying sea urchins that highlights the idea that the time available is not infinite. Researchers wanted to estimate the population’s number and the individuals’ size. They have 100-meter transects and lay down quadrats each meter. Rather than measure every urchin encountered, they have a system: They measure urchins in the first quadrat, skip the next, count the urchins in the third quadrat, skip the next, count the urchins in the fifth quadrat, then skip the next one. Then they repeat. Thus, in the seventh quadrat, the urchins are measured. In this way, they have population and size estimates without the time-consuming method of measuring urchins in every quadrat.
Whole quadrats and points within a quadrat are often used in coastal and intertidal surveys. These work well for abundant species that don’t move much. For species that are not as abundant, the transect and skip quadrat method used for the above-mentioned urchins might be more beneficial. Covering a whole swath, for example, a meter on each side of the line, along transects, would work better for rarer species, particularly those large enough to see easily. This method could be employed in the intertidal or sub-tidal of the Maine coast to assess sea star abundance. Sea stars in both U.S. coasts have recently been devastated by diseases. If the researchers focused on just sea stars, this method would enable coverage of large areas relatively quickly.
The UC Santa Cruz has protocols for their Coastal Biodiversity Survey where they identify each taxa; a swath method would be incredibly time-consuming. They use a point method but gather more data instead of just recording the organism attached to the substrate. If there is layering at the point of sampling, they indicate each organism; those on rock are (R), those on other organisms are (E) for epibiont, and the canopy is indicated with a (C). The host (H) organism of the epibiont is also recorded. The substrate species could be an algae and also indicated as the canopy species.
Here is an example from their guide:
If the point hits a host alga whose holdfast is under the point, it is recorded as both a canopy (C) and a Host (H). The epibiont would be recorded as canopy (C) and Epibiont (E).
The same algae species would be recorded as (R) since the holdfast is on the rock.
A total of up to three taxa are identified under each point.
Other methods are employed for more mobile organisms not hanging out on the rock or moving slowly over the substrate, waiting for researchers to arrive.
The mark-recapture method is an old stand-by to estimate populations of mobile organisms. The basic idea is to capture organisms, fish with a net, mark them somehow, and release them. Later, you recapture fish. The number of fish recaptured, those you marked earlier, can estimate the number of fish in that pond or bay. The well-known method is called the Lincoln–Peterson estimator. There are alterations to this method that are better estimators, but this will give us the idea, and the math is easy:
N = (K*M)/k
N = Number of animals in the population
M = Number of animals marked on the first visit
K = Number of animals captured on the second visit
k = Number of recaptured animals that were marked
Doing multiple capture sessions is one modification to this method that gives a better estimate. Expanding to multiple capture sessions does require more time/energy/money, of course, but the estimates of the population numbers are greatly improved. I won’t write out the math because that also is more complex.
Capturing individuals for this study does not have to be a physical capture. For example, quality photos of the dorsal fins of Blacktip Reef Sharks, Carcharhinus melanopterus, enable the identification of individuals. Therefore, six sessions of photographing sharks’ dorsal fins in a reef system would likely give a good estimate of the number of sharks that use that reef system.
Look what you have done; now I want to snorkel a reef, “tagging” virtually with my camera, Blacktip Reef Sharks.
But back to my whole hunt for a method to sample biofouling organisms. I want to record organisms on an existing dock. I want to have a record of each species, density/abundance, and overall biodiversity (probably two types of biodiversity indices). I think a modification of the UC Santa Cruz protocols for their Coastal Biodiversity Survey will do the trick. I will place mini-transects in 20 random locations along the dock, perpendicular to the dock (into the water). Transects will be PVC pipe with 10 sample points. At each point the method employed above
If there is layering at the point of sampling, each organism in recorded, up to 3 species.
Those on rock are (R). If this is a host organism of the epibiont this is also noted with an (H).
Those on other organisms are (E) for epibiont.
The canopy, if present, is indicated with a (C).
Okay, done; now back to these sharks. :-)
Sources and Further Readings:
James, P., Hannon, C., Þórarinsdóttir, G., Sloane, R., Lochead, J. 2016. Sea Urchin surveying techniques. (Activity A4.1.1 of the NPA URCHIN project). Nofima Report 35.
Powell, L. A., and G. A. Gale. 2015. Estimation of Parameters for Animal Populations: a primer for the rest of us. Caught Napping Publications: Lincoln, NE.
Sutherland, J.P. and Karlson, R.H. 1977. Development and Stability of the Fouling Community at Beaufort, North Carolina. Ecological Monographs, 47: 425-446. https://doi.org/10.2307/1942176
Sutherland, W.J. 2006. Planning a research programme. In Ecological Census Techniques: A Handbook, ed. William J. Sutherland. Cambridge University Press.
University of California Santa Cruz SWAT Team. 2011. Coastal Biodiversity Survey Protocols. May, 2011. http://cbsurveys.ucsc.edu