The Value of Qualitative Data in Describing a Kelp Forest Ecosystem

Catalina Macdonald

BIOE 161 Kelp Forest Ecology

October 10, 2011

The Value of Qualitative Data in Describing a Kelp Forest Ecosystem

Scientific data is either qualitative or quantitative. In this paper I examine the value and cost of using qualitative methods to survey a kelp forest ecosystem. We collected qualitative data in the kelp beds of Hopkins reef. The high difference between buddy pairs indicates that this data was not very reliable.

Introduction

Knowledge of species abundance and distribution can be used both to understand the workings of an ecosystem and to describe how a system changes over time or in response to pressures. For example, a data set giving the abundance of species within a kelp bed can be used to find patterns of species associations, competition, and how species composition varies with depth (Graham 2004, Davenport and Anderson 2007). If the data are taken over a number of years, they can describe how the ecosystem has shifted over time, how marine reserves affect community structure,and how a kelp forest can recover or fail to recover from deforestation (Graham 2004, Paddack and Estes 2000, Dayton el al 1992). Data on species abundance and distribution can be classified into two types, qualitative and quantitative.

Most scientific data is quantitative. Quantitative data is more precise, more rigorous and more objective. We would generally rather hear that someone found 16 blue rockfish (Sebastes mystinus) on their transect than “many.” Quantitative data is better for addressing specific research questions. However, much time and effort must be spent to build up enough quantitative data to show the big picture of species abundance and distribution within an ecosystem. Qualitative data is less precise but may be easier to gather. Some monitoring agencies, such as the Great Annual Fish Count,choose to gather qualitative data on species abundance (Reef 2011). Qualitative surveys, where species are ranked in categories based on their relative abundance, are less objective and less precise than quantitative surveys. But can they still give an accurate picture of species abundance and distribution?

We examine this question in the context of the kelp beds in Hopkins reef. Kelp forests provide shelter and food for a diverse assemblage of fish, algae and invertebrates, and have long been recognized and studied for their ecological importance (Graham 2004). The abundance and diversity of marine life in Hopkins reef, along with the recognized ecological significance of kelp forest communities, make this an appropriate site to address the value of qualitative data in describing marine ecosystems.

We addressed this question by looking at how the individuals in a buddy pair differed from each other while collecting qualitative data on the same transect. A substantial difference between buddy pairs would indicate that the qualitative results weren’t reliable, since two individuals looking at the same species assemblage recorded different results.

The abundance data we collected on species also allowed us to determine which species were more accurately described by this qualitative survey, and analyze what qualities might make a species a better candidate for qualitative sampling.

Methods

To investigate the value of qualitative data in the description of a kelp forest ecosystem, we sent 14 buddy pairs of scientifically certified divers into Hopkins Marine Sanctuary to collect qualitative data on the marine algal, fish and invertebrate assemblages.

To determine the validity of our results, we needed to see if the individuals in a buddy pair taking data on the same transect at the same time would come up with the same species abundances. To this end each buddy pair was assigned a meter mark where they ran a 30m transect offshore and onshore, taking qualitative data in and out for a selection of algae, fish and invertebrates. For each leg a diver would mark a species either absent (1), rare (2), present (3), common (4), or abundant (5). This allowed us to compare the buddy error between different species and between species of different mean abundance, to ascertain which species (if any) are well suited to qualitative analysis.

To retain objectivity in the results, buddy pairs were instructed not to compare their data during or after sampling. Also, they were told not to discuss ahead of time what strategies they would use or how they would define abundance. This allows us to see how well qualitative data taken by buddy pairs agrees when terms have not been defined ahead of time, and leaves room for suggesting standards that might be applied to improve the quality of such data collection.

Results

The ease of qualitative sampling is offset by a high level of inaccuracy. We found that buddy pairs differed markedly in their sampling. The difference between buddy pairs caused more variance in the data than depth, and only slightly less than meter mark (fig 1). If we were looking for patterns of species depth distribution, or other patterns of species association, distribution or abundance, our results would likely be obscured by this level of error.

Figure 1

Some species counts suffered more from buddy pair error than others. Species of medium abundance showed the worst relative difference between buddies, whereas very rare and very abundant and easily recognizable species showed the least (fig 2). Species of apparently medium low abundance, those averaging between rare (2) and present (3), showed the highest relative difference between buddies. This pattern is especially apparent in the invertebrate data. Examples of such species include the orange cup coral (Balanophyllia elegans), the ring topped snail (Calliostoma ligatum), the sand anemone (Pachycerianthus fimbriatus), the great spined star (Pisaster giganteus) and the ball sponge (Tethya aurantia) (figs 3 and 4).

Confusion[jf1] about a specific species wasanothera source of buddy error. The orange cup coral (Balanophyllia elegans) had the highest relative buddy difference of any of the species (fig 4). Since we didn’t establish protocols for counting colonial species like cup coral before our dives, this result is not surprising. Should a large, single colony of cup coral count as abundant, or rare? This discrepancy in our results demonstrates the importance of establishing a set of standards beforeengaging in qualitative sampling, but it also highlights the subjective nature of qualitative data.

The algae with the highest buddy error were Dictyoneurum californicum and Dictyoneurum reticulatum (fig 4). Given both the morphological similarity of these species and the similarity of their names, it seems likely that divers simply confused one species with the other.

Very rare species and very common, easily recognizable species yielded the most reliable qualitative data results. For rarely spotted species like the black and yellow rockfish (Sebastes chrysomelas), the red abalone (Haliotis rufescens), and the white spotted anemone (Urticina lofotensis), this may only be because it was easy for buddy pairs to agree on not having seen any (figs3 and 4).

Giant kelp (Macrocystis pyrifera), chain bladder kelp (Cystoseira osmundacea), and the bat star (Patiria miniata), the three species identified as most abundant, tell a more interesting story. All three are easily recognizable species. Giant kelp is perhaps the most obvious and prominent species in the kelp forest, while bat stars are familiar and abundant throughout Monterey Bay. Cystoseira, a common sight in Hopkins reef,has easily identified chain bladders and oak like leaves. Buddy pairs consistently identified these species as present, common or abundant, with relatively little disagreement between buddies (fig 3 and 4).

Figure 2

Figure 3

Figure 4

Discussion

We found the qualitative method of sampling to be unreliable, based on the high disagreement between buddy pairs. Species of medium abundance (rare to present) showed the worst relative difference between buddies, whereas very rare and very abundant and easily recognizable species showed the least. Colonial species like the orange cup coral and species easily mistaken for each other such as Dictyoneurum californicum and Dictyoneurum reticulatumwere particularly ill suited to qualitative sampling. Abundant and easily recognizable species like Macrocystis, Cystoseira, and Patiria showed relatively little difference between buddy pairs.

Our divers’ problems identifying orange cup coral and Dictyoneurum points to the importance of setting standards and undergoing thorough training prior to qualitative sampling. Protocol must be set to establish what qualitative measures mean. Where do we draw the line between “present” and “common?” Does a single colony of cup coral count as “rare” or “abundant?” These standards must be made clear to divers, and also outlined in any scientific article citing qualitative results, or the data is meaningless. Likewise divers must be trained to easily and correctly identify species that they will be surveying.

Our relative success in surveying Macrocystis, Cystoseira, and Patiria suggests that if divers were trained to the point where all species were as easily recognized as these three, qualitative sampling might be a more valid method. However, it also points to another flaw in qualitative sampling. Inexperienced divers have a tendency to count familiar and easily recognizable species more readily, while skipping species they haven’t learned to look for yet. This bias may be worse in qualitative sampling, since the subjective role of the divers’ judgment is increased, as individuals of a species are not counted but only marked in a category based on how many the diver thought they saw.

I was not surprised by the unreliability of our qualitative data. The low degree of reliability, based on the high relative difference between buddy pairs, implies that the data obtained is not very accurate. Given this, I don’t think that qualitative sampling approach is an appropriate method for describing trends of species abundances through time. With rigorous protocols and diver training, the accuracy of qualitative sampling might be improved. However, due to the subjective nature of qualitative sampling, quantitative methods are preferable. [jf2]

Works Cited

Graham, Mechael H. “Diversity and Structure of Southern California Giant Kelp Forest Food Webs” Ecosystems 2004 pp. 341-357

Davenport and Anderson. “Positive Indirect Effects of Reef Fishes on Kelp Performance: the Importance of Mesograzers” Ecology 88(6) 2007 pp. 1548-1561

Paddack and Estes. “Kelp Forest Fish Population in Marine Reserves and Adjacent Exploited Areas of Central California” Ecological Applications 10(3) 2000 pp. 855-870

Dayton, Tegnar et al. “Temporal and Spatial Patterns of Disturbance and Recovery in a Kelp Forest Community” Ecological Monographs 2(3) 1992 pp. 421-445

REEF 2011. Reef Environmental Education Foundation Volunteer Survey Project Database 10/10/2011

Results (25)

__0__/4 Figure legends Accurate

__0__/4 Figure Legends well composed (complete and concise)

__4__/5 Results organized according to questions

__4__/4 Graphs presented in a logical order, case made for the order

__2__/4 Grammar, sentence structure and spelling

__2__/4 Clarity and conciseness of writing

Discussion (25)

____/9 How well did they answer the questions they present in the Intro?

1)_2___/3 Discuss the results from the specific to the general.

2)__3__/3 Do these results surprise you? In other words, is the qualitative method more or less reliable than you thought it would be, and do you think that degree of reliability (which can be assessed based on relative difference between buddies) implies anything about accuracy?

3)__3__/3 Do you think the qualitative sampling approach is appropriate for describing trends of species abundances through time? Explain your answer

__2__/3 Grammar and Spelling

___2_/2 General Thoughtfulness

__1__/3 Clarity and conciseness

__3__/5 Organization of discussion

__2__/3 Context and Bigger Picture

General Notes: Figures always require legends to give the reader enough information to interpret the graph. Refer to the guidelines. In the results, you want to stick to the results, specifically those that bolster your argument. Often it would have made your points stronger to include values from the graphs.

[jf1]This sort of speculation should be saved for the discussion. Present results in the results….speculation goes into the discussion.

[jf2]You focus a lot on the error associated with different observers based on them having not learned the species. What if observers were able to identify species….then what? Would qualitative be a reliable method? If so, when?