Andrew C. Davenport and Todd W. Anderson. 2007. Positive indirect effects of reef fishes on kelp performance: the importance of mesograzers. Ecology 88:1548–1561.

Appendix A. Procedures for sampling kelp blades, mesograzers, and encrusting epifauna, obtaining estimates of grazing intensity, and validating grazing effects, and a table reporting taxonomic composition of invertebrates collected from kelp blades.

Sampling procedures

Kelp blades and their associated invertebrates were collected by divers using SCUBA to quantify the biomass of mesoinvertebrates and the percentage cover of encrusting epifauna per unit blade surface area. Five samples were taken over a 55-d period every 11 d from 23 Jul to 15 Sep 2003. A baseline sample was collected on 23 Jul 2003 to determine the initial conditions and differences in the invertebrate assemblage between treatments and Habitat Reef, a nearby natural reef. After samples were collected on 5 Sep, all pens and mesh panels were removed from the plots. A final sample was collected on 15 Sep 2003 to determine if the biomass of invertebrates was reduced following the removal of pens.

In the collection of mesograzers, blades of giant kelp were removed from experimental plots or Habitat Reef in two vertical zones: lower (L: 0.5–3.2 m above the holdfast) and upper (U: 3.3–5.7 m above the holdfast). The top 0.3 m of kelp was not sampled because of damage to blades when the canopy was trimmed to standardize the height of fronds. The initial overall mean (± 1 SE) number of fronds per plot across all treatments was 147 ± 5.8, with no significant difference in the number of fronds per treatment (ANOVA: F_{2, 9} = 1.73, P = 0.23).

To collect kelp blades, two giant kelp were selected randomly from the six giant kelp on each plot. Random numbers were then used to select a frond from a held bundle as fronds were released, and a set of random heights was then used to select two kelp blades from the frond, one from each vertical zone (upper and lower zones). From this random height on a frond, the first blade encountered was removed under the following three criteria: (1) the blade was not a reproductive sporophyll, (2) the blade was not within 1-m of an apical meristem, and (3) the blade was not senescent. For a single sampling period, four blades were collected per plot (16 blades per treatment). During the same sampling period, eight individual kelp were randomly selected from the perimeter of the kelp forest at Habitat Reef, with two blades removed per individual as described above. A nytex mesh bag (51 cm long × 25.5 cm wide; 233 µ mesh) was carefully maneuvered over each blade to ensure collection of all motile invertebrates. One bag was used to collect both blades (L and U), combining invertebrates for each individual giant kelp.

In the laboratory, kelp blades were first washed with pressurized salt water to remove invertebrates. The blades then were soaked in fresh water and vigorously agitated to remove any additional invertebrates (Coyer 1979). The invertebrates were washed into a 200-µ sieve and preserved in 10% buffered formaldehyde for 24 h before an additional rinse and preservation in 70% ethanol. All invertebrates were identified to species or lowest taxonomic level and measured to the nearest 0.04 mm using a dissecting microscope and an ocular micrometer. Individual species were categorized into the following invertebrate groups: gammarid amphipods, caprellid amphipods, isopods, molluscs, mysids, carideans, copepods, ostracods, and ‘other’ (Table A1). Invertebrate biomass was determined by length-weight ratios in accordance with Coyer (1979, 1984, 1987). Mesograzers were identified as herbivores from the invertebrate groups according to their natural history (Coyer 1979). Mean total invertebrate biomass and total mesograzer biomass were calculated for each treatment across sampling periods. The biomass (g dry mass) of kelp blades was determined by placing all blades in drying ovens at 70^o C for 5 d. Invertebrate biomass was standardized as milligrams per 100 cm² blade surface area except for comparisons of fish density and invertebrate biomass among reefs (observational component), in which invertebrate biomass was standardized to kelp blade biomass (g). The percentage cover of epiphytes (mm²) was estimated using a clear plastic grid of 3-mm × 3-mm squares; epiphytes were removed and their dry weight obtained to calculate biomass. Because of the overwhelming number of encrusting epifaunal spirorbids, an index was created for their mean weight and percentage cover.

Estimates of grazing intensity and validation of grazing effects

To quantify grazing, the surface area of kelp blades was determined by photographing each blade with a digital camera. Image-editing software (Adobe Photoshop 7.0) was used to convert photographs into black and white images, which were then analyzed with digital-imaging software (Media Cybernetics Image Pro Plus 4.0). The surface area of kelp blades and the area of tissue removed from apparent grazing (holes in the blade) were obtained to calculate the percentage hole area per blade as an estimate of grazing. Although similar holes have been attributed to mesograzers in feeding experiments (Jones 1971), an experiment was conducted to verify that estimates of grazing using hole area were valid and not a result of the senescence of kelp blades or damage caused by other organisms. Twenty-four blades were randomly collected from individual giant kelp at Habitat Reef. In the laboratory, a 3.5 cm × 3.5 cm piece of each blade was removed just above the nematocyst. Initial surface area and wet weight of each blade piece was obtained using the methods described previously. Each piece of blade was placed into a 15-cm long × 3.5-cm diameter plastic cylinder secured on the ends with 200-µm mesh Nytex. Containers were assigned one of two treatments: (1) herbivores absent or (2) herbivores present. Mesograzers were collected from two randomly selected kelp blades using the methods described previously (with the exception of a freshwater rinse) and transferred into a cylinder designated to have herbivores present with a piece of kelp blade. All cylinders were placed in a flow-through seawater system. After 5 d, blade pieces were removed and the surface area and wet weight of pieces were obtained.

Taxonomic composition of invertebrates collected from kelp blades

TABLE A1. Inverterate taxa recorded from samples of kelp collected in three experiments and an observational study. Unknown species are designated by letter groups, and asterisks denote invertebrates that are relatively rare in abundance. Species are denoted as grazers (Y) or non-grazers (N) according to criteria by Coyer (1979).

LITERATURE CITED

Coyer, J. A. 1979. The invertebrate assemblage associated with Macrocystis pyrifera and its utilization as a food source by kelp forest fishes. Ph.D. Dissertation, University of Southern California, Los Angeles, California, USA.

Coyer, J. A. 1984. The invertebrate assemblage associated with giant kelp, Macrocystis pyrifera, at Santa Catalina Island, California: A general description with emphasis of amphipods, copepods, mysids, and shrimps. Fishery Bulletin 82:55–66.

Coyer, J. A. 1987. The mollusk assemblage associated with giant kelp, Macrocystis pyrifera, at Santa Catalina Island, California. Bulletin of Southern California Academy of Science 83:129–138.

Jones, L. G. 1971. Studies on selected small herbivorous invertebrates inhabiting Macrocystis canopies and holdfasts in southern California kelp beds. Nova Hedwigia 32:343–367.