Design of Field and Laboratory Experiments

Using field experiments, I quantified rates of Pisaster predation to test the hypothesis that the strength of the sea star-mussel interaction is reduced during periods of cold-water upwelling. Experiments were conducted at three wave-exposed sites within Neptune State Park. Sites were separated by several hundred meters, and water temperatures varied little (generally < 0.2°C) over these distances. At each site, I identified two large rocky reefs (mean area ± SEM = 132.5 ± 49.7 m2) that were in close proximity to one another yet isolated by surge channels. All sea stars were routinely removed from one reef in each pair and allowed to remain at natural densities on the other.

In April-May 1997, I transplanted 20 clumps of 50 mussels to the low intertidal zone on each of these reefs. Mytilus californianus were used in these experiments to gauge the influence of Pisaster on the competitively dominant space-holder (Paine 1992). Mussels (shell length = 4.5-5.5 cm) were placed in overlapping rows under



Mussel Transplantation

transplanted mussel clump

Figure 4.4. Experimental measurement of predation intensity. To the left is a transplanted mussel clump that has been uncaged and is being approached by several sea stars. To the right is a mussel transplant still protected from predation by a mesh cage.

Caged transplant transplanted mussel clump

Figure 4.4. Experimental measurement of predation intensity. To the left is a transplanted mussel clump that has been uncaged and is being approached by several sea stars. To the right is a mussel transplant still protected from predation by a mesh cage.

Vexar™ mesh cages that were screwed into the rock (Fig. 4.4). Cages held mussels in place, allowed them to reattach firmly to the rock with byssal threads, and protected them from being eaten by sea stars until cages were removed. This technique has been used successfully by other researchers in Washington and Oregon (Paine 1976, Menge et al. 1994, Navarrete and Menge 1996).

Beginning in mid June, I conducted five consecutive experiments to measure the intensity of sea star predation during periods lasting 14 days each. Starting dates were set a priori as the first day of each spring tide series, so that each experiment consisted of a similar 14-day tidal cycle. At the start of each experiment, I randomly selected four mussel transplants on each reef and removed their cages, thereby exposing these mussels to sources of mortality. I then recorded mussel survivorship daily for the first 6 to 7 days, and again on day 14 (study sites were inaccessible during neap tides, days 8-13). On each of these days, I also recorded the local sea star density (defined as the number of sea stars in a 1 m radius around each clump). Uncaged mussels that remained on day 14 were counted and then removed.

I tested whether variation in rates of predation was associated with changes in environmental factors. At each site I installed a data-logger (Optic StowAway™, Onset Computer Corp., Pocasset, MA) in the low intertidal zone that recorded water temperature (during high tide) or air temperature (during low tide) every 30 minutes. Data-loggers were positioned on horizontal surfaces that were exposed to full sunlight during daylight periods of aerial exposure. From these records at each site, I calculated high tide water temperatures, defined as the mean of all readings during a period from 2 hours before to 2 hours after each high tide. I also extracted the maximum temperature recorded during each low tide period of aerial exposure. Temperatures recorded by data-loggers during periods of aerial exposure can be used to accurately predict sea star body temperatures (and thus potential heat stress) at low tide (Sanford 1999a). The time of low and high tides was estimated using National Oceanic and Atmospheric Administration tide charts.

Since movement and feeding of consumers may be reduced during periods of increased wave stress (see references in Menge and Olson 1990), five wave force meters (Bell and Denny 1994) were deployed in the low zone at each site to record variation in maximum wave forces. Meters were read and reset every 24 hours for the first 5 to 7 days of each period.

I also examined the effects of water temperature on Pisaster feeding rates under controlled conditions in the laboratory. Sea stars were maintained under three temperature treatments: (1) constant 12°C, (2) constant 9°C, and (3) a treatment that alternated between two weeks at 12°C, and two weeks at 9°C.The constant 9°C and constant 12°C treatments simulated water temperatures that sea stars would experience during prolonged periods of upwelling or no upwelling, respectively. The alternating treatment simulated the fluctuating water temperatures that are typical of the Oregon coast during upwelling season.

In early June 1996, I collected 48 sea stars (wet weight: range = 118-138 g) from a single site within Neptune State Park. Four individuals were randomly assigned to each of 12 closed 110-liter tanks at Hatfield Marine Science Center in Newport, Oregon. All tanks were held in a cold room at 6 to 8°C, and water temperatures were elevated to set levels and self-regulated to ± 0.1°C using heaters and controllers (Omega Engineering Corp., Controller Model CN76120).Temperatures in treatments were verified independently throughout the experiment by data-loggers submerged in the tanks.

Water was circulated and oxygenated by two pumps in each tank. Water quality was maintained by an in situ wet/dry filter, foam frac-tionater, and routine water changes with filtered seawater (25% of tank volume/week). Salinity was maintained at 36 ± 1%o, and the experiments were conducted under a 12-hour light: 12-hour dark schedule.

All sea stars were initially acclimated for 10 days at 11°C without food. Treatments were then randomly assigned (n = 4 tanks/treatment) and sea stars were provided with mussels ad libitum. Mytilus trossulus (shell length = 3.0-4.0 cm) were selected for these experiments since this species is the most common prey item in the diet of Pisaster at these sites (Navarrete and Menge 1996, Sanford 1999a), occurring in ephemeral low zone patches below the M. californianus beds. The number of mussels consumed per tank was recorded every 14 days.

0 0

Post a comment