The complete data set describes environmentally based variation in size-specific energy storage (dry mass pyloric caeca/dry body mass) and growth of live mass of the intertidal sea star Pisaster ochraceus in British Columbia, Canada. Variation in energy storage and live mass correspond to (1) naturally occurring or (2) experimentally induced changes in the zonation of their prey, the sea mussel Mytilus californianus and associated species. The relevant features of mussel zonation are the depth (vertical distance from a fixed tidal datum) of the lower boundary and the species/size composition of the lower boundary. Downward extensions of the boundaries comprised of masses of relatively young mussels incurred the greatest increases in energy storage and live mass.

Data are presented in 4 files:

(1) Energy storage - Observational study. Variation in energy storage at three sites sampled at irregular intervals from 2008 to 2014. Variables include live mass, dry mass of viscera, and dry mass of the remainder of the body of individual Pisaster ochraceus. Sample dates were chosen to represent natural inter-annual variation in mussel zonation, rather than tracking continuous time courses. 

(2) Storage-boundary depth relationships. Variables derived from observational study and include summary measures of energy storage (slope of the linear regression of dry viscera mass to dry body mass, for dry body masses< 120 g);  boundary depth (vertical distance from tidal datum); and the qualitative composition of the lower boundary.

(3) Energy storage and live mass - Experimental study. Records were made after 1-3 years of experimental manipulation of the mussel boundaries on experimental sites. Variables include initial live mass, final live mass, dry mass of viscera, and dry mass of the remainder of the body of individual Pisaster ochraceus recaptured in the three large-scale replicates of the translocation experiment. 

(4) Boundary depth - Experimental study.  Variables include the type of boundary manipulation, the change in boundary shore levels with respect to conventional tidal data, the resulting composition of the lower boundary, and years of manipulation. 

Further information on-site locations, methods of data acquisition, and definitions of the variables appear in Methods and Usage Notes below.

Data collection in the observational study: Collections of Pisaster for the observational part of the study were made on three sites near Bamfield Marine Sciences Centre, Barkley Sound, British Columbia (Table 1) Over the years of observation, the three sites showed varying degrees of the vertical shift and composition of the lower boundary of their mussel beds. Rather than sample every year at each of the three sites, we chose specific years to represent (1) the inter-annual range of lower boundary shore levels seen at a site; and (2) the variable composition of the lower boundary, from masses of juveniles to predominantly large adults.

Table 1. Observational study sites. Names, location, and years of sampling

for the three sites used to examine the relationship between size-specific energy storage and natural variation in mussel zonation.

For each site-year sample, we collected 25-50 sea stars from the center of the site, making sure to include all size classes from 5 cm to the largest arm lengths present. After collection at low tide and transport, the sea stars were continuously submerged in sea tables for 24-36 H (approximate duration of 2-3 tidal cycles) to allow rehydration before measureing live mass. Each sea star was allowed to drain superficial water for 60-90 seconds before weighing on a tarred electronic balance. Each was then dissected into the compoenents of interest (gonad, pyloric caeca, remainder of body) and the dissections dried to constant weight.We characterized size-specific energy storage as the mass of dry viscera relative to the dry body mass. In the period of sampling (July-August) viscera mass was >95% pyloric caeca, the remainder was gonad.  

Data collection in the experimental study: Pisaster collected from source populations off the experimental sites were tagged with passive internal transponders (PIT tags) and then randomly assigned and translocated to one of three replicated treatment sites: lower boundary expanded vertically downshore, lower recessed upshore, and unaltered (control). Boundary shore levels were manipulated by adding or removing members to the resident sea star populations. A total three such replciates were done during field seasons in 2007 and 2009 (Table 2).

Table 2. Experimental study sites. Names, location, years of data collection, and treatment for three replicates of the experiment.

At the end of the deployment periods (1-2 months depending on the year/relicate, not treatment group) tagged sea stars were recaptured during a 5-6 day spring tide series. The recaptures were held in sea tables for 24-36 hours after a rehydration period as live masses, and dry masses of dissected components were recorded. To accommodate the heavy workload of the 2007 run, energy storage measures were based on partial dissections. The central arm of the trivium was dissected, removing the viscera. The body, with the 4 remaining sets of viscera was cut into sections to promote quicker desiccation, and constant dry weights were taken for both fractions. Four times the dry mass of the trivium viscera was subtracted from dry mass of the partial dissection to estimate the dry body mass. The total dry viscera mass was estimated as 5X the trivium viscera dry mass. To validate this method, another group of sea stars was dissected completely and the dry mass of the trivium viscera recorded separately from the other four sets of viscera. Regression slopes of the actual masses (dry masses from complete dissections) vs the interpolated estimates (5X the masses from partial dissections) deviated from a perfect match (slope=1.0) by <3%. 

Measuring shore levels of the lower boundaries of mussel beds on each study site:  We recorded the shore level of the lower boundary as the distance below the tidal datum +3.05 m Mean Lower Low Water (+10 ft. MLLW), defining the boundary depth. The depth, species, and size composition of the lower boundaries of the mussel beds were estimated for both observational and experimental sites. Stainless steel bolts were set in the rock 5 m apart horizontally, straddling the approximate center of the bed, and along the approximate shore level of the lower boundary at the start of manipulations or observations.  For each survey, a horizontal tape was clipped to the bolts, and using sightlines normal to the tape at 1 m horizontal intervals, temporary markers were placed on the lower boundary of adults. When present as a continuous cover, juveniles were also marked. The locations of the markers in 3D space were measured relative to a permanent georeferenced datum (lat/long records, Tables 1 and 2) in the splash zone of each site. The spatial measurements were made using a TopCon^TM  254 Total Station, and the georeferenced data in Tables 1 and 2 are the Occupied Points of the Total Station.  

Parameters and definitions of variables for each of the four data files included in README file 'Robles_et_al_Readme.txt'