Aranha et al., Primnoa pacifica growth and paleoceanography 

Growth rate variation and potential paleoceanographic proxies in Primnoa pacifica: insights from high-resolution trace element microanalysis.

Renita Aranha^a1,, Evan Edinger^a,b,c2, Graham Layne^a, Glenn Piercey^d

^a Department of Earth Sciences, Memorial University of Newfoundland, St John’s,, NL  A1B 3X5, Canada, renitaaranha@gmail.com, eedinger@mun.ca, gdlayne@mun.ca,

^b Department of Geography, Memorial University of Newfoundland, St John’s, Canada, NL  A1B 3X9

^c Department of  Biology, Memorial University of Newfoundland, St John’s NL  A1B 3X9, Canada

^d CREAIT, Memorial University of Newfoundland, , St John’s, NL A1B 3X9, Canada, glennp@mun.ca

1Current Address: Divestco Inc 400, 520 – 3rd Avenue SW Calgary, Alberta, T2P 0R3

2: Corresponding author. +1-709-864-3233, fax +1-709-864-3019

Abstract:

Red tree coral, Primnoa pacifica, is one of the more common habitat-forming deep-sea gorgonian corals in the Northeast Pacific Ocean, growing in colonies up to 2 m high and living decades to hundreds of years.  Growth characteristics of Primnoa pacifica were studied in Dixon Entrance, northern British Columbia, and the Olympic Coast National Marine Sanctuary, Washington State, USA, based on samples collected in July 2008. To minimize the impact of scientific sampling on coral populations, only dead coral skeletons and dislodged live corals were collected. Ages and growth rates were measured using band counts, and checked against AMS-¹⁴C ages of gorgonin rings. Ba/Ca, Mg/Ca, Na/Ca and Sr/Ca ratios in the calcite cortex were measured using radial Secondary Ion Mass Spectrometer (SIMS) transects with a spot size of < 20 m and separation distance of 25 m. Growth banding was consistent in width between the central mixed zone consisting of calcite and gorgonin and the dominantly calcite cortex. Average annual radial growth rate of the nine corals analysed ranged from 0.23 to 0.58 mm/yr, with an average growth rate of 0.32 mm/yr in Dixon Entrance and 0.36 m/yr in OCNMS.  These growth rates are slightly higher than P. pacifica growth rates from the Gulf of Alaska, and more than 4 times the growth rates of sister species P. resedaeformis in the Northwest Atlantic. Primary productivity is likely a more important driver of geographic variation in Primnoa growth rates than temperature or current strength. Both Dixon Entrance and OCNMS are areas with high primary productivity and strong tidal currents.   Lack of post-Atomic Bomb radiocarbon in all but one of the gorgonin samples, and long radiocarbon reservoir ages in the Northeast Pacific, made radiocarbon-based verification of coral ages and growth rates difficult due to wide errors in calibrated age estimates. Mg/Ca and Sr/Ca ratios were inversely correlated in two of the three corals analyzed, and showed evidence of interannual variation.  Mg/Ca ratios ranged from 70-136 mmol/mol, and Sr/Ca ratios from 2.041–3.14 mmol/mol. Previously published relationships between gorgonian calcite Mg/Ca and seawater temperature yielded average temperatures matching ambient measurements, but the intra- and inter-annual variation in apparent temperature based on Mg/Ca ratios was more than double the observed variation in modern seawater temperature ranges in the region. Annual variation in Mg/Ca and Sr/Ca could be related to seasonal changes in precipitation efficiency, which is likely a function of short-term fluctuations in coral growth rate, in turn related to variation in primary productivity.  Seasonal and interannual variations in food availability, driven by primary productivity, may affect skeletal growth rate, hence in Mg/Ca and Sr/Ca ratios.  Primnoid coral skeletal microgeochemistry probably records temporal changes in both temperature and primary productivity.

1 Introduction:

      Primnoid gorgonians are important long-lived habitat forming corals in cold-water coral provinces world-wide.  Their size and longevity make them highly vulnerable to damage from fisheries and other seafloor disturbances.  Primnoa are of interest biologically for their role in structuring habitat (DuPreez and Tunnicliffe 2011), and geologically for their calcite-protein skeletons, which may be useful as paleoceanographic archives (Sherwood et al., 2005a; b; 2011).  Further knowledge of their growth rates and longevity, and of regional variation in their growth rates, can be important in understanding recovery times from natural or anthropogenic disturbances, and designing appropriate conservation strategies.  High-resolution studies of their skeletal geochemistry can help to evaluate their utility as paleoceanographic archives.  Here we present growth and paleoceanographic records in the most common Primnoa of the northeast Pacific Ocean, Primnoa pacifica.

      Primnoa pacifica, like its Atlantic counterpart Primnoa resedaeformis, has a skeleton composed of three distinct growth zones. It has an inner central rod, a middle zone composed of both calcite and gorgonin, and an outer calcitic zone, the cortex, which contains very little gorgonin (Sherwood et al., 2005b).   

      Smaller colonies only display two growth zones, the calcite cortex is absent. The growth bands can be clearly seen as intercalations of calcite and gorgonin in the second growth zone, and by changes in gorgonin colouration that remain visible after dissolution of the calcite (Risk et al., 2002). More diffuse growth bands can also be recognized in the calcite cortex. These growth bands are known to be annual in Primnoa pacifica (Andrews et al., 2002) and in its Atlantic sister species Primnoa resedaeformis based on radiometric dating (Sherwood, 2005c), and annual banding is also recorded in the calcitic cortex (Sherwood, 2005c). Primnoids are known to have long life spans (up to 700 yrs; Sherwood et al., 2006). Analysis of δ¹⁸O and Sr/Ca in the calcite skeletons of Primnoa resedaeformis has suggested that growth related kinetic effects might have an impact on these isotopic and elemental ratios (Heikoop et al., 2002). Preliminary studies on Primnoa resedaeformis indicate that the variation in Mg/Ca in its skeleton may be controlled by temperature (Sherwood et al, 2005a), and several authors have investigated potential paleoceanographic records in primnoids, bamboo corals (Isididae), and other cold-water gorgonian corals (e.g. Roark et al. 2005, Thresher et al., 2004, 2010, Hill et al., 2011, 2012). Mg/Ca and Sr/Ca in reef-forming tropical coral skeletons are very commonly used as proxies for sea surface temperatures (e.g., Mitsuguchi et al., 1996; Cohen et al., 2006; Azmy et al., 2010). Na/Ca and Ba/Ca also vary systematically in many types of marine biomineralization (Boyle, 1981; Lea et al., 1989; Lea & Boyle, 1990; Amiel et al., 1973), although their behaviour during skeletal growth, and their utility as proxies for seawater temperature or chemical variations, has been examined in far less detail in the available literature.

    Primnoids’ longevity, annual skeletal banding and wide geographic distribution make them potentially interesting from the standpoint of paleoceanography. The purpose of this paper is to measure the growth rates and longevity of P. pacifica in two locations, to investigate regional variation in the growth rate of P. pacifica and P. resedaeformis , and to assess the paleoceanographic significance of  trace element (Mg/Ca, Sr/Ca Ba/Ca, and Na/Ca,) variation in the calcitic portion of Primnoa pacifica skeletons. Since P. pacifica is known to be a habitat forming deep sea coral (Andrews et al., 2002, DuPreez and Tunnicliffe 2011); an accurate determination of its growth rate is extremely important in order to gauge habitat recovery time after damage from deep sea trawling and similar disturbances.

2 Materials and Methods:

2.1. Study areas.

Dixon Entrance is a large elongate strait between Haida Gwaii, northern British Columbia, Canada, and Southeast Alaska, USA (Figure 1).  Dixon Entrance experiences strong tidal currents, and hosts a lush fauna of deep-sea corals including gorgonians, stylasterids, soft corals, sea pens, and rare antipatharians (Edinger et al. 2008; DuPreez and Tunnicliffe 2011, Neves et al, this issue).  Dixon Entrance was scoured during Pleistocene glaciations, and strong tidal currents limit post-glacial sedimentation, thus leaving winnowed glacial deposits on the sea floor in much of Dixon Entrance, along with large glacial erratic boulders and glacially scoured bedrock (Barrie and Conway 1999).  Learmonth Bank is a large granite massif in the western end of Dixon Entrance, which hosts abundant Primnoa pacifica, both on bedrock and on glacial erratic boulders (Figure 2A-B) (Edinger et al., 2008, DuPreez and Tunnicliffe 2011).

The Olympic Coast National Marine Sanctuary (OCNMS) sits at the northwest corner of Washington State, USA, facing Vancouver Island, British Columbia, Canada, across the Strait of Juan de Fuca (Figure 1B). 

         2.2 Field collection of specimens:

      Several dead colonies of Primnoa pacifica were collected in July 2008 from Dixon Entrance and the Olympic Coast National Marine Sanctuary (OCNMS) using the Remotely Operated Vehicle (ROV) ROPOS deployed from the Canadian Coast Guard ship John P. Tully (Table 1 Figures 1, 2). In Dixon Entrance, corals were collected from Learmonth Bank and surrounding glacial deposits (Figure 1A, Table 1).  In OCNMS, corals were collected in 266-312 m water depth near the head of Juan de Fuca Canyon (Figure 1B, Table 1). The specimens were collected using the ROV manipulator arms and were measured, photographed, and frozen immediately after collection (Figure 2).  

      2.3 Growth rate estimation:

      The radial growth banding in the mixed growth zone (consisting of alternating gorgonin and calcite) of P. pacifica can be observed quite clearly in polished cross-sections when photographed under UV light (Sherwood et al., 2005b; c). Since these growth bands are known to be annual (Andrews et al., 2002, Sherwood et al., 2005c) longevity estimates were made by counting the number of bands in the mixed growth zone. Three independent readers counted rings in this zone for each colony. The average number of bands for each colony was then divided by the length of the traversing radius of this zone only to obtain the average annual radial growth rate. In order to determine longevity, the average annual growth rate was extrapolated over the length of the whole radius (including the calcite cortex region). Axial growth rates could not be determined accurately as most colonies were broken and fragmented after death and before collection.

      To better enumerate the banding in the calcite cortex region, cross sections from the base of the colonies were mounted on 25mm x 75mm or 50mm x75mm glass plates, (depending on the size of the specimen) thin sectioned to 100 µm and petrographically polished. These polished thin sections were then scanned with an HP Scanjet 3970 scanner using the greyscale ‘photograph’ setting at 2400 DPI resolution. This procedure allowed growth banding to be observed and counted in the calcite cortexes of two of the specimens used for detailed trace element profiling (R1162-0015, R1156-0016), using essentially the same procedures as for the calcite-gorgonin middle zones. Growth rate in the cortex appeared very similar to that in the calcite-gorgonin middle zone (Table 4), validating the extrapolation of distance vs time in specimens where the cortex growth banding is less clearly evident. Figure 3a shows an original raw photoscanned image of specimen R1162-0015. Digital level pass and gradient filters were then applied to the same image to emphasize the visibility of growth banding in the calcite cortex (Figure 3b). 
 
 

       2.4 Age validation by Radiocarbon dating:

       Three colonies of P. pacifica were chosen for age validation by radiocarbon dating (Figure 2C-H). These colonies were chosen for their relatively large calcite cortex region and macroscopically non-degraded skeletons, features that would permit detailed SIMS analysis of the calcite cortex along with corresponding ¹⁴C dating of the gorgonin rings. Calcite was not directed aged because carbon in the gorgonin layers is derived directly from recently fixed organic carbon, hence tied to surface waters, while carbon in the calcite layers records ¹⁴C of the ambient water mass (Sherwood et al. 2008b). All three colonies selected for ¹⁴C dating were from the OCNMS. The colonies were dead when collected, and had no soft tissue on their outer surface. It was initially assumed that the colonies had died within the last 10 years; thus, it was expected that the colonies would record the radiocarbon bomb spike that occurred between 1958 and the early 1970s. Basal cross-sections of the three colonies were cut using a water-cooled diamond-blade rock saw. Up to seven gorgonin rings were isolated from each cut sample using the method described by Sherwood et al. (2005c). Each isolated gorgonin ring was packaged separately and then ¹⁴C dated at the Center for Accelerator Mass Spectrometry at Lawrence Livermore National Laboratory, California. ¹³C of the gorgonin samples could not be measured concurrently, and the regional average ¹³C value of -15‰ was used for all samples.  The reported radiocarbon ages in Table 5 are in radiocarbon years using the Libby half life (5568 years) and following the conventions of Stuiver and Polach (1977).

      In order to convert the raw radiocarbon data into calendar years, a local radiocarbon marine reservoir effect was calculated. The regional reservoir effect at the OCNMS and southwestern Vancouver Island was calculated as 693±76 ¹⁴C yrs, based on the average reservoir effect documented for seven non-estuarine areas in the Northeast Pacific close to the area of collection (Table 2) (McNeely et al., 2006). This reservoir age

was subtracted from the (δ¹³C -corrected) ¹⁴C age and this resultant age was converted to calendar years using the Intcal09 calibration dataset (Reimer et al., 2009) via the CALIB 6.0 Radiocarbon Calibration Program (http://calib.qub.ac.uk/calib/calib.html).  Ages are reported in both ¹⁴C years and calibrated years relative to 1950. 

             2.5 Sample preparation for trace element analyses:

      Cross sections between 3mm and 5mm in thickness were cut from the base of the same three colonies selected for radiocarbon analyses, using a diamond-blade rock saw cooled with water. These cross sections were then cut into transverse strips using a thin kerf Buehler Isomet low speed diamond blade saw.  The calcitic cortex portion was subsequently isolated by simply breaking the gorgonin portion off by hand from the transverse strips. A corresponding continuous strip of the calcitic cortex could then be mounted into the 1” diameter SIMS sample ring - an aluminum ring with an outer diameter of 1 inch (25.4 mm). The SIMS samples were embedded in the ring using Buehler Epothin Epoxide (Resin:Hardener; 10:3.9).  The casts were polished using silicon carbide wet/dry sandpaper and then subsequently on a lapping wheel using 6µm diamond polish, then manually polished 0.5 µm and 0.03 µm deagglomerated alumina. 

      2.6 Trace element analysis: 

      Trace element analysis followed the approach of Cohen et al. 2001. A Cameca IMS 4f Secondary Ion Mass Spectrometer (SIMS) was used to perform high spatial resolution (<25 µm) spot analyses of Na/Ca, Mg/Ca, Sr/Ca and Ba/Ca in detailed traverses across the polished basal cross-sections from three specimens of P. pacifica. For the Mg/Ca and Sr/Ca spot analyses reported here, typical internal precisions (1) were better than 0.35% for Mg//Ca and 0.1% for Sr/Ca. SIMS also avoids the challenges involving i) high analytical noise and ii) poor signal reproducibility associated with LA-ICP-MS and EPMA analyses documented by Sinclair et al (2005).

      Individual SIMS transects began at the outermost edge of the calcite cortex region and ended at the beginning of the gorgonin/calcite mixed zone. The traverses were 7 mm to 11 mm long and individual SIMS spots were spaced 25 µm apart. Visible gorgonin rings were avoided during analysis.  Each coral consequently had between 303 and 367 SIMS analysis spots, depending on the length of the SIMS transect.

      SIMS analyses utilized bombardment of the samples with primary ¹⁶O^- ions accelerated through a nominal potential of 10 kV. A primary ion current of 3.0-6.0 nA was critically focused on the sample over a spot diameter of 10-20 m.

      Sputtered secondary ions were accelerated into the mass spectrometer through a nominal potential of 4500 V. Secondary ions were energy filtered using a sample offset voltage of -80 V and an energy window of 60eV to suppress isobaric interferences. Prior to each analysis, the spot was pre-sputtered for 120 s. This was designed to eliminate contamination from the 500Å gold coat and also to penetrate the damaged and homogenized surface layer of the mechanically polished sample. Analytical craters were thus typically < 20m diameter and <2 m deep at the completion of each analysis.

      Each analysis involved repeated cycles of peak counting on ²³Na⁺ (2s), ²⁴Mg⁺ (6 s), ⁴²Ca⁺ (2 s), ⁸⁸Sr⁺ (4 s), ¹³⁸Ba⁺ (10 s), as well as counting on a background position (22.67 Da; 1s) to monitor detection noise. A small wait time (0.2 – 0.5 s) was added between each peak switch for magnet settling. For this study, 15 cycles of data were collected over 546 s, for total analysis times of <10 minutes per spot, including pre-sputtering. Typical signal on the ⁴²Ca⁺ reference peak was 10,000-25,000 cps.

      The Memorial IMS 4f is equipped with a High Speed Counting System (Pulse Count Technology Inc.) that produces dark noise background of less than 0.03 cps (2 counts per minute) when used with an ETP133H discrete dynode electron multiplier. Overall system dead time in pulse-counting mode is 14 ns. This system also produces very low detection limits for the elements studied. The elemental detection limits based only on the uncertainty in correcting for detector dark noise (0.03 cps) are typically 1 ng* g^-1 for Na, 2 ng*g^-1 for Mg, 2 ng*g^-1 for Sr and 2 ng*g^-1 for Ba. The error of individual spot analyses was estimated using the standard error of the mean of n cyclical measurements of each ratio during an analysis (internal precision).

      Detailed trace element profiles were plotted against a time axis (years before death; year of death= 0). Time was calculated based on the average annual radial growth rate. The smoothed data lines were produced using a second degree Savitzky-Golay type generalized moving average with filter coefficients determined by an unweighted linear least-squares regression and a second degree polynomial model.

      2.7 Seawater temperature data: 

      Seawater bottom temperature data for the depths within the OCNMS region from which the coral samples analyzed were collected were extracted from the World Ocean Database 2009 (Boyer et al., 2009).  All records within the bounding coordinates 48.0 – 48.5 N and 124.75- 125.5 W were collated, and analyzed within depth intervals of 260-280 m, 280-300 m, and 300-320 m. The mean, 95% confidence limits, minimum and maximum observed temperatures, number of years of observation, and number of data points for each depth class were calculated. 

      2.8 Paleotemperature estimation.

      The temperature of the seawater in which the corals grew was estimated using the empirically defined relationship between Mg/Ca and seawater temperature determined for Primnoa resedaeformis in the NW Atlantic (Sherwood et al., 2005a). 

(1) Mg/Ca (mmol.mol^-1) = 5(±1.4) T (ºC) + 64(±10)

The paleotemperature estimate for all points was calculated using this equation, then the curve was smoothed using a 5-point running average.  

3 Results: 

      3.1 Growth rates:

       All the growth rates listed in Table 3 and summarized in Figure 4 are calculated from band counting results in the gorgonin/calcite mixed zone. The growth banding in the calcite cortex was also clearly imaged in two of the six P. pacifica samples that were thin sectioned and scanned (Figure 3). Therefore a separate growth rate estimate involving the calcite cortex region was also possible for these samples. In both samples, the growth rate in the calcite cortex was quantitatively identical to the growth rate in the gorgonin/calcite zone in these samples (Table 4).

      The average radial growth rate of Primnoa pacifica specimens from the Dixon entrance is 0.323± 0.084 mm*yr^-1 (1σ). The average growth rate of Primnoa pacifica from OCNMS is 0.362±0.023mm*yr^-1 (1σ).  Published radial growth rates for the Gulf of Alaska (Andrews 2002) and for P. resedaeformis in the Atlantic are plotted for comparison (Figure 4)  

      3.2 Specimen age and radiocarbon validation of growth rates

      Radiocarbon ages (Table 5) indicated that all of the rings analyzed, except sample 65-02 (A), were pre-bomb (i.e., formed prior to 1958). Almost all resultant reservoir corrected radiocarbon ages were within error of each other.  Sample 65-02 (A), the outermost ring in specimen R1165-002, was of modern (post-bomb) age. The reservoir corrected radiocarbon ages indicate that none of the samples died more recently than 30 years ago. Specimen # R1165-002 of P. pacifica has been dead for at least 30 years, specimen # R1162-0015 has been dead for at least 45 years, and specimen R1162-0016 has been dead for at least 50 years (Figure 5).

      In aggregate, the samples belonging to coral R1162-0015 showed a discernible growth trend in radiocarbon age- with older radiocarbon ages for samples that were closer to the center of the coral (Table 5, Figures 4, 5a). However, the slope of the growth trajectories obtained through fitting the reservoir corrected radiocarbon ages for this sample is steeper than that obtained from annual band counting (Figure 6a). Sample R1165-0016 and Sample R1165-002 did not show any discernible trend in the radiocarbon ages (Figure 6b,c). 

      3.3 Trace element analysis:

      In the three specimens analyzed in detail by SIMS, the Sr/Ca varied between 2.04 and 3.14 mmol*mol^-1, Mg/Ca varied between 70.6 and 136 mmol*mol^-1, Ba/Ca varied between 0.0036 and 0.0550 mmol*mol^-1 and Na/Ca varied between 16.4 and 53.1 mmol*mol^-1 (Table 6). Mean Mg/Ca and Na/Ca values are similar for Samples R1162-0015 and R1162-0016, but distinctly lower for Sample R1165-002. There was a weak, but statistically valid, inverse correlation between Mg/Ca and Sr/Ca for samples R1165-0015 and R1165-0016, but a positive correlation for sample R1162-00 (Table 7).

      All three samples analyzed showed some degree of cyclicity in the Mg/Ca and Sr/Ca profiles (Figure 7). The Mg/Ca and Sr/Ca ratios in samples R1162-0015 and R1162-0016 (Figure 7a,b) showed an obvious inverse correlation, both short-term and, in R1162-0015, long term, as also reflected in Table 7. In Sample R1162-0015, a major simultaneous spike in Mg/Ca and Sr/Ca occurred at the beginning of the profile (Figure 7a, between year 2 and 3). This spike corresponds to the SIMS analyses spots that passed over a thin gorgonin ring. Because gorgonin has a higher concentration of many elements than calcite, these analyses spots were excluded from the statistical analysis.

      In both samples R1162-0015 and R1162-0016 (Figure 7a-b) a single large dip in Sr/Ca values was observed, with a simultaneous spike in Mg/Ca. This excursion in trace element values noted between year 6 and 7 in sample R1162-0015 and year 7 and 8 in R11162-0016, was not associated with any specific contamination or growth feature in the cross-section. It is very distinct from the profile obtained when the SIMS analyses spot crossed the gorgonin ring; as there is a negative excursion in Sr/Ca values; not a positive excursion as seen when the analyses crossed a gorgonin ring.

      Ba/Ca and Na/Ca did not show any distinct cyclicity in the SIMS profiles of any of the samples analyzed. This lack of cyclicity is probably due to the fact that Na/Ca and Ba/Ca are not only dissolved in the skeletal material but are also present within the skeleton as surface contaminants and/or particulate inclusions. Their patterns do not seem as readily intepretable in P. pacifica as those observed for Mg/Ca and Sr/Ca.

      3.4 Temperature records from the OCNMS region.

      The World Ocean Database temperatrure records for the OCNMS region extend from 1938 to 2006, and included 19 records between 260-279 m, 26 records between 280-299 m, and 38 records between 300-320 m.  The average temperatures ( 95% confidence limits) observed in these three depth classes were 6.60.2, 6.40.2, and 6.40.1, respectively (Figure 8).  The complete temperature ranges recorded in each depth zone were 5.6-7.6, 5.7-7.3, and 5.4-7.0, respectively.  Intra-annual temperature variation in the observed data from 280-320 m depth was less than 1C.

      3.5 Paleotemperatre reconstructions.

      Average calculated paleotemperatures in sample R162-0015 matched observed bottom water temperatures in the 300-320 m depth range, but the range in calculated paleotemperatures exceeded the observed range of temperatures by a factor of two (Figure 9a). Calculated paleotemperatures matched observed temperatures in the first nine years of the coral record in sample R1162-0016, but then dropped below observed temperatures in the last five years (Figure 9b).    The calculated paleotemperatures in sample R1165-0002 were appromiately 3C lower than the observed temperature at the 280-300 m depth range (Figure 9c).  

 

4 Discussion

4.1 Growth rate variations:

        The lowest radial growth rate recorded in this study was 0.22±0.09 mm.yr^-1 for a sample of P. pacifica from the Dixon entrance. This value closely matches the radial growth rate (0.18± 0.03 mm,yr^-1) for a colony of P. pacifica from Alaskan Dixon Entrance studied by Andrews et al., 2002. The growth rates for P. pacifica reported here thus support the conclusions of this previous study that these corals are slow to recover from damage due to trawling and other disturbances.

      Primnoa pacifica was formerly thought to be the same species as Primnoa resedaeformis (Cairns & Bayer, 2009). The growth rates of Primnoa resedaeformis from the Northwest Atlantic (Sherwood & Edinger, 2009) were more than 4 times lower than the growth rate of Primnoa pacifica determined in this study. A log fit titled ‘model maximum’ fitted through the four specimens of P.pacifica with the largest radius in Figure 10; indicates the upper limit of growth rates in our study, along with log fit curves for Dixon Entrance and OCNMS individually.  Comparison of growth characteristics of P. pacifica and P. resedaeformis from various locations, as compiled in Figure 10, indicates that the samples from different locations follow similar logarithmic growth trajectories, but with dramatically different radial growth rates. 

  Sherwood & Edinger (2009) attributed regional differences in growth rate of P. resedaeformis in the North Atlantic to differences in the intensity of tidal currents, suggesting that faster growth rates occur due to stronger tidal currents. Other possible drivers of regional variation in gorgonian coral growth rates include temperature and primary productivity.

      The California undercurrent is active in the OCNMS region. The peak speed of this current is 30 to 50 cm*s^-1, similar to the currents in the Hudson Strait (~40 cm*s^-1), and considerably slower than the tidal currents in Dixon Entrance (90 cm*s^-1, Crawford and Thompson 1991) and the NE Channel (Sherwood & Edinger, 2009). Thus, current velocity alone likely cannot explain the large difference in growth rate between the four locations considered.

      Similarly, temperature variation between the four locations does not appear to explain the observed differences in growth rates. Average bottom temperatures (95% confidence limit) at the depth zone from which corals were collected in the other three areas were 5.60.1ºC (Dixon Entrance, 300 m), 4.00.1ºC (Hudson Strait, 400 m), and 6.10.4ºC (Northeast Channel, 400 m, Boyer 2009).

      Primary productivity, and food supply, may best explain the differences in growth rates among the areas compared.  The California current system is one of the most productive ecosystems in the world (Carr, 2002). Since growth of most deep sea corals is dependent on POM flux, it is likely that the relative higher productivity in the NE Pacific is the dominant driver of the higher growth rates of P. pacifica in comparison to P. resedaeformis from the Hudson strait and NE channel (Jones & Anderson, 1994; Carr, 2002; Thomas et al., 2003b).  Ongoing work quantitatively compares environmental correlates of growth rate variation in several species of cold-water gorgonian corals (Neves and Edinger 2012). 

      4.2 Radiocarbon dating:

        Since the radiocarbon ages of most of the samples were pre-bomb, and within error of each other, a precise age determination was not possible. It can be concluded, however, that none of the samples died recently (within the past 30 years).  The radiocarbon age ranges of R1162-0015 and R1162-0016 are mostly overlapping, suggesting that the two corals may have lived at the same time (Figure 5), but the precise times when the two corals lived are difficult to determine.  The calcite layers of the two corals likely overlapped the early part of the bottom temperature data reported in Figure 8.  High precision dates of pre-bomb radiocarbon are unlikely to yield more precise age estimates from the northeast Pacific, due to the large reservoir ages in that region. 

    4.3 Trace elements:

            4.3.1 Primary controls on trace element variation in P.pacifica:

      The skeletal morphology of Primnoa resedaeformis is indistinguishable from that of Primnoa pacifica.. Sherwood et al. (2005a) analyzed the bulk skeletal composition of several specimens of Primnoa resedaeformis from the North Atlantic and found that the average bulk skeletal Mg/Ca varied between 86 and 118 mmol*mol^-1. Comparison of individual sample values to hydrographic temperature at their sites of collection yielded the relationship Mg/Ca (mmol*mol^-1) = 5(±1.4) T (ºC) + 64(±10).

      Although the calculated average bottom temperature in one of the three samples studied matched observed temperature, the range of variation in calculated bottom temperatures within this coral far exceeded the observed temperature variation (Figure 9a).  The fact that calculated average bottom temperatures matched observed temperatures in the beginning of sample R1162-0016 skeletal Mg/Ca record, but not at its end, is confusing, as no similar large-scale change was recorded in the oceanographic data. The skeletal Mg/Ca in samples of P. pacifica analyzed in this study varied between 70 and 136 mmol.mol^-1, with averages of 103, 106 and 81 mmol.mol^-1in the three samples studied in detail. Applying the Sherwood et al. (2005b) temperature relationship to Mg/Ca profiles obtained in this study implied temperature variation of ~3.5ºC to 12.4ºC in sample R1162-0015, ~2.4ºC to 14.4ºC in sample R1162-0016 and ~1.2ºC to 5.6ºC in sample R1165-0002. In OCNMS the highest ever temperature recorded at 300 m in the past 70 years, by CTD casts, was 6.96ºC, and the lowest was 5.4ºC (Figure 8, Boyer et al., 2009). Observed temperature variation alone cannot account for the range of observed variation in Mg/Ca profiles of the samples studied herein.  Thus subfossil Primnoa coral skeletons may be more useful for indicating average temperatures at some time in the past, and possibly for recording decadal-scale changes. High-amplitude-short wavelength fluctuations in Mg/Ca and/or Sr/Ca ratios likely reflect growth rate variations responding to changes in food availability, rather than highly-resolved annual or subannual temperature variations (cf. Roark et al. 2005; Kuffner et al. 2012, see section 4.4, below).

      The temperature dependence of Sr/Ca in most reef-forming tropical scleractinian corals is between -0.08 and -0.10 mmol*mol^-1/ºC (de Villiers et al., 1995; Gaetani & Cohen, 2006). Cold water scleractinian corals like Lophelia pertusa have a more extreme Sr/Ca sensitivity: approximately -0.18 mmol*mol^-1 (Cohen et al., 2006). The temperature dependence of Sr/Ca in abiogenic carbonate, however, is only -0.039 mmol*mol^-1/º C (Cohen et al., 2006).

      If the variations of Sr/Ca record in the OCNMS Primnoa skeletons were attributed solely to the observed temperature variation in the sampling areas, it would imply a temperature dependence of approximately -0.74 mmol.mol^-1/º C, seven times more sensitive than observed in any tropical coral, and four times higher than that for the cold water scleractinian L. pertusa. Factors other than temperature, most likely short-term fluctuations in growth and calcification rate (cf. Kuffner et al. 2012) evidently exert a major control on the variation in Mg/Ca and Sr/Ca ratios observed in P. pacifica.

      4.3.2 Trace element ratio correlations and possible diagenetic effects

      A significant inverse correlation between Mg/Ca and Sr/Ca was noted in the high resolution trace element traverses of two of the three specimens of P. pacifica analyzed (R1162-0015 and -0016; Figure 7a, b, Table 7).

      In a third specimen (R1165-002; Figure 7c; Table 7) the Mg/Ca and Sr/Ca ratios showed a significant positive correlation. R1165-002 shows several additional differences, relative to R1162-0015 and -0016, in terms of its trace element profile. These differences suggest diagenetic effects involving re-equilibration of the calcite skeleton with seawater/shallow pore water before collection:

      1) The mean Sr/Ca (2.39 mmolmol^-1; Table 6) was discernibly lower. This agrees with the empirical observation (e.g., Carpenter & Lohman, 1992) that Sr/Ca in abiotic marine calcite is consistently lower than for biotic calcite.

      2) The mean Mg/Ca for combined spot analyses of middle zone & calcite cortex of R1162-0015 and -0016 were 103.5 and 106.4 mmolmol^-1, respectively (Table 6). These values are closely comparable to the 100 mmolmol^-1 mean value for bulk skeletal analysis of recent Primnoa by Sherwood et al (2005). The mean Mg/Ca (81.4 mmolmol^-1; Table 6) is more than 20% lower for R1165-002; and more similar to the values expected in abiotic calcite. This value also yields anomalously low calculated temperatures relative to the observed range at the site of collection (Figure 9c). Conversely, specimens R1162-0015 and -0016 yield calculated temperatures that closely overlap those observed at their site of collection.

      Sample 1165-002 did not yield discernibly older radiocarbon ages in its gorgonin layers than the other two corals, nor was its macroscopic taphonomic condition noticeably different. Despite the lack of any obvious “fossilization” or recrystallization textures under the petrographic microscope, diagenesis seems the most likely explanation for the observed trace element pattern in R1165-002. All three specimens were dead for at least 60 years before collection, with potential for the skeletal material of specimen R1165-002 to have experienced a period of shallow burial on the seafloor during that interval, accelerating incipient diagenetic changes in the calcite composition through reaction with shallow marine pore waters. It also implies a need for caution in the use of obviously “sub-fossil” samples of Primnoa for paleoclimate proxies (e.g. Sinclair et al, 2005; Heikoop et al, 2002). Use of SIMS microanalysis to assess the polarity of correlation between Mg/Ca and Sr/Ca in this species appears to be a useful test to quickly assess if individual samples have undergone partial or wholesale diagenetic alteration. If the Mg/Ca and Sr/Ca ratios are positively correlated, then such samples should be avoided in paleoceanographic studies.

      Although the empirical expectation is that biotic marine calcite will also display a positive correlation between Mg/Ca and Sr/Ca (e.g., Carpenter & Lohman, 1992), this is largely based on compiled data on bulk skeletal analyses for shallow water marine species. It is notable that the seemingly intact specimens of calcitic P. pacifica in this study (R1162-0015 and -0016) show the opposite behaviour, mimicking the inverse correlation between Mg/Ca and Sr/Ca observed in aragonitic zooxanthellate tropical reef corals (e.g., Gaetani & Cohen, 2006). 

      4.4 Potential influence of primary productivity and growth rate on Mr/Ca and Sr/Ca ratios.

      Calculations performed by Gaetani & Cohen (2006) indicate that, at a constant temperature, when ‘precipitation efficiency’ (i.e. the mass fraction of carbonate precipitated from the calcifying fluid) increases then Mg/Ca values increase and Sr/Ca and Ba/Ca values simultaneously decrease. In L. pertusa the oscillations in the Sr/Ca ratios could be reproduced by two-fold variation of the assumed precipitation efficiency, coupled with the observed temperature dependence of the partition coefficients determined from the abiogenic aragonite (Cohen at al, 2006).

      Since only about 50% of the variation observed in the Sr/Ca ratios in P. pacifica can be attributed to temperature changes (Figure 9), the observed changes in Sr/Ca and Mg/Ca ratios would call for substantial variations of precipitation efficiency through the annual growth cycle.

      In biogenic carbonate growth, a change in precipitation efficiency (i.e. mass of carbonate precipitated/growth rate of biogenic carbonate) occurs due to changes in the saturation state of the calcifying fluids. In tropical corals this change in saturation state is linked to zooxanthellate photosynthesis which is, in turn, linked to changes in temperature and sunlight (Cohen & McConnaughey, 2003; Cohen et al., 2006). Since any change in precipitation efficiency would be primarily reflected as a change in the observed growth rate of the coral, factors controlling growth rate are reflected in the trace element changes observed in P. pacifica and similar deep sea corals (cf. Kuffner et al. 2012). Observations of Sr/Ca ratios in the cold-water gorgonian Corallium rubrum indicate that Sr/Ca ratios vary with skeletal density; i.e., they are indirectly coupled to growth rate (Weinbauer et al., 2000). Cyclicity of Sr/Ca in deep sea bamboo corals can be used as an indicator of growth rate, rather than being coupled directly to temperature (Roark et al., 2005).

      Since no changes in light occur at depths of ~300m (from where P. Pacifica was collected), and short-term variations in temperature in the area of collection are negligible ,, the primary factor driving changes in the growth rate is most likely a change in food availability (Miller 1995; Ferrier-Pages, 2003; Houlbreque et al., 2003; Houlbreque et al., 2005).

      Several corals, including zooxanthellate species, can meet part of their energy requirements by preying on zooplankton, phytoplankton, pico-nanplankton, dissolved organic matter and particulate organic matter (Tsounis et al., 2010; Ribes et al., 2003; Miller, 1995; Ferrier-Pages, 2003; Houlbrèque et al., 2003; Houlbrèque et al., 2004, Orejas et al., 2011). Stable isotope analysis of two commonly occurring cold water corals, Lophelia pertusa and Madrepora oculata, indicated that they might be omnivores and may primarily feed on mesozooplankton (Duineveld et al., 2004; Kiriakoulakis et al., 2005). Stable isotope (δ¹³C and δ¹⁵N) analysis indicated that P. resedaeformis likely feeds on phytodetritus supplemented by mesozooplankton (Sherwood et al., 2008). A previous study on P. pacifica suggested that it likely feeds on the same trophic level as P. resedaeformis (Sherwood et al., 2005b).

      In controlled laboratory experiments on Stylophora pistillata (a zooxanthellate scleractinian coral) it was noted that an increase in plankton feeding under constant water temperature increased the rate of both skeletal and tissue growth of the coral. This occurred under both light and dark conditions, indicating that feeding has a direct effect on the mass fraction of skeletal material precipitated in the absence of light or temperature changes. (Miller,1995; Ferrier-Pages, 2003; Houlbrèque et al., 2003; Houlbrèque et al., 2005). It was also noted that in Stylophora pistillata the amount of food ingested was proportional to food density and that the coral never reached a saturation of feeding capacity in the experiments (Ferrier-Pages, 2003). Growth rates almost equivalent to tropical corals were noted in Lophelia pertusa and Madrepora oculata specimens stored in dark conditions in aquaria and fed exclusively with zooplankton - with temperature variation during the experiments controlled to ~±0.5°C (Orejas et al., 2008, 2011).

      In aragonitic scleractinian reef corals, Sr/Ca ratio is inversely correlated with calcification rate, and with temperature, thus short-term variation in calcification rate causes high-amplitude fluctuations in Sr/Ca that mimic wide SST variations (Kuffner et al. 2012).  Thus, accurate short-term fluctuations in Sr/Ca-derived paleotemperatures from tropical reef scleractinians should incorporate growth rate into the Sr/Ca-SST calibration (Gaetani, et al. 2011). 

      For Primnoa calcite, in the absence of light, zooxanthellae, or substantial temperature variations, short-term changes in the feeding of the coral should be the major factor modulating the skeletal precipitation efficiency and, consequently, the trace element values. If the mass fraction of carbonate precipitated increases at a fixed temperature then we would expect to observe a decrease in the Sr/Ca ratios and an increase in the Mg/Ca ratios of the carbonate (Gaetani & Cohen, 2006).  By analogy with the tropical scleractinians, our results suggest that measured Mg/Ca and Sr/Ca ratios in cold-water gorgonian calcites may indicate changes in food supply, hence paleoproductivity, if seawater temperature variation is known (cf. Hill et al. 2012). Similarly, Ba/Ca in cold-water gorgonian coral skeletons may serve as a measure of paleoproductivity (Hill et al. 2011), although in our data, Ba/Ca ratios did not yield a readily interpreted signal. 

    

   4.4.1 Oceanography of the Olympic Coast National Marine Sanctuary:

      The waters of OCNMS, are subject to changes in physical, chemical and biological properties due to the California Current system (CCS) (Hickey et al., 2006). The CCS mainly includes the southward California Current, the wintertime northward Davidson Current, and the northward California Undercurrent (Hickey & Banas, 2003). The California Undercurrent (CUC) is of special interest with respect to our samples because it is very active in the area of collection. It is continuous at depths of about 100-400 m and likely carries larval fish, invertebrates and even phytoplankton seed stock (Hickey & Banas, 2003). The intensity of the CUC is known to attain its maximum values in late spring and early autumn (Collins et al., 2003), and is the source of much of the nutrient-rich water supplied to the shelf during coastal upwelling (Hickey & Banas, 2003).

      The seasonal upwelling (Huyer, 1983) in this area favours a large spring plankton bloom, followed by a smaller autumn plankton pulse (Anderson, 1964; Landry et al., 1989; Thomas & Strub, 2001). Landry et al. (1989) have reported an increased concentration in chlorophyll twice a year offshore of Washington State (between 50 to 90 km); one of these episodes occurs between February and April and the other in October. Thomas & Strub (2001) observed that, in the Pacific Northwest, chlorophyll concentrations greater than 2.0 mg m^-3 extend further offshore in late spring-summer (May- June) and that a second offshore extension occurs in late summer (September).

 

  4.4.2 Expected and observed trace element variations in P. pacifca based on 

the oceanography of the region:

      Although the average paleotemperature estimate for one of the corals analyzed matched observed temperatures (Figure 9a), observed temperature variation cannot account for the measured variation in Mg/Ca and Sr/Ca ratios in the Primnoa skeletons analyzed (figure 9 and section 4.3.1). Given that P. pacifica likely feeds on the downward flux of particulate organic matter, plankton and similar sources; the measured Sr/Ca and Mg/Ca ratios could respond to changes in the skeletal growth rate brought about by changes in food availability (Roark et al., 2005, Kuffner et al., 2012, and section 4.3.1, above). In reviews of the primary production, new production and vertical flux of organic carbon in the eastern Pacific it has been argued that production and vertical flux were directly related (Pace et al., 1987; Loubere & Fariduddin, 1999).

      The biannual increase in food availability (due to the two plankton blooms per year) should theoretically be observed as two cycles per year in Mg/Ca and Sr/Ca profiles. While an inverse correlation is apparent in Sr/Ca and Mg/Ca profiles in Figure 7, a biannual cyclicity is not obviously resolved.

      There are suggestions in our data that P. pacifica may also record more abrupt events. A particularly drastic decrease in Sr/Ca values is noted along the SIMS profile in R1162-0016 (between year 6 and 7) and R1162-0015 (between year 7 and 8). This decrease in Sr/Ca is accompanied by a simultaneous increase in the Mg/Ca ratios. This excursion is distinct from spikes caused by gorgonin-rich calcite samples because of the inverse relationship between Mg/Ca and Sr/Ca, as opposed to parallel Mg/Ca and Sr/Ca changes in gorgonin-rich spots. This inverse Mg/Ca-Sr/Ca event could be attributed to a single period of increased growth due to substantial increase in the availability of food. This single period of increased growth occurs ~7 years before death in sample R1162-0015 and ~8 years before death in sample R1162-0016. The corals were collected during the same ROPOS dive from adjacent areas. Thus the trace element profiles may provide a unique method of correlating growth patterns in corals from proximal colonies. While it is possible that both P. pacifica samples have recorded the same event, the age uncertainty of the radiocarbon ages from the two corals is high enough that it is impossible to demonstrate that the two Mg/Ca records precipitated at the same time.

      A substantial, unusual increase in phytoplankton biomass was recorded in the waters off Washington, Oregon and British Columbia in the spring of 2002 (Wheeler et al., 2003; Thomas et al., 2003) due to the invasion of cool, saline subarctic waters (Freeland et al., 2003; Bograd & Lynn, 2003).  Similar events could have occurred during the lifetime of R1162-0016 and -0015, leading to the spikes in the Mg/Ca and decreases in the Sr/Ca. Thus, elemental ratios in the calcite skeletons of Primnoa spp. may be useful in recording short term changes in productivity in the ocean, which reflect basin-scale changes in circulation, similar to the records of oceanographic change preserved in the gorgonin layers (Sherwood et al., 2011). This attribute would be most useful in samples harvested live, where exact year of death is easily established.

      The decreased Mg/Ca ratios, and apparently cooler water temperatures recorded in sample R1162-0016 indicate variation on the same temporal scale as the Pacific Decadal Oscillation (PDO), although with only one shift, a correlation with PDO is impossible to demonstrate.  The seasonal upwelling in this region is also influenced greatly by ENSO events, and a reduction in the nutrients and chlorophyll standing stock is known to occur in association with ENSO events (Carr, 2002; Corwith & Wheeler, 2002).  A strong ENSO event should be recorded as a dramatic dip in the Mg/Ca ratios with a simultaneous spike in the Sr/Ca ratios in response to decreased productivity.  Thus if a coral could survive through a strong ENSO event it might also record changes occurring in surface productivity due to the changes in upwelling. 

5 Conclusions:

      The annual radial growth rate of Primnoa pacifica calculated in this study was 0.23 to 0.58 mm*yr^-1. The lowest reported growth rate closely coincides with the previously reported growth rate for P. pacifica from the Dixon entrance (Andrews el al, 2002). This lends further support to studies indicating that recovery time from damage due to trawling and similar disturbances are very long for this species (Andrews et al., 2002) and other deep-sea gorgonians in Canadian waters (Sherwood and Edinger 2009).

      Primnoa pacifica and P. resedaeformis showed dramatic regional differences in radial growth rate, with growth rates in P. pacifica more than 4 times greater than in its northwest Atlantic sister species. Primary productivity is likely the variable best able to explain these regional differences in growth rate.

      Growth rates in the calcite cortex and gorgonin/calcite mixed zone are virtually identical, thus indicating that the growth rate of the coral remains constant in spite of a change in the skeletal growth mode.

      The previously established paleotemperature equation of Mg/Ca ratios in Primnoa resedaeformis, applied to P. pacfica, yielded average bottom water temperatures matching observed temperatures, but the range of variation in Mg/Ca ratios and calcualted paleotemperatures exceeded the range of variation in observed bottom temperatures.

      The Mg/Ca and Sr/Ca ratios in Primnoa pacifica calcite are likely controlled by surface water productivity changes rather than bottom water temperature alone. The changes in Mg/Ca and Sr/Ca thus have potential to record major changes in productivity, which in turn can provide evidence of other large scale annual and decadal changes in hydrographic conditions.

      The polarity of the correlation between Mg/Ca and Sr/Ca ratios indicated diagenetic change in gorgonian coral skeletons that was otherwise not readily apparent.  Original biogenic calcites retained a negative correlation, while diagenetically altered gorgonian calcite exhibited a positive Mg/Ca – Sr/Ca correlation. 

6. Acknowledgements.

      We thank the crew of the ROV ROPOS and CCGS John P. Tully for making field work possible. Canadian Hydrographic Service, Geological Survey of Canada, and Olympic Coast National Marine Sanctuary provided access to bathymetric datasets. O. Sherwood provided invaluable guidance in sample processing and analysis.  V. Lecours and B. Neves helped with cartography.  Joel Finnis assisted with analysis of archival oceanographc data.  M. Wisshak and D. Scott provided helpful comments on the thesis from which this paper is derived, and O. Sherwood, C. Dullo, and an anonymous reviewer provided helpful comments on the manuscript. This research was sponsored by the NSERC-funded Canadian Healthy Oceans Network - a university-government partnership dedicated to biodiversity science for the sustainability of Canada's three oceans.  Additional funding was provided by NSERC Discovery grants to EE and GL.

 

7 References cited:

Amiel, A. J., Friedman, G. M., and Miller, D. S. (1973) Distribution and nature of incorporation of trace elements in modern aragonitic corals. Sedimentology 20, 1: 47-64. 

Anderson, G. C. (1964) The seasonal and geographic distribution of primary productivity off the Washington and Oregon coasts. Limnol Oceanogr, 9, 284-302. 

Andrews, A.H., Cordes, E., Heifetz, J., Cailliet,G.M., Coale, K.H., Somerton, D., Munk, K. and Mahoney, B. M.(2002) Age and growth of a deep sea, habitat-forming octocorallian (Primnoa sp.) from the Gulf of Alaska, with radiometric age validation. Hydrobiologia 471: 101-111 

Azmy, K., Edinger, E., Lundbergm J., and Diegor, W. (2010) Sea level and paleotemperature records from a mid-Holocene reef on the north coast of Java, Indonesia. Int. J. Earth Sci. 99, 231-244. 

Barrie, J.V., and Conway, K.W. (1999) Late Quaternary glaciation and post-glacial stratigraphy of the Northern Pacific margin of Canada. Quater. Res. 51, 113-123. 

Boyer, T.P., J. I. Antonov, O. K. Baranova, H. E. Garcia, D. R. Johnson, R. A. Locarnini, A. V. Mishonov, D. Seidov, I. V. Smolyar, and M. M. Zweng, 2009: World Ocean Database 2009, Chapter 1: Introduction, NOAA Atlas NESDIS 66, Ed. S. Levitus, U.S. Gov. Printing Office, Wash., D.C. , 216 pp., DVD 

Boyle, E. A., (1981) Cadmium, zinc, copper, and barium in foraminifera tests,” Earth. Planet. Sci. Lett. 53, 1: 11-35. 

Bograd, S. J., and R. J. Lynn. (2003). Anomalous Subarctic influence in the  southern California Current during 2002. Geophys. Res. Lett. 30 (15):8020. 

Cairns S.D. and Bayer F.M. (2009). A generic revision and phylogenetic analysis of the Primnoidae (Cnidaria: Octocorallia). Smithson Contrib Zool 629:1–79. 

Carpenter, J. and Lohman, K.C. (1992). Sr/Mg ratios of modern marine calcite: Empirical indicators of ocean chemistry and precipitation rate. Geochim. Cosmochim. Acta

56: 1837-1849 

Carr, M.-E. (2002), Estimation of potential productivity in Eastern Boundary Currents using remote sensing, Deep Sea Res., Part II, 49, 59– 80. 

Cohen, A.L., Layne, G.D., Hart, S.R., and Lobel, P.S. (2001) Kinetic control of skeletal Sr/Ca in a symbiotic coral; implications for the paleotemperature proxy. Paleoceanography 16, 20-26. 

Cohen, A., McConnaughey, T. (2003) Geochemical perspectives on coral mineralization. Rev Mineral Geochem 54:151-187 

Cohen, A.L., Gaetani, G.A., Lundalv, T., Corliss, B.H., George, R.Y.,( 2006). Compositional variability in a cold-water scleractinian, Lophelia pertusa: new insights into “vital effects”. Geochem. Geophys. Geosyst. 7, Q12004 

Collins, C. A., Ivanov, L.M.  and Melnichenko, O.V. (2003), Seasonal Variability of the

California Undercurrent: Statistical Analysis based on the Trajectories of Floats with

Neutral Buoyancy, Phys. Oceanogr., 13 (3), 135–147 

Corwith, H., and Wheeler, P.A.  (2002), El Niño related variations in nutrient and chlorophyll distributions off Oregon, Progr. Oceanogr., 54, 361–380. 

Crawford, H., and Thompson, R.E. (1991) Physical oceanography of the western Canadian continental shelf. Continent. Shelf Res. 11, 669-683. 

Duineveld, G.C.A., Lavaleye, M.S.S., Berghuis E.M. (2004) Particle flux and food supply to a seamount cold-water coral community (Galicia Bank, NW Spain). Mar Ecol Prog Ser 277:13–23 

DuPreez, C.,Tunnicliffe, V., 2011. Shortspine thornyhead and rockfish (Scorpaenidaea) distirubtion in response to substratum, biogenic structures and trawling. Mar. Ecol. Progr. Ser. 425, 217-231. 
 

Edinger. E., Boutillier. J., Workman, G. Coral Distributions around Learmonth Bank, Northern British Columbia, Canada: Influence of Surficial Geology and Tidal Currents Deep Sea Coral Symposium. 4^th International Symposium on Deep Sea Corals, 2008 Wellington, New Zealand p.98 Abstract nr 6.04 

Edinger, E.N., Sherwood, O.A. (2012).  Applied taphonomy of gorgonian and antipatharian corals in Atlantic Canada: experimental decay rates, field observations, and implications for assessing fisheries damage to deep-se coral habitats. N. Jb. Geol. Palaont. Abh. 265, 199-218. 

Ferrier-Pagès C, Witting J, Tambutté E, Sebens KP (2003) Effects of natural zooplankton feeding on the tissue and skeletal growth of the scleractinian coral Stylophora pistillata. Coral Reefs 22:229–240 

Freeland, H. J., G. Gatien, A. Huyer, and R. L. Smith (2003), Cold halocline in the northern California Current: An invasion of subarctic water, Geophys. Res. Lett., 30(3), 1141. 

Gaetani, G. A., and A. L. Cohen (2006), Element partitioning during precipitation of aragonite from seawater: A framework for understanding paleoproxies, Geochim. Cosmochim. Acta, 70, 4617–463 

Gaetani, G.A., Cohen, A.L., Wang, Z., and Crusius, J. (2011) Rayleigh-based, multi-element coral thermometry: a biomineralization approach to developing climate proxies. Geochim. Cosmochim. Acta 75, 1920-1933 

Heikoop, J.M., Hickmott, D.D., Risk MJ, Shearer CK, Atudorei V (2002) Potential climate signals from the deep-sea gorgonian coral Primnoa resedaeformis. Hydrobiologia 471: 117-124 

Hickey, B.M., Banas, N.S., (2003). Oceanography of the Pacific Northwest coastal ocean and estuaries with application to coastal ecosystems. Estuaries 26 (48), 1010–1031. 

Hickey, B., A. MacFadyen,W. Cochlan, R. Kudela, K. Bruland, and C. Trick (2006), Evolution of chemical, biological, and physical water properties in the northern California Current in 2005: Remote or local wind forcing?, Geophys. Res. Lett., 33, L22S02. 

Hill, T.M., Spero, H.J., Guilderson, T., LaVigne, M., Clague, D., Macalello, S., and Jang, N. (2011).  Temperature and vital effect controls on bamboo coral (Isididae) isotope geochemistry: a test of the "lines method". Geochemistry, Geophysics, Geosystems Q04008, doi:10.1029/2010GC003443. 

Hill, T.M., LaVigne, M., Spero, H.J., Guilderson, T., Gaylord, B., and Clague, D., (2012). Variations in seawater Sr/Ca recorded in deep-sea bamboo corals. Paleoceanography 27, PA3202, doi:10.1029/2011PA002260.  

Houlbrèque, F., Tambutté, E. and Ferrier-Pagès, C. (2003). Effect of zooplankton availability on the rates of photosynthesis, tissue and skeletal growth of the scleractinian coral Stylophora pistillata J. Exp. Mar. Biol. Ecol. 296, 145-166. 

Houlbrèque, F., Tambutté E, Allemand D, Ferrier-Pagès C (2005) Interactions between zooplankton feeding, photosynthesis and skeletal growth in the scleractinian coral Stylophora pistillata. J Exp Biol 207:1461–1469 

Huyer, A. (1983) Coastal Upwelling in the California Current System. Prog. Oceanogr. 12:3 259-284 

Jones, E. P. & Anderson, L. G. (1994). Northern Hudson Bay and Foxe Basin water masses, circulation and productivity. Atmos.Ocean 32: 361-374 

Khripounoff, A., Vangriesheim, A. and Crassous, P. (1998) Vertical and temporal variations of particle fluxes in the deep tropical Atlantic. Deep-Sea Res. I, 45, 193–216. 

Kiriakoulakis, K., Fisher, E., WolV, G.A., Freiwald, A., Grehan, A. and  Roberts, J.M (2005) Lipids and nitrogen isotopes of two deep-water corals from the North-East Atlantic: initial results and implications for their nutrition. In: Freiwald A, Roberts JM (eds) Cold-water corals and ecosystems. Springer, Berlin, pp 715–729 

Kuffner, I.B.,, Jokiel, P.L., Rodgers, K.S., Andersson, A.J., and MacKenzie, F.T. (2012) An apparent "vital effect" of calcification rate on the Sr/Ca temperature proxy in the reef coral Montipora capitata. Geochemistry, Geophysics, Geosystems 13, doi: 10.1029/2012GC004128 

Layne, G.D. and Cohen, A.L.,( 2002), Advantages of Secondary Ion Mass Spectrometry for trace element studies of marine biomineralization, Geochim. Cosmochim. Acta, 66, A436 (abstr.) 

Lea D.W., Shen G.T., Boyle E.A., (1989) “Coralline barium records temporal variability in equatorial Pacific upwelling,” Nature 340, 6232: 373-376. 

Lea, D.W. and Boyle, E.A. (1990), Foraminiferal reconstruction of barium distributions in water masses of the glacial oceans, Paleoceanography 5: 719–742.  

Loubere,P., Fariduddin, M.,(1999). Quantitative estimation of global patterns of surface ocean biological productivity and its seasonal variation on timescales from centuries to millennia. Global Biogeochem. Cycles 13,115-133. 

Landry, M.R., Postel, J.R., Peterson, W.K., Newman, J., (1989). Broad-scale distribution patterns of hydrographic variables on the Washington Oregon shelf. In: Landry, M.R., Hickey, B.M. (Eds.), Coastal Oceanography of Washington and Oregon. Elsevier, Amsterdam (Chapter 1). 

McNeely, R., Dyke, A. S., and Southon, J. R. (2006) Canadian marine reservoir ages, preliminary data assessment, Open File 5049, pp. 3. Geological Survey Canada. 

Miller, M. W. (1995). Growth of a temperate coral: effects of temperature, light, depth, and heterotrophy. Mar. Ecol. Prog. Ser. 122:217-225. 

Mitsuguchi, T., Matsumoto, E., Abe, O., Uchida, T., Isdale, P.J., (1996) Mg/Ca thermometry in coral skeletons. Science 274, 961-963 

Neves, B.M., Du Preez, C., Edinger, E., submitted. Mapping coral and sponge habitats on a shelf-depth environment using multibeam sonar and ROV video observations: Learmonth Bank, Northern British Columbia, Canada. Deep-Sea Research II (this issue) 

Neves, B. de M., Edinger, E.N. (2012) Geographic and environmental variation in gorgonian coral skeletal growth rates. Geological Association of Canada annual meeting, St. Johns, Canada, Abstracts with program, p. 98. 
 

Orejas, C., Gori, A., and Gili, J. M.(2008) Growth rates of live Lophelia pertusa and Madrepora oculata from the Mediterranean Sea maintained in aquaria, Coral Reefs, 27, p. 255. 

Orejas, C.,Ferrier-Pages, C., Reynaud, S., Tsounis, G., Allemand, D., Gili, J.M. (2011). Experimental comparison of skeletal growth rates in the cold-water coral Madrepora oculata Linnaeus, 1758, and three tropical scleractinian corals. J. Exp. Mar. Biol. Ecol., 405, 1-5. 

Pace, M. L., Knauer, G. A., Karl, D. M., and Martin,  J. H.  (1987) Primary production, new production, and vertical flux in the eastern Pacific Ocean. Nature, 325, 803-804. 

Reimer, P.J., Baillie, M.G.L., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Bronk Ramsey , C., Buck, C.E., Burr, G.S., Edwards, R.L., Friedrich, M., Grootes, P.M., Guilderson, T.P., Hajdas, I., Heaton, T.J., Hogg, A.G., Hughen, K.A., Kaiser,K.F.,  Kromer, B.,  McCormac, F.G., Manning, S.W., Reimer, R.W., Richards, D.A., Southon, J.R.,Talamo, S.,  Turney, C.S.M.,  van der Plicht, J. and Weyhenmeyer, C.E. (2009) INTCAL 09 and MARINE09 Raadiocarbon age calibration curves, 0-50,000 years Cal BP .Radiocarbon, 51 (4). 1111-115 
 

Ribes, M., Coma,R.,and Gili,J.M. (1999)Heterogenus feeding n benthic suspension feeders : The natural diet and feeding rate of the temperate gorgonian Paramuricea clavata (Cnidaria:Octocorallia) over a year cycle . Mar Ecol Prog Ser,183,125-137. 

Ribes, M., Coma,R., and Rossi,S. (2003) Natural feeding of the temperate asymbiotic octocoral gorgonian Leptogorgia sarmentosa (Cnidaria:Octocorallia). Mar Ecol Prog Ser 254,141-15 

Risk. M.J., Heikoop, J.M., Snow, M.G., and Beukens, R. (2002) Lifespans and growth patterns of two deep-sea corals: Primnoa resedaeformis and Desmophyllum cristagalli. Hydrobiologia 471, 125-131. 

Roark, E.B., Guilderson, T.P., Flood-Page, S.R., Dunbar, R.B., Ingram, B.L., Fallon, S.J. and  McCulloch, M.T. (2005) Radiocarbon-based ages and growth rates for bamboo corals from the Gulf of Alaska. Geophys Res Lett 32:L04606 

Sherwood, O.A., Heikoop, J.M., Sinclair, D.J., Scott, D.B., Risk, M.J., Shearer, C. and  Azetsu-Scott, K. (2005a) Skeletal Mg/Ca in Primnoa resedaeformis: relationship to temperature. In: Freiwald A, Murray JM (eds) Cold-water corals and ecosystems. Springer, Berlin, p 1061–1079 

Sherwood, O.A., Heikoop, J.M., Scott, D.B., Risk, M.J., Guilderson, T.P. and  McKinney, R.A. (2005b) Stable istopic composition of deep sea gorgonian corals Primnoa spp.: a new archive of surface processes. Mar Ecol Prog Ser 301:135–148 

Sherwood, O.A., Scott, D.B., Risk, M.J., and Guilderson, T.P.. (2005c) Radiocarbon evidence for annual growth rings in the deep-sea octocoral Primnoa resedaeformis. Mar Ecol Prog Ser 301:129–134 

Sherwood, O.A., Scott, D.B. and Risk, M.J. (2006) Late Holocene radiocarbon and aspartic acid racemization dating of deep-sea octocorals. Geochim Cosmochim Acta 70:2806–2814 

Sherwood O.A., Jamieson, R.E., Edinger, E.N. and Wareham, V.E. (2008a) Stable C and N isotopic composition of cold-water corals from the Newfoundland and Labrador continental slope: examination of trophic, depth and spatial effects. Deep-Sea Res I 55:1392–1402 

Sherwood, O.A., Edinger, E.N., Guilderson, T.P., Ghaleb, B., Risk, M.J., Scott, D.B., (2008b) Late Holocene radiocarbon variability in Northwest Atlantic slope waters. Earth Planet. Sci. Lett. 275: 146-153 

Sherwood, O.A. and  Edinger E. (2009) Ages and growth rates of some deep-sea gorgonian and anthipatharian corals of Newfoundland and Labrador. Can J Fish Aquat Sci 66: 142–152 

Sherwood, O.A., Lehmann, M.F., Schubert, C.J., Scott, D.B., McCarthy, M.D., 2011. Nutrient regime shift ni the western North Atlantic indicated by compound-specific d15N of deep-sea gorgonians. Proc. Natl. Acad. Sci. USA 108, 1011-1015 

Sinclair, D.J., Sherwood, O.A., Risk, M.J., Hillaire-Marcel, C., Tubrett, M., Sylvester, P., McCulloch, M., Kinley, L. (2005). Testing the reproducitility of Mg/Ca profiles in the deep-water coral Primnoa resedaeformis: putting the proxy through its paces. In: Freiwald A, Murray JM (eds) Cold-water corals and ecosystems. Springer, Berlin, p 1039–1060. 

Sinclair, D.J., Williams, B., Allard, G., Ghaleb, B., Fallon, S., Ross, S.W., and Risk, M., (2011) Reproducibility of trace element profiles in a specimen of the deep-water bamboo coral Keratoisis sp.. Geochim. Cosmochim. Acta 75, 5101-5121. 

Smith, S.V., Buddemeier, R.W., Redalje, R.C., Houck, J.E., (1979). Strontium-calcium thermometry in coral skeletons. Science 204 (4391), 404–407 
 

Stuiver, M., Polach, H.A., (1977). Discussion: reporting 14 C data. Radiocarbon 19, 355–363 

Tsounis, G., Orejas, C., Reynaud, S., Gili, J.M., Allemand, D., Ferrier-Pagès, C. (2010) Prey-capture rates in four Mediterranean cold water corals. Mar Ecol Prog Ser 398:149-155 

Thomas, A. C., and P. T. Strub, (2001) Cross-shelf phytoplankton pigment variability in the California Current, Cont. Shelf Res., 21, 1157–1190 

Thomas, A. C., Strub, P. T. and Brickley, P. (2003a), Anomalous satellite measured chlorophyll concentrations in the northern California Current in 2001 – 2002, Geophys. Res. Lett., 30(15), 8022 

Thomas, A.C., Townsend, D.W., Weatherbee, R., (2003b). Satellite measured phytoplankton variability in the Gulf of Maine. Cont. Shelf. Res . 23, 971–989 

Thresher, R., Rintoul, S.R., Koslow, J.A., Weidman, C., Adkins, J., and Proctor, C., (2004). Oceanic evidence of climate change in southern Australia over the lsst three centuries. Paleoceangraphy 31, L07212, doi:10.1029/2003GL018869. 

Thresher, R.E., Wilson, N.C., MacRae, C.M., Neil, H., (2010) Temperature effects on the calcite skeletal composition of deep-water gorgonians (Isididae). Geochim. Cosmochim. Acta 74, 4655-4670. 
 

Weber, J. N. & Woodhead, P. M. J. (1972) Stable isotope ratio variations in nonscleractinian coelenterate carbonates as a function of temperature. Mar. Biol., 15, 293- 297 

Weinbauer, M. G., Brandstatter, F., and Velimirov, B. (2000), On the potential use of magnesium and strontium concentrations as ecological indicators in the calcite skeleton of the red coral (Corallium rubrum), Mar. Biol., 137, 801–809. 
 

Wheeler, P. A., Huyer, A. and Fleischbein J. (2003), Cold halocline, increased nutrients, and higher chlorophyll off Oregon in 2002, Geophys. Res. Lett., 30(15), 8021 

 

Figure Captions.

Figure 1: Location map. A. Learmonth Bank, Dixon Entrance, northern British Columbia, Canada.  B. Olympic Coast National Marine Sanctuary, Washington state USA.  Dots indicate locations of coral collections from both study areas.  Inset map shows regional geography.

Figure 2: A. Live Primnoa pacifica as observed in Dixon Entrance, ~350 m water depth, July 2008.  Note standing dead branches, also rockfish resting among branches of coral. B. Dead Primnoa pacifica skeleton as observed on the sea floor, Dixon Entrance, ~300 m water depth, July 2008.  This sample broke during collection. C-H: The three colonies of Primnoa pacifica chosen for radiocarbon dating and SIMS analyses and their corresponding basal cross-sections (photographed under UV light).  C-D: Sample R1162-0015 (top), E-F: sample R1162-0016 (middle) and G-H: sample R1165-0002 (bottom).  Red and yellow markers in cross-section photos (D,F, H) indicate the position of the gorgonin rings which were isolated for ¹⁴C dating.

Figure 3: A.  Original grayscale photoscanned image of thin section of specimen R1162-0015 showing alternating calcite and gorgonin layers in the mixed growth zone. White scale bar = 1 cm.  B. The same image with digital level and gradient filters applied to emphasize banding in the outer calcite cortex. 

Figure 4. Growth rate (mm.yr^-1) vs Age of the coral for P. pacifica. The error bars are 1σ.  Dixon entrance samples are represented by grey markers and OCNMS samples are represented by black markers. The dotted line represents the average growth rate of all samples.  Average growth rate estimates for P. pacifica in the Gulf of Alaska (Andrews et al., 2002) and P. resedaeformis from two sites in the northwest Atlantic plotted for comparison. 

Figure 5. Radiocarbon ages of P. pacifica gorgonin subsamples analyzed, reservoir-corrected ¹⁴C years 1 error.

Figure 6. Radiocarbon validation of growth rates and ages based on ring counts.  (A) Sample R1162-0015 (B) Sample R1162-0016.  (C) Sample R1165-002  Distance (From the center of the coral. Center is at 0 mm) vs reservoir corrected radiocarbon years. The radiocarbon age of the outermost ring was considered valid in plotting the age based on band counting. The error bars (1σ) for the age based on band counts are smaller than the size of the symbol in the graph.

Figure 7: Mg/Ca and Sr/Ca ratios vs. time (years) in the calcite cortex of P. pacifica. Time was calculated by extrapolating average annual radial growth rate over the calcite cortex. Error bars are ±1σ. The profiles start from the time of death of the organism (i.e. the outer edge), hence younger ages are to the left, and older to the right. The data presented has been smoothed using a second degree Savitzky-Golay filter. A. Sample R1162-0015. B. Sample R1162-0016. C. Sample R1165-0002.  In B, data are missing in the areas where visible gorgonin rings cut across the calcite cortex.

Figure 8. Observed bottom temperatures in the Olympic Coast National Marine Sanctuary (OCNMS) region. All data derived from the World Ocean Database 2009, within the bounding coordinates 48-48.5 N and 124.75-125.5 W. Mean temperatures  95% confidence limits, along with minimum and maximum observed temperatures and the duration of observations, in the depth intervals from which corals were collected: 260-280 m, 280-300 m, and 300-320 m. 

Figure 9. Calculated paleotemperatures from P. pacifica in OCNMS region, using the published equation of Sherwood et al. 2005, Mg/Ca (mmol.mol^-1) = 5(±1.4) T (ºC) + 64(±10). A: R1162-0015. B: R1162-0016. C: R1165-0002. In A and B, shaded band indicates maximum temperature range observed in World Ocean Database data for OCNMS region at 300-320 m depth, 1938-1983. In C, shaded band indicates maximum observed temperature range at 280-300 m depth, 1953-2006.

Figure 10: Comparison of radial growth rates of Primnoa resedaeformis from the Hudson strait ([1]: Sherwood & Edinger, 2009) and the Northeast Channel ([2]: Mortensen & Buhl-Mortensen, 2005) with those of P. pacifica from Dixon Entrance (this study, and [3]: Andrews et al, 2002), and P. pacifica from Olympic Coast National Marine Sanctuary (OCNMS; this study). All logarthimic best-fit curves were passed through the youngest and largest coral data points in the Mortensen & Buhl-Mortensen (2005) data set. Andrews et al (2002) provided radial growth data only for one colony of P. pacifica (Gulf of Alaska), thus only one data point for [3] exists in the graph. The fitted line labelled ‘Model Maximum’, is the logarithmic fit for the four coral specimens from this study with the largest radii.