Coralline algal growth-increment widths archive North Atlantic climate variability

This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

Author's personal copy

Coralline algal growth-increment widths archive North Atlantic climate variability

J. Halfar a,⁎, S. Hetzinger a, W. Adey b, T. Zack c, G. Gamboa a, B. Kunz c, B. Williams a, D.E. Jacob c

a CPS-Department, University of Toronto at Mississauga, 3359 Mississauga Rd. N, ON, Mississauga, Canada, L5L 1C6b Department of Botany, Smithsonian Institution, 10th and Constitution Ave, NW, Washington, DC 20560-0166 Washington, USAc Earth System Science Research Centre, Department of Geosciences, Becherweg 21, Johannes Gutenberg-Universität, D-55099 Mainz, Germany

a b s t r a c ta r t i c l e i n f o

Article history:Received 9 November 2009Received in revised form 31 March 2010Accepted 9 April 2010Available online 18 April 2010

Keywords:Coralline algaePaleoclimateNorth AtlanticAtlantic Multidecadal OscillationSclerochronology

Over the past decade coralline algae have increasingly been used as archives of paleoclimate information.Encrusting coralline algae, which deposit annual growth increments in a high Mg-calcite skeleton, areamongst the longest-lived shallow marine organisms. In fact, a live-collected plant has recently beenshown to have lived for at least 850 years based on radiometric dating. While a number of investigationshave successfully used geochemical information of coralline algal skeletons to reconstruct sea surfacetemperatures, less attention has been paid to employ growth increment widths as a temperature proxy. Herewe explore the relationship between growth and environmental parameters in Clathromorphum compactumcollected in the subarctic Northwestern Atlantic. Results indicate that growth-increment widths of individualplants are poorly correlated with instrumental sea surface temperatures (SST). However, an averaged recordof multiple growth increment-width time series from a regional network of C. compactum specimens up to800 km apart reveals strong correlations with annual instrumental SST since 1950. Hence, similar to methodsapplied in dendrochronology, averaging of multiple sclerochronological records of coralline algae providesaccurate climate information. A 115-year growth-increment width master chronology created from modern-collected and museum specimens is highly correlated to multidecadal variability seen in North Atlantic seasurface temperatures. Positive changes in algal growth anomalies record the well-documented regime shiftand warming in the northwestern Atlantic during the 1990s. Large positive changes in algal growthanomalies were also present in the 1920s and 1930s, indicating that the impact of a concurrent large-scaleregime shift throughout the North Atlantic was more strongly felt in the subarctic Northwestern Atlanticthan previously thought, and may have even exceeded the 1990s event with respect to the magnitude of thewarming.

© 2010 Elsevier B.V. All rights reserved.

1. Introduction

Models used to project future climate scenarios rely upon analysesof historical and proxy information and their robustness is directlydependent upon the quality and temporal extent of such information.Instrumental data, particularly from the marine realm, graduallybecome sparse and spatiotemporally incomplete prior to the mid-20th century. These data gaps limit our ability to examine importantlow-frequency climate variations and thus diminish the accuracy offuture climate change projections.

Reconstructions of marine climate and oceanography have totherefore rely on archival information, often derived from long-livedmarine biota (e.g., Pfeiffer and Dullo, 2006; Wanamaker et al., 2008).

While numerous multicentury archival reconstructions of tropicalmarine climate variability have been generated over the past decades(e.g., Dunbar et al., 1994; Saenger et al., 2009), high-resolution andlong-term records of environmental changes in mid- to higherlatitude oceans are less common. Together with geochemicalapproaches, growth-increment-width analyses of bivalve molluskscurrently supply the bulk of annual to subannual resolution extra-tropical marine climate data for near-surface water masses (e.g.,Schöne et al., 2003b; Black et al., 2009; Butler et al., 2009a). Forexample, a number of annual resolution multicentury proxy recordshave been generated using growth-increment widths of the quahogArctica islandica from various locations in the North Atlantic (Schöneet al., 2003b; Schöne et al., 2005a; Schöne et al., 2005b; Scourse et al.,2006; Wanamaker et al., 2008; Butler et al., 2009a). The geoduck(Panopea abrupta) in turn has yielded long-term records of oceanvariability in the extratropical North Pacific (Strom et al., 2004; Stromet al., 2005; Black et al., 2009). While growth-increment widths oflong-lived bivalves have resulted in numerous climate reconstruc-tions, interpretation is complicated by a slowdown of growth withincreasing shell age (Goodwin et al., 2003). This necessitates the

Palaeogeography, Palaeoclimatology, Palaeoecology 302 (2011) 71–80

⁎ Corresponding author. Tel.: +1 905 828 5419; fax: +1 905 828 5425.E-mail addresses: [email protected] (J. Halfar),

[email protected] (S. Hetzinger), [email protected] (W. Adey),[email protected] (T. Zack), [email protected] (G. Gamboa),[email protected] (B. Kunz), [email protected] (B. Williams),[email protected] (D.E. Jacob).

0031-0182/$ – see front matter © 2010 Elsevier B.V. All rights reserved.doi:10.1016/j.palaeo.2010.04.009

Contents lists available at ScienceDirect

Palaeogeography, Palaeoclimatology, Palaeoecology

j ourna l homepage: www.e lsev ie r.com/ locate /pa laeo

Author's personal copy

application of exponential detrending functions that potentiallyremove low-frequency climate oscillations. In addition, records fromlong-lived bivalves are restricted by their biogeographic distribution.For example, the Atlantic A. islandica is absent from most of thesubarctic northwest Atlantic, as it has its northernmost confirmedoccurrences along the southern shore of Newfoundland (Dahlgrenet al., 2000). In fact, the northernmost A. islandica proxy time series inthe Northwest Atlantic are from the Gulf of Maine (Wanamaker et al.,2008).

In contrast, coralline algae are abundant from tropical to polaroceans, sometimes covering up to 100% of the shallow seafloor (Adey,1973; Steneck, 1986; Freiwald andHenrich, 1994; Kuffner et al., 2007).Extratropical coralline algae are particularly important in ecosystemsof the subarctic North Atlantic, the North Pacific and the Arctic Ocean(Adey et al., 2008). A number of studies during the past decade havedemonstrated that mid- and high-latitude crustose coralline algae arean emerging extratropical marine climate archive as they are amongstthe longest-lived shallow marine organisms (Frantz et al., 2005). Infact, U/Th dating of a live-collected coralline alga revealed an age of850±20 yr (Halfar et al., 2007). A number of coralline algal species(e.g., Clathromorphum compactum, Clathromorphum nereostratum,Lithothamnion crassiusculum, Lithothamnion glaciale, Phymatolithoncalcareum) display annual growth increments in a high Mg-calciteskeleton (Frantz et al., 2000; Halfar et al., 2000; Frantz et al., 2005;Halfar et al., 2007; Halfar et al., 2008; Kamenos et al., 2008). Inaddition, they show constant growth over their lifespan and are notsubject to an ontogenetic growth trend with skeletal age. Thefeasibility of using coralline algal oxygen isotope geochemistry aswell as Mg/Ca and Sr/Ca relationships for environmental reconstruc-tions was demonstrated during year-long field calibration studiesconducted in cold-temperate oceans (Gulf of Maine and Scotland —

Halfar et al., 2008; Kamenos et al., 2008). This relationship wasconfirmed by Halfar et al. (2008) and Hetzinger et al. (2009) for twospecies of the genus Clathromorphum. It was further shown that areasof highMg valueswithin the skeleton typically occur during the periodof main growth in summer (Hetzinger et al., 2009). In addition, therobustness of using algalMg/Ca ratios as environmental indicatorswassubsequently substantiated by Synchrotron Mg-X-ray absorbancenear edge structure (XANES) analyses that indicate that Mg is indeedassociated with the calcite lattice (Kamenos et al., 2009), rather thanbeing bound to an organic matrix, as has been suggested for somearagonitic bivalves (Foster et al., 2008). In a multiproxy approach,oxygen isotope time series from coralline algae were significantlycorrelated with a similar A. islandica-derived 30-year time series fromthe Gulf of Maine, additionally highlighting the value of coralline algaeas climate archives (Halfar et al., 2008). The first century-scaleannually-resolved climate records from the western Aleutian Islands(Alaska) were recently generated using oxygen isotopes and Mg/Caratios from the long-lived crustose coralline alga Clathromorphumnereostratum (Halfar et al., 2007; Hetzinger et al., 2009).

While coralline algal proxy archives have yielded a number ofenvironmental reconstructions based on geochemical signals, the useof coralline algal growth-increment widths as archives of past climatehas only recently been investigated (Kamenos and Law, 2010) . Usingaquaculture experiments and information from 50-year long timeseries of the branching species Lithothamnion glaciale, Kamenos andLaw (2010) did not find a relationship between temperature andgrowth increment width, but a negative correlation of temperature todegree of cell wall calcification during the summer months. A strongtemperature control on summer growth is also indicated by Adey(1970) who studied the effects of light and temperature on growth ofseveral species of boreal-subarctic crustose coralline algae. Adey(1970) shows that growth exhibits a hyperbolic curve — increasedrates with rising temperatures and decreased rates with loweringtemperatures, with optimum growth rates for boreal-subarcticcoralline algae between 5 and 10 °C. Growth is strongly light-

dependent at higher temperatures, but shows little dependence atlower temperatures (Adey, 1970). However, very high light values(e.g., summer values at 7 mwater depth inmid-latitude turbid coastalwaters) can have a negative effect on growth in some species (Adey,1970). Light variations in subarctic shallow benthic ecosystems notonly result from annual variations in solar insolation, but also fromlong-term fluctuations of macroalgal (e.g. kelp) shading (Stenecket al., 2002). Annual growth increments are the result of variations incell size and wall thickness and can be grouped into paired dark andlight bands with some coralline algae additionally exhibitingsubannual banding patterns (Freiwald and Henrich, 1994; Basso,1995; Halfar et al., 2000; Kamenos et al., 2008). Yearly growthbanding is usually well developed, although it can be disrupted byinvertebrate grazing, boring and other surface damage (Steneck,1986).

This study focuses on growth of the massive coralline algaClathromorphum compactum from the Northwest Atlantic (Fig. 1),whose significance asapaleoclimatearchivewasdemonstratedduringafield calibration study (Halfar et al., 2008). C. compactum is widelydistributed on rocky substrates along coastlines of the boreal-subarcticNorth Atlantic, the northern Pacific, and even extending into the ArcticOcean (Adey, 1965; Adey and Steneck, 2001). C. compactum thrives inwater temperatures b16 °C and occurs from 1 to 40 m water depthreaching its maximum abundance around 8 m (Adey, 1965). Meanannual growth-increment widths of numerous specimens collected inNew England (USA) measured from 230 to 330 μm (Adey, 1965). Asindividual dome-shaped plants attain a thickness of up to 3 cm,C. compactum can reach a lifespan of ∼100 years (Fig. 2). Verticalgrowth occurs in the perithallus, where large and poorly calcified cellsforming in the spring overlay short and heavily calcified cells havingdeveloped in the previous year (Moberly, 1968). This leads to theformation of conspicuous growth lines separating growth incrementsthat contain uncalcified conceptacles (Fig. 2). Conceptacles arespheroidal cavities accommodating reproductive sporangia of corallinealga. In C. compactum they form only in the autumn and winter (Adey,1965), but decalcification extends down into vegetative tissue con-structed during the previous summer. Since the mean conceptacleheight is 185 µm (130–230 µm) and the mean yearly growth rates arebetween 220 and 330 µm (Adey, 1965), the conceptacles are usuallyentirely constrained within yearly layers of accretion, and the cavitiesserve to identify the yearly growth cycles. However, during a season oflow growth, conceptacle bases may erode into the previous year'scalcified tissue. Conceptacles are not necessarily formed in the samearea of a plant in every year, and in some years, probably in part due tolow productivity for that year, they may not form at all. The abovecharacteristics and the fact that C. compactum forms massive, damage-resistant plantsmake C. compactum an ideal species for studying growthincrement width–temperature relationships.

Hence, the objective of this research is to demonstrate the potentialof coralline algal sclerochronology for reconstructing century-scaleclimate variability in the Canadian Northwestern Atlantic, a region fromwhich annual resolution proxy records are currently unavailable. Byevaluating intra-specimen and inter-specimen growth variability weshow that algal growth-increment width — sea surface temperaturerelationships can be improved by combining multiple growth records.Furthermore, using a network of modern-collected and museumspecimens of the coralline alga C. compactum from the shelf regions ofNewfoundland and the Gulf of St. Lawrence, Canada, we present a 115-year annual-resolution growth-increment width master chronology.This chronology is statistically compared to large-scale climate patternsin the northern Atlantic.

2. Methods

Living plants of C. compactum were collected from hard substratevia SCUBA at several localities along the eastern coastlines of

72 J. Halfar et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 302 (2011) 71–80

Author's personal copy

Fig. 1. Location map showing sampling locations: (1) Bay Bulls, Newfoundland; (2) Cape St. Martin, Newfoundland; (3) Quirpon Island, Newfoundland; (4) Outer Wapitagun Island,Gulf of St. Lawrence; and (5) Ile Longue, Quetachu Bay, Gulf of St. Lawrence. Heavy lines associated with LC indicate deep offshore branch, thin lines shallow inshore branch. Grey boxindicates area used for calculating instrumental ERSST time series. Flow path of LC from Colbourne et al. (1997) and Han (2004).

Fig. 2. Clathromorphum compactum from Northwest Atlantic. A) C. compactum (arrow) habitat at 10 m water depth; B) Photomosaic of polished cross section indicating LA-ICP-MStransects (dark lines) and conceptacle cavities (grey circles); shown are ∼70 years of growth; C) Detailed view of B showing faint growth increments and conceptacles; and D) BackScatter Electron image of uppermost three years of growth; note absence of subannual growth bands.

73J. Halfar et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 302 (2011) 71–80

Author's personal copy

Newfoundland and Quebec (Table 1). All locations are within theinfluence of the Labrador Current (LC). Specimens were a) obtainedfrom an extensive museum collection established in 1964 (U.S.National Herbarium of the Smithsonian Institution, Washington DC—

3 specimens) and b) collected in August 2008 at two locations inNewfoundland (4 specimens — Fig. 1, Table 1). Sampling sites wereheavily covered by coralline algae and modern sites are characterizedby less than 30% understory kelp (i.e. kelp growing close to theseafloor) providing shading (Fig. 2A). No kelp abundance informationis available from the museum-collected specimens. 2008 sampleswere collected at 10 m depth, whereas the 1964 samples werecollected at 3–10 m depth.

The air-dried C. compactum specimens were sectioned verticallyparallel to the direction of growth using a rock saw. Thick sections(2 mm) were cut from the slabs while preserving entire growthinformation and polished to 1 µm. High-resolution digital images ofthe polished surface (1 µm finish) were produced using an Olympusreflected light microscope (BX51) attached to an automated samplingstage/imaging system equipped with the software geo.TS (OlympusSoft Imaging Systems). This setup allows 2-dimensional mapping ofthe surfaces of polished specimens at various magnifications. Theresulting high-resolution photomosaics enabled the identification andlateral mapping of growth patterns over the entire sample (Fig. 2B).While the general growth pattern was visible on microscopic images,annual banding was occasionally difficult to identify with optical(Fig. 2C) or even scanning electron microscopy (Fig. 2D). Hence,yearly growth-increment widths were calculated from annual Mg/Ca-element cycle widths that were obtained by laser ablation inductivelycoupled plasma mass spectrometry (LA-ICP-MS). Results of Mg/Ca-based climate interpretations from C. compactum are presented inHetzinger et al. (2011-this issue). In preparation for LA-ICP-MSanalysis, paths for laser line transects as well as two reference pointsper sample were digitized on high-resolution photomosaics along thedirection of growth using geo.TS software (Fig. 2B). Care was taken toavoid areas of growth disruptions and small unconformities. Coordi-nates of digitized paths and reference points were subsequentlytransferred to the LA-ICP-MS system. After recoordination of thesample, laser line transects could be precisely positioned and analyzedalong previously digitized paths.

Mg and Ca concentrations were measured using an Agilent 7500ceQuadrupole ICP-MS coupled to a New Wave Research UP-213 laserablation system(213 nmwavelength,Nd:YAGLaser) at theDepartmentof Geosciences, Johannes Gutenberg-Universität Mainz, Germany.Measurements were carried out with laser energy densities of 6 J/cm2

and helium as carrier gas. Transects measuring up to 6000 µm in lengthwere analyzed with a scan speed of 10 µm/s, a spot size of 65 µm and

10 Hz pulse rate. 43Ca was used as the internal standard with calciumconcentrations measured by ICP-OES (Hetzinger et al., 2009). NIST (USNational Institute of Standard and Technology Standard ReferenceMaterial) glass reference material NIST SRM 610 was used as externalstandard. Independent measurements of NIST SRM 610 were used toconfirm that instrument drift was insignificant. Data reduction wascarried out with the commercial software GLITTER 4.4.2 (MacquarieUniversity, Sydney). Data for NIST SRM 610 were taken from theGeoReM database (Jochum and Nehring, 2006). Detection limits were:24 Mg=0.16 ppm, 43Ca=54.9 ppm; 1-sigma error from repeatedanalysis of NBS610: Mg=3.4%, Ca=3.1%. Mg/Ca ratios measured byLA-ICP-MS and microprobe agree within 4%.

Age models were established based on the pronounced seasonalcycle in algal Mg/Ca. Minimum Mg/Ca values were tied to March(Fig. 3), which is on average the coolest month at the study sites. Oncean age model was finalized, widths of annual Mg/Ca cycles werecalculated by multiplying the number of time-stamped ICP-MSmeasurements between two adjacent Mg/Ca minima with the laser-scan distance between each measurement. The latter is a function ofICP-MS cycling time and the laser-scan speed was set at 3.044 µm/s.Widths of annual Mg/Ca cycles are equivalent to annual verticalgrowth rates. As collections took place during the summer months,the year of collection is incomplete and counting was done startingfrom the year prior to collection.

Where possible, two Mg/Ca transects were analyzed on eachsample and statistically compared in order to test intraspecimenvariability. The two transects were then averaged to create acombined transect for that specimen. Combined and single transectswere compared to instrumental sea surface temperatures (see nextparagraph). Combined transects as well as single transects (fromspecimens where only one transect was available) of both modernand museum-collected specimens from all five sites were incorpo-rated into a single growth-increment width master chronology bycalculating the arithmetic mean for each year. Number of specimensin the master chronology was limited by cost of LA-ICP-MS analysis.Standard correlative statistics were used to estimate the similaritybetween the individual time series, the growth-increment widthmaster chronology, and instrumental observations. In addition, thegrowth-increment width master chronology was subjected to MTManalysis (Multitaper method, analysis done using KSpectra (Spectra-Works version 2.2) against a red-noise background spectrum - Mannand Lees, 1996). MTM provides a means for spectral estimation of atime series which is believed to exhibit a spectrum containing bothcontinuous and singular components (Thompson, 1982).

Instrumental observations were obtained from gridded SST dataavailable from the National Oceanic and Atmospheric Administration

Table 1Correlation coefficients and significance levels of annually resolved C. compactum growth time series from the Northwest Atlantic, Canada. Note that correlations improve when twotransects within same sample are averaged (e.g. SJ28-L1+2), and when averaging all transects (Master). Correlation coefficients (r) and significance levels (p) from 1950 to end ofrecord with annually averaged ERSSTJul–Sept (Smith et al., 2008); n = number of years used in correlation.

Sample Site Location Latitude(°N)

Longitude(°W)

Period ofGrowth

Annual Growth (μm) r p n

Average 1 Stand. Dev.

SJ28-L1 Bay Bulls, St. Johns Newfoundland 47°18.496′ 52°47.354′ 1933–2007 276 44 0.13 0.333 58SJ28-L2 Bay Bulls, St. Johns Newfoundland 47°18.496′ 52°47.354′ 1920–2007 251 39 0.19 0.155 58SJ21-L2 Bay Bulls, St. Johns Newfoundland 47°18.496′ 52°47.354′ 1914–2007 335 80 0.26 0.044 58QP4-4-L1 Quirpon Island Newfoundland 51°35.135′ 55°25.490′ 1941–2007 303 75 0.34 0.009 58QP4-4-L2 Quirpon Island Newfoundland 51°35.135′ 55°25.490′ 1941–2007 335 77 0.28 0.036 58QP4-3-L1 Quirpon Island Newfoundland 51°35.135′ 55°25.490′ 1933–2007 326 105 0.20 0.133 58QP4-3-L2 Quirpon Island Newfoundland 51°35.135′ 55°25.490′ 1941–2007 358 114 0.20 0.127 58170176 Cape St. Martin Newfoundland 50°1.5′ 55°53′ 1871–1963 298 84 0.09 0.761 1464-27-C Outer Wapitagun Island Gulf St. Lawrence 50°11.7′ 60°0.8′ 1893–1963 265 73 0.07 0.824 1464-22-H Ile Longue, Quetachu Bay Gulf St. Lawrence 50°16.4′ 62°48′ 1872–1963 220 65 0.15 0.606 14SJ28-L1+2 Bay Bulls, St. Johns Newfoundland 47°18.496′ 52°47.354′ 1920–2007 264 42 0.23 0.076 58QP4-4-L1+2 Quirpon Island Newfoundland 51°35.135′ 55°25.490′ 1941–2007 319 76 0.38 0.003 58QP4-3-L1+2 Quirpon Island Newfoundland 51°35.135′ 55°25.490′ 1933–2007 342 110 0.27 0.037 58Master All Sites 1893–2007 296 43 0.48 0.0001 58

74 J. Halfar et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 302 (2011) 71–80

Author's personal copy

Extended Reconstructed Sea Surface Temperature version 3 data set(NOAA ERSST.v3— Smith et al., 2008) for the 47–53°N, 53–55°W grid(Fig. 1). Average annual minimum (maximum) temperatures (1950–2007) for this grid are−0.8 °C (11.3 °C). The gridwas selected as it is aclose representation of water mass characteristics of the inshorebranch of the LC, which has a dominant influence on individual algalcollection sites on the shelf of Newfoundland and the southern coastof Quebec. Number of observations for the selected grid were sparse(b20/month) prior to 1950 (Fig. 4). Hence, correlations of algalgrowth-increment width time series with instrumental data wereonly calculated for the post-1950 period. Spatial correlations wereconducted using the tool Climate Explorer (http://climexp.knmi.nl/—Oldenborgh et al., 2009) and the NOAA ERSST.v3 data set. Similarly, areconstruction of the Atlantic Multidecadal Oscillation (AMO —

Enfield et al., 2001) was compared to algal proxy data.

3. Results

3.1. Growth

Based on the arithmetic mean of annual growth-increment widthsof all 7 specimens analyzed in this study, the average growth rate overthe lifetimes of all specimens is 297 µm±76 µm. Maximum age of a

specimen included in this study is 93 years. Average annual growthbetween sites ranges from a minimum of 220 µm in the Gulf of St.Lawrence to 358 µm at Quirpon (Table 1). There is no latitudinal orother obvious environmental trend in average annual growth ratesbetween different samples. For example, two samples analyzed fromthe site Bay Bulls (SJ28 and SJ21) show distinctly different averagegrowth rates ranging from 251 to 335 µm. With growth being afunction of light and temperature (Adey, 1970), algal growth rates arelikely to vary throughout an annual light and temperature cycle. Bycomparing the arithmetic mean of annual growth rates of allspecimens to monthly instrumental SST anomalies between 1950and 2007, a subannual growth model was established to determineduringwhichmonths the algae grow (Fig. 5). Themodel indicates thatsignificant growth (N95%) takes place during the warm and well-litmonths of July, August and September. Hence, growth-incrementwidths will be most strongly influenced by summer temperatures andare therefore compared to July–September SSTs only.

3.2. Individual chronologies and ERSST

Average annual growth rates within the same specimen can bedifferent depending on the position of the laser line transect relativeto the dome shaped thallus (Fig.6, Table 1). Multiple transects within

Fig. 4. Number of monthly observations in ERSST instrumental data for temperature field used in this study 47–53°N, 53–55°W. From 1950 onwards, number of observationsconsistently are N20/month; a further increase in number of observations takes place after 1964.

Fig. 3. LA-ICP-MS measured Mg/Ca cycles used for calculating annual growth-increment widths. Lower panel shows Mg/Ca cycles superimposed on cross-section view ofC. compactum. White line indicates location of laser transect along growth axis.

75J. Halfar et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 302 (2011) 71–80

Author's personal copy

the same specimen begin in different years as a result of thegrowth record extending further in certain regions of the specimen.Growth rates between two different Mg/Ca transects within the samespecimen are poorly correlated (sample SJ28: r=0.028, p=0.8,n=75; QP4-4: r=0.15, p=0.2, n=67; QP4-3: r=0.05, p=0.7,n=67), indicating a high degree of intra-specimen variability. Intra-specimen variability can be reduced by averaging growth rates fromtwo transects within the same specimen. For example, correlationsbetween ERSSTJul–Sep and growth-increment widths (1950–2007) intransects 1 and 2 of sample SJ28 are poor (r=0.13 and 0.19,respectively), but increase when measurements from both transectsare combined (SJ28-L1+2; r=0.23; Table 1). Similarly, inter-specimen variability is high, as correlations of individual transectswith ERSSTJul–Sep from 1950 to 2007 are generally weak (Table 1).While correlations are poor, the general trend of SST evolution since1950 is well reflected in all transects (Fig. 6).

3.3. Master chronology and ERSST

In order to compile a growth-increment width master chronologyextending to 1893 the arithmetic mean of chronologies from 7samples was calculated, including data from 3 specimens collectedin 1964 and 4 modern-collected specimens. Three of the modern-collected specimens yielded two uninterrupted chronologies each.The two chronologies were averaged for each sample before beingintegrated into the growth-increment width master chronology(Fig. 7A). Museum specimens date back to 1871, 1872 and 1893,however, due to only two records being available pre-1893, datavariance is higher before 1893. The master chronology was thereforeconfined to the period 1893–2007 when at least 3 common recordswere available (Table 1). When comparing normalized growth ratesof the master chronology to ERSSTJul–Sep (1950–2007) correlationsimprove drastically in contrast to individual records (r=0.48,p=0.0002, n=58 — Fig. 8; Table 1). While the quantity ofobservational data used in the ERSST reconstruction increases from1950 on, detailed instrumental data coverage is only available after1964 in the study region (Fig. 4). Accordingly, post-1964 correlationsbetween the master chronology and ERSSTJul–Sep improve further tor=0.63, p=0.00006, n=43 years. In addition, annual growth rates ofthe 115-year C. compactum master chronology track the ERSST-derived AMO index after 1900 and both are significantly correlated forthe entire length of the record (n=115 years; pb0.00001 — Fig. 7B).

4. Discussion

4.1. Growth

Average vertical growth rates of all specimens were found to be297 µm per year and therefore within the range of growth ratesreported earlier for C. compactum in the northwest Atlantic (Adey,1965). Moberly (1968) proposed that vertical growth in C. compactumis continuous throughout the year, albeit extremely slow during thewinter months. However, the presence of a distinct growth line

Fig. 5. Correlation of annually averaged growth incrementwidths ofC. compactum fromallsites with monthly instrumental sea-surface temperatures (NOAA ERSST.v3 — 47–53°N;53–55°W) from 1950 to 2001 (vertical bars). Significant correlations (95%) indicate mainperiod of vertical growth (=calcification) rates occur between July and September, whencombined influence of temperature and solar insolation is highest. Also shown aremonthly averaged ERSST (1950–2001) for same region (black line).

Fig. 6. Annual growth rates for individual samples (solid line) and ERSSTJul–Sep temperatures from 1950 to 2007 (dashed line). Note poor correlations of individual time series withtemperature. Correlations improve when growth rates of two transects within an individual sample are averaged (shaded box).

76 J. Halfar et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 302 (2011) 71–80

Author's personal copy

underlying large spring cells suggests that cell division comes to ahalt in the winter. The absence of a significant relationship betweengrowth and temperature during the dark and cold months of Januarythrough May further supports a halt in growth during winter andspring (Fig. 5). This pattern is similar to observations by Halfar et al.(2008) in an oxygen isotope time series from the same species, wherepoor correspondence between oxygen isotopes and temperature wasfound between January and April and highest agreements in thesummer months. Interestingly, both the oxygen isotope and growthrecords show weak correlations with temperature in November,while exhibiting stronger, yet insignificant correlations in October andDecember. This might be due to energy being used for conceptacleformation in November, when sporangia and conceptacles go into asignificant enlargement phase. Hence, during this time energy is notinvested in vertical accretion. This is similar to other marine calcifiersthat show a slowdown or interruption in growth during the repro-duction period (Schöne et al., 2004). Growth picks up again duringDecember when SSTs of N1.5 °C are still above the winter minimum.As indicated in Fig. 5, maximum growth takes place when tempera-tures are above N6 °C, which is similar to the findings of Adey (1970)for boreal-subarctic coralline algae.

As shown above, there is a high degree of intra- and inter-specimen variability in annual growth-increment widths, withindividual growth-increment width transects generally being poorlycorrelated to each other and to instrumental sea surface tempera-tures. Correlations improve by averaging annual growth-incrementwidths of two transects within the same specimen. Hence, similarto methodology developed in dendro- and sclerochronology measur-ing increment-width time series at multiple sites within the sameindividual significantly reduces variability (Schweingruber, 1988;Scourse et al., 2006; Butler et al., 2009b). Variability between theincrement-width series of biogenic proxies represents a source ofnoise in relation to any population-wide environmental signal (Butleret al., 2009a). By calculating the arithmetic mean of multiple algalrecords from different sites variability is further decreased, resultingin strong relationships with instrumental sea surface temperatures(Fig. 8).

In addition to temperature, light conditions influence growthrates. Light can vary locally according to sample position (e.g., under aledge versus exposed on a horizontal surface) and shading providedbymacroalgal cover (e.g., understory kelp). Long-term average annualgrowth rates in different samples fluctuate not only within anindividual site but also between sites. At the time of collection, theBay Bulls sampling site was characterized by up to ∼30% cover ofunderstory kelp, while macroalgae were absent from the QuirponIsland site. Average growth rates in Quirpon Island were higher thanin Bay Bulls, even though the latter site was characterized by higherwater temperatures. In summary, local factors can influence theamount of light at a given site or a given specimen, and this explainsdifferences in growth rates between individuals. Macroalgal cover isnot necessarily stable on multidecadal time scales at individual sites,but depends on many regional and local factors such as grazingpressure (Scheibling, 1986). Hence, we argue that long-term fluctua-tions in local shading increase the noise in the overall data set, whichmight in part be responsible for poor correlations observed betweenindividual records and SST (Fig. 6).

A local shading signal can be suppressed by averaging multiple timeseries from distant sites exhibiting diverse shading histories. In thepresent study, records spanning a distance of 800 km within the sameoceanographic regime (the LC) were averaged to generate a reliablelong-term proxy temperature record. Similarly, sclerochronological

Fig. 7. Growth-increment width master chronology from 1893 to 2007. A) Growth-increment widths of individual samples and average of all samples. B) 10-year moving average of115-year growth-increment width master chronology compared to Atlantic Multidecadal Oscillation (r=0.74).

Fig. 8. Correlation of growth-increment width master chronology (all samplesaveraged) with ERSSTJul–Sep instrumental temperature anomalies from 1950 to 2007.When used in combination, the 7 samples of C. compactum closely track SST. Three-yearmoving average: radj=0.57, p=0.005, n=58; data adjusted for loss of degrees offreedom.

77J. Halfar et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 302 (2011) 71–80

Author's personal copy

records frommultiple specimens of A. islandica between 80 and 200 kmapart have been shown to increase the robustness of masterchronologies (Schöne et al., 2003a; Butler et al., 2009a). Approachestaken in dendrochronology combine networks of records over distantgeographical areas N1000 kmapart to express ocean-basinwide signals(Gedalof and Smith, 2001; Gray et al., 2004). In the case of the marineenvironment, the spatial extent of the common signal is likely to beconstrained by water mass mixing patterns forced by wind fields,topography and stratification dynamics (Butler et al., 2009a). Asdemonstrated here, distant marine coralline algal time series can yielda common signal as long as they are influencedby similar oceanographicfeatures, such as the LC. The fact that the consistency of the commonsignal in this study is maintained across significant distances highlightsthe suitability of C. compactum as a proxy archive for reconstructingLC dynamics and therefore marine climate of the Northwest Atlanticregion. Hence, complementing other century-scale mid- and high-latitude marine climate reconstructions, such as those generated fromlong-lived bivalves, coralline algal time series can serve to establish anetwork of extratropicalmarine proxy-based chronologies analogous tonetworks of tree-ring chronologies (Briffa et al., 2002; Butler et al.,2009b). This is especially important for the subarctic Northwest Atlanticand similar regions where other century-scale proxy archives areabsent.

A recent study by Kamenos and Law (2010) did not find significantcorrelations between growth-increment width and water temperature.However, Kamenos and Law (2010) studied finely branched and free-living specimens (e.g., maerl or rhodolith forming) of Lithothamnionglaciale. Attached living and massive growing algae such asC. compactum can be expected to yield more consistent signals, sincerhodoliths or maerl forming species are subject to partial and temporalburial, and additionally might be moved across depth (e.g., above orbelow a shallow thermocline) during their multidecadal life span. Inaddition, branched growth morphologies are more susceptible todamage or breakage than the massive C. compactum. In fact, damagethrough storms and grazing is common in branched coralline algae(Freiwald and Henrich, 1994) resulting not only in frequent changes inorientation of branch growth axes, but also leading to an irregularinternal structure and ultimately an interrupted environmental record.In C. compactum growth disruptions from breakage or grazing areminimized due to the massive and internally regular growth architec-ture. In a similar fashion, corals exhibitingmassive growth have yieldedthe best and longest environmental records, when compared tobranching corals (e.g., Dunbar and Wellington, 1981; Watanabe et al.,2003; Pfeiffer et al., 2004).

4.2. Clathromorphum compactum as recorder of Northwest-Atlanticclimate

Growth rates of C. compactum reliably record SST wheremeasurements are available and can thus provide data from timeintervals and in regions for which instrumental measurements areincomplete. This is particularly important prior to the mid-20thcentury where scarcity of temperature data is common in many timeseries of instrumental observations (Black et al., 2009), especially inmore remote extratropical regions. To date no high-resolution proxyrecords have been provided from the shallow subarctic CanadianAtlantic, even though this region is of basin-scale oceanographicimportance due to the influence of the LC System (Colbourne, 2004).Hence, the oceanography along the slope of the Eastern Canadiancoast is strongly influenced by the position, strength, and properties ofthe LC (Mertz et al., 1993; Colbourne et al., 1997), which has a netcooling effect on both air and coastal water temperatures along theCanadian Atlantic provinces (Drinkwater et al., 1999). After passingthe Labrador shelf, the inshore portion of the LC travels south alongthe east coast of Newfoundland and branches out into small flows(Fig. 1). Approximately 15% of the transport heads to the southeast

and enters the Gulf of St. Lawrence through the Strait of Belle Islewhich separates Newfoundland and Labrador (Petrie and Anderson,1983; Han, 2004). In the meantime, the main LC inshore branchmaintains its path south over the continental shelf of Newfoundland(Lazier and Wright, 1993). Hence, specimens of C. compactum fromthe southern coast of Quebec in the Gulf of St. Lawrence record asimilar signal as the spatially distant time series from the eastern shelf ofNewfoundland (Fig. 1). The C. compactum growth-increment widthmaster chronology thereforedocuments large-scale spatial temperaturepatterns associated with the LC in the northwestern Atlantic (Fig. 9). Amulti-taper power spectrum of the C. compactum master chronologyexhibits commonmaximumspectral power atNorthAtlantic Oscillation(NAO)-type periods (∼8 and ∼2.5–3 years — Hurrell et al., 2003) andlow frequency multidecadal oscillations at 40–60 years. The latteroscillations are similar to hemispheric multidecadal temperatureanomalies that may be related to meridional overturning (e.g., AMO —

Figs. 7b and 10; Delworth and Mann, 2000). Multidecadal temperatureanomalies influence the subarctic Northwest Atlantic with watertemperatures decreasing during AMO negative modes (Wanamakeret al., 2008). Hence, cold water temperatures associated with strong LCtransport are reflected by negative C. compactum growth anomalies.Negative growth anomalies are the result of reduced calcification rateswith lower water temperature. A tendency of increasing watertemperatures in the LC region occurring during positive AMO phasesis in agreementwith previous observations (Delworth andMann, 2000;Wanamaker et al., 2008) and demonstrated by increased algal growthrates (Fig. 7b). The nature of the opposite trends pre-1900has to remainunexplained until longer records become available.

Recently, dramatic oceanographic and ecosystem changes in theNorthwestern Atlantic have been observed (e.g., Greene et al., 2004;Greene and Pershing, 2007). A large-scale regime shift during the early1990s has been attributed to a freshening trend changing circulationand stratification patterns on the Canadian shelf (Frank et al., 2005). Theecosystem shift resulted in changes in the abundance of phytoplankton,zooplankton, and fish populations, and together with overfishing,contributed to the collapse of cod stocks in the early 1990s (Greene andPershing, 2007). This is reflected by a strong increase in growth rates of

Fig. 9. Spatial correlation of growth-increment width master chronology against NOAAERSSTJul–Sep data from 1950 to 2007 indicating that algae record large-scale NorthwestAtlantic temperature patterns in association with the Labrador Current. Circles indicatelocation of individual records. Note that locations 2, 4, and 5 only contribute data before1963. Analysis conducted using Climate Explorer (http://climexp.knmi.nl/ — Old-enborgh et al., 2009). Current patterns according to Colbourne et al. (1997) and Han(2004).

78 J. Halfar et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 302 (2011) 71–80

Author's personal copy

C. compactum from an all-time low in 1992 tomaximumvalues in 2000.The positive growth anomalies of the year 2000were only surpassed byeven higher algal growth rates in the mid 1930s (Fig. 7A). During thistime an Atlantic-wide regime shifts accompanied by dramatic warmingof air and ocean temperatures has taken place. This constitutes themostsignificant regime shift experienced in the North Atlantic in the 20thcentury (Drinkwater, 2006). The high temperatures recordedduring thewarm period of the 1930smatch, and in some cases exceed, the presentday warming (Johannessen et al., 2004). Drinkwater (2006), however,argues that evidence from the Northwest Atlantic is limited, as no signsof significant warming during the 1920s and 1930s are reported fromthat region. The network of C. compactum growth records presentedhere clearly indicates thatwarmingwasmore extensive than previouslythought as it was clearly recorded by positive anomalies in algal growthrates. Hence, Northwest Atlantic ecosystem responses to the basinwidewarming trend might have been stronger than previously assumed. Infact, as indicated by algal growth rates, themid 1920s to 1930swarmingin the subarctic Northwest Atlantic might have been stronger or ofsimilar magnitude than the recent warming trend.

5. Conclusions

Our study demonstrates that variations in annual growth ratesof C. compactum reliably record SST. This is particularly importantprior to 1950 where scarcity of temperature data is common in manytime series of instrumental observations especially in more remoteextratropical regions. Using a regional network of specimens ofthe coralline alga C. compactum spanning portions of the LC inshorebranch from the Gulf of St. Lawrence to both latitudinal extremes of theeastern Newfoundland shelf, it was possible to generate a 115-year longgrowth-increment-width based record of subarctic Northwest Atlanticsurface temperatures. The record closely reflects multidecadal oscilla-tions of the AMO, and higher frequency variability in the interannualbands that is similar to the frequency bands associated with NAO. Inaddition, a regime shift in the 1920s and 1930s is recorded by positivealgal growth anomalies. This suggests that warming during the early20th century has possibly affected theNorthwest Atlanticmore stronglythan previously thought. In addition, algal growth anomalies indicatethat the shift was of similar magnitude or even more pronouncedthan the regime shift observed during the 1990s, as has recently beensuggested for the central and easternNorthAtlantic (Drinkwater, 2006).

Coralline algal sclerochronology is a powerful approach for gener-ating mid- and high-latitude North Atlantic marine proxy records andcan contribute to spatially increase the network of existing extratropicalhigh-resolution reconstructions of North Atlantic Ocean variability overthepast centuries. The coralline algal growth-increment-width networkhas to now be expanded to encompass a larger geographical region and

extended back in time by analyzing further specimens from museumcollections or pre-modern sediments. In addition, the coralline algalgrowth increment width - climate relationship needs to be tested forother species and inexpensive methods developed to obtain growthincrement width information from coralline algae.

Acknowledgments

We thank Wade Saunders, Bob Hooper from Bonne Bay MarineStation and Philip Sargent from Memorial University of Newfoundlandfor logistical support andadvice on sampling locations. AndreasKronzatUniversität Göttingen, Germany, provided backscatter electron images.J.H. was supported by a Natural Sciences and Engineering ResearchCouncil of Canada Discovery Grant, and a Canadian Foundation forClimate and Atmospheric Sciences Grant (Gr-7004). S.H. acknowledgessupport from the Alexander von Humboldt Foundation (Feodor LynenFellowship). This is a Geocycles contribution. Two anonymousreviewers provided excellent input.

References

Adey, W.H., 1965. The genus Clathromorphum (Corallinaceae) in the Gulf of Maine.Hydrobiologia 26 (3–4), 539–573.

Adey, W.H., 1970. The effects of light and temperature on growth rates in boreal-subarctic crustose corallines. Journal of Phycology 6, 269–276.

Adey, W.H., 1973. Temperature control of reproduction and productivity in a subarcticcoralline alga. Phycologia 12 (3/4), 111–118.

Adey, W.H., Steneck, R., 2001. Thermogeography over time creates biogeographicregions: a temperature/space/time-integrated model and an abundance-weightedtest for benthic marine algae. Journal of Phycology 37, 677–698.

Adey, W.H., Lindstrom, S.C., Hommersand, M.H., Müller, K.M., 2008. The biogeographicorigin of Arctic endemic seaweeds: a thermogeographic view. Journal of Phycology44, 1384–1394.

Basso, D., 1995. Study of living calcareous algae by a paleontological approach: the non-geniculate corallinaceae (Rhodophyta) of the soft bottoms of the Tyrrhenian Sea(Western Mediterranean). The genera Phymatolithon Foslie and MesophyllumLemoine. Rivista Italiana di Paleontologia e Stratigrafia 100 (4), 575–596.

Black, B.A., Copenheaver, C.A., Frank, D., Stuckey, M.J., Kormanyos, R.E., 2009. Multi-proxy reconstructions of northeastern Pacific sea surface temperature data fromtrees and Pacific geoduck. Palaeogeography, Palaeoclimatology, Palaeoecology 278,40–47.

Briffa, K., Osborn, T.J., Schweingruber, F.H., Jones, P.D., Shiyatov, S.G., Vaganov, E.A.,2002. Tree-ring width and density data around the Northern Hemisphere: part 1Local and regional climate signals. Holocene 12 (6), 737–757.

Butler, P.G., Richardson, C., Scourse, J.D., Witbaard, R., Schöne, B.R., Fraser, N.M.,Wanamaker Jr, A.D., Bryant, C.L., Harris, I., Robertson, I., 2009a. Accurate incrementidentification and the spatial extent of the common signal in five Arctica islandicachronologies from the Fladen Ground, northern North Sea. Paleoceanography 24.doi:10.1029/2008PA001715 PA2210.

Butler, P.G., Scourse, J.D., Richardson, C., Wanamaker Jr, A.D., Bryant, C.L., Bennell, J.D.,2009b. Continuous marine radiocarbon reservoir calibration and the 13C Suesseffect in the Irish Sea: results from the first multi-centennial shell-based marinemaster chronology. Earth and Planetary Science Letters 279, 230–241.

Colbourne, E.B., 2004. Decadal changes in the ocean climate in Newfoundland andLabrador waters from the 1950s to the 1990s. Journal of Northwestern AtlanticFisheries Science 34, 43–61.

Colbourne, E.B., deYoung, B., Narayanan, S., Helbig, J., 1997. Comparison of hydrographyand circulation on the Newfoundland Shelf during 1990–1993 with the long-termmean. Canadian Journal of Fisheries and Aquatic Sciences 54, 68–80.

Dahlgren, T.G., Weinberg, J.R., Halanych, K.M., 2000. Phylogeography of the oceanquahog (Arctica islandica): influences of paleoclimate on genetic diversity andspecies range. Marine Biology 137, 487–495.

Delworth, T.L., Mann, M.E., 2000. Observed and simulated multidecadal variability inthe Northern Hemisphere. Climate Dynamics 16, 661–676.

Drinkwater, K., 2006. The regime shift of the 1920s and 1930s in the North Atlantic.Progress in Oceanography 68, 134–151.

Drinkwater, K.F., Mountain, D.B., Herman, A., 1999. Variability in the slope waterproperties off eastern North America and their effects on adjacent shelves. ICES CM8, 26 1999/O.

Dunbar, R.B., Wellington, G.M., 1981. Stable isotopes in a branching coral monitorseasonal temperature variation. Nature 293, 453–455.

Dunbar, R.B., Wellington, G.M., Colgan, M.W., Glynn, P.W., 1994. Eastern Pacific seasurface temperatures since 1600 A.D.: the δ18O record of climate variability inGalápagos corals. Paleoceanography 9 (2), 291–315.

Enfield, D.B., Mestas-Nunez, A.M., Trimble, P.J., 2001. The AtlanticMultidecadal Oscillationand its relation to rainfall and river flows in the continental U.S. Geophysical ResearchLetters 28 (10), 2077–2080.

Foster, L.C., Finch, A.A., Allison, N., Andersson, C., Clarke, L.J., 2008. Mg in aragoniticbivalve shells: seasonal variations and mode of incorporation in Arctica islandica.Chemical Geology 254, 113–119.

Fig. 10. Multi-taper power spectrum of annual growth-increment width masterchronology (1893–2007) showing significant power at ∼60, 8, and 3 years (shadedregions) highlighting interannual and multidecadal variability.

79J. Halfar et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 302 (2011) 71–80

Author's personal copy

Frank, K.T., Petrie, B., Choi, J.S., Legett, W.C., 2005. Trophic cascades in a formerly cod-dominated ecosystem. Science 308 (5728), 1621–1623.

Frantz, B.R., Kashgarian, M., Coale, K.H., Foster, M.S., 2000. Growth rate and potentialclimate record from a rhodolith using 14C accelerator mass spectrometry.Limnology and Oceanography 45 (8), 1773–1777.

Frantz, B.R., Foster, M.S., Riosmena-Rodríguez, R., 2005. Clathromorphum nereostratum(Corallinales, Rhodophyta): the oldest alga? Journal of Phycology 41 (4), 770–773.

Freiwald, A., Henrich, R., 1994. Reefal coralline algal build-ups within the Arctic Circle:morphology and sedimentary dynamics under extreme environmental seasonality.Sedimentology 41, 963–984.

Gedalof, Z., Smith, D.J., 2001. Interdecadal climate variability and regime-scale shifts inPacific North America. Geophysical Research Letters 28, 1515–1518.

Goodwin, D.H., Schöne, B.R., Dettman, D.L., 2003. Resolution and fidelity of oxygenisotopes as paleotemperature proxies in bivalve mollusk shells: models andobservations. Palaios 18, 110–125.

Gray, S.T., Graumlich, L.J., Betancourt, J.L., Pederson, G.T., 2004. A tree-ring basedreconstruction of the Atlantic Multidecadal Oscillation since 1567 A.D. GeophysicalResearch Letters 31 (L12205).

Greene, C.H., Pershing, A.J., 2007. Climate drives sea change. Science 315, 1084–1085.Greene, C.H., Pershing, A.J., Monger, B.C., Benfield, M.C., Durbin, E.G., Casas, M.C., 2004.

Supply-side ecology response of climate-driven changes in North Atlantic Ocean.Oceanography 17 (3), 61–71.

Halfar, J., Zack, T., Kronz, A., Zachos, J., 2000. Geochemical signals of rhodoliths(coralline red algae)— a new biogenic archive. Journal of Geophysical Research 105(C9), 22107–22116.

Halfar, J., Steneck, R., Schöne, B.R., Moore, G.W.K., Joachimski, M.M., Kronz, A., Fietzke, J.,Estes, J.A., 2007. Coralline alga reveals first marine record of subarctic North Pacificclimate change. Geophysical Research Letters 34, L07702. doi:10.1029/2006GL028811.

Halfar, J., Steneck, R.S., Joachimski, M., Kronz, A., Wanamaker Jr., A.D., 2008. Corallinered algae as high-resolution climate recorders. Geology 36, 463–466.

Han, G., 2004. Sea level and surface current variability in the Gulf of St. Lawrence fromsatellite altimetry. International Journal of Remote Sensing 25 (22), 5069–5088.

Hetzinger, S., Halfar, J., Zack, T., Jacob, D.E., Kunz, B., Gamboa, G., Simon, K., Kronz, A., Adey,W.H., Lebednik, P.A., Steneck, R., 2011. High-resolution analysis of trace elementsfrom theNorthAtlantic andNorthPacific inencrusting coralline algaeby laser ablationICP-MS. Palaeogeography, Palaeoclimataology, Palaeoecology 302, 81–94 (this issue).

Hetzinger, S., Halfar, J., Kronz, A., Steneck, R., Adey, W.H., Lebednik, P.A., Schöne, B.R.,2009. High-resolution Mg/Ca ratios in a coralline red algae as a proxy for Bering Seatemperature variations from 1902 to 1967. Palaios 24, 406–412.

Hurrell, J.W., Kushnir, Y., Ottersen, G., Visbeck, M., 2003. An overview of the NorthAtlantic Oscillation. In: Hurrell, J.W., Kushnir, Y., Ottersen, G., Visbeck, M. (Eds.), TheNorth Atlantic Oscillation: Climate Significance and Environmental Impact.Geophysical Monograph Series. American Geophysical Union, pp. 1–35.

Jochum, K.P. Nehring, F., 2006. GeoReM preferred values. http://georem.mpch-mainz.gwdg.de/sample_query_pref.asp.

Johannessen, O.M., Bengtsson, L., Miles, M.W., Kuzmina, S.I., Semenov, V.A., Alekseev, G.V.,Nagurnyi, A.P., Zakharov,V.F., Bobylev, L.P., Pettersson, L.H.,Hasselmann, K., Cattle, H.P.,2004. Arctic climate change: observed and modelled temperature and sea-icevariability. Tellus 56A, 328–341.

Kamenos, N., Law, A., 2010. Temperature controls on coralline algal skeletal growth.Journal of Psychology 46 (2), 331–335.

Kamenos, N., Cusack, M., Moore, P.G., 2008. Coralline algae are global paleothermometerswith bi-weekly resolution. Geochimica et Cosmochimica Acta 72, 771–779.

Kamenos, N., Cusack, M., Huthwelker, T., Lagarde, P., Scheibling, R.E., 2009. Mg-latticeassociations in red coralline algae. Geochimica et Cosmochimica Acta 73, 1901–1907.

Kuffner, I.B., Andersson, A.J., Jokiel, P.L., Rodgers, K.S., Mackenzie, F.T., 2007. Decreasedabundance of crustose coralline algae due to ocean acidification. Nature Geoscience1, 114–117.

Lazier, J.R.N., Wright, D.G., 1993. Annual velocity variations in the Labrador Current.Journal of Physical Oceanography 23 (659–678).

Mann, M.E., Lees, J., 1996. Robust estimation of background noise and signal detectionin climatic time series. Climate Change 33, 409–445.

Mertz, G., Narayanan, S., Helbig, J., 1993. The freshwater transport in the LabradorCurrent. Atmosphere-Ocean 31 (2), 281–295.

Moberly Jr., R., 1968. Composition of magnesian calcites of algae and pelecypods byelectron microprobe analysis. Sedimentology 11, 61–82.

Oldenborgh, G.J., Drijfhout, S.S., van Ulden, R., Haarsma, R., Sterl, A., Severijns, W.,Hazeleger, W., Dijkstra, H.A., 2009. Western Europe is warming much faster thanexpected. Climate of the Past 5 (1), 1–12.

Petrie, B., Anderson, C., 1983. Circulation on the Newfoundland Continental Shelf.Atmosphere-Ocean 21 (2), 207–226.

Pfeiffer, M., Dullo, W.C., 2006. Monsoon-induced cooling of the western equatorialIndian Ocean as recorded in coral oxygen isotope records from the Seychellescovering the period of 1840–1994 AD. Quaternary Science Reviews 25, 993–1009.

Pfeiffer, M., Timm, O., Dullo, W.C., Podlech, S., 2004. Oceanic forcing of interannal andmultidecadal climate variability in the southwestern Indian Ocean: evidence from a160-year coral isotopic record (La Réunion, 55°E, 21°S). Paleoceanography 19.doi:10.1029/2003PA000964 PA4006.

Saenger, C., Cohen, A.L., Oppo, D.W., Halley, R.B., Carilli, J.E., 2009. Surface-temperaturetrends and variability in the low-latitude North Atlantic since 1552. NatureGeoscience 2, 492–495.

Scheibling, R.E., 1986. Increased macroalgal abundance following mass mortalities ofsea urchins (Stronglyocentrotus droebachiensis) along the Atlantic coast of NovaScotia. Oecologia 68, 186–198.

Schöne, B.R., Kröncke, I., Houk, S.D., Castro, D.F., 2003a. The cornucopia of chilly winters:ocean quahog (Arctica islandica L., Mollusca) master chronology reveals bottomwater nutrient enrichment during colder winters (North Sea). SenckenbergianaMaritima 32, 165–175.

Schöne, B.R., Oschmann, W., Rössler, J., Freyre Castro, A.D., Houk, S.D., Kröncke, I., Dreyer,W., Janssen, R., Rumohr, H., Dunca, E., 2003b. North Atlantic Oscillation dynamicsrecorded in shells of a long-lived bivalve mollusk. Geology 31 (12), 1037–1040.

Schöne, B.R., Freyre Castro, A.D., Fiebig, J., Houk, S.D., Oschmann, W., Kröncke, I., 2004.Sea surface temperatures over the period 1884–1983 reconstructed from oxygenisotope ratios of a bivalve mollusk shell (Arctica islandica, southern North Sea).Palaeogeography, Palaeoclimatology, Palaeoecology 212, 215–232.

Schöne, B.R., Fiebig, J., Pfeiffer, M., Gleß, R., Hickson, J., Johnson, A.L.A., Dreyer, W.,Oschmann, W., 2005a. Climate records from a bivalved Methuselah (Arcticaislandica, Mollusca; Iceland). Palaeogeography, Palaeoclimatology, Palaeoecology228, 130–148.

Schöne, B.R., Houk, S.D., FreyreCastro, A.D., Fiebig, J., Oschmann,W., Kröncke, I., Dreyer,W.,Gosselck, F., 2005b. Daily growth rates in shells of Arctica islandica: assessing sub-seasonal environmental controls on a long-lived bivalve mollusk. Palaios 20, 78–92.

Schweingruber, F.H., 1988. Tree rings: basics and applications of dendrochronology.Kluwer Academic, Dordrecht, Netherlands. 276 pp.

Scourse, J.D., Richardson, C., Forsythe, G., Harris, I., Heinemeier, J., Fraser, N.M., Briffa, K.,Jones, P., 2006. First cross-matched floating chronology from the marine fossilrecord: data from growth lines of the long-lived bivalve mollusc Arctica islandica.Holocene 16 (7), 967–974.

Smith, T.M., Reynolds, R.W., Peterson, T.C., Lawrimore, J., 2008. Improvements toNOAA's historical merged land-ocean surface temperature analysis (1880-2006).Journal of Climate 21, 2283–2296.

Steneck, R., 1986. The ecology of coralline algal crusts: convergent patterns andadaptive strategies. Annual Review of Ecology and Systematics 17, 273–303.

Steneck, R.S., Graham, M.H., Bourque, B.J., Corbett, D., Erlandson, J.M., Estes, J.A., Tegner,M.J., 2002. Kelp forest ecosystems: biodiversity, stability, resilience and future.Environmental Conservation 29 (4), 436–459.

Strom, A., Francis, R.C., Mantua, N., Miles, E.L., Peterson, D.L., 2004. North Pacific climaterecorded in growth rings of geoduck clams: a new tool for paleoenvironmentalreconstruction. Geophysical Research Letters 31, 1–4 L06206.

Strom, A., Francis, R.C., Mantua, N.J., Miles, E.L., Peterson, D.L., 2005. Preserving low-frequency climate signals in growth records of geoduck clams (Panopea abrupta).Palaeogeography, Palaeoclimatology, Palaeoecology 228, 167–178.

Thompson, D.J., 1982. Spectrum estimation and harmonic analysis. IEEE Proceedings 70(1055–1096).

Wanamaker Jr., A.D., Kreutz, K.J., Schöne, B.R., Pettigrew, N.R., Borns Jr., H.W., Introne, D.S.,Belknap,D.,Maasch, K.A., Feindel, S., 2008. CoupledNorthAtlantic SlopeWater forcingon Gulf of Maine temperatures over the past millennium. Climate Dynamics.doi:10.1007/s00382-007-0344-8.

Watanabe, T., Gagan, M.K., Corrége, T., Scott-Gagan, H., Cowley, J., Hantoro, W.S., 2003.Oxygen isotope systematics in Diploastrea heliopora: new coral archive of tropicalpaleoclimate. Geochimica et Cosmochimica Acta 67 (7), 1349–1358.

80 J. Halfar et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 302 (2011) 71–80