Using Dendrochronological Methods With Bivale Mollusc Shells Biology Essay


Dendrochronological methods have been used as a base to produce techniques enabling bivalve mollusc shells to be aged. This has been achieved through a comparison of increment growth and structure; methods of data collection and analysis have also been modified. Along with the construction of chronologies showing variation in temperature and isotope ratios which can be compared to tree-ring studies in addition to the similar palaeoclimatic applications.

Dendrochronology is the scientific method of chronology building based on the analysis of patterns of tree-rings widths (Stokes 1968). Wider increments are created during periods of optimum conditions, and narrower increments during periods of harsh environments i.e. the limiting growth factor has a greater influence. The resultant ring-widths that have been recorded from one year to the next have long been recognised as an invaluable source for climatic and chronological information (Bradley 1999). Dendrochronology is used as a proxy of palaeoclimatic information, which is the study of climatic conditions, together with their causes and effects in the geologic past. In many European countries, chronologies produced from tree ring-widths have been able to reach back to the early eighteenth century when there were several accounts of the unusual narrowness of tree rings and frost damage dating from the harsh winter during 1708-1709 (Derham 1709, Pain 2009). This strong ring-width variation has also been recorded during 1816, known as "The Year Without a Summer", in which a severe and intense climate abnormalities resulted in average global temperatures to decrease by about 0.4-0.7oC causing major food shortages across the Northern Hemisphere, along with a narrowed annual ring-width. The abnormality in climate was caused by low solar activity and a series of major volcanic eruptions which finished with the eruption of Mt. Tambora in Indonesia (Oppenheimer 2003).

Lady using a tablet
Lady using a tablet


Essay Writers

Lady Using Tablet

Get your grade
or your money back

using our Essay Writing Service!

Essay Writing Service

The "father of tree ring studies" is mostly considered to be A.E Douglass (1867-1962) who was an American astronomer studying the relationships between sunspot activity and rainfall. To fully test the sunspot-climate relationship, Douglass required long climatic records, which were supplied by the ring-width variation in trees of the arid areas of South-Western US which also provided a long proxy record of rainfall variation (Douglass 1914, 1919). By using proxies such as beams from old buildings along with Sequoias and other long lived trees, he noticed that the rings were thinner during dry years, particularly in connection with the 17th century lack of sunspots, known as the "Maunder Minimum" (Reid 1997, Sinclair 1993). However scientists found they could not trust this source of data outside the region studied and the value of tree rings for climate studies was not fully established until 1960 (Bradley 1999).

Douglass' early work was crucial for the development of dendrochronology, the use of tree rings for dating and for dendroclimatology, the use of tree rings as an indicator of climate.

Sclerochronology is "the study of physical and chemical variations on the accretionary hard tissue of organisms, and the temporal context in which they formed" (Jones 2009), and has much in common with dendrochronology, which is considered the forbearer of this field of research.

Knutson and Buddemeier used Douglass' ideas and pioneered work on growth bands in coral around nuclear test atolls (Marshall Islands, Pacific Ocean) in the early 1970s but the term Sclerochronology was not used until 1974 (Buddemeier 1974). From their studies it was found that, similar to tree growth, the cyclic variations in density bands revealed by x-radiography were annual (Knutson 1972). This research was furthered by Rob Witbaard (1990, 1994), from his research, it was found that the annual banding patterns found in the shells from the bivalve mollusc Arctica islandica can be used to create a past record of marine conditions in a similar way to how the growth-rings in trees are being used to construct a chronology of past terrestrial conditions (Witbaard 1997, 2003), these chronologies can be compared to create an overall chronology of overarching climatic conditions.

These two sets of chronologies can be compared to build a more complete climatic record encompassing both atmospheric and oceanic climates at similar climate signals during overlapping chronologies. Knowing the date of cessation of growth and the periodicity allows for dates to be allocated to increments. Both tree growth-rings and mollusc shell growth-bands varying in width depending on the environmental conditions, growing wider in more optimum conditions and the climate and as reference points or marker years, for new sequences of rings and bands to be added to the record.

Lady using a tablet
Lady using a tablet


Writing Services

Lady Using Tablet

Always on Time

Marked to Standard

Order Now

To construct a chronology, firstly the raw data is crossdated which is the matching of marker years in one specimen or sample with corresponding ring patterns in another sample or previous record via the recognisable sequence of varying width of tree growth-rings and shell-bands that makes cross dating possible and thus master chronologies (Scourse 2006). Once crossdated, standardisation or de-trending is required to remove the trends of decreasing variance in growth with age along with certain periodical variations produced which are not of interest to the study. Whilst both crossdating and detrending is a very similar process carried out in both sclerochronology and dendrochronology however the analysis process and preparation work is different for shells and trees, when using density x-rays, isotope ratios and electron backscatter analysis.

Biochronicle Requirements

Sclerochronology as defined by Jones (2009) as "the study of the chemical and physical changes in the accretion of hard tissues in organisms, along with the environmental context when formed" which is similar to the definition of dendrochronology by Jacoby (in Noller 2000) as the "dating method based on variations in annual growth rings of trees in response to climatic variations".

Shell cross-section studies of banding patterns can be taken from collections of shell samples in any part of the world where the periodicity of molluscan growth is known. There are four main criteria that need to be passed before a shell cross-section can be analysed as described (below) by Thompson and Jones (1977);

(1) Synchronised response by the individuals of a population to environmental conditions.

(2) Consistent periodicity of the growth unit being measured.

(3) Long lived.

(4) Indeterminate growth that does not stop throughout its life.

These are similar to the requirements for creating a dendrochronology time series. There are four main requirements that must be met before a tree-ring section can be analysed as described by (Stokes 1968);

(1) Tree must add only one ring for each growing season; hence annual rings

(2) One environmental factor must dominate in growth limitation. For example in South Western American regions, the growth limiting factor is precipitation whereas in Alaska, temperature is the limiting factor.

(3) The growth limiting environmental factor must have a recordable variation in intensity from year to year and the resulting growth rings of the trees have to reflect such variation with a change in width.

(4) The variable environmental growth limiting factor must be measurable and uniformly effective over a large geographical area.

Arctica islandica

The commonly called Ocean Quahog, Arctica islandica fulfils the stated criteria and therefore sclerochronological methods can be used to with shells samples of Arctica, along with other bivalve species such as S.solidissima and M.mercenaria (Jones 1980, 1981, 1990) along with exoskeletons of corals (Insalaco 1996, Marschal 2004), encrusting calcareous algae (Backman 1983) and fish otoliths (ear-bones) (Dove 1996, Fenton 1992). Arctica islandica shells are widely used to study the marine palaeoclimate around the North Atlantic Coast, this is due to its distribution commonly being between 35-70oN and following the 10-280m depth contour which is likely to be set by the maximum bottom water temperature of 16oC (Nicol 1951, Witbaard 1997) (Figure 1). Implying that for different latitudes and areas different congruent data-sets can be obtained, for example chronologies from the Baltic Sea and Icelandic waters can be compared. Arctica islandica is the only living species of a bivalve genus originating in the early Cretaceous period, and has not significantly physiologically or physically changed since that time (Nicol 1951, Begum 2009), individuals can exceed 200 years, and one found to be over 400 years old (Ridgway 2010) with easily identifiable incremental growth on an annual basis (Ropes 1985; Jones 1980; Thompson 1980).

This longevity requires fewer specimens to span a given time interval also there is a longer overlapping series of years which allows for much more robust correlation tests between individuals (Marchitto 2000). There is a wealth of knowledge on the physiology (Mann 1982,Winter 1978), anatomy (Palmer 1979) and behaviour (Taylor 1976) of Arctica islandica. Many scientists use Arctica in experiments because of its attractive size and easiness to handle as the maximum height of the shell is approximately 10cm with females being slightly larger than their male counterparts (Fritz 1991). Arctica has a highly patchy distribution with densities of 0.1individuals/m2 found in the Eastern North Sea, whereas in the Northern North Sea the maximum density found was 16 individuals/m2 as found by Witbaard (1997).

Lady using a tablet
Lady using a tablet

This Essay is

a Student's Work

Lady Using Tablet

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Examples of our work

Arctica is classed as a K-selected species (Pianka 1970) due to its slow growth, long living species with an irregular recruitment. Once an Arctica larva has settled to the seabed, it remains in a very small vicinity and is unable to escape from adverse conditions. It is these aspects that make Arctica growth increments sensitive to disturbance. Changes to the environmental conditions will either benefit or limit survival, reproduction and growth (Witbaard 1997). During shell development and growth specific information, such as oxygen and carbon isotope ratios, provide understanding about the environment and is integrated into the annual deposition of carbonate increments therefore the growth record of Arctica creates a bio-chronicle which enables retrospective assessment of climatic change (Witbaard 1997). Weidman (1994a) used a similar microsampling technique to Dettman and Lohmann (1993) to produce a very high resolution d18O record from a living Arctica collected from a depth of 60m on the Nantucket shoals (Massachusetts, Atlantic Ocean). Using comparisons with nearby instrumental measurements of salinity and bottom water temperature during the same time interval showed that the oxygen composition of carbonate in an Arctica islandica shell is in isotopic equilibrium with ambient seawater. This provided all the data required for Weidman (1994a) to construct a 109 year record of bottom temperatures during A.D. 1875 - 1983, using d18O measurements from live specimens. To further this research Weidman and Jones (1993, 1994) constructed a 52 year d14C record using Arctica islandica shells sampled from South East Georges Bank, and it was established that the record of the d14C bomb pulse which was caused by nuclear bomb testing in the Pacific Ocean increasing d14C in the atmosphere and subsequently the ocean which corresponded to mollusc shell and coral exoskeleton records created by Weidman (1994b) from Iceland, Norway and the North Sea with similar records constructed from Florida and Bermuda by Druffel in 1989.

Growth of Tree Rings and Shell Bands

Trees grow by increasing their radial (breadth) and apical (height) size, which is a result of cell growth in the meristem tissue located in two crucial regions of the tree. The apical growth is a product of primary tissue growth in the apical meristem causing the tree to extend the length of its branches and main trunk. Lateral growth is a consequence of vascular cambium differentiating into xylem producing the woody section on the inside of the cambium and those cells forming the phloem outside the cambium (Stokes 1968). After the primary tissue has been formed, the annual xylem growth is laid down outside the previous year's growth and thus creates the ring structure.

From studying cross sections of a tree, the xylem is marked by visible annual growth rings with darker rings of phloem. Phloem, which is fundamental for the movement of nutrients to cells, is of no real use in dating the wood, except that its presence assures no xylem is missing. The protective bark is continually shed from the outside of the tree as growth occurs so that only a few years of growth remain which is also of no use to dating due to there being no distinguishable ring structure (Stokes 1968).

Bivalve annual growth has been supported by analysis of 14C which is incorporated into the band structures (Witbaard 1997). The term increment corresponds to the amount of calcium carbonate deposited during one year's growth. It is also known as the growth band or in a descriptive manner the wide or light band and is delineated by a narrow growth line known as the dark band due to its greater density (Witbaard 1997).

Annual Shell Band and Tree Ring Structure

Witbaard (1997) described the optimal section of the shell to sample to show a clear section of the banding pattern (Figure 2) and the shell cross section structure of Arctica islandica (Figure 3) consisting of three layers of a mixture of calcium carbonate and aragonite; the increment, a thin prismatic myostracum which separates the outer and inner layers and the narrow growth line and all have different mircotextual features, for example the growth line consists of irregular prisms whereas the increment has a more homogenous structure allowing for differentiation (Jones 1980).

Figure 2: Arctica islandica; Inside view of the left-hand valve. The lines "A" and "B" respectively represent direction of the maximum shell height and maximum shell length. Line A corresponds to the direction of sectioning (Witbaard 1997).

Figure 3: Arctica islandica; Shell cross-section along the line of maximum height (line A in figure 2) of an 8 year old specimen which was collected in March 1991. Growth lines are indicated by black lines. Most recently deposited increment is on the left side (Witbaard 1997).

Ropes (1984) presents a more detailed description of the crystal morphotypes that can be seen in a shell cross section. From discovering the differences in crystal types and the desire to visualise the internal growth lines Ropes (1985) implemented the acetate peel technique from the research in palaeontology which is the study of life forms that existed in prehistoric or geological times represented by the fossils of flora and fauna (Kummel 1965) which is described in the methods of data analysis section.

The ring structure of tree cross sections can be studied in a similar way due to the density differences between earlywood and latewood (Figure 5). Earlywood is formed during the period of rapid radial growth associated with the beginning of a growth season and is only alive for a few days until each cell divides and rapidly increases in size then begins to function as a conducting element (Schweingruber 1987). Whereas latewood is formed towards the end of a growing season as the cambial growth and activity slows down. These cell walls are thicker and stronger and appear much darker in colour due to their continual slow growth until the end of the vegetation period (Stokes 1968)

It is the sharp contrast between the early and latewood and their consecutive growing seasons that forms the well known and easily identifiable annual ring patterns that can be seen in tree cross sections without magnification (Figure 4). In addition to the radial rings, there may be horizontal bands called rays present. These are rows of cells radiating out from the centre of the stem at right angles to the rest of the cells. Their function is of lateral conduction and is only of interest to dendrochronologists during archaeological excavations, when these rays can be used to identify the type of wood for example if the piece of excavated wood has 'travelled' or if it was made, buried and lost where it was discovered (Brothwell 2001, Schweingruber 1987).

Environmental Factors Influencing Growth of Rings and Bands

The environment supplies all the necessary components required for metabolic process such as growth as well as for photosynthesis for trees, either via gyres, climate fronts determined by prevailing atmospheric and ocean circulation patterns. Any changes, either short term such as seasonal or long term such as decadal growth and recession of ice ages, of any of these parameters will be reflected in chemical changes in the water column and surrounding air masses and thus will be recorded in growth patterns (Lowe 1997). The abundance or lack of any or all of these contributing elements will determine if the tree or mollusc will grow to the limits of it genetic potential rather than have a growth limiting factor. The genotype which is the genetic make-up of the tree or mollusc will determine the environment ranges in which the individual can tolerate and prosper; the genotype also controls the response given to the influences of the environmental conditions (Stokes 1968). This affects the phenotype which is the observable physical or biochemical characteristics of an organism this may be caused by limiting or allowing maximum growth (Smithgill 1983).

In marginal environments where stress reduces yield (Byerlee 1993), there are two types of climatic stress that are commonly registered in tree ring data, these are moisture and temperature stresses. Tree that grow in semi-arid areas are often limited by the availability of water which is primarily reflected in the ring-width variations (Schweingruber 1987). Trees that grow near their species latitudinal or altitudinal maximum tree-line extent display temperature limitations on their growth. However, in 1971 Fritts described that there are several other climatic factors that may well be indirectly involved along with biological processes within the tree which are highly complex and are often modelled (Godin 2000, Sievanen 2000) and similar growth series may be a result of a significantly different combination of influential climatic conditions (Schweingruber 1987).

In South Western American regions the dominant growth limitation is precipitation, growth varies with the amount of precipitation received and utilised by the tree. Precipitation levels affect the soil moisture content, for example; the more precipitation the higher the soil moisture content, therefore there is a greater amount of water available for use by the tree which results in a wider growth ring (Bradley 1999). The effective soil moisture content is controlled by the amount, type and timing of precipitation but crucially also by the texture, drainage and composition of the soil (Figure 7) (Stokes 1968). Ring-widths of bald cypress conifer trees (Taxodium distichum) from swamps in the South Western US have been sampled and used to construct a chronology demonstrating the drought and precipitation history of the area over the last 1000 years (Stahle and Cleveland 1988, 1992).

If local underground water is available or losses from runoff are low, the soil moisture content will be sufficient in most years for optimum growth of a tree. This period of optimum growth creates a 'complacent' ring pattern (Figure 7) which can be uniformly narrow or wide rings, with insufficient variation to produce a recognisable sequence to a dendrochronogologist however may supply an excellent botanical specimen (Stokes 1968). This type of ring pattern is found in trees growing near lakes, in a river valley or on a roadside location. If sample sites have no permanent underground water available and the soil drainage is good, often found along rocky hillsides and steep slope. Radial growth is significantly proportional to total precipitation levels, and the resultant 'sensitive' ring pattern produces datable ring patterns (Stokes 1968, Schweingruber 1987).

In contrast, temperature is the dominant growth limiting factor of the Taiga, which is also known as the boreal forest and is dominated by coniferous trees such as larch, fir, and pine the White Spruce (Picea glauca) of Central Alaska (Chapin 1987). The sun is low in the horizon for the majority of the year, there is an added difficultly of receiving enough light energy for photosynthesis. Pine, fir and spruce are evergreen and photosynthesis with their older leaves in late winter and into spring when light levels have increased, but the temperature is still too cold for new growth (Lloyd 2005). Chapin (1987) described that the effect of temperature upon growth of taiga flora is the cumulative result of the effect of temperature upon all physiological processes. That is that the present in-situ growth rate is close to the maximum that can possibly be supported by the obtainable resources such as light and nutrients, therefore a more rapid growth rate would either require more resources or a warmer temperature to allow more resources to be made available.

As trees age there growth rate slows, this also happens to molluscs where new growth increments are deposited along the margin of the shell become too crowded to separate the increments externally from each other (Witbaard 1997). As trees reduce in growth rate, there is a morphological change, many species primary branches become thicker and the thinner branches become gnarled and twisted due to the unequal extension of branches (Ryan 1997). This reduction in growth rate is taken into account through detrending methods during the analysis of the primary data.

Furthermore, preconditioning of physiological processes the tree may occur, for example advantageous climatic conditions prior to the growing season will strongly facilitate subsequent growth (Schweingruber 1987). This principle could also be applied to molluscs, growth and food production in one year may influence performance in the following year, which leads to strong serial correlation in records.

Marine molluscan growth is affected by a number of varying influential environmental factors such as substrate, nutrient supply from the water column, temperature, salinity and oxygen levels, light levels and depth which itself also affects a large proportion of the factors (Lowe 1997). However the main controls over their distribution, abundance and life strategies are the currents and water temperature (Peacock 1989, 1993). Marine molluscs are grouped into their major zoogeographical provinces (Figure 1) which are ocean zones that echo gradual or sudden changes in water temperature. Living, and fossil, assemblages can consequently be categorised as Boreal, Lusitanian or Arctic depending on their ecological affinities. Major currents shown in Figure 1 contribute by affecting the essential nutrient supply along with maintaining water temperatures and also influence the dispersal of larval stages (Lowe 1997, Witbaard 1997). Salinity variations may also have an important control of growth in shallow or enclosed sea, as indicated by carbon and oxygen isotope ratios attained from mollusc shells in the Baltic Sea (Punning 1988).

Witbaard (1998) carried out experiments that showed that Arctica islandica shell growth begins early in the year even in areas at temperatures of 1oC. This suggested that small differences in bottom water temperature that coincide with the springbloom down flux of sediments towards the seabed might significantly influence shell growth (Winter 1969). However there is a tenfold increase of shell growth at temperatures between 1o and 12oC. The greatest change in shell growth rate occurred between o and 6oC with average shell growth varying between 0.0003mm/day at 1oC to 0.0032mm/day at 12oC (Witbaard 1998).The carbon conversion efficiency calculated at a temperature of 9oC is between 11-14% for Arctica with a shell height of between 10-23 mm (Witbaard 1997, 1998). These results imply that temperature has a small effect on the time spent in filtration, compared to particle density which strongly influences that response. Which is also suggested during periods of high particle density Arctica balances lower filtration rates with extended periods of activity (Witbaard 1997). However on a larger scale, temperature has a significant effect on growth when seen over the entire latitudinal range (52o-66oN) although the effect on North Sea organisms is small.

The inflow of Atlantic water along the Eastern side of the Shetland Islands establishes a topographically generated eddy overlying the area. The central section of the eddy coincides with high densities of Arctica as found by Witbaard in 1997. It was hypothesised that these Arctica beds are situated in that location due to the water circulation within the eddy which leads to enhanced deposition of phytodetrius increasing the nutrient and food supply to the bivalve molluscs. This overall implies that shell growth variations may reflect variations in the type, quality and amount of influx of Atlantic water into the North Sea (Witbaard 1997).

Methods of Sample Preparation

Schweingruber (1987) described the optimal cross-sections of tree ring series should be taken from fresh wood; however fossilised and dead trees can also be sampled in a similar manner but need to be treated with greater care as not to damage the fragile samples. Extracts of wood are sampled with a metal increment borer from the tree trunk and can be up to 1 metre in one sample length. These samples are then mounted and polished prior to further examination which includes counting and visual inspection of the ring series measurements are carried out under normal magnification (Lowe 1997).

Marine mollusc sections are sectioned as previously described in Figure 2. The surface of the cross-section is polished and etched in a 1% solution of Hydrochloric acid which dissolves the carbonate sections but the carbonate matrix is conserved. This procedure results in a cross-section that highlights the structural difference between the increment and the growth line and is transformed into a micro-relief (Witbaard 1997).

X-Ray Densitometry

Once tree ring samples have been collected, X-ray densitometry is carried out, in which the sample sections of wood are x-rayed and the negatives produced are scanned by a beam of light and a photocell. The amount of transmitted light through the negative is determined by the density of the wood (Figure 8). These density variations are often more reliable than ring widths as climatic indicators (Bradley 1999, Schweingruber 1987). This is due to wood density being an integrated measure of several properties including thickness of the cell walls along with the size and density of vessels and ducts (Polge 1970). It is the variations between early and late wood average density that can be used to identify annual growth rings and to cross sample data (Parker 1971). It was shown by Schweingruber (1979, 1993) that these density variations contain a strong climatic signal and can be used to estimate long-term variations in climate over wide areas.

Figure 8: An example of a tree-ring density plot based on an x-ray negative of a section of wood (top). Minimum and maximum densities of each ring (bottom) are then used to measure the annual width as well as the width of early and late wood (Bradley 1999).

Density variations are valuable in dendroclimatology due to their simple growth function which is often close to linear with age (Schweingruber 1987). Standardisation of this density data would allow for more low-frequency climatic changes to be retained compared to standardised ring-width data. The minimum and maximum densities are measured representing locations in the early and latewood sections, however Schweingruber (1993) demonstrated that maximum density values were of more analytical which was proved by D'Arrigo (1992) who correlated maximum densities with April-August mean temperatures in trees from boreal regions. Whereas the minimum and mean density values along with ring-widths had a weaker relationship with summer temperatures. The calibration of maximum latewood density values is achieved in the same way was ring-width data using various statistical procedures as described under the standardisation section. Optimum climatic chronologies can be constructed by using both ring-widths and the densitometric data to maximise climatic signals (Bradley 1999, Briffa 1995).

Acetate Peel Technique

Earlier studies showed that growth variations in the hinge and valve of Arctica shells correspond closely, however the hinge band is used because the band is condensed, well defined and less susceptible to short term environmental disturbances. This is because the growth of the hinge band occurs in an area that is under maximum shielding conditions (Thompson 1980). The left hand valve was used due to the large hinge tooth which is positioned in the direction of maximum shell height which crosses the umbo. It is for these reasons that Witbaard (1997) used this section for study because growth increments can be easily traced back.

d18O and d14C Isotope Analysis

In sclerochronology the two main stable isotope ratios used are the d18O and the d14C whereas in dendrochronology radiocarbon dating is mainly used to analyse the d14C content. The oxygen isotope composition of a molluscan shell is a function of the water temperature at which carbonate (CaCO3) was deposited into the increment (Weber 1970). The d18O/ d16O isotope ratio varies locally due to local temperature influences for example currents bringing warmer or colder waters, and globally with variations in continental ice volume (Bradley 1999).

The procedure to measure the isotopic composotion of a particular increment of Arctica shell, as described by Jones in 1996 outlined that the shell must be sectioned as outlined previously and then sampled by using a small (<0.5 mm) drill to grind shallow grooves into the outer shell layer, parallel to external growth lines, across at least two major growth increments (Krantz 1984). Organic contaminants were removed from the CaCO3 samples by using the H2O2 procedure outlined by Allmon (1992).The powdered samples were then analysed to concur with the standard techniques involving a reaction in a vacuum with 100% orthophosphoric acid at 90oC for 0.25hr. Computer software facilitated the production and purification of the carbon dioxide (CO2) gas produced. The PDB standard, which is a reference used for d13O marine CaCO3 standard obtained from a Cretaceous marine fossil, Belemnitella americana, from the PeeDee Formation in South Carolina. PDB is used because it has a higher d13C/ d12C ratio compared to all other natural carbon-based substances and has been assigned a value of zero, giving almost all other samples a positive value (Craig 1957). The isotopic differences between the sample CO2 and the PDB standard were calculated using a mass spectrometer (Jones 1996).

Changes in isotopic composition can be complicated by;

(1) The metabolic production of CO2. During the formation of CaCo3 bands metabolic produced CO2 may be incorporated, this would affect the isotopic equilibrium with the water and the resultant isotopic composition would be different to the thermodynamically prediction. This would result in lower d18O and d14C values than the expected equilibrium vales (Bradley 1999, Duplessy 1970, Vinot-Bertouille 1973).

Figure 10: The relationship between d18O and salinity in oceanic surface waters, based on modern water samples by Broecker (1989) using data of H.Craig.

(2) d18O values are strongly related to salinity. As salinity increases so does the concentration of d18O in the carbonate shell (Figure 10), this is a consequence of continental ice sheet formation during glacial periods due to the removal of isotopically light water from the oceans which involves the rejection of salts into the water column which increases the salinity of the water column (Bradley 1999).

This is recorded in the carbonate band as a section of a high d18O/ d16O ratio, i.e. there is a greater concentration of d18O than d16O. The opposite effect occurs during a warmer interglacial period which results in an abundance of d16O due to diluting effect of the meltwater discharge into the oceans which creates a smaller d18O/ d16O ratio (Bradley 1999). This can be used to estimate the salinity changes at the sea surface, though this assumes all other effects can be determined independently.

d14C analysis of Arctica islandica growth bands support the hypothesis of an annual growth increment deposition (Witbaard 1997). The observed cyclic variation in the d18O and the d14C values coincided with growth bands i.e. the variation in d18O and the d14C integrated in each increment matches the expected variation predicted by seasonal variation in temperature, primary production and ice formation. An example of this is the previously mentioned 1960s pulse of nuclear bomb d14C can be seen in Figure x along with the 1816 "Year Without a Summer".

Variations in oxygen and carbon in wood have been studied as a possible proxy of temperature variations throughout geological time periods; however the associated complexities of isotope fractionation both in the trees and in the hydrological system make interpretations difficult to calculate (Bradley 1999). For isotopic dendroclimatic studies the sensitivity requirements are not as critical, and it may be preferable to use complacent tree-ring series for analysis due to different species of tree with differences in width of early and latewood causing different isotope ratios (Gray 1978, Wigley 1978).

Radiocarbon Dating

(FIND section (Bradley 1999)

Methods of Data Analysis


Ring and band-width series from a limited geographical area can be matched through crossdating. This is possible due to trees and mollusc shells reflecting climatic variations in characteristic and recognisable patterns i.e. growth series must be identifiable to enable matching to an existing chronology. This forms the basis for cross-matching (Lowe 1997) between trees and shells of overlapping age range. Where fossilised logs are available for example in peat bogs, or shells from Arctica samples that had previously died, crossdating these specimens enable a back chronology to be created spanning thousands of years. Several computer aided techniques have been developed for this purpose (Bradley 1999, Lowe 1997).

Spatial coherence of a common signal will determine the geographical area the chronology constructed would include (Butler 2009). The similarities in time series recorded at sites that are distant from each other are known as teleconnections (Rolland 2002). In tree-ring data the common signal can be detected across a distance of 1200km in the French Alps by using Mountain and Stone Pine, European Larch and Norway Spruce (Rolland 2002). Large scale climate reconstructions using a system of tree-ring chronologies require records from neighbouring areas to have a smooth transition so that a regional record can be constructed (Briffa 2002).

Trees are deliberately selected from stressful situations so that sensitive ring series are formed, there is the possibility that the tree may fail to produced new cells or produce unequal growth producing missing rings., which can also occur during the growth of bivalve molluscs (Bradley 1999). If there is more than one growth layer per annum this is known as a false ring and is produced by another short growth season during late spring or early summer, after the spring cells have commenced growth. Missing and false shell and ring increments are more than likely to be locally produced, so that as networks of chronologies are constructed, the alignment of distinctive increment widths can be matched using computer software thus identifying the anomalous increments (Butler 2009, Holmes 1983). Great care is required when crossdating tree-ring patterns due to the possibility of false or missing rings which are identified by;

(1) Comparing several tree ring series samples from the same specimen or population (Douglass 1934)

(2) Anchoring the sample to another wood sample of a known date obtained from an archaeological excavation (Douglass 1934)

(3) By comparison with a neighbouring, similar, population (Fritts 1976)

In the development of shell-based chronologies only the first method is usually viable whereas the second is used in a limited extent. However, the identification of historic specimens of A. islandica obtained from museum collections of shallow water specimens with a known date of death can be used for this purpose. The third method is is becoming more widely used due to the chronologies constructed so far have been isolated in space samples and do not cover large regions (Butler 2009), however as more research is carried out and more chronologies constructed the more likely this method is to cover larger areas.

In the marine environment these teleconnections are inhibited by currents and water mass mixing forced by topography, wind forcing or stratification dynamics (Butler 2009). A study by Witbaard in 1997 compared two Arctica chronologies from samples taken 75km apart in the Northern North Sea during 1890-1990. These sites showed synchronous responses in growth rates until 1960, and then became negatively correlated. Witbaard (1997) theorised that this was due to changes in hydrography of one of the sites which affected the food supply to the most Northern site.

Data series can be disrupted due to physical damage to the shell or tree. Tree ring data may be damaged due to several factors such as fire (Zackrisson 1997), disease and insect activity (Cochran 1998, Cook 1990). The most common cause of Arctica shell damage is trawling. The South Eastern North Sea densities of Arctica decreased between the 1970s to the 1990s, with the mortality rate in the South Eastern North Sea being twice as high as that in the Northern North Sea (Witbaard 1997). Small injuries which are non-fatal caused by the tickler chains may result in scar damage and the frequency of damaged shells sampled have increased since the 1970s. This corresponds with the increasing size and total engine capacity of the Dutch beam trawl fleet which normally trawls in this area (Witbaard 1994, 1997).


Master Chronologies


The major application for chronologies constructed, both sclerochronological and dendrochronological time series, is to provide useable proxies for palaeoclimate reconstruction. This allows scientists to understand the variations and the frequency of major events in the geological past which allows for prediction and preparation.

Dendrochronology can be used, as previously mentioned, as an exceptional archaeological tool. Robinson in 1976 used beams and logs of wood from Indian pueblos located in many places in South West US to construct chronologies of up to 2000 years. Similarly, important archaeological chronologies have been established by Hoffsummer (1996) in Western Europe, by using supporting beams of oak from buildings in South Eastern Belgium which extends back to A.D. 672. In several regions of France, Lambert (1996) constructed chronologies from construction timbers dating back over 1000 years. Socially, dendrochronology has also been used to date important works of art, for example the wooden panels used in paintings, furniture and coverboards of early books (Lavier and Lambert 1996). Recovered tree stumps from alluvial sediments and bog areas have been cross-dated allowing composite chronologies extending back through the entire Holocene (Bradley 1999). These long tree ring series have been used to calibrate radiocarbon timescales (FIND section FROM BRADLEY 1999).

Tree-ring series and shell-based chronologies are unique within palaeoclimatic proxies due to through cross-dating multiple samples, an absolute age of a sample can be found. This distinguishes tree-ring and shell increment data from other proxies of high resolution such as ice cores and corals because comparable replication of these types of records is rarely possible (Bradley 1999). Therefore, with the exception of specific marker years such as those associated with volcanic events, for example the Mt. Tambora event in 1816, dating of other proxies is more uncertain than chronologies from trees and mollusc shells.



The bivalve mollusc Arctica islandica is known as the "Tree of the Sea" due to the methods and techniques of marine palaeoclimatic chronology construction being near identical to terrestrial chronology created from dendrochronological studies.