Journal of Petrology | Volume 43 | Number 11 | Pages 1987-2012 | 2002
© Oxford University Press 2002
Plume–Ridge Interaction: a Geochemical Perspective from the Reykjanes Ridge
1SOUTHAMPTON OCEANOGRAPHY CENTRE, EMPRESS DOCK, SOUTHAMPTON SO14 3ZH, UK
2DEPARTMENT OF GEOLOGY, ROYAL HOLLOWAY, UNIVERSITY OF LONDON, EGHAM TW20 0EX, UK
Received April 17, 2001; Revised typescript accepted May 3, 2002
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSINTERPRETATION AND DISCUSSIONSCHEMATIC MODEL FOR A...CONCLUSIONSREFERENCES |
|---|
Plume–ridge interaction in the Reykjanes Ridge and Iceland region is graphically demonstrated by several V-shaped ridges surrounding the spreading axis, indicating mantle flow away from Iceland. It also has significant geochemical effects. Regionally, incompatible element concentrations increase northwards coinciding with decreasing depth and increasing crustal thickness, depth of melting and proximity to Iceland. Major and trace element data show that isolated magma bodies feed individual volcanic systems along the ridge. Fractionation within these systems increases towards 60–61°N, where it coincides with the intersection of a V-shaped ridge, thicker crust and more abundant seamounts. Trace element, Nd and Sr isotopic data reveal dynamic melting and mixing within a southward-thinning, heterogeneous mantle wedge beneath the Reykjanes Ridge. Melting is variable and locally enhanced at 58°N, 59°N, 60°N and 61°N. A total of six mantle components are identified. Some are specific to Iceland whereas others are found only beneath the ridge axis. The geographical distribution of these components reflects their origin within the deep upper and lower mantle and subsequent translation by plume outflow along the entire length of the ridge.
KEY WORDS: plume–ridge interaction; Iceland; Reykjanes Ridge; dynamic mantle mixing and melting
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSINTERPRETATION AND DISCUSSIONSCHEMATIC MODEL FOR A...CONCLUSIONSREFERENCES |
|---|
Mantle plumes and hotspots affect large portions of the oceanic lithosphere. However, there are many unknown aspects to how plumes interact with spreading ridges. For example, it is generally thought that segmentation of slow-spreading ridges reflects organization in upwelling of the underlying asthenosphere and focusing of melt towards segment centres (Crane, 1985; Schouten et al., 1985; Macdonald et al., 1988; Schouten & Whitehead, 1992). Yet despite a half-spreading rate of 10 mm/yr (DeMets et al., 1994), the Reykjanes Ridge south of Iceland is continuous along strike for almost 1000 km (Talwani et al., 1971; Applegate & Shor, 1994). Thus its proximity to the Iceland plume appears to suppress ridge segmentation. One explanation for this may be that sub-horizontal asthenospheric flow beneath the ridge, driven away from the Iceland plume by local buoyancy forces (Sleep, 1996), prevents organization in upwelling and melting. Such outflow, at a rate of 10–15 cm/yr, is indicated by the presence of southward-closing V-shaped ridges located symmetrically about the Reykjanes Ridge (Vogt, 1971, 1974; Vogt & Avery, 1974; Sandwell & Smith, 1992; White et al., 1995; Searle et al., 1998). Alternatively, the Reykjanes Ridge may have segmentation similar to other slow-spreading ridges but its tectonic and bathymetric expression is obscured by along-axis flow of the lower crust resulting from higher mantle temperatures and thicker crust (Bell & Buck, 1992).
Here we investigate how the upwelling Iceland plume interacts with the neighbouring Reykjanes Ridge and whether it affects organization in sub-ridge mantle upwelling, melting and melt delivery. In particular, we explore to what extent the lack of tectonic segmentation of the ridge is reflected in its geochemical organization and the fate of plume material as it interacts with the mantle beneath the North Atlantic. We pursue these objectives using a comprehensive suite of geochemical analyses for samples recovered from 186 sites between 57·5°N and 63°N along the neovolcanic axis of the Reykjanes Ridge (Fig. 1).
|
Geological setting
The slow-spreading Reykjanes Ridge is part of the Mid-Atlantic Ridge, which forms the American–Eurasian plate boundary to the south of Iceland. It is oriented at 035°N, oblique to the plate separation trend of 092°N (Talwani et al., 1971; Applegate & Shor, 1994; Searle et al., 1998). To the south of 59°N, the ridge has an axial valley similar to other slow-spreading ridges, whereas north of 59°N it forms an axial rise (Talwani et al., 1971). The ridge shallows northwards and is exposed on the Reykjanes Peninsula in SW Iceland.
Although unsegmented, the Reykjanes Ridge expresses subtle morphological variations and regional slopes related to the proximity of the Iceland plume. These include a shallow rise (located between 60°N and 61°N), bathymetric undulations or swells with an 
100 km wavelength (centred on approximately 58°N, 59°N and 60°N) and 20–50 km long axial volcanic ridges (AVRs) that are the primary volcanic systems of accretion on the ridge axis (Laughton et al., 1979; Searle & Laughton, 1981; Murton & Parson, 1993; Parson et al., 1993; Applegate & Shor, 1994; Searle et al., 1998).
The AVRs are akin to en echelon fissure volcanoes on Iceland (Jacoby, 1980). They strike oblique to the trend of the Reykjanes Ridge and nearly orthogonal to the plate separation direction of 092°N (Searle & Laughton, 1981). On the basis of high-resolution sonar data, Parson et al. (1993) related variations in AVR morphology to cycles of volcanic accretion and tectonic dismemberment. Their interpretation is supported by seismic and electromagnetic observations of an ephemeral magma body beneath a morphologically ‘young’ AVR at 57·75°N (Sinha et al., 1998).
To each side of the spreading axis are a series of near-parallel ridges that converge to the south, forming a nested V-shaped pattern. The most northerly V-shaped ridge intersects the Reykjanes Ridge between 60°N and 61°N, where it coincides with a two-fold increase in the abundance of circular seamounts (Madge & Smith, 1994; Searle et al., 1998), a 50% increase in the size of AVRs (Murton & Parson, 1993; Searle et al., 1998) and a 20% increase in crustal thickness (Smallwood & White, 1998). These changes are thought to reflect higher melt supply associated with hotter mantle (e.g. up to 30°C warmer than ambient), migrating south at a rate of 10–15 cm/yr, forming the V-shaped ridges (Murton & Parson, 1993; Madge & Smith, 1994; White et al., 1995; Yale & Morgan, 1998; Albers & Christensen, 2001).
Northward shoaling of the Reykjanes Ridge towards Iceland is accompanied by increasing 87Sr/86Sr ratios, lower 143Nd/144Nd ratios and enrichment in incompatible trace element concentrations (Schilling, 1973). This is most noticeable to the north of 61°N (i.e. within 400 km of Iceland), where the composition of the ridge changes rapidly from a depleted, mid-ocean ridge basalt (MORB)-like composition to enriched compositions characteristic of the Iceland plume. In contrast, high 3He/4He ratios extend over a much greater distance, to at least 1000 km south of Iceland (Poreda et al., 1986; Hilton et al., 2000).
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSINTERPRETATION AND DISCUSSIONSCHEMATIC MODEL FOR A...CONCLUSIONSREFERENCES |
|---|
Sample selection and preparation
The samples used in this study were recovered during R.R.S. Charles Darwin cruise CD80 (Murton, 1994). Recovery of volcanic material was made from >180 stations. Average spacing between stations is 3·5 km, but 2 km in more detailed sampling areas. Fresh volcanic rock targets were identified from multibeam bathymetric data, 30 kHz sidescan sonar and a 3·5 kHz vertical-incidence echo. A short-haul dredging technique was used, ensuring the length of seafloor covered during sampling was <500 m. Nearly all of the dredges recovered a mixture of crystalline basaltic material with glassy rinds, as either pillow or sheet flows. Also recovered were hyaloclastite, pelagic sediment and biological material (Murton, 1994). Samples suspected of being ice-rafted (e.g. rounded, weathered or of non-volcanic composition) were rejected from this study.
Major and trace element analyses were made on predominantly crystalline material. In addition, selected glass samples were analysed for comparison. After removing exposed, altered and sediment-stained faces, samples were cleaned by washing several times in deionized water. After drying, they were crushed in a steel press. The fragments were again soaked in warm deionized water for 1 h to remove any vestiges of contamination before being powdered. Radiogenic isotopic analyses were performed on visually unaltered subsamples that were crushed but not ground.
Analytical procedures
Major element analyses were performed by X-ray fluorescence (XRF) on fused glass discs using a Philips PW1400 system at the University of Southampton and trace elements were determined by XRF on pressed powder pellets with a Philips PW1480 instrument at Royal Holloway, University of London, using the method described by Jochum et al. (1990) & Thirlwall et al. (1994).
In light of the low concentrations of trace elements expected in these basalts, count times were set to give a two times standard deviation (2 SD) reproducibility and comparable accuracy of ±0·6 ppm for Zr, ±0·4 ppm for Y and ±0·1 ppm for Nb and Rb. In addition, concentrations of trace elements and rare earth elements (REE) were measured using inductively coupled plasma mass spectrometry (ICP-MS) at the University of Southampton. Concentrations were determined by calibration of isotope count levels with matrix-matched international standards BIR-1, JB-2, BHVO-1, BCR-1, BE-N and JB-1a. ICP-MS REE data were further calibrated using 11 analyses for REE performed using isotope dilution (ID). For these, 95% of the ICP-MS REE values were within 3·5% of the ID results. Ratios agree (at 2 SD) to 1%, 2% and 2·4% for Sm/Nd, La/Sm and Dy/Yb, respectively. A comparison of trace element results by XRF and ICP-MS for an internal standard and the international standard BIR-1 has been given by Taylor et al. (1995, table 1) and shows XRF vs ICP-MS differences to be small: Rb, Nb and Ba agree within 0·1, 0·2 and 3 ppm, respectively, at low levels, although ICP-MS Nb is 10% higher at >6 ppm. XRF and ICP-MS determinations for Zr, Sr and Y agree within ±6% (2
). Although the ratios SrXRF/SrICP-MS, ZrXRF/ZrICP-MS and YXRF/YICP-MS are strongly correlated, samples from the same dredge haul frequently show near-identical XRF but disparate ICP-MS results, indicating that the ICP-MS data are substantially less reproducible than XRF, although more precise at low levels. Consequently the ICP-MS data have been adjusted by multiplication by the ratio ZrXRF/ZrICP-MS, which preserves the higher precision of the ICP-MS data for some elements but eliminates the poorer reproducibility owing to minor ICP-MS sensitivity drift.
|
Radiogenic isotope ratios for Sr and Nd were measured by thermal ionization mass spectrometry (TIMS) on a VG354 system at Royal Holloway, University of London (Thirlwall, 1991a, 1991b). Subsamples for Sr isotopic analyses were leached in dilute HCl for 5 h. Within-run standard errors (2 SE) are less than the last two significant figures quoted for each ratio [see Taylor et al. (1997) for details]. External reproducibility is better than ±0·000015 for 87Sr/86Sr and ±0·000008 for 143Nd/144Nd. Sr and Nd isotopes are normalized to 87Sr/86Sr = 0·1194 and 143Nd/144Nd = 0·7219. Pb isotope data, determined by double-spike techniques, will be published together with improved determinations for some of the Sr–Nd isotope data of Thirlwall et al. (in preparation).
Our results, summarized in Table 1 and Fig. 2, show that samples from north of 61°N are enriched in selected incompatible elements relative to those further to the south and to average N-MORB (Sun & McDonough, 1989). The full data are available as an electronic appendix, which may be downloaded from the Journal of Petrology website at http://www.petrology.oupjournals.org.
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSINTERPRETATION AND DISCUSSIONSCHEMATIC MODEL FOR A...CONCLUSIONSREFERENCES |
|---|
Major element variations
The major element data display coherent trends with distance along the ridge (Fig. 3). For the more incompatible elements, K2O and P2O5, there is a general decrease in average concentration northwards from 57·5°N to a minimum at 60·75°N. This behaviour is also shown, to a lesser extent, by TiO2. In contrast, the average concentration of Na2O remains almost unchanged along the ridge. North of 61°N, K2O, P2O5 and TiO2 concentrations are significantly richer. Compared with samples from 60–61°N, those from 63°N are richer by a factor of 5·5 for K2O, 2 for P2O5 and 1·5 for TiO2. To the south of 61°N, low concentrations of these elements occur systematically at 58°, 59° and 60°N, where the relative change in concentration is about 80% for K2O, 50% for TiO2, 20% for P2O5 and 10% for Na2O. Conversely, MgO monotonically decreases in concentration northwards from an average of 8·6 wt % MgO at 57·75°N to 7·5 wt % at 63°N. Its average of 8·1 wt % is close to the global average for MORB (Langmuir et al., 1992). Only nine samples have >9 wt % MgO and of these, two samples with >10 wt % MgO have an excess of 8% (by volume) of olivine phenocryst (e.g. samples from sites 17D and 153D). Similarly Mg-number [i.e. expressed as 100Mg2+/(Mg2+ + Fe2+)] decreases northwards along the ridge from an average value of 74 to 68.
|
In the southern part of the Reykjanes Ridge, at 57·52°N, samples from site 14D, an exceptionally large seamount (referred to here as the Mega-seamount) located in a non-transform offset basin (at 57·52°N), are anomalously enriched. Compared with its immediate neighbours, the Mega-seamount has 70% more K2O, 25% more P2O5 and 12% more TiO2 but only average MgO. Other samples also have anomalous compositions compared with their neighbours. For example, site 17D (from 57·89°N) is depleted by 11% for Na2O, 27% for TiO2 and 45% for P2O5. Although this coincides with slightly higher concentrations of MgO, consistent with the presence of >8% by volume of olivine phenocrysts, the depletion in incompatible major elements is too large to be the result of olivine dilution alone.
Trace element variations
The incompatible major and trace elements show distributions similar to K2O, P2O5 and TiO2, having minima in concentrations at about 60·25°N and increasing concentrations northwards (Fig. 4). Most of these elements are enriched relative to average N-MORB, except for Sr (Fig. 2). Nb, Ba, Rb, Zr and La show the greatest increase in concentration northwards, whereas Y and Yb have similar concentrations along the entire length of the ridge. Compared with 60·25°N, the average concentration at 63°N is higher by a factor of 11 for Ba, 10·8 for Nb, 9·5 for Rb, 5 for La, 2·1 for Sr, 1·8 for Zr and 1·2 for both Y and Yb. The average concentration at 58°N is also higher by a factor of 2 for Ba and Rb, 1·5 for Nb and La, 1·1 for Sr and 1·15 for Zr, and is similar for Y and Yb, compared with 60·25°N.
|
In addition to their regional distribution, unusually low concentrations of incompatible trace elements occur systematically at 58°N, 59°N and 60·25°N (Fig. 4). The amplitude of these undulations in concentration exceeds analytical error and hence they are considered significant. Because the data coverage extends south only to 57·5°N, where there is considerable compositional variance, the minimum observed at 58°N might be part of a longer-wavelength decline in concentration of which the southern and lower limit is not observed.
The Mega-seamount (site 14D), site 17D and site 78D all have anomalous compositions for their latitude. The Mega-seamount is richer in the more incompatible trace elements (Nb, Ba, Rb and Sr) and closely resembles samples located at 62°N. Site 78D is also relatively rich with high Nb, Ba, Rb and La concentrations. In contrast, site 17D is relatively poor compared with the average for its latitude: by 55% for Zr, 50% for Yb, 45% for La, 37% for Nb, 30% for Y, 16% for Ba and 8% for Sr.
With the exception of the most northerly site (185D), all samples show chondrite-normalized, middle to light rare earth element (MREE and LREE) depleted profiles. The extent of depletion increases towards the south (Fig. 5). All profiles also show negative Eu anomalies, becoming more negative with increasing HREE concentrations (Fig. 6). Compared with the northern suite, the southern suite (south of 61°N) has lower LREE concentrations. Most samples have convex-upward profiles (between Lu and Tb) at values of about 10 times chondrite and values that are progressively more depleted towards the LREE end (i.e. from Dy to La; Fig. 5a). These REE profiles are generally parallel and show Lu[N] values ranging from 10·3 to 13·2. However, the Mega-seamount and site 78D have profiles with higher concentrations of LREE compared with the average for the southern suite. The Mega-seamount has an almost flat REE profile, with chondrite-normalized REE values of 10.
|
|
|
) and 57·45°N (
). These trends are reproducible by modelling Rayleigh fractional crystallization of olivine, plagioclase and clinopyroxene in proportions ranging from 1:2:1 to 1:4:1 and using electron microprobe data for individual phases in these rocks reported by In contrast, REE profiles for samples forming the northern suite are less parallel to one another (Fig. 5b). Several cross the average trend and show variable degrees of LREE depletion. The most northerly sample (site 185D) has LREE enrichment and the highest value of La[N] (16·2). Yet nearby, site 147D (from 61·46°N) is strongly LREE depleted with a low value of La[N] (5·61) but one of the highest values of Lu[N] (14·28).
Sites located between 60° and 61°N have some of the lowest REE concentrations. For example, REE profiles for sites 108D and 122D are parallel to the trend for the majority of data, but have the lowest La[N] (2·88–2·10), Dy[N] (8·81–10·18) and Lu[N] (8·35–9·83) values found anywhere along the ridge. They also have the weakest negative Eu anomalies. The location of these samples also coincides with the lowest concentrations of incompatible major and trace elements found anywhere along the ridge.
Isotopic variations
Along the entire length of the ridge, 87Sr/86Sr and 143Nd/144Nd ratios generally display coherent variations with latitude (Fig. 7). For 87Sr/86Sr the ratios are generally constant between 57·5°N and 59°N, falling to a minimum near 61°N and then rising sharply towards a maximum at 63°N. The Mega-seamount (i.e. site 14D) and sites 17D, 12aD and 78D deviate from the regional trend with 87Sr/86Sr ratios that are significantly higher than their neighbours. However, whereas the Mega-seamount is incompatible major, trace and light rare earth element rich, its neighbouring site 17D is relatively poor.
|
|
Generally, 143Nd/144Nd ratios display a reciprocal pattern to 87Sr/86Sr. With the exception of a few samples, they also lie within a narrow range. However, unlike 87Sr/86Sr, 143Nd/144Nd decreases slightly between 57·5°N and 59·7°N, rises towards 60·83°N, then decreases steeply to towards 63°N. At 57·5°N the average ratio is 0·51318 whereas at 59·7° it is 0·51314. At 60·83°N the ratio has increased to 0·51318 but falls northwards to a minimum of 0·51306 at 63°N. Some samples have 143Nd/144Nd ratios that fall outside of these general trends, coinciding with anomalous 87Sr/86Sr. In particular, they include the Mega-seamount (site 14D) and sites 17D, 78D and 143D.
Although the 87Sr/86Sr data generally show an inverse relationship with 143Nd/144Nd, there are systematic differences between the northern and southern parts of the ridge (Fig. 8). Between 57·5°N and 58°N the data form a field with higher 143Nd/144Nd for a given 87Sr/86Sr compared with those samples located further to the north. The trace-element-poor samples from site 17D plot at the high 87Sr/86Sr, intermediate 143Nd/144Nd end of this field. Between 58°N and 60°N the ridge has lower 143Nd/144Nd at a given 87Sr/86Sr. The group of data from 60°N to 61°N coincide with the field containing data from north of 61°N. Compared with those samples located to the south, the data from north of 60°N form an array that has lower 143Nd/144Nd for a given 87Sr/86Sr, extending towards even higher 87Sr/86Sr with increasing distance north. The Mega-seamount plots within the northern array, where it is located at the higher 87Sr/86Sr, lower 143Nd/144Nd end.
|
Relationship between major and trace elements, and Nd and Sr isotope ratios
In general, the incompatible major elements correlate positively with the incompatible trace elements and 87Sr/86Sr but inversely with 143Nd/144Nd. However, there are systematic differences along the ridge axis, especially between samples located to the north and south of 61°N.
Incompatible major and trace element co-variation is demonstrated by K2O/TiO2, which increases with both Nb/Zr and Rb/Sr (Fig. 9a). Although samples to the south of 60°N have lower incompatible element ratios and concentrations than those to the north of 61°N, they form an array with higher K2O/TiO2 for a given Nb/Zr. Samples from between 60°N and 61°N have amongst the lowest incompatible element ratios, forming a field that trends parallel to, and overlaps with, the southern suite. Samples from site 17D plot within the field formed by the southern samples whereas the Mega-seamount plots above the northern array with even higher K2O/TiO2 for a given Nb/Zr ratio.
|
For K2O/TiO2 and Rb/Sr, both the northern and southern suites form a single trend. Again, the Mega-seamount plots above the northern field with higher K2O/TiO2 for a given Rb/Sr ratio (Fig. 9b). A similar relationship is observed for La/Yb vs Nb/Zr and La/Yb vs Zr/Y, with the southern suite having lower incompatible element ratios compared with the northern suite (Fig. 9c and d). Samples from between 60°N and 61°N form an array, with the lowest La/Yb, Nb/Zr and Zr/Y values, that overlaps and parallels the southern suite. However, whereas the southern suite plots on the same Nb/Zr vs La/ Yb trend as the northern suite, it forms a separate trend towards higher Zr/Y for a given La/Yb ratio. Again, samples from the Mega-seamount plot within the field defined by the northern suite whereas site 17D is one of the most depleted.
Co-variation of La/Yb and Nb/Zr with 87Sr/86Sr and 143Nd/144Nd also shows similar differentiation between the northern and southern suites (Fig. 10). Compared with the northern suite, the southern suite forms an array with lower La/Yb and Nb/Zr for a given 87Sr/86Sr and 143Nd/144Nd, which extends towards higher 87Sr/86Sr and across a range of 143Nd/144Nd. The northern suite forms a steeper array with higher La/Yb and Nb/Zr for a given 87Sr/86Sr and 143Nd/144Nd, which also extends towards higher 87Sr/86Sr and lower 143Nd/144Nd.
|
In all cases, samples from between 60°N and 61°N form a discrete field with low La/Yb, low 87Sr/86Sr and high 143Nd/144Nd that overlaps and extends parallel to the southern suite. The Mega-seamount is anomalous and lies outside the southern group to which it belongs geographically, falling instead within the northern array. Site 17D plots with low Nb/Zr and La/Yb, at the high 87Sr/86Sr and low 143Nd/144Nd ends of the southern array.
|
INTERPRETATION AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSINTERPRETATION AND DISCUSSIONSCHEMATIC MODEL FOR A...CONCLUSIONSREFERENCES |
|---|
Volcanic processes
The progressive decrease northwards along the ridge in MgO content and Mg-number is compatible with increasing total fractional crystallization. The trend is also concurrent with the northwards shoaling of the ridge axis and increasing crustal thickness, implying increased melt supply associated with the Iceland plume (White, 1997). Because fractional crystallization requires sustained residence of magma below its liquidus temperature before eruption, the increase in total fractional crystallization northwards along the ridge indicates an increase in magma chamber control. However, the limited range in MgO for individual AVRs is also evidence for magma chamber replenishment (Gee et al., 1998a, 2000). Hence the increase in total fractionation northwards suggests a greater interval between replenishment events, possibly as a result of an increase in the size of magma chambers beneath individual AVRs. Alternatively, fractional crystallization may occur during ascent of magma through the crust, such that increasing crustal thickness results in greater ascent times and hence more fractionation before eruption.
Most samples show phenocrysts of olivine and plagioclase, and occasionally clinopyroxene, consistent with low-pressure cotectic crystallization [see Taylor et al. (1995) for a fuller description]. As a whole, the data from the ridge form a large range in CaO, Al2O3, TiO2 and Fe2O3 vs MgO (Fig. 11). The general trends show an increase in CaO and Al2O3 as MgO decreases from 9·3 wt % to 8 wt %, below which the decrease in CaO and Al2O3 is consistent with increasing plagioclase removal. In contrast, TiO2 and Fe2O3 increase with decreasing MgO, indicating little or no magnetite fractionation. Within these broad arrays, data for individual AVRs show discrete liquid lines of descent with different slopes away from different compositions, demonstrating both parental melt diversity and a lack of connectivity between magma systems.
These liquid lines of descent are modelled by simple fractional crystallization of between 10 and 15% removal of olivine, plagioclase and clinopyroxene, in proportions ranging from 1:2:1 to 1:4:1 [using electron microprobe data for individual phases in these rocks, reported by Nichols (2000)]. The ubiquitous presence of a negative Eu anomaly in most REE profiles also supports evidence for the removal of plagioclase during crystal fractionation.
We interpret the individual fractionation trends as evidence that AVRs are underlain and fed by isolated magma chambers with little communication between one another. Because the AVRs form a closely spaced, overlapping, en echelon pattern within the ridge axis, the lack of magmatic mixing suggests it is unlikely that adjacent magma chambers are active simultaneously. This interpretation is supported by the observation of systematic variations in morphology between adjacent AVRs, interpreted as evidence for cycles of volcanic accretion and tectonic dismemberment (Murton & Parson, 1993; Parson et al., 1993). Although there is no correlation between morphological ‘age’ of the AVRs and their geochemistry, the lack of magma mixing suggests that magma chambers beneath adjacent AVRs are out of phase with one another.
Assuming Ti behaves incompatibly during fractionation [with a bulk partition coefficient (KD) of 0·047 (McKenzie & O’Nions, 1991)], we can model the local extent of fractional crystallization within each AVR. The local extent of fractionation represents the diversity of magma compositions erupted at any single AVR that can be related by crystal fractionation, but is not the total extent of fractionation. The local extent of fractionation is represented by the difference between the maximum and minimum TiO2 concentration of lavas erupted at a single AVR. Figure 12 shows that the local extent of fractional crystallization increases northwards, from 57·5°N to a maximum at 60°N, falling to a low at 61°N, then increasing towards 61·7°N, before finally decreasing towards 63°N. South of 60°N, the pattern is in phase with the variation in abundance of circular seamounts observed along the ridge axis (Searle et al., 1998). However, north of 60°N, the amplitudes of the two patterns are less similar. Assuming the correlation is more than merely coincidental, especially to the south of 60°N, it indicates that an increase in the abundance of volcanoes occurs where conditions are favourable for magma chambers to form beneath AVRs. Further north, the local extent of fractionation at individual AVRs decreases, despite an increase in the total extent of fractional crystallization (i.e. a decrease in Mg-number or MgO). An explanation for this might be that, with thicker crust northwards, magmas undergo more fractionation at deeper levels. With increasing melt supply, these magmas more frequently replenish higher-level magma chambers beneath AVRs, thus reducing the diversity of compositions erupted subsequently.
|
Influence of the Iceland plume: melting vs source composition
An increase in incompatible element enrichment and crustal thickness, and decrease in depth associated with Iceland has been linked to the presence of a plume of hot mantle material rising beneath the centre of the Central Rift Zone (Morgan, 1971; Schilling, 1973; McKenzie, 1984; Smallwood & White, 1998). Although the seismic velocity anomaly defining the rising plume is thought to be only a few hundred kilometres in diameter (Wolfe et al., 1997; Bijwaard & Spakman, 1999; Foulger & Pearson, 2001), the geographical extent of anomalously thick oceanic crust, shallow bathymetry and elevated incompatible element abundances is considerably greater, extending 1000 km to the south of the plume’s centre. The petrogenetic effect of elevated mantle temperature beneath a spreading ridge is two-fold: it increases both the depth of initial melting and the average melt fraction (McKenzie, 1984). These have predictable geochemical effects that can be used to identify the extent of influence of the plume.
Because the partitioning of Si and Fe is affected by the depth of melting within the spinel lherzolite stability field, their concentrations can be used as proxies for the relative depth of the solidus (Klein & Langmuir, 1987). To account for the effects of low-pressure crystal fractionation, Klein & Langmuir (1987) used major element abundance regressed to a common concentration of MgO.
Our data show a relatively simple inverse relationship between Fe2O3 and MgO concentrations, from which there is no evidence for magnetite fractionation (Fig. 11). Therefore, we are able to determine values of both Fe2O3 and SiO2 at an equivalent MgO concentration, in accordance with the procedure described by Klein & Langmuir (1987). Both Si[8] and Fe[8] show reciprocal behaviour. Si[8] falls by an average of 0·8% northwards along the ridge whereas Fe[8] increases by an average of 1% (Fig. 13). This is evidence for an increase in the initial depth of mantle melting towards Iceland and hence an inferred rise in mantle temperature or fall in solidus temperature (Klein & Langmuir, 1987; Langmuir et al., 1992). It is also consistent with the general decrease in incompatible element concentrations (e.g. K2O, TiO2, P2O5 and the incompatible trace elements), reflecting greater total melt fraction, at least as far north as 61°N.
|
The regional variation in Fe[8] and Si[8] values is evidence for a broad positive temperature anomaly that decreases from the north, spans the entire length of the ridge and coincides with the regional-scale depth and crustal thickness variations. Significantly, there are no marked changes in values for Fe[8] or Si[8] north of 61°N, despite the ridge arguably showing the strongest compositional effect of the Iceland plume in this region. This suggests that the temperature anomaly associated with the plume is decoupled from its most enriched chemical signature (Fitton et al., 1997).
The ratios of Zr/Y, Nb/Zr, Sr/Rb and La/Yb are also sensitive to variations in the degree of fractional melting in the spinel lherzolite field, but relatively insensitive to low-pressure crystal fractionation. Modelling allows the regional trends in the observed trace element ratios to be compared with ratio variations predicted by simple Rayleigh crystal fractionation and fractional melting (Fig. 14), in an attempt to distinguish variations in melting.
|
The distributions of Zr/Y, Nb/Zr, Rb/Sr and La/Yb with latitude along the ridge axis are similar to one another and to the incompatible element concentrations (Fig. 4). All show a regional-scale decrease northwards from 57·5°N, towards a minimum at 60–61°N, followed by an increase towards maximum ratios at 63°N. South of 61°N, low ratios are located consistently at 58°N, 59·2°N and 60·2°N. These undulations are especially marked for Zr/Y and Rb/Sr. The sites that have anomalous incompatible element concentrations also have anomalous trace element ratios. For example, the Mega-seamount and site 78D have high ratios whereas site 17D has relatively low ones, at least for Zr/Y.
These local and regional variations in incompatible element ratios exceed those produced by fractional crystallization alone. For example, north from 60°N, Zr/Y ratios increase by 80% whereas to the south of 61°N the ratios vary by 20%. However, even 45% fractional crystallization can only account for a variation in Zr/Y of up to 1·5%. The variation in Rb/Sr is even more extreme, with a factor of 10 increase between 60·2° and 63°N and a factor of two variation forming the undulations south of 61°N.
The low trace element ratios at 58°N, 59·2°N and 60·2°N are independent of any systematic changes in isotopic ratios (Fig. 7). This is consistent with locally higher degrees of melting of a homogeneous mantle source. Modelling indicates a 10–20% variation in melt fraction (i.e. ranging from a nominal 10% to 12% total fusion; Fig. 14), possibly reflecting focused regions of upwelling mantle in the form of diapirs. Although this process is similar to that inferred as a cause of second-order segmentation at slow-spreading ridge elsewhere (Crane, 1985; Macdonald et al., 1988), on the Reykjanes Ridge its expression is more subtle, forming undulations in depth rather than tectonic offsets of the spreading axis (Murton & Parson, 1993).
Despite having compositions similar to those of samples located at 62°N, the Mega-seamount has relatively lower Fe[8], suggesting its source is shallower and relatively cooler than at 62°N. The tectonic setting of the Mega-seamount, within a non-transform offset basin, supports this interpretation. Here, the locally deep bathymetry indicates thin volcanic crust and a relatively low magma budget (Searle et al., 1994). The similarity in composition between the Mega-seamount and sites 500 km to the north is also evidence that the volcano is fed by a local compositional heterogeneity in the sub-ridge mantle. Although probably derived from the Iceland plume, it is unclear how this local source reached so far south without being depleted en route by melting beneath the ridge axis.
In summary, to the north of 60°N the composition of the ridge becomes increasingly diverse with higher incompatible element concentrations, higher 87Sr/86Sr, lower 143Nd/144Nd, higher Fe[8] and lower Si[8]. These changes in composition coincide with decreasing water depth and increasing crustal thickness. Together, these feature indicate increasing depth of melting and melt fraction from an increasingly enriched mantle source with distance towards Iceland. South of 60°N, the composition of the ridge is less diverse, having lower incompatible element concentrations, lower 87Sr/86Sr, higher 143Nd/144Nd, lower Fe[8] and higher Si[8]. These features indicate a southward decrease in depth of melting and melt fraction from an increasingly depleted mantle source. Superimposed on the southern part of the ridge are low-amplitude undulations in incompatible element concentrations and ratios that reflect subtle changes in the degree of melting. Although we can identify systematic changes in melt fraction, the main compositional variations are dominated by changes in mantle composition.
Multi-component mantle interaction
The enrichment in incompatible trace element concentrations and ratios, and concomitant increase in 87Sr/86Sr and decrease in 143Nd/144Nd isotopic ratios, to the north of 61°N have been interpreted as evidence for binary mixing between a trace-element-rich Iceland plume and an N-MORB source (Schilling, 1973; Schilling et al., 1983). However, we have already shown that the southern part of the ridge is distinct from the northern part in having, at similar ratios of 143Nd/144Nd, slightly elevated ratios of 87Sr/86Sr. Thus the Sr and Nd isotopic relationships along the ridge indicate more than simple binary mixing. Although there is good evidence for mixing along the ridge, the mixing process is less clear. Magma mingling by along-axis melt transport is unlikely, as neighbouring AVRs have different fractional crystallization histories and parental melt compositions indicating little communication between magma bodies. Therefore, mixing is more likely to involve different mantle sources and be mediated by differential partial melting.
Separation of arrays for the northern and southern suites, in space defined by incompatible element and isotopic ratios, demonstrates at least four compositional domains. The northern end of the ridge is characterized by high 87Sr/86Sr, low 143Nd/144Nd, high Zr/Y and high Nb/Zr. The region between 60°N and 61°N is characterized by relatively low 87Sr/86Sr, high 143Nd/144Nd, low Zr/Y and low Nb/Zr. A separate, trace-element-poor type is represented by site 17D, which is characterized by relatively high 87Sr/86Sr, moderate 143Nd/144Nd, low Zr/Y and low Nb/Zr. The fourth compositional domain, represented by the southern end of the ridge, is characterized by relatively moderate 87Sr/86Sr, high 143Nd/144Nd, moderate Zr/Y and low Nb/Zr.
Dispersion of the different regional suites of samples between these four compositional types, in trace element and isotopic space, is evidence for mixing along the ridge between at least three end-members. For example, in Fig. 9d, the northern suite plots on a mixing line between a high La/Yb, high Zr/Y component (E1) and the 60–61°N group. The southern suite lies on a mixing line between a low La/Yb, high Zr/Y component (D1) and a low La/Yb, low Zr/Y component (D2), with the 60–61°N group between. In Fig. 10a–c, the northern suite also lies on a mixing line between a high Nb/Zr, high La/Yb, high 87Sr/86Sr and low 143Nd/144Nd component (E1) and the 60–61°N group. The southern suite also lies on a mixing line between a low Nb/Zr, low La/Yb, low 87Sr/86Sr and high 143Nd/144Nd component (D1) and a low Nb/Zr, low La/Yb, high 87Sr/86Sr and intermediate 143Nd/144Nd component (D2), with the 60–61°N group between. However, in Fig. 10c the samples from south of 60°N plot on the D1 side of the 60–61°N group, whereas in the other diagrams they plot on the D2 side. This indicates that the D1 component comprises a range in 143Nd/144Nd, or even possibly a mixture of two separate components.
The data require at least three separate mantle components, of which D2 and the high 143Nd/144Nd version of D1 are dominant in the southern end of the Reykjanes Ridge. Furthermore, in Fig. 15 (i.e. in Nd vs Sr isotope space), the 60–61°N group overlaps the northern suite and plots on a mixing line between D1 and E1. Yet this contradicts the evidence presented in Figs 9 and 10, where the 60–61°N group plots on a mixing line between D1 and D2. This requires a fourth component (E2) that has low Nb/Zr and La/Yb (i.e. with trace element ratios similar to D1 and D2), but high 87Sr/86Sr and low 143Nd/144Nd (i.e. isotopically similar to E1). Although E2 may be an integral part of the original plume, it is also possible to derive it from E1 by partial melting. For example, simple modelling of incremental fractional melting [using bulk KD values and mantle modal mineralogy from McKenzie (1984)] demonstrates that a further 9% melting of the trace element rich source, E1 (i.e. after it has already produced a typical Iceland basalt, e.g. Eldgja), can yield a melt with E1 isotopic characteristics but that is trace element poor. A 1:1 mixture of E2 and D2, subsequently mixed with up to 70% D1, successfully recovers the field and slope of the 60–61°N group in 143Nd/144Nd, 87Sr/86Sr and trace element space (Fig. 15).
|
Comparison with other models
We have identified a number of components in the Iceland–Reykjanes Ridge system and can compare them with components identified from other studies. In common with our interpretation, Langmuir et al. (1978), Thirlwall (1995), Fitton et al. (1997), Taylor et al. (1997) and Kempton et al. (1998, 2000) also argued against simple binary mixing along the ridge. In general, they all required both enriched and depleted plume components supplying melts to the Reykjanes Ridge. The depleted components are distinct from the depleted Iceland plume as well as distinct from an N-MORB source. On the basis of 143Nd/144Nd and 87Sr/86Sr relationships, Taylor et al. (1997) suggested that the contribution from at least three end-members changes proportionally along the Reykjanes Ridge, from a mixture of enriched plume and depleted mantle north of 61°N, to a mixture of depleted plume and depleted mantle south of 61°N.
A comparison of the Reykjanes Ridge with alkalic Iceland and basaltic Iceland (represented by Snaefellsnes and Eldgja, respectively; Fig. 16) indicates the northern suite plots on a mixing trajectory with Eldgja, the common trace-element-rich basaltic component in Iceland (Hémond et al., 1993; Chauvel & Hémond, 2000). This suggests that our component E1 is similar to the trace-element-rich Iceland plume component identified previously (Schilling, 1973; Schilling et al., 1983; Fitton et al., 1997; Taylor et al., 1997; Kempton et al., 2000). Fitton et al. (1997) also showed that, whereas the depleted plume component on Iceland (termed ‘DP’) has a positive 
Nb signature, the ridge south of 61°N has a negative Nb signature similar to N-MORB. Therefore, although our depleted plume component, D2, beneath the southern Reykjanes Ridge appears similar to ‘DP’ in its 143Nd/144Nd and 87Sr/86Sr ratios and some trace element compositions (Figs 10 and 15), the two components cannot be the same. The ridge also has higher 
Hf for a given Nd compared with North Atlantic N-MORB (Kempton et al., 2000), as well as higher 3He/4He at least as far south as 58°N (Poreda et al., 1986; Hilton et al., 2000). Therefore, although our component D1 has similar 143Nd/144Nd and 87Sr/86Sr isotopic compositions to the depleted upper-mantle source for N-MORB, the two sources must be different.
|
In their interpretation, both Fitton et al. (1997) and Kempton et al. (2000) argued that the mantle beneath the southern Reykjanes Ridge is an ‘N-MORB-like’ source, entrained by and surrounding the Iceland plume as a ‘plume sheath’, that has had a long history of trace element depletion and isolation from the convecting upper mantle (i.e. producing its high Hf signature). This ‘plume sheath’ is also distinguished from the Iceland plume by having negative Nb. Their model interprets the high 3He/4He signature as infusion of primordial helium from the lower mantle into the ‘N-MORB-like’ source when it resided at the upper–lower mantle boundary, before being entrained by the Iceland plume.
Our trace-element-poor component, D1, falls spatially within the negative Nb, positive Hf ‘plume sheath’ identified by Fitton et al. (1997) and Kempton et al. (2000). However, the progressive decrease in 3He/4He (Poreda et al., 1987; Hilton et al., 2000), and increase in 143Nd/144Nd for a given87Sr/86Sr with distance south of 61°N (this study) demonstrates that this ‘plume sheath’ must be compositionally zoned. On this basis, we differentiate D1 into two components: a high 3He/4He, intermediate 143Nd/144Nd component (D1(He)) beneath the northern end of the southern ridge and a low 3He/4He, high 143Nd/144Nd component (D1(He)) beneath the southern end of the ridge.
Although D1(He) is distinguished from an N-MORB source by its positive Hf (Kempton et al., 2000), it remains possible that it is representative of plume-free upper mantle in the North Atlantic region of this study. However, this would require a significant difference in upper-mantle domains between this region south of Iceland and plume-free North Atlantic elsewhere.
Identification of new source components
Considering the available data, there appear to be six source components in the Iceland–Reykjanes Ridge system: (1) E1 (high 3He/4He, high 87Sr/86Sr, low 143Nd/144Nd, positive Nb, trace-element-rich Iceland plume); (2) E2 (a trace-element-poor version of E1); (3) DP (high 3He/4He, high 87Sr/86Sr, low 143Nd/144Nd, positive Nb, trace-element-poor Iceland plume); (4) D1(He) (high 3He/4He, low 87Sr/86Sr, intermediate 143Nd/144Nd, negative Nb, trace-element-poor ‘plume sheath’); (5) D2 (high 3He/4He, high 87Sr/86Sr, low 143Nd/144Nd, negative Nb, trace-element-poor Iceland plume); (6) D1(He) (low 3He/4He, low 87Sr/86Sr, high 143Nd/144Nd, negative Nb, trace-element-poor ‘plume sheath’).
These components are mixed systematically along the ridge with distance from Iceland. On Iceland, the lavas are characterized by a heterogeneous mixture of enriched and depleted components (Hémond et al., 1993; Chauvel & Hémond, 2000) of which E1 and DP are significant. Between 63°N and 61°N, the ridge comprises a mixture of E1 + D1(He). The region between 61°N and 60°N comprises a mixture of D1(He) + E2 ± D2. With increasing distance south of 60°N the mixture graduates towards D2 + D1(He).
|
SCHEMATIC MODEL FOR A ZONED AND DYNAMIC MANTLE |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSINTERPRETATION AND DISCUSSIONSCHEMATIC MODEL FOR A...CONCLUSIONSREFERENCES |
|---|
The variation in source components along the Reykjanes Ridge may reflect random changes in the composition of the plume and ‘sheath’ through time. Alternatively, it might reflect vertical and horizontal zoning in the plume and ‘plume sheath’ that is translated beneath the ridge by vertical upwelling and lateral transport.
We interpret the distribution and diversity of these components along the Reykjanes Ridge in terms similar to those described by Fitton et al. (1987) and later developed by Kempton et al. (2000). As the Iceland plume rises through the 670 km discontinuity (Fig. 17a), it entrains a sheath of material that has a long history of trace element depletion and isolation from the convecting upper mantle (i.e. resulting in a high Hf signature; Kempton et al., 2000). Infusion of primordial He into the base of this material creates a gradient in 3He/4He, from values more typical of the lower mantle to values more typical of the convecting upper mantle. This causes the sheath material to be vertically zoned between D1(He) near its base to D1(He) at its top. The rising plume is also zoned. Its core comprises a heterogeneous mixture of enriched and depleted components (including the trace-element-rich component, E1 and the positive Nb trace-element-poor component, DP) whereas an outer zone comprises a mixture of E1 and the negative Nb, trace-element-poor component, D2. Together, the plume and its entrained sheath form a rising column of heterogeneous mantle. As the column reaches the base of the lithosphere beneath Iceland, it inverts and spreads horizontally under the adjacent Reykjanes Ridge.
|
With increasing distance away from the centre of the plume, melting samples different depths and components within the zoned mantle beneath the ridge (Fig. 17b). Melting beneath the northern ridge involves predominantly the high 3He/4He plume component, D1(He), and the trace-element-rich component, E1. By the time the outflow reaches 60–61°N, the E1 component has become trace element poor (i.e. it is now E2) and further south it becomes completely consumed. South of 60°N, melting involves the high 3He/4He, trace-element-poor plume component, D2, and the low 3He/4He component, D1(He).
Survival of enriched components in the southward-migrating mantle beneath the ridge depends on their having a migration path that remains below the solidus. Those that are deepest will survive furthest south. This explains the presence of enriched compositions as far south as the Mega-seamount.
The dominance of the D1(He) component between 60°N and 61°N may be a result of locally higher temperatures causing an increase in the depth of the solidus and hence a greater proportion of D1(He) in the melt. This is supported by the relatively shallow depth of the ridge in this region (indicating higher melt flux). Alternatively, it may be a result of variations in the relative thickness of the original plume and plume sheath. The question of whether D1(He) is part of the ‘plume sheath’ or representative of the plume-free N-MORB source in the region south of Iceland remains to be resolved.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSINTERPRETATION AND DISCUSSIONSCHEMATIC MODEL FOR A...CONCLUSIONSREFERENCES |
|---|
The Reykjanes Ridge is tectonically continuous over a distance of
1100 km between Iceland and the Bight Fracture Zone. Volcanic accretion is focused in a series of en echelon axial volcanic ridges (AVRs). Detailed sampling from over 180 sites between 57·5°N and 63°N reveals a variety of mantle and crustal processes involved in the genesis of this plume-influenced ridge. The AVRs are fed by discrete magma bodies with little communication between one another. Crystal fractionation, which affects AVRs individually, ranges from 10 to 15% removal of olivine, plagioclase and clinopyroxene in the proportions ranging between 1:2:1 to 1:4:1. Regionally, the total extent of fractionation increases northwards along the ridge and is a linked to changes in crustal thickness, melt supply and magma residence time. For individual AVRs, the local extent of fractionation correlates with the abundance of seamounts, at least south of 60°N.
Melt fraction varies systematically on both a regional and local scale. An increase in Fe[8] and decrease in Si[8] northwards indicates a progressive increase in the initial depth of melting towards Iceland. This is compatible with a rise in mantle temperature, or fall in solidus temperature, associated with the Iceland plume. In addition, there are locally up to 20% variations in melting (i.e. changes in total melt fraction from 10% to 12%). Greater degrees of melting supply elevated regions of the ridge spaced 100 km apart and located approximately at 58°N, 59°N and 60°N. These may indicate regions of diapiric upwelling mantle beneath the ridge, analogues to three-dimensional upwelling inferred for other slow-spreading ridges. Between 60 and 61°N, the spreading axis intersects a southward-closing V-shaped ridge. This is associated with locally thicker crust, indicating that pulses of enhanced melt supply migrate southwards along the ridge axis.
The mantle involved in melting beneath Iceland and the Reykjanes Ridge comprises at least six components, all distinct from an average plume-free, North Atlantic N-MORB source. Some components are specific to Iceland whereas others are found only on the ridge axis. Together, these components form a southward-flowing and -thinning mantle wedge beneath the spreading axis. The geographical distribution of these components reflects their origin within the deep-upper and lower mantle and subsequent translation by plume outflow and mixing along the entire length of the ridge. Changes in the temperature of the outflowing mantle result in variations in the relative proportions of these components contributing to magmas and to the melt flux. Migration of hot pulses of outflowing mantle leaves a V-shaped trail of thicker crust symmetrically about the ridge axis.
|
|
|
ACKNOWLEDGEMENTS |
|---|
We are indebted to Bob Nesbitt and Mary Gee for helpful discussions about the ideas presented here. We also thank Andy Saunders, Godfrey Fitton, Catherine Chauval and Folkmar Hauff for their reviews and constructive comments, which significantly improved earlier drafts of this contribution. The sampling work involved all members of the shipboard party of cruise CD80. We are particularly grateful to the master, Mr Richard Bourne, officers and crew of the R.R.S. Charles Darwin for their seamanship and persistence with the sampling programme in the face of often hostile sea conditions. This work was funded by the Natural Environment Research Council, grant GR3/8757A.
|
FOOTNOTES |
|---|
*Corresponding author. Telephone: +44 (0)23 8059 6666. E-mail: bjm{at}soc.soton.ac.uk
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSINTERPRETATION AND DISCUSSIONSCHEMATIC MODEL FOR A...CONCLUSIONSREFERENCES |
|---|
Albers, M. & Christensen, U. R. (2001). Channeling of plume flow beneath mid-ocean ridges. Earth and Planetary Science Letters 187, 207–220.[Web of Science]
Anders, E. & Grevesse, N. (1989). Abundances of the elements: meteoric and solar. Geochimica et Cosmochimica Acta 53, 197–214.[Web of Science]
Applegate, B. & Shor, A. N. (1994). The northern Mid-Atlantic and Reykjanes Ridges: spreading centre morphology between 55°50'N and 63°N. Journal of Geophysical Research 99, 17935–17952.
Bell, R. E. & Buck, W. R. (1992). Crustal control of ridge segmentation inferred from observation of the Reykjanes Ridge. Nature 357, 583–586.
Bijwaard, H. & Spakman, W. (1999). Tomographic evidence for a narrow whole mantle plume below Iceland. Earth and Planetary Science Letters 166, 121–126.[Web of Science]
Bindeman, I. N., Davis, A. M. & Drake, M. J. (1998). Ion microprobe study of plagioclase–basalt partition experiments at natural concentration levels of trace elements. Geochimica et Cosmochimica Acta 62, 1175–1193.[Web of Science]
Bougault, H., Dimitiriev, L., Schilling, J.-G., Sobolev, A. & Joron, J.-L. (1988). Mantle heterogeneity from trace elements: Mid-Atlantic Ridge triple junction near 14°N. Earth and Planetary Science Letters 88, 27–36.[Web of Science]
Chauvel, C. & Hémond, C. (2000). Melting of a complete section of recycled oceanic crust: trace element and Pb isotopic evidence from Iceland. G3 1999GC000002.
Crane, K. (1985). The spacing of rift axis highs: dependence upon diapiric processes in the underlying asthenosphere? Earth and Planetary Science Letters 72, 405–414.[Web of Science]
DeMets, C., Gordan, D. E., Argus, D. F. & Stein, S. (1994). Effects of recent revisions to the geomagnetic reversal timescale on estimates of current plate motions. Geophysical Research Letters 21, 2191–2194.[Web of Science]
Dosso, L., Hanan, B. B., Bougault, H., Schilling, J.-G. & Joron, J.-L. (1991). Sr–Nd–Pb geochemical morphology between 10° and 17°N on the Mid Atlantic Ridge: a new MORB isotope signature. Earth and Planetary Science Letters 106, 29–43.[Web of Science]
Elliot, T. R., Hawkesworth, C. J. & Gronvold, K. (1991). Dynamic melting of the Iceland Plume. Nature, 351, 201–206.
Fitton, J. G., Saunders, A. D., Norry, M. J., Hardarson, B. S. & Taylor, R. N. (1997). Thermal and chemical structure of the Iceland plume. Earth and Planetary Science Letters 153, 197–208.[Web of Science]
Foulger, G. R. & Pearson, D. G. (2001). Is Iceland underlain by a plume in the lower mantle? Seismology and helium isotopes. Geophysical Journal International 145, 1–5.[Web of Science]
Furman, T., Frey, F. & Park, K.-H. (1995). The scale and source of heterogeneity beneath the Eastern neovolcanic zone, Iceland. Journal of the Geological Society, London 152, 997–1002.
Gee, M. A. M., Taylor, R. N., Thirlwall, M. F. & Murton, B. J. (1998a). Glacioisostacy controls chemical and iostopic characteristics of tholeiites from the Reykjanes Peninsula, SW Iceland. Earth and Planetary Science Letters 164, 1–5.[Web of Science]
Gee, M. A. M., Thirlwall, M. F., Taylor, R. N., Lowry, D. & Murton, B. J. (1998b). Crustal processes: major controls on Reykjanes Peninsula lava chemistry, SW Iceland. Journal of Petrology 39, 819–839.[Web of Science]
Gee, M. A. M., Taylor, R. N., Thirlwall, M. F. & Murton, B. J. (2000). Axial magma reservoirs located by variation in lava chemistry along Iceland’s mid-ocean ridge. Geology 28, 699–702.
Hards, V. L., Kempton, P. D. & Thompson, R. N. (1995). The heterogeneous Iceland plume: new insights from the alkaline basalts of the Snaefell volcanic centre. Journal of the Geological Society, London 152, 1003–1009.
Hémond, C., Arndt, N. T., Lichtenstein, U., Hofmann, A. W., Oskarsson, N. & Steinthorsson, S. (1993). The heterogeneous Iceland plume—Nd–Sr–O isotopes and trace-element constraints. Journal of Geophysical Research 98, 15833–15850.
Hilton, D. R., Thirlwall, M. F., Taylor, R. N., Murton, B. J. & Nichols, A. (2000). Controls on magmatic degassing along the Reykjanes Ridge with implications for the helium paradox. Earth and Planetary Science Letters 183, 43–50.[Web of Science]
Jacoby, W. R. (1980). Morphology of the Reykjanes Ridge crest near 62°N. Journal of Geophysical Research 47, 81–85.
Jochum, K. P., Seufert, H. M. & Thirlwall, M. F. (1990). High sensitivity Nb analysis by spark-source mass spectrometry (SSMS) and calibration of XRF Nb and Zr. Chemical Geology 81, 1–16.[Web of Science]
Kempton, P. D., Nowell, G. M., Fitton, J. G., Saunders, A. D. & Taylor, R. N. (1998). The Iceland plume in space and time: a Sr–Nd–Pb–Hf study of the North Atlantic rifted margin. Mineralogical Magazine 62A, 759–760.[Abstract]
Kempton, P. D., Fitton, J. G., Saunders, A. D., Nowell, G. M., Taylor, R. N., Hardarson, B. S. & Pearson, G. (2000). The Iceland plume in space and time: a Sr–Nd–Pb–Hf study of the North Atlantic rifted margin. Earth and Planetary Science Letters 177, 255–271.[Web of Science]
Klein, E. M. & Langmuir, C. H. (1987). Global correlations of oceanic ridge basalt chemistry with axial depth and crustal thickness. Journal of Geophysical Research 92, 8098–8115.
Langmuir, C. H., Vocke, R. D., Jr & Hanson, G. N. (1978). A general mixing equation with application to Icelandic basalts. Earth and Planetary Science Letters 37, 380–392.[Web of Science]
Langmuir, C. H., Klein, E. M. & Plank, T. (1992). Petrological constraints of mid-ocean ridge basalts: constraints on melt generation beneath ocean ridges. In: Morgan, J. P., Blackman, D. K. & Sinton, J. M. (eds) Mantle Flow and Melt Generation at Mid-ocean Ridges. Geophysical Monograph, American Geophysical Union 71, 183–280.
Laughton, A. S., Searle, R. C. & Roberts, D. G. (1979). The Reykjanes Ridge crest and the transition between its rifted and non-rifted regions. Tectonophysics 55, 173–177.[Web of Science]
Macdonald, K. C., Fox, P. J., Parram, L. J., Eisen, M. F., Hasman, R. M., Miller, S. P., Corbotte, S. M., Cormier, M.-H. & Shor, A. N. (1988). A new view of the mid ocean ridge from the behaviour of ridge-axis discontinuities. Nature 355, 217–222.
Madge, L. S. & Smith, D. K. (1994). Seamount volcanism along the Reykjanes Ridge: relationship to the Iceland hot spot. Journal of Geophysical Research 100, 8449–8468.
McKenzie, D. (1984). The generation and compaction of partially molten rock. Journal of Petrology 25, 713–765.
McKenzie, D. & O’Nions, R. K. (1991). Partial melt distributions from inversion of rare earth element concentrations. Journal of Petrology 32, 1021–1091.
Morgan, W. J. (1971). Convection plumes in the lower mantle. Nature 230, 42–43.
Murton, B. J. (1994). RRS Charles Darwin Cruise 80, 01 September to 01 October 1993. The PETROS Programme: geological sampling and bathymetric surveying of the Reykjanes Ridge between 57°N and 63°N, southwest of Iceland. Institute of Oceanographic Sciences, Deacon Laboratory Cruise Report 236, 32 pp.
Murton, B. J. & Parson, L. M. (1993). Segmentation, volcanism and deformation of oblique spreading centres: a quantitative study of the Reykjanes Ridge. Tectonophysics 222, 237–257.[Web of Science]
Nichols, A. (2000). Is Iceland a wet spot? Ph.D. thesis, Bristol University, 215 pp.
Parson, L. M., Murton, B. J., Searle, R. C., Booth, D., Evans, J., Field, P., Keeton, J., Laughton, A., McAllister, E., Millard, N., Redbourne, L., Rouse, I., Shor, A., Smith, D., Spencer, S., Summerhayes, C. & Walker, C. (1993). En echelon axial volcanic ridges at the Reykjanes Ridge: a life cycle of volcanism and tectonics. Earth and Planetary Science Letters 117, 73–87.[Web of Science]
Poreda, R., Schilling, J.-G. & Craig, H. (1986). Helium and hydrogen isotopes in ocean ridge basalts north and south of Iceland. Earth and Planetary Science Letters 78, 1–17.[Web of Science]
Sandwell, D. T. & Smith, W. H. F. (1992). Global marine gravity from ERS-1, Geosat and Seasat reveals new tectonic fabric. EOS Transactions, Americal Geophysical Union 73, 133.
Schilling, J.-G. (1973). The Icelandic Plume, geochemical evidence along the Reykjanes Ridge. Nature 242, 565–571.
Schilling, J.-G., Zajak, M., Evans, R., Johnson, T., White, W., Devine, J. D. & Kingsly, R. (1983). Petrologic and geochemical variations along the Mid-Atlantic Ridge from 29°N to 73°N. American Journal of Science 283, 510–586.
Schilling, J.-G., Hannan, B. B., McCully, B., Kingsly, R. H. & Fontignie, D. (1994). Influences of the Sierra Leone mantle plume on the equatorial Mid-Atlantic Ridge: a Nd–Sr–Pb isotopic study. Journal of Geophysical Research 99, 12005–12028.
Schouten, H. & Whitehead, J. (1992). Modeling ridge segmentation, a possible mechanism. Oceanus 34, 19–20.
Schouten, H., Klitgord, K. D. & Whitehead, J. (1985). Segmentation of mid-ocean ridges. Nature 317, 225–229.
Searle, R. C. & Laughton, A. S. (1981). Fine-scale sonar study of tectonics and volcanism on the Reykjanes Ridge. Oceanologica Acta 4, 5–13.
Searle, R. C., Field, P. R. & Owens, R. B. (1994). Segmentation and nontransform ridge offsets on the Reykjanes Ridge near 58°N. Journal of Geophysical Research 99, 24159–24172.
Searle, R. C., Keeton, J. A., Owens, R. B., White, R. S., Mecklenburgh, R., Parsons, B. & Lee, S. M. (1998). The Reykjanes Ridge: structure and tectonic evolution of a hot-spot-influenced, slow-spreading ridge, from multibeam, gravity and magnetic investigations. Earth and Planetary Science Letters 160, 463–478.[Web of Science]
Sinha, M. C., Constable, S. C., Pierce, C., White, A., Heinson, G., MacGregor, L. M. & Navin, D. A. (1998). Magmatic processes at slow spreading ridges: implications of the RAMASSES experiment at 57°45'N on the Mid-Atlantic Ridge. Geophysics Journal International 135, 731–745.
Sleep, N. H. (1996). Lateral flow of hot plume material ponded at sublithospheric depths. Journal of Geophysical Research 101, 28065–28083.[Web of Science]
Smallwood, J. R. & White, R. S. (1998). Crustal accretion at the Reykjanes Ridge, 61°N–62°N. Journal of Geophysical Research 103, 5185–5201.
Sun, S.-S. & Jahn, B.-M. (1975). Lead and strontium isotopes in post-glacial basalts from Iceland. Nature 255, 527–530.
Sun, S.-S. & McDonough, W. F. (1989). Chemical and isotopic systematics of ocean basalts. In: Saunders, A. D. & Norry, M. J. (eds) Magmatism in the Ocean Basins. Geological Society, London, Special Publications 42, 313–347.
Talwani, M., Windisch, W. W. & Langseth, M. G. (1971). The Reykjanes Ridge crest: a detailed geophysical study. Journal of Geophysical Research 76, 473–577.
Taylor, R. N., Murton, B. J. & Thirlwall, M. F. (1995). Petrographic and geochemical variation along the Reykjanes Ridge, 57°–59°N. Journal of the Geological Society, London 152, 1031–1037.
Taylor, R. N., Thirlwall, M. F., Murton, B. J., Hilton, D. R. & Gee, M. A. M. (1997). Isotopic constraints on the influence of the Icelandic Plume. Earth and Planetary Science Letters 148, 1–8.[Web of Science]
Thirlwall, M. F. (1991a). High precision multicollector isotope analysis of low levels of Nd as oxide. Chemical Geology 94, 13–22.
Thirlwall, M. F. (1991b). Long-term reproducibility of multicollector Sr and Nd isotope ratio analysis. Chemical Geology 94, 85–104.[Web of Science]
Thirlwall, M. F. (1995). Generation of the Pb isotopic characteristics of the Iceland plume. Journal of the Geological Society, London 152, 991–996.
Thirlwall, M. F., Upton, B. G. J. & Jenkins, C. (1994). Interaction between continental lithosphere and the Iceland plume—Sr–Nd–Pb isotope geochemistry of Tertiary basalts, NE Greenland. Journal of Petrology 35, 839–879.
Villemant, B., Jaffrezic, H., Jordon, J. L. & Treuil, M. (1981). Distribution coefficients of major and trace-elements—fractional crystallization in the alkali basalt series of Chaîne-des-Puys (Massif Central, France). Geochimica et Cosmochimica Acta 45, 1997–2016.[Web of Science]
Vogt, P. R. (1971). Asthenospheric motion recorded by the ocean floor south of Iceland. Earth and Planetary Science Letters 13, 153–164.[Web of Science]
Vogt, P. R. (1974). The Icelandic Phenomenon: imprints of hot-spot on the ocean crust, implications for flow beneath plates. In: Kristjonsson, L. (ed.) Geodynamics of Iceland and the North Atlantic Area. Dordrecht: Reidel, pp. 105–126.
Vogt, P. R. & Avery, O. E. (1974). Detailed magnetic surveys in the northeast Atlantic and Labrador Sea. Journal of Geophysical Research 79, 342–363.
White, R. S. (1997). Rift–plume interaction in the North Atlantic. Philosophical Transactions of the Royal Society of London 355, 319–339.
White, W. M. & Hofmann, A. (1982). Mantle heterogeneity in isotopes in oceanic basalts. Nature 295, 363–364.
White, R. S., Brown, J. W. & Smallwood, J. R. (1995). The temperature of the Iceland plume and origin of outward-propagating V-shaped ridges. Journal of the Geological Society, London 152, 1039–1045.
Wolfe, C. J., Gjarnason, I. T., VanDecar, J. C. & Solomon, S. C. (1997). Seismic structure of the Iceland mantle plume. Nature 385, 245–247.
Yale, M. M. & Morgan, J. P. (1998). Asthenosphere flow model of hotspot–ridge interactions: a comparison of Iceland and Kerguelen. Earth and Planetary Science Letters 161, 45–56.[Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  T. F. KOKFELT, K. HOERNLE, F. HAUFF, J. FIEBIG, R. WERNER, and D. GARBE-SCHONBERG Combined Trace Element and Pb-Nd-Sr-O Isotope Evidence for Recycled Oceanic Crust (Upper and Lower) in the Iceland Mantle Plume J. Petrology, September 1, 2006; 47(9): 1705 - 1749. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. A. PEARCE Mantle Preconditioning by Melt Extraction during Flow: Theory and Petrogenetic Implications J. Petrology, May 1, 2005; 46(5): 973 - 997. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. HERZBERG Partial Crystallization of Mid-Ocean Ridge Basalts in the Crust and Mantle J. Petrology, December 1, 2004; 45(12): 2389 - 2405. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||