GEOCHEMISTRY OF HEARD-ISLAND (SOUTHERN INDIAN-OCEAN) - CHARACTERIZATION OF AN ENRICHED MANTLE COMPONENT AND IMPLICATIONS FOR ENRICHMENT OF THE SUB-INDIAN OCEAN MANTLE
GEOCHEMISTRY OF HEARD-ISLAND (SOUTHERN INDIAN-OCEAN) - CHARACTERIZATION OF AN ENRICHED MANTLE COMPONENT AND IMPLICATIONS FOR ENRICHMENT OF THE SUB-INDIAN OCEAN MANTLE
复制标题
DOI：
10.1093/petrology/35.4.1017
复制
发表时间：
1994-08-01
影响因子：
3.9
通讯作者：
NICHOLLS, IA
中科院分区：
文献类型：
--
作者：
BARLING, J;GOLDSTEIN, SL;NICHOLLS, IA
位于南印度洋克尔盖伦高原的赫德岛的熔岩显示出最大的同位素组成范围（例如，Sr-87/Sr-86 = 0.7047-0.7079）。同位素组成具有良好的相关性，并伴随着不相容微量元素比例的系统变化，特别是涉及Nb的比例。这些变化可以用不相容的微量元素比例来解释，特别是那些涉及Nb的比例。这些变化被解释为两种成分混合的结果。一类具有Sr-87/Sr-86高、Pb-206/Pb-204低、Nd-143/Nd-144低、Nb和Eu负异常等特征，最终来源于上陆地壳；另一种具有较低的Sr-87/Sr-86，较高的Pb-206/Pb-204和Nd-143/N-144，缺乏对Nb和Eu的消耗。考虑了源的低sr -87/Sr-86成分的两种可能组成。第一个是在希尔德岛数据阵列的低sr -87/Sr-86端，最接近于劳伦斯半岛的熔岩。然而，微量元素的变化表明，这些熔岩可能不是希尔德羽流的代表。第二个接近主火山熔岩同位素阵列的低sr -87/Sr-86端。在这种情况下，岩石圈地幔被认为是劳伦斯半岛熔岩的起源。同位素数据、主要元素组成和不相容的微量元素比例之间的关系表明，大陆源物质可能存在于地幔源中，除了赫德岛样品外，它对地幔源的最大贡献都小于4 wt.%。然而，如果克格伦高原是一个淹没的大陆块体，就不能排除浅层污染的可能性。将用于解释赫德岛地球化学变化的二元混合模型扩展到其他印度洋岛屿和洋中脊玄武岩（OIB和MORB）。我们发现印度洋OIB的同位素组成与一个区域储层的样品一致，在该储层中，相同的两种组分以不同的比例存在（通常为1-5 wt.%的大陆源组分）。印度洋MORB独特的同位素组成与类似大西洋或太平洋的MORB地幔源的混合成分相一致。这些“富集”的印度洋地幔组分的相对非放射性成因的Pb-206/Pb-204同位素组成与当今任何海洋沉积物不同，表明它们的来源的U-238/Pb-204比率（mu）远低于bbbb1 Ga的典型上大陆地壳。这些年龄早于冈瓦纳大陆的形成（大约600- 130 Ma），因此不支持冈瓦纳大陆下的沉积物俯冲是次印度洋地幔富集的原因。我们认为印度洋OIB源的富集是由于大约1-2 Ga的冈瓦纳元古代前体下地壳上物质的俯冲作用。印度洋MORB源的富集可能有类似的起源，也可能是在Gondwana分裂（200-130 Ma）期间，由岩石圈地幔返回软流圈地幔的次大陆岩石圈地幔引起的。
Lavas from Heard Island, located on the Kerguelen Plateau in the southern Indian Ocean, exhibit the largest range (e.g., Sr-87/Sr-86 = 0.7047-0.7079) of isotopic compositions yet observed on a single oceanic island. Isotopic compositions are well correlated and are accompanied by systematic changes in incompatible trace element ratios, particularly those involving Nb. These variations are interpreted in incompatible trace element ratios, particularly those involving Nb. These variations are interpreted as resulting from mixing between two components. One is characterized by high Sr-87/Sr-86, low Pb-206/Pb-204 and Nd-143/Nd-144 ratios, and negative Nb and Eu anomalies, and is derived ultimately from the upper continental crust. The other has lower Sr-87/Sr-86, and higher Pb-206/Pb-204 and Nd-143/N-144 ratios, and lacks the depletions in Nb and Eu. Two possible compositions are considered for the low-Sr-87/Sr-86 component of the source. The first is at the low-Sr-87/Sr-86 end of the Heard Island data array, represented most closely by lavas from the Laurens Peninsula. However, trace element variations suggest that these lavas might not be representive of the Heard plume. The second is close to the low-Sr-87/Sr-86 end of the isotopic array for lavas from the main volcano. In this case a lithospheric mantle origin is suggested for the Laurens Peninsula lavas. The relationships between isotopic data, major element compositions, and incompatible trace element ratios indicate that the continent-derived material is probably present in the mantle source, where it makes a maximum contribution of < 4 wt.% for all but one Heard Island sample. However, if the Kerguelen Plateau is a submerged continental block, shallow-level contamination cannot be ruled out.The binary mixing model developed to explain the Heard Island geochemical variations is extended to include other Indian Ocean oceanic island and mid-ocean ridge basalts (OIB and MORB). We show that isotopic compositions of Indian Ocean OIB are consistent with sampling of a regional reservior in which the same two components exist in variable proportions (generally 1-5 wt.% of the continent-derived component). The distinctive isotopic compositions of Indian Ocean MORB are consistent with mixing of a similar component into an Atlantic- or Pacific-like MORB mantle source. The relatively unradiogenic Pb-206/Pb-204 isotopic compositions of these 'enriched' Indian Ocean mantle components are unlike any present-day marine sediments and indicate that their source has had U-238/Pb-204 ratios (mu) much lower than typical upper continental crust for > 1 Ga. These ages pre-date the formation of Gondwana (approximately 600-approximately 130 Ma) and therefore do not support sediment subduction beneath Gondwana as the cause of enrichment in the sub-Indian Ocean mantle. We propose that the enrichment of Indian Ocean OIB sources was due to subduction of upper-crustal material beneath a Proterozoic precursor of Gondwana at approximately 1-2 Ga. The enrichment of the Indian Ocean MORB sources could have had a similar origin, or could have been derived from sub-continental lithospheric mantle returned to the asthenospheric mantle, perhaps during the break-up of Gondwana (200-130 Ma).