DETAILS OF DAY 2

The aim of day two is to examine the various elements of the upper oceanic crustal sequence of the Troodos ophiolite. This should combine with day one to give a reasonably complete picture of the nature of the oceanic crust as displayed by ophiolites. This will be the main opportunity on the trip to study the upper parts of the oceanic crust intrusives, and to examine the nature of the volcanic aspects of oceanic crustal accretion. The Akaki River canyon will be visited, a spectacular and classic locality giving probably the best view of the extrusive sequence in the whole Troodos. There will also be an opportunity to examine some tectonic features related to the on-going convergence between Africa and Eurasia displayed superbly in the frontal aspect of the Cyprus Arc, and developed from Mid to Late Miocene times.

The day will be spent in the Troodos mountains along the ridges of Pitsylia, east of Mt. Olympus. Although being in the high Troodos, we will be at a lower altitude than on day one, and today will be below 1200m.

We will be looking at the nature and disposition of high level magma chambers within the crust, and seeing how they link to intrusive and extrusive igneous products. In so doing, we will be looking at the role of these magma chambers in the accretion of neovolcanic crust. The layered gabbro bodies that we saw yesterday are the lower parts of crustal magma chambers. Above the layered cumulates, unlayered gabbros and norites are extensively developed. In many parts of the ophiolite the gabbro bodies show feeders to the upper parts of the crustal sequence, establishing these bodies as magma reservoirs feeding higher level extrusives and minor intrusives. Gabbro bodies are also notably crosscutting in their relationship with adjacent plutons, which suggest multiple high level magma chambers under the ridge crest. The axial nature of many of these high level plutons can be inferred from their general lack of chilling relationships with neighbouring bodies and with the overlying minor intrusives. Also, stoped contacts and xenolith inclusions are observed for some of the bodies. This indicates very little temperature difference between intrusion and host. Finally, spreading deformation fabrics are noticed in the high level plutons, while cross-cutting and thus somewhat later bodies are sometimes undeformed.

Bodies of plagiogranite also occur as high level intrusives. These too can cross-cut gabbros, and are sometimes themselves cross-cut by other gabbros or plagiogranites. These plagioclase-quartz-hornblende rocks represent the final stages of fractionation of mafic magmas. Their occurrence adjacent to and cross-cut by gabbros again indicates that magmatic activity was via numerous independent magma chambers within which replenishment or fractionation was taking place independently, as shown in figure 4.

In some sections of the crustal sequence, gabbro bodies show feeders to the sheeted dyke complex above. In other parts, dykes cross-cut gabbro bodies, and gabbros intrude the dyke complex, reinforcing the idea of multiple magma chambers with multiple intrusions of dykes occurring within the area under the ridge crest. Within the SDC itself, dykes are seen intruded by other, later dykes, again suggesting that crustal accretion took place via multiple phases of intrusion. The geochemical evidence supports these ideas in identifying 3 relatively distinct magma compositions from which dykes, lavas, and in cases plutonics were derived. None of these compositions can be related to each other via established evolution or differentiation trends, so the conclusion drawn is one of independent, localised magma sources. Where plumbing allowed pooling of these products from local source regions in mid-crustal magma chambers prior to eruption, geochemical distinctions would be lost, as only composite magmas would be extruded (Coogan et al., 2003).

The pillow lavas form the upper part of the crustal sequence. In places the pillow lavas can be divided into two groups, upper and lower, on the basis of field relationships, geochemistry, and alteration. However, these divisions are not readily applicable over the whole of the ophiolite complex, nor where they are applicable do the distinguishing criteria always coincide. What is noticeable however are structural domains where both dyke dips and pillow lava orientations are consistent, and the dykes all dip away from definable axes. Sets of faults are also associated with the dykes and lavas that too have consistent direction, and dip towards these defined axes. This has allowed the identification of ridge crest axial grabens, and three main spreading axes to be defined for the ophiolite, shown in figure 5 ( Varga and Moores, 1985, 1990; Moores et al., 1990). Spreading is believed to have progressed eastwards (present day orientation) by ridge jumping. These axial grabens define episodes of nonmagmatic spreading, where the movement was accommodated by faulting and graben development. An intermittent magma supply is thus indicated.

Overall the lavas can be divided into two broad compositional groups, irrespective of magma ponding and homogenisation variations. A series of primitive depleted boninitic melts are identified, interspersed with lavas of more tholeiitic composition. The boninites are low Ti while the arc-tholeiites are higher Ti. Other fore-arc settings also show this intercalation of boninites and tholeiites, for example the IBM system (Crawford et al., 1981), and in the Oman ophiolite (Arai et al., 2006).

Upwards from the base of the SDC hydrothermal circulation has caused significant alteration to the mineral composition of dykes and lavas, especially close to hydrothermal vents. Generally, sea water circulation through the lavas has resulted in low temperature zeolite facies metamorphism, with carbonates, clay minerals, and zeolites produced. The SDC, being deeper in the igneous succession, has undergone higher temperature metamorphism to greenschist facies, with actinolite and plagioclase developed, chlorite and quartz less abundant, and rare epidote. In other areas, particularly those close to the massive sulphide deposits, presumed higher temperature hydrothermal alteration has converted the dykes to a quartz/epidote composition. These epidosites are sometimes associated with significant brecciation and mineralization, which can be related to hydrothermal circulation pathways, and are believed to have occurred early in the spreading history before major dyke-tilting and faulting took place (Varga et al., 1999).