GEOCHEMISTRY AND AGE RELATIONSHIPS OF METAMORPHOSED MAFIC SILLS FROM SIERRA DE ENMEDIO AND SIERRA DE CARRASCOY (EASTERN BETIC ZONE, SOUTHEASTERN SPAIN)

The presence of fairly abundant shallow-intrusive mafic bodies is a common feature of the almagride units, a recently defined group of tectonic elements of the Eastern Betic Zone whose paleogeograpbic and teclOnic interpretation is particuIarly controversia\. In this paper we focuse on the bulk geochemistry (including KIAr data) and possible age relationships of tbese rocks, and discuss them in view of tbeir significance regarding possible tectonic framework of emplacement and subsequent crustal evolution. The analyses (49-55% SiO,) point to a derivation from tholeiitic magmas that interacted with continental crust, as indicated by enrichment in the less compatible elements, such as Ba, Rb, Th and K, whose amounts are similar lO those of well known continental tholeiitic provinces. Thus, the magmatic event is inferred to have been generated in a tectonic environment comparable to that of continental rifts. Emplacement may have taken place in Upper Triassic to Middle Jurassic times, as bracketed by their intrusion within Triassic beds and one whole-rock KIAr date of 178±4 Ma from a less altered sample. Much younger analytical KIAr ages, scattering between 41±5 and 57±2 Ma, are obtained from common samples, however, reflecting an Eocene or younger metamorphic event that caused partial recrystallisation under low-grade greenchist, to actinolite-pumpelIyite facies conditions. As compared to other mafic complexes in the Betics, the observed chemical evidence for crustal contamination makes the almagride metabasites more similar to those in the external rona (e.g. the so-calIed «opbites») than in tbe Nevado-Filábride Ensemble, tbus being consistent with proposals that correlate these units with the Subbetic in the Murcia area.


Introduction
The presence of fairly abundant metabasite bodies is a conspicuous feature of a number of tectonic units of the Eastern Betic Zone forming the massives of Sierra de Almagro, Sierra de Enmedio and Sierra de Carrascoy.The c1assification oC these elements and their paleogeographic interpretation within c1assical schemes oC the Betic Cordilleras have had a particularly controversial development.Until recentIy, these units were thought to form part oC the so-called Ballabona-Cucharon Complex (Ege1er and Simón, 1969), which would include units with seemingly intermediate lithostratisgraphic and tectono-metamorphic development as compared with neighbouring Nevado-Filabride or Alpujarride sequences (see also Simon 1963;Simon el aL, 1976;Kozur el aL, 1974).
However, the distinction of the Ballabona-Cucharón Complex as a major ensemble within the Betic Zone (in addition to the three c1assical Nevado-Filabride, Alpujarride and Malaguide Ensembles), as proposed by Egeler and Simon (1969), has been.a matter of disagreement.Aldaya el al. (1979), fm lDs.tan~, suggested instead its adscription to the AlpuJarnde Ensemble (i.e., within their Lújar Group oC alpujárride nappes).Still more recentIy, Simon and Visscher (1983) dropped the mere notion oC the Ballabona-Cucharón Complex, and regrouped its e1ements into a newly proposed «Almagride Complex» and the Alpujarride Ensemble.In doing so, they also emphasized that the new group oC «almagride» units (always tectonically underlying the alpujarrides but with.an uncertain structural relationship to the Nevado-FIlábrides) bear marked lithostratigraphic affinities with Triassic sequences belonging to the Subbetic ~f the province oC Murcia (see also Besems andSimon, 1982 andKozur el al., 1985).In consequence, two largely contrasting hypotheses 00 the paleogreogra~hic provenance oC these units (i.e. the part oC the previous Ballabona-Cucharón group not reassigned to The Alpujarride Ensemble) would now coexist, involving implicit interpretations that regard them either as a southern prolongation of the Subbetic (external) realm, or as a part of the northern margin of the Alpujarride crustal domain.
With regard to these problerns, a study of the mafic bodies present in these units was deemed promising, to compare with other well-hnown occurrences of mafic rocks within the Betic Cordilleras, i.e., in the Subbetic and in the Nevado-Filabride Ensemble.The purpose of the present paper is to discuss the tectonic framework of emplacement of these mafic intrusives, as inferred from their major and trace-element geochemistry, as well as age relationships based on field evidence and four whole-rock K/Ar determinations.

Field Setting
Figure 1 shows the location of the two selected study areas in Sierra de Enmedio and Sie~ra de Carrascoy, respectively.The sampled mafic bodies are intercalated in the Sierra de Enmedio Uoit (Espinosa el aL, 1974;Kozur el aL, 1974) of the Sierra de Enmedio (fig.2), and within the Romero and Carrascoy Units (Kampschuur, 1972) oC the Sierra de Carrascoy (fig.3).
The field appearance of the mafic bodies in both areas is analogous.Most oC these do probably represent the remains of discrete sill-like intrusions, with thicknesses that range from metres up to more than 100 metres.Their location in the respective stratigraphic columns (composed basically of Permian-(?)(1).Triassic suites with a predominantly quartzopelitic lower formation and a predominantIy carbonatic upper formation' see reCerences aboye) is variable, although commonl~coincides with the occurrence oC vertical lithologic discontinuities such as transitions from carbonatic to pelitic members.In fact, only one body was found to be emplaced within a re1atively homogeneous sequence oC presumable Early Triassi~quar~ zites and phyllitic schists oC the S. de EnmedlO Umt (fig.2).
In all instances, the Permian (?)-Triassic sequences enc10sing the mafic intrusives bear the imprint of deCormation (ofien complex) and low-grade metamorphic recrystallisation, both oC which obviously postdate magmatic emplacement.With regard to de-Cormation most oC the bodies seem to have acted as essentiallY competent blocks, in comparison with host metasediments, which explains the common disturbance °Orihuela MURCIA e
of original contact relationships.One notable exception is a rather large sill (now forming several discrete outcrops) located within the Romero Unit in the Sierra de Carrascoy (lig.3), whose emplacement pro-duced well preserved small-scale contact effects (porphyritic «chilled» borders, baking of host pelitic rack, local Na-metasomatism, etc. (see Kampschuur, 1972, for more details).lithotypes where local granophyric textures may develop.Apatite is always present as an accessory constituent.Near intrusive contacts, however, the samples show porphyritic to glomeroporphyritic textures, with olivine and orthopyroxene phenocrysts within a microcrystalline matrix of plagioclase, augite and Fe-Ti oxides.

Mineral Assemblages
As a rule, the aboye mentioned igneous assemblages are partly replaced by metamorphic minerals of different nature, although in most cases the original magmatic texture is well preserved.Thus, serpentine and chlorite pseudomorphs formed after olivine, which in tum have later been replaced by epidote and actinolite.Orthopyroxene shows nearly complete replacement by tale and serpentine.Augite is transformed into variable proportions of homblendic and actinolitic amphiboles, with exsolution of rutile needles, and/or may be rimmed by a metamorphic greenish clinopyroxene.Most plagioclases have been nearly completely altered and replaced by assemblages of sericitic micas, prehnite, clinozoisite and/or pumpellyite.These metamorphic assemblages are also found witbin small veinlets, associated with sphene, quartz, caleite and albite.A stronger degree of textural change is shown by sorne dolerites, grading into decussate amphibolite.In such cases neoformation of albite is common, associated to epidote and actinolitee and, less frequently, to crossite.
Although it is not the purpose of this report to discuss the petrogenesis of the aboye mentioned secondary assemblages, it is to be noted that the coexistence of actinolite and epidote with pumpellyite indicates metamorphic conditions near the boundary of the greenchist facies with Hashimoto's (1966)

LEGEND
The most abundant type of rock forming the mafic bodies in both Sierra de Enmedio and Sierra de Carrascoy is fine, to coarse-grained dolerite, with textures ranging from doleritic to ophitic or subophitic.In these rocks the main igneous mineralogical constituents are plagioclase (andesine-Iabradorite) and augite, in sorne instances accompanied by bronzite.Interstitial micrographic intergrowths of quartz and alkali feldspar are also present, especially in the more differentiated pumpellyite subfacies.The generation of this assernblage, together with the occasional presence of crossite (cf.also Wood, 1979), would point to environmental pressures between 3.5 and 5.5 kbar, and temperatures around 350°C.In Sierra de Enmedio the temperature range might have been somewhat higher, as far as in these samples greenchist facies assemblages predominate (devoid of or with scarce pumpellyite), and metamorphic textures are also better developed.

Whole-rock geochemistry
Tbe data corresponding to 19 samples from the Sierra de Enmedio and 16 samples from the Sierra de Carrascoy have been plotted in figures 5, 6, 8 and 12. Frorn these, a selection of 15 major and traceelernent analyses, together with their corresponding CIPW norrns, are given in table 1   Figure 4 is a silica versus Zr/Ti0 2 diagram, which allows classifying the whole group of rocks as belonging to the family of subalkaline basalts.Individual lithotypes, as defined in the silica versus alkalies diagram (figure 5), may nevertheless range compositionally from subalkaline basalt to basaltic andesite and andesite.Differentiation trends appear to have been clearly tholeiitic (Le., with strong iron enrichment in the more differentiated rocks; see figure 6) in both series of samples, and was in all probability connected to the early crystallisation of both olivine and bronzite.These are found as phenocrysts in the porphyritic samples found along well preserved intrusive contacts of sorne bodies.Strong Cr and Ni depletion accompanied the segregation of these early crystallizing phases (figure 7), which was coupled with enrichment in sorne elements like Cu and V.The latter observation, particularly the «incompatible» behaviour of Cu, are presently regarded as an exclusive feature of continental tholeiites (Dupuy and Dostal, 1984).
Other compositional features point to minor differential aspects between the rocks from Sierra de Enmedio and Sierra de Carrascoy, notably the fact that, for a given silica range, the rocks from S. de Carrascoy have somewhat higher Al z 0 3 , KzO  CaO contents and lower TiO z than those from S. de EnmOOio.These differences are more clearly illustratOO in a plot of KzO versus Na 2 0+CaO (fig.8).Within this diagram, according to the models of differentiation proposOO by Puga and Portugal (1989) for Mesozoic magmatic rocks in the Subbetic Zone, the compositions plotting in the FC field (see fig. 8; in our case most of the samples from S. de Enmedio and less than a This diagram separales suites wilh varying degrees of assimilation versus fractional crystallisation as componental mechanisms oC chemical differentialion.Fields after Puga and Portugal (1989).Key to abbreviations refer to inferred main differentiation mechanisms: FC = fractional crystallisalion; AFC = fraetional crystallisalion, plus assimilation; HAFC=bigh assimilation, plus fractional crystaUisation.Symbols as in figure 4.
Tectonic environment of magma generation and emplacement half of those from S. de Carrascoy) would have followed mainly a trend of differentiation by fractio-natOO crystal settling.On the other hand, those within the AFC field would also have differentiatOO mainly through fractionated crystallisation but accompaniOO by noticeable crustal contamination.In particular, assimilation of pelitic rock is thought to explain the relatively high KzO, SiOz and Alz03 contents in many of the sample, whereas that of small proportions of carbonates and/or gypsum should result in the observed less pronounced decrease of CaO, with differentiation, especially in the rocks from S. de Carrascoy.
The pattern of normalized abundances of incompatible elements is analogous to that of well known continental tholeiitic provinces (figure 9), whose magmas are ofien enrichOO in the more incompatible elements (e.g.Ha, Rb, Rb, Th and K; see Dupuy and Dostal, 1984).The Th maximum is particularly typical of these tectonic settings, due to Th leaching froro deep crustal rocks while magma is transportOO towards the surface (cf.Thompson el al., 1982).For comparison purposes, figure 9 also ineludes the corresponding patterns of mafic rocks (<<ophites») from the Subbetic Zone, also belonging to a continental province and that of edogites from the Nevado-Filabride Ensemble from the subbetic (triangles) and eclogitized basalts from the Nevado-Filábride Ensemble (solid squares).Values for the Columbia River basa1t were computed after the international standard BCR-I (Govindaraju, 1984), and those of the subbetic «ophites» and nevado-filábride eclogites after Puga and Diaz de Federico (1988).
Normalizing factors used, aceording to Thompson (1982).are given below each element symbol.
of the Betic Zone, having a more plain oceanic character (Puga el a1., 1988).A further indication of overall intraplate tectonic environment is given by the Zr/Y versus Zr plot (figure 10).As seen, only two (non cumulitic) samples plot outside the within-plate (WPB) field, due to particularly low Zr contents.
Another useful elemental index of tectonic setting is tbe VITi ratio, whose value does not change easily as the result of either alteration or metamorphism (Shervais, 1982).V/Ti ratios of the rocks investigated are rather homogenous in both Sierras (aH values comprised betwen 20 and 40), and are in the range found in both rnid-ocean and flood continental tholeütes (figure 11), hence proper of crustal extensional environments.Furthermore, the presence of high ahsolute amounts of V and Ti as found in one of our samples (a coarser differentiate from S. de Enmedio, with interstitial granophyric intergrowths) appear to be exclusive of continental basalts, such as those of the Columbia River Province.

K/Ar data
In a preliminary attempt to assess age relationships, four total-rock samples from Sierra de Enmedio have been subjected to K/Ar analysis, the data being summarized in table 2. Higbly discordant results were obtained, with apparent ages ranging from about 178 Ma (sample E9) to 59-40 Ma (samples E37, E21 and Ell).
For one interpretation of these discordant ages, however, it is important to remark that K/Ar results appear to be correlated to the degree of replacement of plagioclases by sericite.Thus, the oldest result is given by sample E9 (l78±4 Ma), which consists of an andesine-clinopyroxene coarse-grained differentiate, with granophyric interstitial intergrowths and abundant ilmenite.In tbis rock the pyroxene appears extensively replaced by actinolite, but plagioclases and K-feldspar granophyric intergrowths are to a large extent weH preserved.This contrasts with the other three dated samples where plagioclases show rather advanced degrees of replacement by sericite.Hence, and opposite lo tbe obvious alternative explanation involving possible radiogenic Ar enrichment (e.g., Hebeda el al., 1980), we think that recrystallisation accompanying lowgrade metamorphism was in aH probability responsible for the observed discrepancies among the four samples dated, via differential argon loss and the concomitant 600 The tme significance of the remaining much younger estimates is more difficult to assess without additional evidence.The relatively fair concordance among the remaining three dates would however suggest that the metamorphic event(s) (and associated deformation) affecting these units might have taken place during the Middle Eocene, but also in younger times if Ar loss was incomplete during low-grade metamorphism.An Eocene event might be better correlated with similar alpine metamorphic ages from nevado-filabride sequences (cf.Portugal Ferreira et a1., 1988; «eoalpine event»), as well as with possible significant tectonic activity in the area (cf.Paquet, 1966Paquet, , 1974)), but again younger metamorphic ages and strong tectonicism are also well known lo have taken place in the Betic Zone up to the Early Miocene.Shervais, 1982), showing the plotung of the investigated metabasics with that of subbetic «ophites» and continental and mid-ocean ridge tholeiites.See text for additional comment.Fields for rocks from the Columbia River province are indicated, as well as for average basalt from the Red Sea axis (asterisk; see sources in Shervais, 1982).Other symbols as in figure 4.
rejuvenation of radiometric dates.The degree to which radiogenic argon was liberated from these systems would hence be related to the intensity of recrystallisation, which we have noted is fairly variable even for samples coming from one and the same outcrop.The 178±4 Ma datum might thus be taken as a minimum estimate for magmatic emplacement, indicating that intrusion of these bodies took place during Early Middle Jurassic or somewhat older times.A maximum bracket is given by stratigraphic location, within Upper Triassic beds.This interpretation, as opossed to any hipothesis involving Ar enrichment, is also consistent with field evidence as summarized by Kampschuur (1972, p. 43), and interpreted in the

Comparison with other mafie eomplexes in the Betie Cordilleras
Evidence of significant Mesozoic basaltic activity is known from several places within the Betic Cordilleras, notably in the Subbetic Zone and the Nevado-Filábride Ensemble (see Puga et a1., 1988a, b, and references therein).A comparison with these instances is made in figure 12.This plot further illustrates the already noted general trend of potassium enrichment of the rocks investigated (that is even more pronounced in samples from S. de Carrascoy; cf. also figure 8), which we relate to the occurrence of assimilation of pelitic rock.This enrichment in K (and other lithophile elements; see fig. 9) is hence regarded as evidence in Concluding remarks basites from the Nevado-Filabride Ensemble, whose generally lower lithophile-element contents are more consistent with an emplacement within an open oceanic environment (Puga and Díaz de Federico, 1988).Fig. 12.-Comparison of the alkali ratio versus total alkali content of the investigated metabasics with that of subbetic «ophites» and nevado-filabride eclogites (data from Puga and Díaz de Federico, 1988).Key to fields: discontinuous line=nevado-filabride eclogites; lines and dots=subbetic «ophites».Plotting symbols as in figure 4.
See text for additional comments.
tectonic scenario of magmatic emplacement.AIso, the existence of differences in the degree of ensialic contamination between the two groups of outcrops investigated might be interpreted in terms of variations in the degree of attenuation of continental crust in the corresponding domains.Crustal thinning (related to the extensional event) could have been less pronounced under the depositional area of the Romero unit (S. de Carrascoy) than under the one of the S. de Enmedio Unit, possibly meaning that the two domains were not arranged parallel to the rifting axis.
With regard to the problem of the provenance of the enclosing almagride units (refer to the introductory section), one significant point to be remarked is that these mafic intrusives bear much rassemblance with known hypabissal metabasites from the external zone of the Betic Cordilleras (see aboye), which is consistent with proposals that correlate these units with the Subbetic in the Murcia area.As a note of caution, nonetheless, it should be remembered that within the subbetic realm submarine volcanics played a significant role, that is absent here, and that the alternative possibility that the mafic bodies investigated could witness either (a) the existence of an independent zone of aborted rifting in the surroundings of the inferred oceanic gap associated to the Nevado-Filabride Ensemble, or (b) the remains of an initial stage of opening of the latter, preserved within one of the separating thinned and faulted continental margins, cannot be completely discarded.The latter hypotheses implying a closer plaleogeographic relationship to the Alpujarride or Nevado-Filabride crustal realms, would in turn be supported by the relatively older time of emplacement oC the investigated mafic bodies, as cornpared with inferred maximum extensional dynamics in the western Tethys during the Middle Jurassic to the Early Cretaceous (cf., e.g., Lemoine el al., 1987).
During the EarIy Tertiary, both the mafic intrusions and host sedimentary sequences were affected by low grade metamorphism, whose characteristics are not easily explained unless sorne kind of underthrusting, or analogous process causing deep burial under a moderate temperature regime, is invoked.Torres-Roldán (1979, p. 38) suggested that one possible explanation might be their dragging down (or underplating) in relation with one possible Late Cretaceous to EarIy Tertiary subduction event, that was postulated responsible for the eoalpine high-pressure assemblages in the Nevado-Filabride Ensemble (cf.also Díaz de Federico el al., 1979).This notion would appear consistent with our metamorphic K/Ar dates, although the precise tectonic setting of the metamorphism remains now more open given the fact that LT-HP assemblages are presently known to occurr also in both the subbetic (Puga el al., 1983(Puga el al., , 1988b) ) and alpujarride (Goffé el al., 1988) sequences.The study of the mafic complex in the alrnagride units of the Eastern Betic Zone reveals a two-stage history, related to original igneous emplacement and subsequent metamorphism, respectively.During the EarIy Mesozoic, i.e., within a period stretching from the Late Trias to the EarIy Middle Jurassic, the paleogeographic site of these units was the locus of crustal extension and subcrustal magma generation.The geochemical characters of the ensuing mafic intrusives, especially their relatively high lithophile-element contents, point to sorne kind of aborted or limited continental rifting as the more probable the sense that the corresponding magmas must have traversed sorne extent of continental crust while in their way to final emplacement.These chemical indices of assimilation make these rocks comparable to sorne of the mafic intrusives (the so-called «ophites») from the Subbetic Zone, whose magmas, also tholeiitic in nature, do bear similar chemical signatures of contamination by continental crustal material.As shown in figure 12, however, this was not the case with meta-

FFigMFig. 7 .
Fig.6.-AFMplot (An=Na,O+K,O; F=FeO,,, ,; M=MgO), showing the overall iron-enrichment (tholeiitic) trend of the mafic suites from Sierra de Enmedio and Sierra de Carrascoy.The noticeable scattering of points towards the A apex is thougth to be at least partly related to heterogenous assimilation of crustal material and/or syn-metamorphic element mobilisation.Note also the wider range of differentiation of the rocks from S. de Enmedio, Symbols as in figure 4.

Fig
Fig. 1O.-Plot of the investigated rocks within the Zr versus Zr/Y diagram, proposed by Pearce and Norry (1979) as indicative of tectonic environment of generation of basalts.Key to abbreviations: WPB = within-plate basalts; MORB = mid-oceanic ridge basalts; IAB=island arc basalts.Symbols as in figure 4.
Fig. ll.-V/Ti diagram (afterShervais, 1982), showing the plotung of the investigated metabasics with that of subbetic «ophites» and continental and mid-ocean ridge tholeiites.See text for additional comment.Fields for rocks from the Columbia River province are indicated, as well as for average basalt from the Red Sea axis (asterisk; see sources inShervais, 1982).Other symbols as in figure4.

Table 2 .
-Analytic results of K/Ar dating (*) Oetails on the analytica1 procedures and constants used are given in Puga et al. (1 988b).