ATTEMPTS OF WHOLE-ROCK K / AR DATING OF MESOZOIC VOLCANIC AND HYPABISSAL IGNEOUS ROCKS FROM THE CENTRAL SUBBETIC ( SOUTHERN SPAIN ) : A CASE OF DIFFERENTIAL ARGON LOSS RELATED TO VERY LOW-GRADE METAMORPHISM

Doce muestras de cuerpos básicos intrusivos en rocas triásicas (<<ofitas») y 11 muestras de volcanitas y rocas intrusivas asociadas en secuencias jurásico-cretáceas de la rona Subbética han sido objeto de datación radiométrica K/Ar (roca total) en combinación con análisis químico-petrográfico. Las edades analíticas obtenidas son 'sólo satisfactorias en relación con un cierto número de muestras de rocas volcánicas, no obstante, con una concordancia respecto de las edades estratigráficas en el orden del 10% relativo; las del resto, fundamentalmente muestras de «ofitas», presentan una fuerte dispersión, con diferencias que pueden alcanzar valores superiores a los 30 Ma, incluso en conjuntos provinientes del mismo afloramiento o localidad. Se concluye que las pérdidas de Ar causantes de los rejuvenecimientos de las edades analíticas observados son debidos a la existencia de transformaciones metamórficas alpinas, de muy bajo grado, que afectaron con mayor intensidad a las ofitas que a las rocas volcánicas presentes en niveles estratigráficos más altos. Otros cambios posteriores al emplazamiento magmático, tales como el grado de oxidación secundaria, son así mismo distintivos en ambos grupos de muestras al tiempo que proveen de soporte adicional al concepto de que el medio de alteración de las «ofitas» debió haber producido condiciones favorables a una interacción roca/fluidos más penetrativa y, por tanto, una recristalización más homogénea. En conjunto, la actividad magmática de la que derivaron las «ofitas» podría haber comenzado en el Trias terminal y continuado en el Jurásico Inferior. Tanto las «ofitas» como las volcanitas se consideran el resultado de eventos magmáticos ligados a movimientos de distensivos a transtensivos que afectaron a las cuencas externas de las Cordilleras Béticas desde el Trías terminal hasta el Cretáceo inferior.


Introduction
The external zones of the Betic Cordilleras represent various sOOimentary realros of rocks that accumulatOO in basins adjacent to the Iberian subplate from Triassic to Miocene times.A bipartite subdivision is usually made on the basis of the inferrOO relative location of the depositional areas: the Prebetic series (corresponding to the Prebetic Zone) were depositOO directly adjarent to the Iberic (continental) margin and the Subbetic series (in the Subbetic Zone )more to the S or SW, farther away from the former Iberic coastline.
During the alpine tectogenesis (with important deformational events up lo the Middle -and even Late--;-Miocene) most of this ensemble of Mesoroic or younger sedimentary sequences was detachOO from its Paleoroic to Lower (most) Triassic substratum.The ensemble was then broken up into a great number of tectonic elements that, according to c1assieal concepts, resulted essentially or at least primarily from thrusting towards the N or NW (e.g.: Durand-De1ga, 1980; Durand-Delga and Fontboté, 1980;García-Hernández, el al., 1980), or otherwise intrineately mixOO (e.g.: Bourgois, 1978;Hermes, 1978;van de Fliert el al., 1980;de Smet, 1984).The extreme structural complexity has loo to considerable difticulty in attempts at paleogeographic and evolutionary reconstruction of the original areas of deposition, particularIy of the subbetic sequences.In the central segment of the subbetic belt (between the meridians of Antequera and Poro Alc6n, fig. 1) an <externa1», <<median» and <<IDternal» subdomain were distinguishOO (cf.García-Dueñas, 1967).These subdomains can be partly extrapolated to adjacent segments where similar subdivisions have been proposed.
The difl'erentiation of and within the Subbetic domain originated in the EarIy Jurassic, at the end of Carixian and beginning of Domerian times (about 200 Ma ago) and was caused by the development of a difIerential bathimetry, relatOO to basement fracturing.That event is now generally interpreted as marking the onset of an extensional to trascurrent regiBle, in connection with tbe separation of Laurasia and Gondwanaland.Stratigraphit? development indieates a possible continuation of this regime into the latest Early Cretaceous (Aptian-Albian).
Apart from the above-mentionOO lavas and spatialIy associatOO intrusives, frequent bodies of dolerite (the so-caIIed «ophites» (l) are also found as more or less isolated masses within (often chaotic) Triassic sediments of germano-andalusian facies, eitber when they form the stratigraphic basis of the Subbetic units or within independent tectonic elements dominated by this type of facies (e.g., the so-called «Trias de Antequera» and the «Unidad de Cambih>.
Because the «ophites» appear to have been intrusive and have not so far been found accompaniOO by indisputable extrusives within the Triassic rocks, their interpretation in terros botb of age and possible genetic relationship witb the volcanites and other magrnatic bodies locatOO at bigher levels has remained uncertain.On the basis of tbeir remarkable local abundance and quite uniform mode of appearance in the field, these «ophitic» bodies have often been interpreted as resulting from an older and distinct magmatic event, possibly connected with precursory tectonic instability within the corresponding Triassic basins.
PreIiminary KJAr dating of several samples and field observations in the Cantar area (see fig. 1), however, loo van de F1iert el al. (1979) to discuss the allegOO Triassic age and distinct origin and to conclude that most (1) The term «ophite», introduced Cor comparabale rocks in the Pyrenean mountains by the Rev. Father Palassou as Car back as 1798, has since been defined in various ways.As a rock name it did not receive general acceptance, in contrast with the concepl oC ophitic texture (Rinne, 1921).In the Western Mediterranean area (Pyrenean mountains, Spain, Morocco, Algeria and Tunisia) the term «ophite» is still commonly used Cor blocks and larger bodies oC basic rocks oC varying composition and texture encountered in -mostly diapiric-Triassic sediments.«ophites» could in Cact be nearly contemporaneous (within about 10-20 Ma) with the Jurassic and younger lavas.
In the Archidona region, on the other hand, the K/Ar dating oC amphiboles from similar «opbitic» bodies resolted in apparent ages between 78 and 90 Ma, interpreted by Puga et al. (1983) as an indication oC postemplacement metamorphic recrystallisation.
On tbe present paper we report tbe results of attempts oC K/Ar dating oC an additional set oC samples oC volcanic and intrusive rocks (incIuding «opbites») Crom tbe Subbetic series oC tbe western and central segments oC tbe Subbetic Belt, and discuss tbeir significance as regards true ages oC emplacement and possible later dJangcs, in oonnectioo with sorne aspeds of the evolution oC the Subbetic realm and of tbe External Zone oC tbe Betic Cordillera in general.

Field setting and description of the investigated samples
Tbe bypabyssal dolerites (<<ophites») intruded into sedimentary rocks oC Tl'ia$ic age (plotted as squares in figs.2-4) do mostly Corm small stocks with diameters oC up to several hundred metres.The original relation witb tbe encIosing strata is usually obscured as a resolt oC secondary tectonic disturbance.Less frequentIy they Corm sills and discordant dikes witb thicknesses up to sorne teos oC metres.In most iostances these outcrops do not reveal any cIear spatial realtiooship witb the volcanics oC Jurassic or younger age, even when corresponding tectonic units bear complete Mesozoic sequences.
The boundary between fields 3 and 4 is aíter Irvine and &ragar (1971).Two samples from!he Cantar area were also plotted in this figure (Spa 1 and 6).intrude either the lava flows or the sedimentary strata altemating with these.
The more remarkable petrographic features of the dated samples have been summarized in tables 1-4, whilst major-element analyses are given in tables 5-6.From these data, as wen as from a study of other samples from the same outcropi, it might be concluded that both groups are made essentially of similar basie rocks, in the range of alkali or tholeiitie basalt/dolerite (plus hawaüte, mugearite or basaltie andesite) (fig.2); although a note oí caution should be made about some samples whose original mineralogy, and hence possibly overall ehemistry, might have been modified by secondary recrystallisation.As a whole, nonetheless, both ehemical and primary mineralogical compositions might be regarded as transitional between the tholeütie and alkaline sodie magmatie series.Apart from indicating a mantle origin, the petrographic and cbemical charaeters oC tbese rocks are consistent with the view that their emplacement or extrusion should have taken place under a relatively extensional erustal regime.In this regard these data agree well with existing interpretations in whieh the contemporaneous crustal dynamics are connected with the operation of deep NNE-SSW directed faults, whose mechanism woud have been mostly transtensive (Hermes,  1978; Comas et al., 1986).Median Subbetic F1attened pi-Fine sized intergranular, Alkali ba-Uow lavas up witb vesicles salt to 1 m in díameter Besides these similarities, however, there are also some differential features between the rocks from each group of outcrops tbat must be taken into account when interpreting the results of K/Ar analysis.Many of these distinctive features, part of which are reflected in tables 1-4 and figs.2-3, were already pointed out by Puga and Ruiz-Cruz (1980).For each group of samples they may be summarized as follows:

«Opbites», intruded witbin Triassic sediments
Altbough, as in the case of tbe Jurassic lavas, these rocks appear lo have been derived m~y from transitional magmas, their chemistIy bear a definite tendeocy towards tboleütic rather than alkaline (see fig. 2 and table 5).Normative hypersthene (witb or witbout quartz) is more frequent than nepheline, and there is also a nearIy ubiquitous presence of small proportions of modal quartz, in some instance forming granophyric intergrowths witb alkali feldspar (table 1).Anotber primary mineralogic feature of this group of samples is the presence of a titanium-poor augite as the more common pyroxene, while plagioclases are in the range of labradorite to oligoclase.
Medium sized subophitic textures are prevalent in tbese rocks.Olivine usually appears as inelusions within augite crystals and is systematically replaced by chIorite pseudomorphs.Common secondary assemblages inelude chIorite, sericite, prehnite and pumpellyite (table 1), that in some ÍIlStances are accompanied by a slightly jadeitic pyroxene (Jd < 20%), actinolite, crossite (GI38Rieb62) and epidotes.The samples of tbis group are also charac-terlUd by more homogeneous and lower total volatile quartz is present and the more abundant primary pyroxene is a titaniferous augite.Plagioclase, moreover, is generally more ca1cic, in the range of bitownite to andesine.Tbe volcanic varieties within this group have a sma1ler grain si7.e, vesides, and oommon intersertal lo inteIgranular textures, and hence bear considerable petrograpbic differences as compared witb the «ophites».Nonetheless, the hypabissal varieties of this group also differ from tbeir «ophitic» counterparts, inc1uding the prevalence of opbitic or porphyritic ratber tban subopbitic textures.Additionally, olivine crystals may show on1y partial replacement by smectite-ch1orite assemblages and are less frequentIy found as inc1usions in the pyroxene.
Plagioclases sometimes have a «c1oudy» appearance (due to incipient alteration to c1ay minerals), but no sericite is seen to replace them as in the case of most «opbites».In general, the more significant secondary mineral assemblages inc1ude saponite, mixed-layer smectite-chlorite, zeolites and stilpnomelane, although these may in sorne instances be also accompanied by

Volcanic and Íntrusive rocks within Jurassic to Cretaceous strata
Tbeir chemistry has a tendency towards alkaline sodic rather tban tholeiitic, in contrast with «ophite» samples, as retlected in the more common appearance of nepheline instead of hyperstbene when norms are ca1culated (see fig. 2 mi table 6).In good agreement with tlm, no modal

Interpretation olthe chemical and petrographic differences between the two groups ol maJic rocks
Tbe observed differences concero either chemical trends (major elements), mineralogy and textures, and might receive diverse explanations depending on the accepted relationship between their respective primary magmas.As regards primary chemical trends, however, our major-element data deserve further scrutiny in the light of a more comprehensive trace-element data set, especia1ly if it is taken into account the noticeable degree of secondary alterations of sorne samples.Otber cbemical and mineralogical distinctive features between tbe two groups, on tbe otber band, sucb as tbe different -or inequally developed-secondary assemblages, degree of secondary oxidization and total volatile content merit sorne special consideration, particularly witb regard to tbe occurrence of low-, to very-low «burial» metamorpbic recrystallisation.Tbus, tbe comparatively bigber degrees of recrystallisation found in most «opbites» would suggest tbe deeper burial of these bodies, as compared witb tbe Jurassic and younger lavas.This would also agree witb tbeir more bomogeneous and moderate degrees of secondary oxidization, inasmucb as recrystallisation under somewbat bigber temperatures and/or witbin better confined systems sbould bave resulted in improved internal buffering of oxigen fugacity, and, in general, in a more advanced condition of chemical equilibrium.In our samples, tbe resulting metamorpbic assemblages do generally belong to tbe prehnite-pumpellyite (in some instances 10 tbe aetinolitepumpellyite) facies, in tbe case of tbe «opbites», and to tbe zeolite facies in tbe case of the Jurassic or younger lavas and associated intrusives.This agrees witb previous observations by Puga et al., (1983), wbose estimates suggest conditions up to 3 kbar and 300°C for tbe prehnite-pumpellyite assemblage (somewbat bigber for tbe aetinolite-pumpellyite assemblages found in sorne «opbites»).For additional details and discussion of tbese metamorpbic assemblages see Puga et al. (1983) and Morten and Puga (1983).

K/Ar Data
Experimental procedures and constants Sieve fraetions (125-250 ~) of the wbole rocks were analysed.Potassium was analysed by fiame pbotometry with litbium internal stantard and CsA1 buffer.Argon was extraeted in a glass vacuum apparatus and determíned by stable isotope dilution tcchniques, usiog "Ar as a tracer witb a Varian GD-150 mass spectrometer.All measurements were made by static mode.Analytical errors are estimated to be within 1% for K and 2% for Ar.Tbe following constants were used for age calculations: L.=0.581•1Q-1O a -1, L~=4.962•1Q-lo a-I , and tbe abundance for «>Ar=0.01l67atom percent of total K.

Analytical results
Tbe results of K/Ar analysis of 23 samples are presentfd in table 7 aro fig. 4. In order 10 ~that the calculated dates do not necessarily correspond witb true age of rocks, tbe term «analytic age» is used.
Regarding tbese data, it sbould be remarked first tbat attempts of wbole-rock K/Ar dating of basic intrusive and extrusive rocks in tbe external realm of tbe Betic Cordilleras meet witb serious difficulties, tbe causes of wbicb are only partly understood so faro When tbis program was initiated, bowever, more bopeful results were to be expected, tbe zone being considered to be composed of completely unmetamorpbosed rocks.A first attempt was reported from tbe Cantar area (Van de Fliert et al., 1979), based on tbe analysis of 6 samples: two from extrusive rocks oontaining fragments oC Titbonian limestone and covered by Lower Cretaceous strata, and 4 from dolerites in tbe surrounding Trias.In tbat case tbe analytic age oC tbe pillow-Iavas corresponded closely witb stratigrapbic location, altbougb tbe dolerites produced somewbat younger analytic ages in tbe range oC 120 to 100 Ma.Sorne indications oC metamorpbism in tbe dolerites raised a problem in an admitted non metamorpbic zone, at tbat time, so tbat tbe interpretation tben advanced regarded comagmatic alteration (deuteresis) as tbe cause oC tbe observed discrepancies.Tbe origin oC tbe pillow-lavas and dolerites was tbus bypotbetically attributed to tbe same magmatic activity in tbe Latest Jurassic to Early Cretaceous times.
At present, however, tbe analysis oC additional samples over a mucb wider area witbin tbe Subbetic Zone, as with a more or less pronounced but general loss of bere reported, witb mucb greater discrepancies between radiogenic Ar, affecting especially tbe «ophites».Effectitbe analytical ages given by tbe two groups of basic vely, tbe analytic ages of most «ophites», witb tbe only rocks, as well as the description of unmistakebly meta-exception of sample SPA-24, are systematically lower, morpbosed «opbites» in tbe Archidona region (Puga et often considerably, tban tbose of tbe volcanics and al., 1983), suggest that the occurrence of a postassociated intrusives, while their strong scattering malees emplacement very-low-grade metamorphism should insthem obviously more suspect of having suffered partial tead be taken into consideration.This means that the isotopic resetting.Tbe analytic ages of the volcanic and hypothesis of one and the same magmatic event for both associated intrusive rocks, on the other hand, can be groups of rocks is to be reappraised, as far as the analytic evaluated much more precisely because of stratigraphic ages obtained, even in the case of apparent1y unweathered control In table 8 they are compared with the values of fresh rock samples, cannot be safely regarded as direct1y the Harland et al. (1982) time sca1e, as appropiate for related to true emplacement age.each stratigrapbic location.Looking table 8, it can be Contrary to the excess Ar (or loss of K?) encountered seen tbat the analytic ages of tbree samples lie within the in sorne metamorpbosed basic rocks from the Betic Zone accuracy limits ofHarland's time scale (SPA 11,13 and (see Hebeda et al., 1980), hence, we seem to meet here 23; the same appears to hold for samples SPA 20 and 21, but tbe exaet stratigraphic location of tbese is more uncertain).Tbe otber six samples sbow deviations from 8 to 15 relative percent, an average of about 10% too low wben compared witb tbe time scale.analytic error) both mutually and with the mínimum given by stratigrapbic location.Vo/canic and associated intrusives in Jurassic-Cretaceous strata

Intrusive rocks in
As follows from table 8, tbe analytic ages of some of tbe submarine lavas agree witbin reasonable limits (i.e., less tban about 10 relative percent) witb estimates based on tbeir stratigraphic location.Tbe remaining appear lo bave been somewbat rejuvenated, in tbe range of 10-20 Ma, most probably as tbe result of the comparativ~ly higber degree of wheathering of the corresponding samples.This is well illustrated, for instance, by the difIerent analytic ages obtained from samples SPA 9, 10, 11 and 12, in spite of too fact that tbese were taken at ~he same locality and witbin as approximately 20 m thlCk lava pile near Montejícar (see fig. 1).In tbis respect it may be remarked tbat tbe oldest estimate obtained from this particular lava tlow is given by sample 11, whicb also bears the lowermost modal percentage of secondary minerals witbin the group.Moreover, considering tbe whole set of samples this ti~e! it seems not a coincidence that the two samples grvmg the lowest analytic ages (SPA 12 and 14) have ratber high potassium contents (table 6), retleeting probably a more advanced degree oí secondary clay formation from tbe primary assemblage.Tbe samples giving the oldest analytic ages, on tbe otber hand (SPA 20-21) are exceptionally fresh subvolcanic rocks (from a sill-like body), wOO;e e&imated K/Ar ages are quite consistent (within the limits of

Re/ation olana/ytic age and roek alteration
Our interpretation of the discordant ages ~ve~by ~e samples of ophitic rocks, in terms of a partIal lSOtop~C resetting by a Late Eocene or younger metamorphlc Tbe analytic ages obtained from tbis group of samples are much more scattered, with differences that may exceed 100 Ma.Marked age discrepancies are aIso obtained from samples that were collected at the same locality or outcrop.One extreme example is given by SPA 19 and 24, from tbe Trias ofthe Guadiana Menor, near Ceal, whose K/Ar ages differ by about 80 Ma, althought it should be noted that this case might no be representative as far as they were collected f~om a sedimentary accumulation rich in blocks of dolente that permitted tbe sampling of particularly fresb-looking pieces of rock, but that might not have originat~d from tbe same body.Tbe samples from the Cambll area, midway between Cambil and Huelma (SPA 25 and 27-29 witb a maximum internal discrepancy of about 27 M~), do belong to tbe same intrusive body.
Tbus, and especially taking into account tbat no stratigraphic control on these samples can be made o.t~er tban their being Triassic or younger, any defiDltl~e conclusion regarding too true age of tbe related magmatlc event (or events) is precluded on the basis of our K/Ar data alone.Tbe only significant constraint in tbis regard is perbaps tbat such true ages should have been older tban tbe highest analytic estimate for each oatcrop or locality (about 160 Ma and 74 Ma, respec~ively,. in the examples discussed aboye; tbe same reasonmg migh~be applied lo otber single-sample outcrops as summarized in fig. 1, with «minimum» ages in the range of98 to 129 M~. .Tbe strong scattering of the analytlc ag~obtam~d with this group of samples is nonetheless conslS~ent wlth tbe already noted higher degree of metamorphlc recrystallization in most «opbites», and more particularly with tbe great extent to which primary plagioclases of these rocks bave been replaced by sericite (cf.table 1).Tbis observation might be taken as an additional indication in tbe sense that, regarded as very-low-grade metamorphic rocks tbe «ophites» probably attained a higber -altough still ~ariabledegree of chemical equilibrium witb conditions prevailing during deep burial, as evidenced by secondary míneralogy.From tbis standpoint, our K/Ar data would suggest that sucb metamorphic event sbould have taken place during tbe Late Eocene or younger times, as indicated by the mínimum 47 Ma datum obtained froro sample SPA 29.An interesting featore of the «ophite» negative correlation to analytic age is that, in spite of the great uncertainty about the exact position of the regression line (tbe one shown in Fig. 4 was estimated visually), it points to the possibility tbat the intrusion of at least some «ophite» bodies preceded tbe oldest presently dated lava flow (around 200 Ma ago; samples SPA 20-21).This foHows from the observation, in fig.4, that tbe extrapolation of the «regression line» up to a point with an oxidization ratio of about 0.2, taken as representative of many fresh basaltic rocks (cf.e.g., Irvine and Baragar, 1971; see also sorne additional analyses of unaltered volcanic rocks from tbe Subbetic Zone in Puga and Ruiz-Cruz, 1980), would suggest an initial magmatic emplacement around 190 Ma (or even earlier) on tbe time scale.Hence, if the assumptions underlying tbis plot are indeed approximately correct, one could speculate with tbe possibility tbat magmatic activity witbin tbe Subbetic basins bad started early, in tbe Lias and perhaps even in the Late Triassic (<<ophites»), reaching a maximum development later in Late Jurassic to Early Cretaceous times (volcanics and associated intrusives).Such possibility would agree with field observations by García-Cervigón et al., (1976), in relation with «opbite» outcrops some 150 km to the E, near Cehegin.Tbere, a Late Triassic age of magmatic emplacement was su~ested on the basis of the interbedding, within Upper Triassic rocks, of placer magnetite depar;its supposedly derived from an already weathered «opbitic» body.Although outside the Subbetic realm, it is interesting to note aIso that the existence of pillow-breccias within Triassic sediments was aIso reported by Soediono (1971), in the malaguide sequence outcropping near Ciudad Granada (W of Vélez-Rubio).

Metamorphism in the External Zone of the Betic Cordilleras
Until recently, hardly anything was known about metamorpbism in the Subbetic Zone.Helmers (1978) reported in the Explanatory text of the Metamorphic Map of Europe (p.152): «Apart from prehnitization or pumpellyitization of a few basic igneous rocks no metamorpbic recrystallization has been reported from the Subbetic Zone».After recent confirmation of the presence of true metamorpbic assemblages in «opbites» from the Archidona region, near the boundary with the Betic Zone (Puga et al., 1983), further petrologic research now reveals that very-Iow-grade metamorpbic transformations affect aIso basic rocks over a much wider zone, including, although to a lesser extenl, some volcanites and associated intrusives.In the case of the «ophites», this metamorphic event is now additionally revealed by the more or less pronounced occurrence of Argon loss within those systems, affecting the results of K/Ar dating attempts.
As mentioned before, the related assemblage are thought to be indicative of pressures in the range of 3 Kbar and temperatures around 300°C at the climax.Conditions may have been somewhat higher in the Archidona region.The causes of this metamorpbism are still to be completely elucidated bul, in the case of tbe «opbites», tbe neccesary conditions mighthave been easily reached taking into account that the original stratigraphic overburden can be estimated in the range of 3000 m.From this, any further buriaI might have been related to the ocurrence of deep going transpressive fauItzones and associated tectonic Ioading, perhaps as one result of over-/under-thrusting reIated to them.
Additiooal research is undoubtely needed to, far instana; confirm the presumable existence of metamorpbic transformations within the enclosing sedimentary rack suites bul, anyhow, our data do already suggest that the metamorphic event must bave taken place in relatively recent times, perhaps during the Middle Tertiary, when the southern margin of tbe Iberic subplate collided with the Alboran Block.Tbus the external Zone of tbe Betic

Table l .
-«Opbites».Mineralogy and modal estimates of the samples dated.

Table 4 .
-Volcanics and assodated intrusives witbin Jurassic-Cretaceous sequences.Location and main features of the sampled outcrops.

Table 8 .
-Volcanics and assoclated lntrusives within Jurassic-Cretaceous sequences.Comparison of radiometric results with values deduced from stratigrapbic location.
Plot of K/Ar ages versus degree of secondary oxidization of the dated samples.the latter taken as an indicator of secondary a1teration environment.See text for implications and explanation.