Bruno Turi

Oxygen isotope studies of potassic volcanic rocks of the Roman Province, Central Italy

The 18O/16O ratios of rocks and coexisting minerals were measured for 93 samples of leucite-bearing lavas, pyroclastics, and related volcanic rocks from the Quaternary Roman Co-Magmatic Province, Italy. The δ18O values were found to generally increase northward in the sequence: Ischia (5.8 to 7.0); Somma-Vesuvius and Phlegrean Fields (7.3 to 8.3); Alban Hills (7.3 to 8.7); M. Sabatini (7.3 to 9.7); Vico Volcano (7.4 to 10.2); and M. Vulsini (8.1 to 11.7). The northward increase in δ18O parallels a similar increase in 87Sr/86Sr, and these data indicate that the Roman magmas have interacted strongly with high-18O continental crust. A marked increase in δ18O occurs just north of Rome where the Roman Province begins to overlap the calc-alkaline, oversaturated Tuscan Magmatic Province. Therefore, some of the observed 18O/16O and 87Sr/86Sr enrichments in the Roman magmas may have been facilitated by direct mixing with the high-18O Tuscan magmas or because the high-18O country rocks underwent widespread heating during a couple of million years of Tuscan igneous activity. Although many of the Roman magmas underwent fractional crystallization without appreciable change in δ18O, contamination has produced a correlation between δ18O and SiO2 content at several of the volcanic centers; thus the trachytes are typically higher in 18O than the undersaturated rocks. The major features of the oxygen isotope data can be explained in terms of a simple two-component mixing model in which one end-member was a primary, strongly undersaturated magma derived from the upper mantle, with δ18O≈+6, 87Sr/86Sr≈0.704 to 0.705, and SiO2<44wt.%. However, none of the analyzed samples have these values, as they have all been contaminated to some extent. The closest approach is found in some of the leucitepyroxenite ejecta from the Alban Hills. The second end-member, derived from the continental crust, had a variable composition with δ18O≈+12 to +20, 87Sr/86Sr≧0.712 to 0.720, and SiO2≧65wt.%, and it mixed in much greater proportions in the volcanoes north of Rome than in those of the Alban Hills or the Naples area. The widespread interactions between the Roman magmas and the continental crust are probably due to (1) the fact that such low-SiO2 magmas always have a very strong tendency to interact with quartz-bearing rocks of the continental crust, and (2) in Italy, these magmas were emplaced into a tectonically very active area containing poorly consolidated sedimentary rocks, and in the northern part of the belt there had been a prior history of extensive calc-alkaline igneous activity.

Isotopic study of the origin of sulfur and carbon in Solfatara fumaroles, Campi Flegrei caldera

Abstract Isotopic study of the origin of sulfur and carbon in the hottest (Solfatara) fumaroles of Campi Flegrei caldera, Southern Italy, was carried out on gas samples collected between 1983 and 1988, i.e. during and after the 1982–1984 seismo-volcanic crisis. The results for sulfur (H2S), the first ever reported on these gases, indicate a mean ∂ 34 S of −0.3±0.3‰ (range: −0.7 to +0.1‰ ) versus Canyon Diablo Troilite standard, consistent with an igneous derivation of this element, from either active magma degassing or/and leaching of reduced sulfur-bearing minerals in the volcanic layers. The lack of peculiar ∂34S variation during and after the crisis suggests that the chemical variation of H2S and S/C ratio in the fumaroles (increase and then decrease by a factor 3) were not due to a changing origin of sulfur. The mean ∂13C of carbon (CO2) over the period of survey, −1.6±0.2‰ (range: −1.9 to −1.3‰ ) versus PDB standard, is similar to the values obtained before the crisis (since 1970). Such an isotopic constancy requires a large and stable source of carbon feeding the fumaroles. The measured ∂13C values are much higher than those typical of primary mantle-magmatic carbon ( −6±2‰ ) and plot within the ∂13C range for marine carbonates ( 0±2‰ ). Such high values may reflect either (a) 13C-fractionation during degassing of CO2 from the underlying (⩽5 km depth) magma chamber or (b) the contribution of heavy CO2 of sedimentary origin, derived from either thermometamorphism of Mesozoic limestone series embedding the magma chamber or, possibly, past contamination of the local mantle by subducted sediments. Various arguments, among which volcanological evidence of an isolated and cooling magma reservoir (which would have been extensively degassed and, so, depleted in 13C along with time), the low 3He/4He ratios and the broad 13C-enrichment of volcanic fluids in the region, and geochemical evidence of crust-magma fluid interactions, suggest that a considerable fraction (⩾60%) of CO2 in Solfatara fumaroles derives from carbonate sediments in the basement. The contribution of magma-derived CO2 may be higher within the central part of the caldera (including Solfatara crater) than toward its western margin, where fumarolic and geothermal well gases exhibit lower 3He/4He ratios and still higher ∂13C values. Such a geochemical pattern is consistent with the central distribution of ground deformation and seismicity during the 1982–1984 crisis and with the idea of a residual magma body, confined beneath the central part of the structure. Alternatively, higher ∂13C and lower 3He/4He ratios toward the western margin may result from dilution of Solfatara-type gas during progressively deeper water boiling. Finally, accepting that Solfatara CO2 derives from simple crustal mixing between magmatic and sedimentary carbon, its constant isotopic composition (together with the constant He isotope ratio) would restrict the possibility of magma intrusion and/or higher magmatic gas input as mechanisms responsible for the 1982–1984 events. However, this conclusion would no more hold true if the magma itself, or even its mantle source, were previously contaminated by crustal carbon.

Oxygen isotope geochemistry of the potassic igneous rocks from the Roccamonfina volcano, Roman comagmatic region, Italy

Abstract 18O/16O measurements were made on 80 samples of lavas, pyroclastics, and xenolithic nodules from the 0.05- to 1.54-m.y.-old Roccamonfina volcano. The δ18O values of these magmas varied from +5.6 to +10.1 in the High-K Series (HKS) and +6.6 to +8.8 in the Low-K Series (LKS). The xenolithic nodules, many of which are HKS cumulates, crystallized from magmas with δ18O= +7.8to+10.3. These data indicate that assimilation of high-18O country rocks (pelitic schists, flysch, and some carbonates) was an important petrologic process during fractional crystallization of both the early, HKS magmas (leucite-bearing tephrites, leucitites, and phonolites) and the later, LKS magmas (trachybasalts, trachyandesites, and trachytes). Combined with87Sr/86Sr, Pb isotope, and143Nd/144Nd data obtained by other workers from the same outcrop localities, the18O/16O data prove that: (1)| The assimilated country rocks had relatively high87Sr/86Sr values (0.710–0.720), low143Nd/144Nd values (eNd ≈ −10to−15), and206Pb/204Pb values of about 18.70. (2)| The HKS and LKS represent completely separate trends of magmatic differentiation and different parent magmas. (3)| The HKS parent magma (leucite basanite?) probably had δ18O≈ +6.0and87Sr/86Sr≈ 0.708, while the LKS parent magma (trachybasalt?) had δ18O≈ 6.5and87Sr/86Sr≈ 0.706. (4)| The HKS and LKS parent magmas are only indirectly related genetically, perhpps in their source regions in the upper mantle, or during an earlier episode of wall rock-magma interaction in the lower crust. (5)| Although wall rock assimilation has played an important role in determining the isotopic and trace element characteristics of these magmas, the liquid line-of-descent and the changes in major element chemistry were dominated by fractional crystallization. (6)| The energy required to heat up and dissolve the assimilated rocks was provided by the latent heat associated with concurrent fractional crystallization of the cumulates (as represented by some of the plutonic nodules). (7)| The oxygen and strontium isotopic changes at Roccamonfina, although very significant, are much less than in the undersaturated magmas of the Roman Province further to the north. This is apparently because during and prior to the Quaternary potassic vulcanism, the latter terrane was invaded and heated by anatectic magmas associated with the Tuscan rhyolitic event, thus creating conditions favorable for much greater crustal assimilation.

High-18O igneous rocks from the Tuscan Magmatic Province, Italy

The 18O/16O ratios were measured for 60 rocks and coexisting minerals from the Plio-Pleistocene, calc-alkaline, Tuscan Magmatic Province, Italy. The δ18O values of these magmas were as follows: Elba, Giglio, and Montecristo granodiorites (11.4 to 12.1); M. Cimini rhyolites and trachytes (11.2 to 11.7); Roccastrada, S. Vincenzo, and M. Amiata rhyolites (12.3 to 13.4); and the Tolfa rhyolites and quartz latites (15.3 to 16.4). The latter are by far the highest δ18O values yet reported for primary volcanic rocks. The extremely high δ18O values in the Tuscan Province indicate that these magmas all formed by melting or large-scale assimilation of high-18O argillaceous sedimentary rocks, in agreement with previous evidence based on the high 87Sr/86Sr ratios and ubiquitous presence of cordierite. Low δ18O values (+5 to +6) were found in the feldspar, but not the quartz, of the small Porto Azzuro stock on Elba, as well as in the marginal facies of the M. Capanne stock, indicating that these bodies interacted with low-18O meteoric-hydrothermal fluids subsequent to solidification. The M. Cimini lavas apparently formed about 1 m.y. ago by mixing of high-18O Tuscan magmas containing large K feldspar crystals with potassic, undersaturated, low-18O magmas of the adjacent Roman Co-Magmatic Province. These large sanidine crystals did not completely equilibrate 18O/16O with the coexisting magma.

Oxygen and strontium isotope studies of K-rich volcanic rocks from the Alban Hills, Italy

18O/16O and 87Sr/86Sr ratios and major and trace element contents were measured for 33 leucite-bearing lavas from the Alban Hills, located just south of Rome. Petrologically, this volcanic center is the least complex of all the Pleistocene to Holocene volcanoes in the Roman Comagmatic Province, and the 87Sr/86Sr uniformity reflects this (0.71024–0.71091). Whole-rock δ18O = +5.6 to +9.8, but many samples were enriched in18O by post-eruption hydration, evidenced by the good correlation between H2O content (up to 5 wt. %) and δ18O. Correcting for these effects, we obtain a δ18O range of +5.6 to +7.8 for the original magmas. No other volcanic center in the Roman Province displays such uniform strontium and oxygen isotopic compositions; thus, this volcano provides special insight into the origin of the High-K Series magma end-member in Central Italy. Three groups of lavas are recognized on 87Sr/86Sr-1/Sr, δ18O-87Sr/86Sr, and K2O-SiO2 graphs; all of these groups, as well as the major caldera-forming eruption (Villa Senni Tuff), are derived from a very uniform, LIL-enriched, metasomatized source in the upper mantle. This source had δ18O = +6.5 ± 1.0 and 87Sr/86Sr= 0.71030 ± 0.00010. The δ18O values correlate positively with 87Sr/86Sr, indicating minor interaction with the continental crust. Essentially all chemical and isotopic variations in the primitive (low-SiO2, high-Ca, high-Sr) potassic lavas in Italy can be explained by mixing in the upper mantle between this Alban Hills end-member and a Low-K Series Roccamonfina-type end-member.

Evidence for crustal assimilation, mixing of magmas, and a87Sr-rich upper mantle: An oxygen and strontium isotope study of the M. Vulsini volcanic area, Central Italy

Abstract18O/16O,87Sr/86Sr and chemical analyses were made on 39 lavas and ignimbrites from M. Vulsini, the most northerly district of the K-rich Quaternary Roman Province of Italy. These rocks belong mainly to the undersaturated, leucite-bearing (High-K) series, but also included are samples from the less abundant, SiO2-saturated, hypersthene-(quartz)-normative (Low-K) series. The effects of post-eruption alteration on theδ18O of these lavas were taken into account by analyzing phenocrysts or by using the extrapolation procedure developed for the nearby Alban Hills center. Because of the high Sr contents (500–2400 ppm), the87Sr/86Sr ratios of these rocks were little affected by such alteration processes. The M. Vulsini volcanics have Sr- and O-isotopic ratios much less uniform, and on the average much higher, than at any of the other volcanic centers of the province:87Sr/86Sr=0.7097 to 0.7168;δ18O=6.5 to 13.8. This is attributable to the fact that M. Vulsini is one of the sites of greatest crustal assimilation and hybridism between K-rich Roman magmas and SiO2-rich Tuscan anatectic magmas. The High-K series parent magmas at M. Vulsini had a very high and uniform87Sr/86Sr=0.7102 to 0.7104, and a somewhat more variableδ18O=+5.5 to +7.5; they must have come from an upper mantle source region previously metasomatically enriched in87Sr and LIL elements. These18O/16O and87Sr/86Sr ratios are identical to the parent magma at the Alban Hills, 120 km to the south, where Low-K lavas are absent. Low-K series magmas at M. Vulsini originated from a lower-87Sr source region than the High-K series (<0.7097); a similar relationship is observed in all of the other localities in Italy where the two magma series coexist.

18O/16O and chemical relationships in K-rich volcanic rocks from Australia, East Africa, Antarctica, and San Venanzo-Cupaello, Italy

18O/16O analyses were made on a set of leucite-bearing igneous rocks from a variety of localities around the world. The samples chosen for study all have relatively primitive characteristics, such as low SiO2 contents (42–52 wt.%). Whole-rock δ18O values from the 10–15 m.y. old Australian leucitites range from +7.1 to +11.0; however, none of these values represents the original δ18O of the erupted lava, as every sample has undergone some subsolidus18O enrichment. The δ18O values of the primary magmas (≈ +6.5) can be accurately calculated from the δ18O of clinopyroxene mineral separates. However, the leucite mineral separates (pure, stoichiometric KA1Si2O6) have δ18O = +8.8to+10.6, and clearly have not retained their primary igneous oxygen isotopic compositions. The whole-rock δ18O values exhibit a positive correlation with H2O+ contents, and extrapolation to a plausible water content for such a subaerially erupted lava ( <0.5 wt.% H2O) gives a range of primary δ18O values of +5.8 to +6.8. The much less hydrated Bufumbira, Africa and Gaussberg, Antarctica samples have δ18O = +6.7to+7.9and+6.1to+7.0, respectively. These have probably undergone only minor subsolidus18O enrichments. The San Venanzo, Italy, samples are even less hydrated, so their extremely high δ18O values of +10.8 to +12.0 are definitely magmatic values. Those data confirm our earlier conclusions that the leucite-bearing magmas erupted in north-central Italy are unique in having much higher δ18O values than potassic rocks from southern Italy or anywhere else in the world. The northward increase in δ18O in the leucite-bearing rocks of Italy is thus clearly a manifestation of progressively greater interaction between mantle-derived magmas (δ18O = +5.5to+7.5) and the high-18O sedimentary and metasedimentary rocks of the Italian continental crust (δ18O = +15to+25), which were strongly heated during the episode of Tuscan anatectic granitic magmatism.

Marbles from Roman Hispania: stable isotope and cathodoluminescence characterization

Abstract Pure white marble has been considered a valuable ornamental and architectural material since ancient times. Many scientific techniques have been used to create an extensive data base of “finger-prints” characterizing white marbles from the major classical quarries. However, determining the provenance of white marbles is a difficult task due to their similarity in physical and chemical parameters. Three techniques (petrography, cathodoluminescence and stable C and O isotopes) have been used to characterize white marbles from the ancient quarries of the Iberian Peninsula. Maximum grain size, texture and isotopic composition can be used to identify the different quarries. Each area is generally represented by several cathodomicrofacies, but quantitative CL analysis is also helpful in distinguishing those quarries for which the data provided by other techniques are not sufficiently diagnostic. The database and the discriminating criteria presented in this study have been tentatively applied to some ancient sculptures from the National Museum of Roman Art in Merida (Spain).

40Ar-39Ar and K-Ar dating of K-rich rocks from the Roccamonfina Volcano, Roman comagmatic Region, Italy

Abstract Roccamonfina is the northernmost Volcano of the Campanian area of the K-rich Roman comagmatic Region of Italy. It erupted a huge amount of pyroclastics and lavas belonging to both the Leucite-Basanite and Leucitite Series (LBLS) and the Shoshonite Series (SS), spread over an area of about 300 km2. The above series correspond to the High-K. Series (HKS) and Low-K Series (LKS) of Appleton (1972), respectively. 40Ar-39Ar and K-Ar dating of samples from both series gave ages ranging from 0.656 to 0.096 Ma for the SS and from 1.03(?) to 0.053 Ma for the LBLS. These results indicate that the products of the two series were outpoured together at least between 0.7 and 0.1 Ma ago, i.e. during both the so-called pre-caldera phase and the post-caldera phase of activity. The latest products of the volcanism at Roccamonfina were erupted just before the deposition of the Grey Campanian Ignimbrite, which erupted from vents located about 50 km to the south in the Phlegrean Fields near Naples and has an age of about 33,000 years. Taking into account all the available radiometric data, we conclude that Roccamonfina was active between 1.5 and 0.05 Ma ago, in excellent agreement with the stratigraphie evidence. In this same time span is concentrated the activity of all the centers of the Roman Region north of Naples.

Calcite in fractures in a volcanic environment (Vulcano Island, Italy): contribution of geochronological and isotopic studies to volcanotectonics

Abstract The southern and southeastern rim of the present caldera (La Fossa caldera) at Vulcano Island (Aeolian Islands, Italy) is crosscut by a network of fractures filled with calcite and minor chalcedony. Fluid inclusion studies indicate that both mineral phases were deposited from hydrothermal solutions. Several calcite samples were analyzed and their U, Sr, Nd, O and C isotopic composition was determined. U Th geochronological data suggest two discrete depositional episodes: one at 50 ka (southern rim) and the other at 25 ka (southeastern rim). Oxygen isotopic compositions, 87 Sr 86 Sr and 234 U 238 U ratios exclude that seawater was involved significantly in the hydrothermal system, in agreement with fluid inclusion studies which did not document the presence of NaCl. The Nd and Sr isotopic composition of calcite reflects the composition of the host rocks. The oxygen isotopic composition excludes a post-depositional interaction with meteoric water and suggests a depositional temperature consistent with that obtained from the fluid inclusion study (160–170 °C), assuming equilibrium conditions and a fluid isotopic composition close to present-day steam emissions. The positive correlation between U content and C isotopic composition indicates a rapid change in CO 2 pressure during calcite deposition. This observation is consistent with a rapid fluid ascent, likely connected to two phases of collapse of the La Fossa caldera. The ages of the calcites which fill the fractures at the La Fossa caldera indicate the activation time of the fracture systems which, in turn, are connected to the above mentioned phases of the caldera collapse.