Geochemistry and volatile content of magmas feeding explosive eruptions at Telica volcano (Nicaragua)

Accepted Manuscript Geochemistry and volatile content of magmas feeding explosive eruptions at Telica volcano (Nicaragua)

P. Robidoux, S.G. Rotolo, A. Aiuppa, G. Lanzo, E.H. Hauri PII: DOI: Reference:

S0377-0273(16)30513-3 doi: 10.1016/j.jvolgeores.2017.05.007 VOLGEO 6092

To appear in:

Journal of Volcanology and Geothermal Research

Received date: Revised date: Accepted date:

17 December 2016 5 May 2017 8 May 2017

Please cite this article as: P. Robidoux, S.G. Rotolo, A. Aiuppa, G. Lanzo, E.H. Hauri , Geochemistry and volatile content of magmas feeding explosive eruptions at Telica volcano (Nicaragua), Journal of Volcanology and Geothermal Research (2017), doi: 10.1016/j.jvolgeores.2017.05.007

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Geochemistry and volatile content of magmas feeding explosive eruptions at Telica volcano (Nicaragua)

IP

T

Robidoux, P.,*,a,bRotolo, S.G.a,b, Aiuppa, A.a,b, Lanzo, G.a, Hauri, E.H.c

a

CR

Dipartimento di Scienze della Terra e del Mare (DiSTeM), Università di Palermo, Via Archirafi 36, 90123 Palermo, Italy b

US

Istituto Nazionale di Geofisica e Vulcanologia (INGV), Via Ugo La Malfa 153, 90146 Palermo, Italy. c

AN

Department of Terrestrial Magnetism, Carnegie Institution of Washington, 5241 Broad Branch Rd. NW, Washington, DC 20015, USA * Corresponding author at: Dipartimento di Scienze della Terra e del Mare (DiSTeM), Università

M

di Palermo, Via Archirafi 36, 90123 Palermo, Italy

ED

Canada Tel.: +001 450-649-7618

PT

E-mail address: [email protected] (P. Robidoux).

CE

ABSTRACT

Telica volcano, in north-west Nicaragua, is a young stratovolcano of intermediate magma

AC

composition producing frequent Vulcanian to phreatic explosive eruptions. The Telica stratigraphic record also includes examples of (pre)historic sub-Plinian activity. To refine our knowledge of this very active volcano, we analyzed major element composition and volatile content of melt inclusions from some stratigraphically significant Telica tephra deposits. These include: (1) the Scoria Telica Superior (STS) deposit (2000 to 200 years Before Present; Volcanic Explosive Index, VEI, of 2-3) and (2) pyroclasts from the post-1970s eruptive cycle (1982; 2011). Based on measurements with nanoscale secondary ion mass spectrometry, olivine-hosted (forsterite [Fo] >80) glass inclusions fall into 2 distinct clusters: a group of H2O-rich (1.8–5.2 wt%) inclusions, similar to those of nearby Cerro Negro volcano, and a second group of CO2-rich

ACCEPTED MANUSCRIPT (360–1,700 µg/g CO2) inclusions (Nejapa, Granada). Model calculations show that CO2 dominates the equilibrium magmatic vapor phase in the majority of the primitive inclusions (XCO2 > 0.62–0.95). CO2, sulfur (generally <2,000 µg/g) and H2O are lost to the vapor phase during deep decompression (P >400 MPa) and early crystallization of magmas. Chlorine exhibits a wide concentration range (400–2,300 µg/g) in primitive olivine-entrapped melts (likely suggesting variable source heterogeneity) and is typically enriched in the most differentiated

T

melts (1,000–3,000 µg/g). Primitive, volatile-rich olivine-hosted melt inclusions (entrapment

IP

pressures, 5–15 km depth) are exclusively found in the largest-scale Telica eruptions

CR

(exemplified by STS in our study). These eruptions are thus tentatively explained as due to injection of deep CO2-rich mafic magma into the shallow crustal plumbing system. More recent

US

(post-1970), milder (VEI 1-2) eruptions, instead, do only exhibit evidence for low-pressure (P <50–60 MPa), volatile-poor (H2O <0.3–1.7 wt%; CO2 <23–308 µg/g) magmatic conditions.

AN

These are manifested as andesitic magmas, recording multiple magma mixing events, in pyroxene inclusions. We propose that post-1970s eruptions are possibly related to the high

M

viscosity of resident magma in shallow plumbing system (<2.4 km), due to crystallization and

AC

CE

PT

ED

degassing.

ACCEPTED MANUSCRIPT 1. Introduction Telica is one of the most active open-vent degassing volcanoes of basaltic-andesite composition in the Central American Volcanic Arc (CAVA). Seismic and degassing activity have been closely examined as the volcano recently (during the 2000s) entered a phase of intense Strombolian to Vulcanian activity, and minor phreatomagmatic events (Geirsson et al., 2014;

T

Rodgers et al., 2015). Understanding evolution of the ongoing volcanic unrest is key to improved

IP

volcanic hazard assessment. Several large-scale explosive eruptions, ranging from Vulcanian,

CR

sub-Plinian to Plinian, and phreatomagmatic, have been identified in the Telica stratigraphic record (Navarro, 1994; Havlicek et al. 1999, 2000; Novák and Přichystal, 2006), with recurrence times similar to that of the Masaya complex (40,000 – 20,000 to less than 2,000 yrs; Navarro,

US

1994). A similar paroxysmal event could thus occur again in the future, potentially impacting all the north-west Nicaragua. The last of such large scale events (Volcanic Explosive Index, VEI =

AN

3) occured in 1529 (0.1 km3; reaching 6 km from the vent; Navarro, 1994; GVP). The best preserved and widespread deposit clearly identified in the field is attributed to the Scoria Telica

M

Superior (STS), the youngest sequence of the Telica Superior Group (VEI ~2-3; Hradecký et al.,

ED

2007).

Regional geological mapping studies in the area were conducted by the Czech Geological

PT

Survey – CGS (Hradecký et al., 2007). Havlicek et al. (1999, 2000) examined stratigraphy and bulk rock compositions of Telica pyroclastic deposits, assessing the dispersion area of STS and

CE

post-1970s explosive activity, and identifying the key compositional features of Telica magma (and older centers as well). Lava and scoria samples erupted from different emission centers, all

AC

contemporaneous to the present-day Telica complex, helped to define the compositional range (from basaltic to basaltic andesite; see CAVA database in Carr et al., 2014). The regional studies of Carr (1984), Patino et al. (2000) and Bolge et al. (2009) highlighted the peculiar geochemical signature (trace elements and REE) of Telica (and north-west Nicaragua) magmas, within the CAVA volcanism. More recently, Sadofsky et al. (2008) and Wehrmann et al. (2011) discussed regional patterns of major/trace/REE compositions and volatile abundances of CAVA melt inclusions, but they included only one sample from Telica in their datasets. Therefore, our current knowledge of pre-eruptive volatiles in Telica magmas is incomplete, and prevents us from obtaining a better understanding of the volcano’s plumbing system.

ACCEPTED MANUSCRIPT We present here a first comprehensive melt-inclusion characterization of pre-eruptive volatiles in Telica magmas, which we use to yield novel information on the volcano’s magmatic plumbing system. Volatile contents in melt inclusions, combined with results of equilibrium saturation models, are used to numerically reproduce magmatic degassing trends, and therefore yield conditions of magma storage, ascent and decompression. This novel information is finally integrated in the attempt to better constrain the explosive character of Telica volcano. We discuss

T

the potential processes responsible for triggering the explosive eruptions, with special focus on

IP

the role played by the liquid melt composition, dissolved volatile contents and crystallization on

CR

magma viscosity. Our results contribute an improved knowledge of eruptive dynamic for the

US

large STS eruptive sequence and smaller explosive eruptions from the post-1970s period.

AN

2. Regional setting

M

2.1 Telica volcano in the Cordillera Los Maribios

Telica is a 1,061m a.s.l. high basaltic-andesitic stratovolcano located at 12°60’ N and

ED

86°85’ W in the central part of the Cordillera Los Maribios, 20 km east of the Pacific Coast (Havlicek et al., 1999) and extending along the Pacific coast-line between Guatemala and Costa

PT

Rica (Fig. 1). This Cordillera is composed of a northwest-southeast-trending chain of volcanoes within the Nicaragua Depression, interpreted as the result of northeasterly Cocos Plate’s

CE

subduction beneath the Caribbean Plate (LaFemina et al., 2009). The volcanic chain includes (from north-west to south-east) San Cristóbal/Casita, Telica, La Rota, Cerro Negro, El Hoyo,

AC

Asososca and Momotombo (Mc Birney and Williams, 1965; Walker et al., 2001). Telica stratovolcano is part of the Telica Complex, which is composed of a series of voluminous pyroclastic deposits (areal extension > 81 km2; Navarro, 1994) and surrounded by extinct satellite cones from the Plio-Pleistocene (Fig. 1). In a geochronological scale, the Telica Complex encloses the (i) Group of Pre-Telica Caldera, (ii) the group of satellite volcanoes, and (iii) the Group of Telica (Havlicek et al., 2000). Stratigraphic units include pyroclastic flows, phreatomagmatic deposits and thick fallout deposits covering the flanks and superior sections of the volcano. The caldera shaped morphology could represent the remnant of the Pre-Telica

ACCEPTED MANUSCRIPT structures and testify for larger sub-Plinian to Plinian caldera-forming eruptions (Navarro, 1994; Havlicek et al., 2000). 2.2 Group of Telica According to Havlicek et al. (2000), the Group of Telica is composed of five sequences: the

T

Scoria Telica Valles (STV), the Phreatomagmatic Sequence (FS) (minimum age, 2,150 +/- 150

IP

yrs B.P.), the Scoria Najanjada Sequence (SNS), the Cañadas Sequence (CS) (younger than 2,300 yrs B.P.; Hradecký et al., 2000) and the Scoria Telica Superior (STS). The whole group is

CR

illustrated in figure 1. The group is dominated by thick scoria sequences (as a whole at least 23 m thick), covered by alluvial deposits, and interbedded with altered tuffs (e.g.,the Caňada

US

Sequence), and thin altered scoriae and block layers resulting from phreatomagmatic activity

AN

(Havlicek et al., 2000). 2.3 Scoria Telica superior (STS)

M

The most recent member of the Group of Telica is the Scoria Telica Superior (STS), a deposit

ED

formed by a Vulcanian to sub-Plinian eruption covering the area between the summit Crater and Las Colimas (Fig. 1). STS is composed of six scoriae layers separated by thin laminations of beige tuff layers, ranging in thickness from millimeters to few centimeters. Depending on the

PT

locality, the basal sequence is overlying the Phreatomagmatic Sequence (FS), the Scoria Najanjada Sequence (SNS), or the Cañadas Sequence (CS), with a sharp erosional contact. The

AC

al., 1999, 2000).

CE

deposit is therefore younger than 2,150 +/- 150 yrs B.P. (the age of the FS deposit; Havlicek et

Upper STS sequences have been associated to sub-Plinian eruptive style (Navarro, 1994) using the Walker et al. (1981) classification scheme. This classification also concords with the asymmetric elliptic area of tephra dispersion and the thick horizontal fallout deposit that testify sustained eruptive clouds at high altitudes (Havlicek et al., 2000; Sigurdsson et al., 2015).

2.4 Observed historical activity at Telica

ACCEPTED MANUSCRIPT At least 10 important eruptions of Vulcanian/Strombolian style have been recorded at Telica since 1527 (Novák and Přichystal, 2006). The 1982 eruption produced a 3.7-4.3 km high column, and the related volcanic ash covered an area extending from León to Corinto (Stoiber et al., 1982; Navarro, 1994; Novák and Přichystal, 2006). The 1982 eruption impacted the western and northern flanks of Telica, where according to reports from Dartmouth College geologists (GVP), angular blocks and bread-crusted bombs of andesite-basaltic composition and metric size were

T

deposited (Novák and Přichystal, 2006). Since then, several explosions of Volcanic Explosive

IP

Index (VEI) between 0 and 2 have occurred at the summit producing highly fragmented and

CR

altered material, mostly related to small and moderate phreatic explosions which were associated

US

to an increase of seismic activity (Rodgers et al., 2013, 2015; Geirsson et al. 2014).

AN

3. Material and methods

M

3.1 Field work

ED

The recent Telica activity is investigated from lithic and juvenile fragments collected in the present active crater area (Fig. 1). The 1982 eruption (samples TEL01, TEL02) was Vulcanian in nature and erupted pyroclasts of variable vesicularity including proximal metric size

PT

blocks and bread-crusted bombs (see also Navarro, 1994). In 2011, frequent short-lived jet explosions have been interpreted as evidence of a phreatic eruptive episode and those produced

CE

various pyroclastic fragments (this study; Tenorio, 2011d; Geirsson et al., 2014). Some of the lithic fragments (TEL03), erupted during the Telica eruption(s) in 2011, were collected during the

AC

field work. They have vesicular textures and are not affected by rain (samples TEL03A, TEL03B; Fig. 1; Robidoux, 2016). Large bombs and lapilli-sized scoriae from the 12-20 February 1982 explosions probably represent the only remaining fresh juvenile materials from recent eruptive periods. Several of these bombs were sampled; TEL01 is an elliptic-shaped bomb with dimensions of 82x52cm, with bread-crusted fractured layers aligned parallel to the long axis. TEL02 has similar shape and structure, and impressive dimensions (392x195cm). Rock samples were also taken from some of the most recent large-scale eruptions recorded in the stratigraphic sequence. Stratigraphic columns are described in Appendix 1, while

ACCEPTED MANUSCRIPT mineral abundance and mineral chemistry shown in Appendix 2. Scoriae from the STS (Group of Telica, GT), were collected at ~5.5 km west from the crater. STS sequence is covered by a soil layer, and is composed of layers of scoriae and tuff (Havlicek et al., 2000), these latter interpreted here as lahar deposits (given their resistant and compact nature, and the presence of small porous textures with frequent oxidized organic imprints and fossils). Depending on the locality, the thickness and presence of such reworked fluvial layers and lahar deposits can vary. No charcoal

T

or burned pieces of wood were found for dating, and attempts to date oxidized leaf imprints with

CR

18th century, at the top of the sequence (Appendix 1a-b-c-d).

IP

accelerator mass spectrometry (AMS) gave a minimum approximated limit corresponding to the

The stratigraphic section TEL04 corresponds to different parts of the lithostratigraphic

US

column STS (Scoria Telica Superior) of Havlicek et al. (1999) (see Appendix 1d). At the bottom, TEL04B (Appendix 1c) is made of coarse ash with 3 thin layers. The STS section continues with

AN

a fluvial deposit eroding the top of TEL04C. This is followed by several reworked fine tuff layers, cutting off and separating several distinct layers of fine to coarse ash and lapilli-size

M

scoriae. These tuffs, here interpreted as lahars, probably represent several single events or rainy seasons, and therefore make geochronology difficult to distinguish between each scoriae layer

ED

(Appendix 1a).

PT

In a different STS outcrop, ~4 km west from the crater of Telica, an aa lava flow unit from the same STS group is also encountered. This lava flow (sample TEL07E; Appendix 1b),

CE

clearly identified in Havlicek et al. (1999), serves as the best stratigraphic marker of the entire sequence, because it shows a direct contact with scoriae from STS’ pulse III. This section

AC

includes layer TEL07A, which is made of medium lapilli and coarse ashes (Bevacqua, 2014). The last studied layer is TEL04K and it corresponds to the pulse V among the whole STS sequence which is composed of coarse to medium lapilli and medium ashes. Textural observations (Appendix 1-2) show that each STS layer consist of a normally graded scoriae bed, with very few (<10%) lithic fragments, and total metric grain size distribution (TGSD; Logd ~ -2 to -3; Bevacqua, 2014) is in the range of Vulcanian-violent Strombolian deposits (Rust and Cashman, 2011). Alternatively, the whole STS group (total volume, >0.31 km3) could have formed after a single, long-lasting sub-Plinian eruption, composed of a sequence of explosive pulses (e.g. Navarro, 1994; Havlicek et al., 2000). The STS deposit(s) would rank as sub-Plinian based on

ACCEPTED MANUSCRIPT analysis of the few known outcrops (in the literature and this study; Fig. 1) using the Pyle et al. (1989) method for tephra volume estimation. Single «pulses» III and IV (0.07 and 0.06 km3) are already >3 times more voluminous than the Cerro Negro sub-Plinian eruption in 1992 (0.026 km3; Hill et al., 1998; Connor and Connor, 2006). Such rough estimation is determined after measuring layer thickness in Havlicek et al. (2000) and an approximate area of deposition (Fig.

IP

T

1), but this exercise require future extensive field sampling of STS.

CR

3.2 Petrography and mineral chemistry

US

Whole-rock major element compositions were analyzed for pyroclasts and lava by wavelengthdispersive X-ray fluorescence (XRF) (Rigaku ZSX Primus) at DiSTeM (Univ. of Palermo)

AN

(Table 1)(standards WS-E and JB-1a available in Robidoux et al., 2016) on pressed powder pellets.

M

Olivine, clinopyroxene and glass fragments were handpicked from ash and scoriae

ED

samples of Holocene deposits (TEL04, TEL07) for inspection and preparation for melt inclusion study. Careful mineral selection was performed on the 0.5–2.0 mm sieve fraction of crushed bombs and scoriae from post-1970s eruptions (TEL01 – 2 – 3) for spherical and glassy inclusions

PT

only. In order to avoid effects from differential cooling rates and hydrogen lost during melt inclusion formation (Gaetani et al., 2012; Loyd et al., 2013), all those minerals hosting glass

CE

inclusions with daughter minerals and particular phases/textures linked to post-entrapment effects (bubble, oxides, microlites, etc.) were discarded for inclusion analysis. Plagioclase phenocrysts

AC

are found along all tephra and a small fraction of amphibole phenocrysts (traces to 4 vol %) are only present in STS Holocene deposit (Appendix 2). Minerals were prepared on Crystal BondTM resin, then polished over one side for glass fragments and glass inclusions intersection. Major elements, Cl, and S contents in glass inclusions, groundmass glass, glass phenocryst rims and crystals were determined using a JXA-8200 WD/ED combined electron microprobe (EMPA; at INGV-Roma) with instrumental settings explained in Robidoux (2016). Table 2 lists average matrix glass compositions, while Table 3 presents EMPA compositions of 41 glass inclusions from 13 different olivine crystals and 20

ACCEPTED MANUSCRIPT clinopyroxenes. Additional 50 EMPA spots were performed on cores and rims from 18 olivine phenocrysts (Appendix 3). 27 spots on 17 clinopyroxenes were performed including SEM-EDS runs on sample TEL02 and TEL03A that were not studied for their inclusions, because the trapped inclusions are heavily crystallized (Appendix 4). CO2, H2O, S and Cl concentrations (Table 4) were determined using a Cameca NanoSIMS 50/50L ion probe at the Carnegie Institution of Science with preparation, information on standards and instrumental setting

IP

T

explained in details in Robidoux (2016).

CR

The compositions of melt inclusions in olivines were corrected for post-entrapment crystallization (PEC) and for Fe-loss by using Petrolog3 code (Danyushevsky and Plechov, 2011). We used an equilibrium Fe-Mg distribution coefficient between olivine and liquid (KdFeof 0.30 (Toplis, 2005), adding incrementally the olivine composition to the melt inclusion

US

Mg ol-liq)

until the equilibrium Kd value (0.30) with the host olivine was reached (Fig. 2). In order to better

AN

compare with other data set, lines representing different KdFe-Mgol-liq (0.27 and 0.33), were also calculated. The melt Fe3+/∑Fe ratio was fixed at ~ 0.24, as derived at the Ni–NiO buffer (NNO)

M

(Kilinc et al., 1983), commonly considered the most realitsic fO2 for arc magmas (e.g. Jugo et al.,

ED

2009), at the temperature calculated according to the Sugawara (2000) geothermometer. The major elements concentrations, after PEC were then recalculated on a 100% volatile

PT

free basis. PEC corrections depended on a Fe2+/Fe3+ ratio based on Kilinc et al. (1983) model for each inclusion, but the major element concentrations were recalculated in Petrolog on a volatile

CE

free basis with FeOT, as total FeO taking into account the average Fe3+/∑Fe ratio of the corresponding layer. Uncorrected values are listed in Appendix 5 with their mineral host

AC

compositions. Mineral standards are shown in Appendix 6. Similarly, melt inclusions hosted in clinopyroxene phenocrysts were corrected for PEC with an Excel custom made spreadsheet, adding clinopyroxene component to the melt inclusions until the equilibrium KdFe-Mgcpx-liq = 0.26, was reached , at an fO2 = NNO.

4. Results

ACCEPTED MANUSCRIPT 4.1 Petrography and mineral chemistry The studied rocks are variably porphyritic. Scoriae fragments from STS deposit are dark to grey in colour, highly vesicular, and show mineral assemblage similar to the rest of our samples in thin section. All the studied rocks are variably porphyritic with phenocryst abundance as high as 40– 45 vol % phenocryst (plagioclase > clinopyroxene > olivine > magnetite; Fig. 3; Appendix 2)

T

with dominant equigranular plagioclase microphenocrysts (>5–12 mm) that show frequent

IP

twinning intergrowth textures (Fig. 3b). TEL04K and TEL07A have similar characteristics (see

CR

textural details in Bevacqua, 2014), but TEL04B has larger olivine phenocrysts (>4–12 mm) and also clinopyroxene phenocrysts (~10 mm vs 8 mm in other scoriae).

US

Bombs TEL01-TEL02 from the February 1982 eruption are porphyritic and contain large plagioclase phenocrysts (up to 15 mm in length). Phenocryst abundance varies from 6 vol % in

AN

the quenched rims to more than 40 vol % in the core of the bombs. Phenocryst assemblages are dominated by clinopyroxene over olivine (Fig. 3a). Fragments are dark and frequently highly

M

vesicular and glassy along the external parts of the bombs, and the groundmass is generally microcrystalline. Samples from the May 2011 eruption (TEL03A, TEL03B) are highly vesicular

ED

(vesicles >60 vol %), crystal-poor scoriae, with similar mineral assemblage to TEL01-TEL02. Olivine, clinopyroxene and plagioclase phenocryst chemistry data are illustrated in figures 4-5

CE

PT

and described in details with Appendix 2.

AC

4.2 Whole rock geochemistry Whole-rock compositions range from basaltic to andesitic (SiO2 = 50.9–58.6 wt%; K2O= 0.84– 1.43 wt%) and plot in the field from calc-alkaline to HK-CA rocks (Fig. 6a). The most primitive sample is the lava flow TEL07E, with 50.9 wt% SiO2 and MgO= 6.02 wt% (Table 3) with relatively low K2O (0.82 wt%). This lava is similar in composition to the Telica basaltic lava flows available in the literature (50.8–53.5 wt% SiO2; 0.9–1.1 wt% K2O; e.g. Patino et al., 2000; Heydolph et al., 2012; Carr et al., 2014).

ACCEPTED MANUSCRIPT The STS scoriae samples have variable compositions, in the range of SiO2= 50.9–55.5 wt% , and are therefore more evolved than the associated lava flow (TEL07E). The metric-sized bombs emitted during February 1982 (TEL01-TEL02) are basaltic andesites, with 1.0% K2O. Lithic lapilli-sized fragments emitted during May 2011 (TEL03A, TEL03B) are more evolved andesites, with SiO2=58.3–58.6 wt% and total alkalis between 4.0 and 4.2 wt%. They appear slightly more evolved than the juvenile-poor ashes analyzed during the same period by Geirsson

T

et al. (2014) (SiO2= 55.6–57.6 wt%; K2O =0.9–1.2 wt%). Scoriae and ash from the post-1970s

CR

IP

period also spread between basalts and andesites (SiO2= 49.6–57.6 wt% ; K2O = 0.9–1.3 wt%). STS bulk rock samples are richer in MgO (4.0–6.2 wt%) than pyroclasts of the post-1970s period (2.0–4.4 wt%) (Fig. 6e). Similar bulk-rock compositions are observed for STS and 1982

US

eruption products for CaO (9.8–11.0 wt%), Al2O3 (17.0–18.2 wt%), TiO2 (0.8–1.6 wt%) and FeOT (8.4–10.0 wt%) (see Figs. 6c and d-f). Samples from the May 2011 eruption fall outside

AN

this group. P2O5 concentrations show positive dependence on SiO2 (Fig. 6b), while CaO, Al2O3,

ED

M

MgO and TiO2 are negatively correlated with SiO2 (Fig. 6c-d-e).

4.3 Glass inclusions and matrix glass composition

liq

PT

Corrected melt inclusion compositions (Fig. 2; Table 3) generally follow the modeled KdFe-Mgxtal= 0.30 trend for olivine in all cases except for three samples with extreme MgO contrast (i.e.

CE

low MgO content in glass inclusions and, Fo% in the host olivine, or vice-versa). In five melt inclusions, important corrections ( > 2 % olivine, to a maximum of 9 %) were applied

AC

(Danyushevsky and Plechov, 2011). The inclusions were preserved in moderate to Mg-rich olivines and all of them (15) had original FeOT <0.2–2.8 below the corresponding host rock FeOT vs MgO trending, so they were also corrected for Fe-loss by using Petrolog3 (Fig. 2). The glass inclusions were also preserved in moderate to Fe-rich clinopyroxenes. Several (15/31) inclusions had FeOT >0.1–2.4 %, or higher than their host rocks FeOT vs MgO trending, while the rest (16/31) had FeOT <0.1–7.5 %.

ACCEPTED MANUSCRIPT Fifteen over 31 inclusions in clinopyroxene needed >2% correction to re-equilibrate their FeOT vs MgO concentrations for PEC. Pyroxene glassy rims plot along the minimum Mg# liquid vs Mg# crystal array (Kd = 0.24). Glass wetting olivine phenocryst rims and olivine rims have a KdFe-Mgxtal-liq = 0.27, in the range of equilibrium Kd as explained above. Recalculated compositions from glass inclusions in clinopyroxenes are rich in K2O (0.8–

T

4.0 wt%), and they are classified mostly as andesites, while olivine-hosted melt inclusions are

IP

mostly basaltic (Fig. 6a), with a K2O = 0.2–0.9 wt%. Glass inclusions are SiO2-richer in

CR

pyroxenes (55–66 wt%) than in olivines (47–52 wt%). Similarly, olivine and pyroxene-hosted inclusions show distinct P2O5 populations (0.03–0.41 wt%; average 0.19 wt%). Matrix glasses have high K2O (>1.8 wt%) and SiO2 (>54 wt%) compositions (Table 2). STS olivines (TEL04B,

AN

US

TEL04K) trapped inclusions of basaltic compositions.

M

4.4 Volatile contents

ED

The highest water contents (determined by NanoSIMS; Table 4) in the olivine-hosted melt inclusions were obtained for TEL04B samples (H2O= 5.0 ± 0.2 wt%), while TEL04K series were somewhat H2O-poorer (between 1.58 to 4.07 wt%). Two inclusions in one pyroxene crystal

PT

(TEL01) were also quantified (H2O= 0.25 and 0.46 wt%).

CE

Olivine phenocrysts of sample TEL04K contain the inclusions with the highest CO2 contents (36–1,642 µg/g). In comparison TEL04B has low-to-moderate CO2 contents (329–360

AC

µg/g). CO2 contents were also determined by NanoSIMS in clinopyroxene-hosted inclusions from TEL07 (56 to 98 µg/g CO2) and TEL04K (5 to 309 µg/g CO2). CO2 contents are lower in glass inclusions of TEL01, but in the same pyroxene the fully enclosed melt inclusion has 65 µg/g (a reentrant inclusion, in a crystal embayment, has 17 µg/g). Melt inclusions in olivine and pyroxene contain variable S and Cl contents (Table 3 and 4), and yielded consistent analytical results when measured by both EMPA and NanoSIMS (Robidoux, 2016). A larger quantity of EMPA results were obtained and much less for NanoSIMS considering that many crystals were lost during manipulation before their insertion

ACCEPTED MANUSCRIPT on indium mounts, required for the latter method. For this reason EMPA S-Cl data are used here for. Olivine-hosted TEL04B inclusions contain 780–1,020 µg/g S (mean 913 ± 78 µg/g S), or twice as much as that measured in inclusions in pyroxenes (S= 196–977 µg/g ; mean 502 ± 310 µg/g). In TEL04K, olivine-hosted inclusions have S contents between 810 and 1,253 µg/g (mean

T

1,213 ± 220 µg/g S), with the exception of the extremely high values of TEL04Kol3 (1,985 µg/g

IP

S). Pyroxene-hosted inclusions of TEL04K have S contents between 72–705 µg/g (mean 272

CR

±176 µg/g). Low S contents were detected in TEL01 (332 µg/g) and TEL03B (695 µg/g S). Overall, the primitive inclusions trapped in olivines of TEL04K (lowest wt% SiO2) contain higher sulfur contents than more evolved inclusions trapped in pyroxenes. For comparison,

US

matrix glasses contain S= 140–216 µg/g .

AN

TEL04B inclusions contain Cl= 875–1,754 µg/g (in olivines) to 1,285–2,440 µg/g (in pyroxenes). In olivine phenocryst TEL04B2 there is a clear core to rim increase in Cl: the glass

M

inclusion closest to the core has Cl= 875 µg/g, the intermediate core-rim (1,133 µg/g) and the 2 inclusions in proximity of the phenocryst rim have the highest chlorine contents (1,542 to 1,754

ED

µg/g) which is defining a trend of increasing chlorine concentration in the melt, parallel to the olivine growth. In olivine phenocrysts of the sample TEL04K, the 8 inclusions have Cl= 468–

PT

1,906 µg/g (mean 1,467 ± 472 µg/g), while pyroxenes have between Cl= 1,785 to 2,900 µg/g (mean 2,378 ± 509 µg/g). For glass inclusions in pyroxenes of TEL01, EMPA reveals an average

CE

chlorine content of 3,030 ± 108 Cl (n = 3). In TEL03B, one inclusion has 4,285 µg/g Cl. Matrix glasses chlorine contents were found within Cl= 1,480 ± 50 µg/g in sample TEL04K and Cl=

AC

1,791 ± 256 µg/g in sample TEL03A.

5. Discussion 5.1 Magma composition preserved in olivines and pyroxenes Volatile contents in melt inclusions from olivine and pyroxene phenocrysts vary widely. Olivine phenocrysts are systematically higher in Mg/Mg+Fe than pyroxenes (Fig. 4a-b, 7a),

ACCEPTED MANUSCRIPT except for a few Mg-poor olivines in TEL03 fragments. In view of our crystallization models (below), these olivines are interpreted as formed in evolved melts resulting from late crystallization steps. Both olivines and pyroxenes are found to trap melt inclusions slightly less differentiated than the corresponding bulk rocks (e.g., with higher CaO/Al2O3) (Fig. 7a; Fig. 8a-c; Fig. 9a-b). Finally, melt inclusions hosted in olivine phenocrysts are generally less degassed (higher S/Cl) than those in pyroxenes in any of the study samples (Fig. 7c), hence they represent

IP

T

the best candidates to probe parental melt compositions at Telica.

CR

One significant observation is that at least two inclusion groups, with distinct volatile contents, have been trapped in Telica olivines. One group has moderate water contents (H2O= 1.5-2.0 wt%), but is high in CO2, with maximum detected CO2 content >1,700 µg/g, similarly to

US

those measured in the CO2-rich, primitive magmas of Central Nicaragua volcanoes (Granada and Nejapa; Wehrmann et al., 2011). A second group of inclusions is water-rich (H2O= 4.0–6.1 %),

AN

similar to the hydrous magma compositions seen at nearby Cerro Negro (Roggensack et al.,

M

1997; Portnyagin et al. 2014).

The solubility model of Papale et al (2006) was used to estimate the entrapment pressure

ED

conditions of both groups of inclusions, and to calculate the equilibrium compositions of coexisting melt and vapour phase along inferred degassing paths (Fig. 8a). Two distinct

PT

degassing paths were calculated (in either closed or open system conditions), using different initial conditions. A first degassing path, representative of the population of CO2-rich inclusions,

CE

is initialised from composition of TEL04Kol2a inclusions (H2O= 1.7 %, CO2= 1,583 µg/g CO2), from an initial pressure of 164 MPa (at 1110 °C temperature) (Fig. 8a). The second path is

AC

initialized with inclusion TEL04B1 (H2O= 5.2 %, CO2= 360 µg/g), with an initial pressure fixed at 198 MPa (at T = 1143 °C). The model degassing path could even be started at a maximum pressure of ~400 MPa (the entrapment pressure of a hypothetical inclusion with the highest water (5.2 %) and CO2 (1,641 µg/g) content in our dataset), but an identical degassing trend would be obtained. The simulated degassing paths, although poorly constrained by the limited number of inclusions with CO2 contents available, reasonably fit the natural volatile contents, including the volatile-poor melts recorded by pyroxenes (Fig. 8a). It remains unclear, given the limited data

ACCEPTED MANUSCRIPT available, if the presence of two melt inclusion clusters means that two distinct magmas, with remarkably different H2O contents, have been feeding the Telica plumbing system during the STS phase. The Telica water-rich degassing path shows similarities with that exhibited by melt inclusions from Cerro Negro, that are similarly water-rich (>6.0 % H2O; Roggensack et al. 1997; Sadofsky et al., 2008; Wehrman et al., 2011; Portnyagin et al., 2014). The H2O vs. CO2/S plot (Fig. 8b) confirms this observation, and also accounts for melt differentiation using Mg#

T

isopleths. The water-poor Telica cluster of inclusions mimics the degassing trend of Granada-

IP

Nejapa’s CO2-rich inclusions (Sadofsky et al., 2008; Wehrmann et al., 2011), but starting at

CR

lower pressures (<164 MPa). These low magmatic water contents (Fig. 8a-b) imply either (i) a water-poor magma source, or (ii) disequilibrium degassing due to delayed bubble nucleation in

US

the magma (Gonnerman and Manga, 2005; Pichavant et al., 2013), or (iii) magma fluxing by CO2-rich bubbles from a deep magmatic source (Blundy et al. 2010). We cannot exclude that the

AN

H2O-poor inclusions may also have suffered post-entrapment protonic diffusion (i.e. loss of H+) to the olivine host (Gaetani et al., 2012). Finally, we exclude that magma mixing might have

M

played a role at the olivine-melt trapping conditions, given the similarity of major element and SCl compositions of the two melt inclusion groups.

ED

Sulfur contents are higher in inclusions in olivines than in pyroxenes (Fig. 8c), which supports the late-stage formation of the latter mineral. In addition to distinct entrapment

PT

pressures, olivines and pyroxenes may eventually record dissimilar redox conditions (fO2 is well known to severely affect S solubility; Moretti and Papale, 2003). Pyroxenes demonstrate a

CE

variety of melt inclusion formation conditions, as seen by their textures and variable compositions. The low S content in pyroxene-hosted melt inclusions is associated with a

AC

generally low CaO/Al2O3 ratio (Fig. 9a), confirming that pyroxenes record in these magmas the shallow crystallization and degassing conditions. However, the crystallization sequence is composition-specific: in other high potassic-calc-alkaline arc-basalts, clinopyroxene is an early liquidus phase (Di Carlo et al., 2006; Lanzo et al. 2016), hence suitable to track the early crystallization-degassing magma history. Another possibility is that inclusions in pyroxenes keep record of volatile-depleted melts simply due to diffusive H2O-loss, which is known to depend on distinct cooling histories of differently sized bombs and lapilli (i.e. on their thermal inertia, Lloyd et al., 2013).

ACCEPTED MANUSCRIPT In contrast, Cl is enriched in the most differentiated inclusions in a single crystal (Fig. 3c), and increases with decreasing CaO/Al2O3 ratios, i.e. with magma differentiation (Fig. 9b). The contrasting degassing behavior of the two elements (Signorelli and Carroll, 2000; Spilliaert et al., 2006) can be tracked from the melt inclusions’ S/Cl ratios (Robidoux, 2016). This shows that melts trapped in pyroxenes are even Cl-richer (and with lower S/Cl ratios) than recorded by matrix glass; clearly, substantial S loss has occurred before trapping (Stix et al., 1993) and Cl

CR

IP

T

enrichment is pointing out to increasing effects of crystallization (Fig. 8c).

5.2 Steps in magma differentiation

US

The trapped melt inclusions of both olivine and pyroxenes exhibit negative dependences of Al2O3 and CaO contents on SiO2 that are likely caused by fractional crystallization (Fig. 6c-d).

AN

To illustrate these trends, we schematically identify in figures 10a two crystallization steps, STEP

M

1 and STEP 2.

Crystallization steps are initially graphically obtained by simply connecting together the

ED

compositional end member compositions of some selected samples (Figs. 10a); these compositional tie-lines therefore only have a qualitative (illustrative) nature. To provide more

PT

support to our arguments, we additionally quantitatively modeled steps of fractional crystallization using MELTS (Gualda and Ghiorso, 2015 and references therein). The starting

CE

point of the MELTS simulations was fixed using the primitive melt inclusion composition of sample TEL04B1 and maximum (hypothetical) volatile content recorded with NanoSIMS in all

AC

olivine-hosted melt inclusions (H2O = 5.2 wt%; CO2 = 1,642 µg/g), which would correspond to an entrapment pressure of ~400 MPa with the solubility model of Papale et al. (2006) (Fig. 10a). Importantly, this pressure is much higher than the saturation pressures derived from solubility models applied to the measured MI compositions (P= 164–198 MPa), which in turn are in good agreement with T and P constraints placed by clinopyroxene thermo-barometers (Putirka et al., 2008). These latter indicate a range of pressure from 103 to 198 MPa, and temperatures of 1090 to 1153 °C. By using an initial pressure of 400 MPa, we implicitly target the deepest parts of the plumbing system, eventually not recorded by the melt inclusions/crystal record.

ACCEPTED MANUSCRIPT STEP 1 is represented by the tie-lines connecting compositions of (i) the most primitive PEC melt inclusion composition (average of TEL04B) and (ii) the bulk rock composition of TEL04B (end of arrow Fig. 10a), representing the residual liquid. Several data-points cluster along this STEP 1 vector. To validate the simple vector illustration, we used MELTS with melt inclusion sample TEL04B1’s composition to simulate the magma crystallization path in a K2O vs MgO wt% diagram (Fig. 10a), and determined that measured melt inclusion compositions are

T

best reproduced by a first stage of fractional crystallization until a pressure of 150 MPa is

IP

reached, and fractionating the following cumulative mineral percentages: 24.8% clinopyroxene,

CR

22.9 % plagioclase, 10.5 % olivine, 2.4% magnetite (liquid= 39.4 %). We caution that phase equilibria experiments would be required to validate our 400–150 MPa pressures, high water

US

conditions for STEP 1, and specifically to test if crystallization at mid-crustal levels (where hydrous mineral stability is expanded) is consistent with the few observed amphibole

AN

phenocrysts. We stress amphiboles may also form during near-solidus crystallization in cooler portions of the magma reservoir (Bolge et al., 2009; Simakin et al., 2009).

M

Several glass inclusions in pyroxenes are more evolved than STEP1 stage would predict. To explain this, a second crystallization stage (illustrated by STEP 2 vector) is implicated (Fig.

ED

6`; Fig. 10a). In this specific case, we use the bulk rock composition of TEL03B as a proxy for the daughter liquid. By following the trendline connected until this daughter endmember,

PT

MELTS predicts crystallization evolving down to 50-60 MPa pressure, with the following mass fractions of fractionated minerals: 26.4% clinopyroxene - 37.5 % plagioclase - 10.6 % olivine -

CE

3.6% magnetite – (liquid= 18.0 %). This crystallization step successfully reproduces the highest P2O5 and K2O enrichments (Fig. 6a-b), and also explains the relatively higher modal vol %

AC

plagioclase distribution in pyroclasts from the post-1970s period (Appendix 2). Overall, the different tephra samples analyzed at Telica, including the most recent products

of

post-1970s

activity,

record

similar

evolving

stages

of

magma

decompression/crystallization, as demonstrated by their overlapping crystal (Fo% content), bulkrock (CaO/Al2O3), and volatile (S/Cl) compositions (Fig. 9a-b-c). The scatter plots of major elements vs silica content (Fig. 6) of bulk rock shows that lava samples from Telica (this study and literature sources) also share the same evolutionary trend defined by pyroclastic deposits. Even the most evolved melt inclusions (hosted in augite phenocrysts), reaching andesitic to

ACCEPTED MANUSCRIPT dacitic compositions (Ibid. to andesitic melt inclusions in olivines at Cerro Negro, Portnyagin et al., 2014, and/or dacitic melt inclusions found in pyroxenes at Cosigüina volcano, Longpré et al., 2014), can be explained by polybaric fractional crystallization behaviors in two generalized depth zones (Section 5.3). Similar crystallization processes, with the addition of some extent of heterogeneity in the deep mantle source, can explain variations in bulk rock trace and rare earth elements (see Robidoux, 2016), for example the range of Ba/La ratios in effusive and explosive

CR

IP

T

products, as observed in figure 9d (Patino et al., 2000; Bolge et al., 2009; Robidoux, 2016).

US

5.3 The Telica plumbing system

According to our MELTS and solubility models, primitive Telica magmas start crystallizing olivines and degassing at 15 km (Fig. 10a, 11). The presence of poorly evolved

AN

olivine-hosted melt inclusions, with entrapment pressures corresponding to equivalent depths as low as 5 km (Fig. 10a, 11), implies relatively limited magma evolution in the 5–15 km depth

M

range. Thus, the “deep” magmatic plumbing system is “sampled” only by primitive melt

ED

inclusions in STS olivines.

The later (lower-pressure) magma crystallization and differentiation sequence at Telica is

PT

more complex, but preserved in magmas erupted during the post-1970s period and in STS clinopyroxene-hosted melt inclusions. Since entrapment pressures, volatile contents and degree

CE

of differentiation are very distinct for inclusions trapped in olivine and pyroxene phenocrysts, it is arguable that the two mineral phases did not crystallize at the same time during magma ascent,

AC

or possibly that they stabilized in two different areas of Telica’s plumbing system. These distinct conditions may also be reflected by the two crystal fractionation steps, discussed above. Many lines of evidence support complex crystal growth history during magma ascent in the shallow plumbing system. The low, but variable, volatile contents registered in pyroxenes, plus their numerous series of concentric inclusion assemblages, suggest long residence time and recycling in the shallow magmatic plumbing system, and their zoning may reflect crystal growth during ascent (Cashman and Blundy, 2000). In turn, reverse Mg# zoning in pyroxenes (Fig. 3d; Fig. 4b), and the presence of both Cl-poor and Cl-rich differentiation trends (Fig. 9b), suggests a

ACCEPTED MANUSCRIPT heterogeneous magmatic system, and possibly mixing of compositionally distinct magma batches at shallow depths (e.g. nearby Cerro Negro Volcano and Las Pilas-El Hoyo Complex, Venugopal et al., 2016). Mixing is likely to occur during the second modeled step of crystallization (STEP 2; Fig. 6, Fig. 10a), whereby variably degassed (Fig. 9a-b) melts do interact in the shallow depth reservoirs (<5 km). At this level, large augite phenocrysts are considered to crystallize (Mg/Mg+Fe= 60–80 mol.%), while both olivine and augite Fe-rich microphenocrysts nucleate

T

together, as indicated by SEM-EDS analysis in Bevacqua (2014). Olivine phenocrysts from

IP

evolved scoriae of TEL03 (Mg/Mg+Fe <62 mol. %), and olivine microphenocrysts (50–62 mol.

CR

%) in the matrix of STS scoriae are thus representative of daughter liquids in the crystallization model above (Fig. 6; Fig. 10).

US

According to our entrapment pressure data, we infer that most of this late crystallization sequence occurred at crustal depths shallower than 5 km, and that most pyroxenes may have

AN

continued to grow and trap variably degassed liquids, in the 0.3 and 2.4 km depth range. It is unclear if a large magma chamber exists below Telica, but seismic records (Rodgers et al. 2015)

M

evidence a shallow cluster of seismic events below the summit of Telica, between 0.5 and 2 km depth (e.g. Geirsson et al., 2014; Fig. 11). These events most probably locate the upper part of the

ED

shallow magma ponding zone. This depth may correspond to the source area where the final

PT

stages of magma differentiation occur.

Our results can tentatively be used to draw some preliminary conclusions on the source

CE

mechanisms driving Telica explosive eruptions. Based on our melt inclusion data, we argue that distinct portions of the Telica plumbing system are involved in the generation of large-scale (VEI

AC

 3) and historical (VEI 0-2) eruptions, respectively. We find that only rocks erupted in the more explosive Telica eruptions (exemplified by STS in our study) bring mineralogical and petrological evidence for the involvement of the deep (> 5 km) plumbing system (Fig. 11). We therefore suggest that VEI  3 events at Telica were triggered by injection of deep mafic magma, rich in CO2, into the superficial crustal magma network. This is supported by the olivine melt inclusions being in equilibrium with a magmatic gas phase with XCO2 > 0.62–0.95 (calculated with the model of Papale et al., 2006). Mixing between deep rising magma and more differentiated, volatile depleted resident magma (as supported by pyroxene inclusion data), in the 2.4–5.0 km depth range, was likely the driver for overpressure, and rapid gas-magma ascent,

ACCEPTED MANUSCRIPT during paroxysmal Vulcanian to sub-Plinian explosions (e.g. Masaya, Stix, 2007; Popocatépetl, Roberge et al., 2009; Soufrière, Edmonds et al., 2010, etc.). In contrast, only the shallowest portions of the magmatic system (Fig. 11), as recorded by differentiated volatile-poor inclusions, appears to be involved in the more modest (VEI 0-2) post1970s explosive behaviour of Telica. Shallow magmas still have the potential to produce

T

explosive eruptions, in view of the strong impact of degassing and crystallization on magma

IP

viscosity (e.g. Cerro Negro, Portnyagin et al., 2014). Magma viscosity is known to strongly

CR

increase when magma critical crystallinity is reached (Fig. 10b; Marsh, 1981). According to our MELT models (Fig. 10a), residual liquids formed after STEP 1 of fractional crystallization account for only ~41% of the total magmatic mass (or 37 vol %; Fig. 10 a-b), which already

US

attains the critical crystallinity condition for basalts (i.e. liquid < 40–50 vol %; Marsh, 1981). At Telica, this condition is achieved at 0.52 – 0.65 wt. % K2O contents (Fig. 10b). As this critical

AN

point is reached, or at even higher crystal (<60 wt. %) and K2O (>0.65 wt. %) contents, the magma becomes uneruptible due to the highly viscous behavior. The most evolved of such

M

liquids, perhaps squeezed in a crystal-mush matrix, are recorded in the end-member composition melts trapped in some pyroxene-hosted inclusions (1.6 – 4.0 wt. % K2O; probably affected by

ED

strong post entrapment crystallization at K2O > 2.6 wt. %). We argue that magmatic gas bubble supply, perhaps from the deep plumbing system, may be a recurrent casual factor for achieving

PT

over-pressuring, and ultimately eruption, of this viscous magma. We yet caution that additional work is required to determine the driving mechanisms leading to brittle breakage at Telica (e.g.

CE

Gonnerman and Manga, 2007; Vetere et al., 2010; Rust and Cashman, 2011). We cannot exclude, for example, a phreatic or phreato-magmatic trigger for some of the most recent Telica eruptions

AC

(e.g. Geirsson et al. 2014; de Moor et al., 2016 and references therein). A phreatic mechanism has been invoked for the 2011 eruption, but no clear evidence of interaction between liquid water and fresh magma has been found in support to phreatic or phreatomagmatic activity in the March 1982 pyroclasts.

ACCEPTED MANUSCRIPT 6. Conclusions The combined interpretation of volatile and major element compositions of melt inclusions and whole rocks contribute first models of fractional crystallization and decompression for magmas feeding explosive eruptions from Telica volcano, in Nicaragua. Olivines were probed to define the most primitive end-members, while pyroxenes of old and

T

recent explosive eruptions, concurred to characterize the intermediate/shallow levels of the

IP

plumbing system.

CR

Based on our geochemical analysis of Telica pyroclasts, we conclude that:

US

(1) Volatile contents in olivine-hosted melt inclusions indicate high entrapment pressures (<198–400 MPa), and are therefore representative of a deep (5–15 km) magmatic system.

AN

Two magma types, either water-rich (e.g. similar to the nearby Cerro Negro system) or water-poor (similar to Nejapa and Granada magmatic systems), may have

M

alternated/coexisted in the deep Telica magmatic system; (2) Pyroxenes of any Telica eruptions show a remarkable range of magma compositions,

ED

from basalts to dacites, and record shallow degassing processes linked to differentiation of residual chlorine-rich melts. Well preserved STS pyroxene inclusions record

PT

entrapment depths of 0.3–2.4 km (from 130 MPa to < 40 MPa), while pyroxene inclusions in 1982 and 2011 Telica eruptions are strongly affected by post entrapment

AC

pyroclasts;

CE

water diffusion, likely due to their occurence in relatively slowly-cooled large-size

(3) A model of two reservoirs at different depths evolving in two stages (STEP 1 and 2) is supported by mineral phase stability conditions, and is constrained by MELTS simulations starting from a parental magma at 400 MPa, ΔNNO = 0, and temperature of 1153 °C. During the first stage, volatile loss (CO2-H2O-S) to bubbles mainly occurs by decompression from ~300 MPa down to 130 MPa. Water loss, from 5.2 to < 1.8 wt%, may cause increasing olivine and plagioclase crystallization during decompression. At shallow depth, second stage of crystallization is recorded in pyroxene melt inclusions.

ACCEPTED MANUSCRIPT Rapid cooling rates may accentuate magma differentiation in such shallow (<130 MPa) levels. (4) According to MELTS calculations and solubility models, primitive Telica magmas initiate olivine crystallization and degassing at 15 km, until reaching 5 km depth. Clinopyroxene and plagioclase nucleate and accelerate magma crystallization between

T

2.4–5.0 km.

IP

(5) Large-scale (STS-like) explosive eruptions likely involve injections of deep primitive

CR

volatile-rich magmas into the shallow crustal plumbing system. Inverse zoning, variable pyroxene compositions and low volatile contents in their glass inclusions, do reflect a

US

shallow crustal reservoir where mixing between shallow resident and deeply intruding magma may have triggered STS explosive eruptions.

AN

(6) Magma differentiation and degassing in the shallow plumbing system, increasing melt viscosity, are likely to have played a major control on post-1970s magmatic explosive

M

eruptions.

ED

Our study here contributes novel information of the chemical structure of the Telica plumbing system. Additional chronostratigraphic and physical volcanology studies are now

PT

required to better characterise magnitude, duration, recurrence time, style and dynamics of the volcano past eruptions. At a shorter time scale, studies on the ongoing degassing/eruptive

CE

unrests are vital for improved volcanic hazard assessment (Geirsson et al. 2014; Conde et al.,

AC

2014; Rodgers et al., 2015).

FUNDING

This work was supported by the DECADE research initiative of the Deep Carbon Observatory and received funding from the European Research Council under the European Union’s Seventh Framework Program (FP7/2007/2013)/ERC grant agreement n 305377.

ACCEPTED MANUSCRIPT ACKNOWLEDGMENTS

US

REFERENCES

CR

IP

T

This work was reviewed as part of the Ph.D thesis of P. Robidoux by Prof. Y. Taran (UNAM) and Prof. J. Stix (McGill). We are indebted to the two official Reviewers and the Editor whose thorough reviews and comments gave a great help in improving the paper. Special thanks to A. Cavallo (INGV-Roma) for electron microprobe analysis. to F. Furnari (DiSTeM) for SEM analysis, to G. Bevacqua for helping with glass inclusion preparation, to J. Wang for NanoSIMS sample preparation at Carnegie Institution, to Russel Rogers for English revision. It.is important to mention the field participation of V. Conde and B. Galle (Chalmers U. Technology) and the INETER member and staff involvement in field planning and sampling during 2013-2014 (A. Ceballos, D. Chavarria, H. Sanchez, E. Espinosa, A. Muñoz).

M

AN

1. Aiuppa, A., Robidoux, P., Tamburello, G., Conde, V., Galle, B., Avard, G. , Bagnato, E., DeMoor, J. M. , Martinez, M. & Munoz, A. (2014). The Costa Rica – Nicaragua volcanic segment: along arc variations in volcanic gas chemistry and improved CO2 budget. Earth and Planetary Science Letters 407, 134-147

ED

2. Atlas, Z. D. (2008). Volatiles in melt inclusions from Mexican and Nicaraguan volcanoes: implications for complex degassing processes, Open Access Dissertations, Paper 142, 1131

CE

PT

3. Bevacqua, G. (2014). Studio delle inclusioni vetrose del Vulcano Telica (Nicaragua), Tesis esperimental di laurea di Giovanna Bevacqua, Anno 2013-2014, DISTEM, U. Palermo, Corr. S. Rotolo, G. Lanzo, 91 p.

AC

4. Blundy, J., Cashman, K. V., Rust, A., & Witham, F. (2010). A case for CO2-rich arc magmas. Earth and Planetary Science Letters 290 (3–4), 289-301 5. Bolge, L. L., Carr, M. J., Milidakis, K. I., Lindsay, F. N. & Feigenson, M. D. (2009). Correlating Geochemistry, Tectonics, and Volcanic volume along the Central American Volcanic Front. Geochemistry Geophysics Geosystems 10 (12), 1-15 6. Carr, M. J. (1984). Symmetrical and segmented variation of physical and geochemical characteristics of the Central American volcanic front. Journal of Volcanology and Geothermal Research 20, 231-252 7. Carr, M. J., Feigenson, M. D. & Bennett, E. A. (1990). Incompatible element and isotopic evidence for tectonic control of source mixing and melt extraction along the Central American arc. Contributions to Mineralogy and Petrology, 105 (4), 369-380

ACCEPTED MANUSCRIPT 8. Carr, M. J., Feigenson, M. D., Bolge, L. L., Walker, J. A., & Gazel, E. (2014). RU_CAGeochem, a database and sample repository for Central American volcanic rocks at Rutgers University. Geoscience Data Journal 1(1), 43-48 9. Cashman, K. & Blundy, J. (2000). Degassing and crystallization of ascending andesite and dacite. Philosophical Transactions of the Royal Society of London A: Mathematical, Physical and Engineering Sciences, 358 (1770), 1487-1513

IP

T

10. Conde V., (2015). Application of DOAS, FTIR and Multi-GAS for studies of evolving volcanic activity, 2015, Department of Earth and Space Sciences, Chalmers University of Technology; Gothenburg, Sweden, 198 pp.

US

CR

11. Conde, V., Robidoux, P., Avard, G., Galle, B., Aiuppa, A.& Muñoz, A. (2014). Measurements of SO2 and CO2 by combining DOAS, Multi-GAS and FTIR observations: a case study from Turrialba and Telica volcanoes, International Journal of Earth Sciences, June 2014, 1437-3254

AN

12. Connor, L. J., & Connor, C. B. (2006). Inversion is the key to dispersion: understanding eruption dynamics by inverting tephra fallout. Statistics in Volcanology 1, 231242

M

13. Deer W.A., Howie R.A., Zussman J. (1992). An introduction to the rock-forming ninerals, 688 Longman, Harlow, UK, 696 p.

PT

ED

14. De Moor, J. M., Aiuppa, A., Pacheco, J., Avard, G., Kern, C., Liuzzo, M., Martinez, M., Giudice, G. & Fischer, T. P. (2016). Short-period volcanic gas precursors to phreatic eruptions: Insights from Poás Volcano, Costa Rica. Earth and Planetary Science Letters 442, 218-227

CE

15. Danyushevsky, L. V. & Plechov, P. (2011). Petrolog3: Integrated software for modeling crystallization processes. Geochemistry Geophysics Geosystems 12 (7), 1-32

AC

16. Di Carlo, I., Pichavant, M., Rotolo, S. G., & Scaillet, B. (2006). Experimental crystallization of a high-K arc basalt: the golden pumice, Stromboli volcano (Italy). Journal of Petrology 47(7), 1317-1343 17. Edmonds, M., Aiuppa, A., Humphreys, M., Moretti, R., Giudice, G., Martin, R. S., & Christopher, T. (2010). Excess volatiles supplied by mingling of mafic magma at an andesite arc volcano. Geochemistry, Geophysics, Geosystems 11(4) 18. Gale, A., Dalton, C. A., Langmuir, C. H., Su, Y. & Schilling, J. G. (2013). The mean composition of ocean ridge basalts. Geochemistry, Geophysics, Geosystems 14 (3), 489518 19. Geirsson, H., Rodgers, M., LaFemina, P., Witter, M., Roman, D., Muñoz, A., Tenorio, V., Alvarez, J., Jacobo, V. C., Nilsson, D., Galle, B., Feineman, M. D., Furman, T. &

ACCEPTED MANUSCRIPT Morales, A. (2014). Multidisciplinary observations of the 2011 explosive eruption of Telica volcano, Nicaragua: Implications for the dynamics of low explosivity ash eruptions. Journal of Volcanology and Geothermal Research 271, 55-69 20. Global Volcanism Program, National Museum of Natural History (GVP). Retrieved 2014, http://volcano.si.edu/

IP

T

21. Gonnerman, H.M. & Manga, M. (2005). Nonequilibrium magma degssing : Results from modeling of the ca. 1340 A.D. eruption of Mono Craters, California, Earth and Planetary Science Letters 238, 1-16

CR

22. Gonnermann, H. M., & Manga, M. (2007). The fluid mechanics inside a volcano. Annu. Rev. Fluid Mech. 39, 321-356

US

23. Gualda, J., & Ghiorso, M. S. (2015). MELTS_Excel: A Microsoft Excel-based MELTS interface for research and teaching of magma properties and evolution. Geochemistry, Geophysics, Geosystems 16, 315-324

M

AN

24. Hauri, E. H. Wang, J., Dixon, J. E., King, P. L., Mandeville, C. & Newman, S. (2002). SIMS analysis of volatiles in silicate glasses: 1. Calibration, matrix effects and comparisons with FTIR. Chemical Geology 183, 99-114

ED

25. Hauri, E. H. , Shaw, A. M., Wang, J., Dixon, J. E., King, P. L., & Mandeville, C. (2006). Matrix effects in hydrogen isotope analysis of silicate glasses by SIMS, Chemical Geology 235, 352-365

CE

PT

26. Havlicek, P., Hradecky, P., Hrubes, M., Mlcoch, B., Opletal, M., Sebesta, J. (CGU), Buitrago & N., Strauch, W. (1999). Estudio Geológico Y Reconocimiento De La Amenaza Natural - Zona Chinandega-Leon , Nicaragua. Praga-Managua 1999, Resumen Ejecutivo (Servicio Geológico Checo, CGU, en cooperación con Instituto Nicaragüense de Estudios Territoriales, INETER), 23 p.

AC

27. Havlicek, P., Hradecky, P., Hrubes, M., Kyel, P., Mlcoch, B., Mrazova, S., Novak, Z., Opletal, M., Prichystal, A., Sebesta, J., Seveik, J., Bustillo, M. & Rizo, Z. (2000). Estudio Geológico Y Reconocimiento De La Amenaza Geológica En El Área De León - La Paz Centro Y Malpasillo, Praga-Managua 2000, Reporte Final (Servicio Geológico Checo, CGU, en cooperación con Instituto Nicaragüense de Estudios Territoriales, INETER), 244 p. 28. Heydolph, K., Hoernle, K., Hauff, F., van den Bogaard, P., Portnyagin, M., Bindeman, I. & Garbe-Schonberg, D. (2012). Along and across arc geochemical variations in NW Central America: Evidence for involvement of lithospheric pyroxenite, Geochimica et Cosmochimica Acta 84, p.459-491

ACCEPTED MANUSCRIPT 29. Hradecký, P., Śebesta, J., Havlíček, P., Hrubeš, M., Kycl, P., Mlčoch, B., ... & Rapprich, V. (2007). Hradecký, P.-Šebesta J.(eds.): Mapa de la Cadena volcánica Nicaraguense/Map of the Nicaraguan Volcanic Chain 1: 200 000. Czech Geological Survey

T

30. Hill, B. E., Connor, C. B., Jarzemba, M. S., La Femina, P. C., Navarro, M., & Strauch, W. (1998). 1995 eruptions of Cerro Negro volcano, Nicaragua, and risk assessment for future eruptions. Geological Society of America Bulletin 110 (10), 1231-1241

IP

31. INETER, Recursos Naturales (1972). Estructuras (Catastro e Inventario de Recursos Naturales-INETER,1972). Mapa Geologica de Telica, codigo 2853-V

US

CR

32. INETER (Instituto Nicaragüense de Estudios Territoriales) (2014). Geofísica en Nicaragua – Boletines Sismológicos, Weblink: http://webserver2.ineter.gob.ni/sis/bolsis/bolsis.html

AN

33. Jugo, P. J., (2009). Sulfur content at sulfide saturation in oxidized magmas. Geology 37 (5), 415-418

M

34. Kilinc, A., Carmichael, I. S. E., Rivers, M., & Sack, R. O. (1983). The Ferric–Ferrous Ratio of Natural Silicate Liquids Equilibrated in Air. Contributions to Mineralogy and Petrology 83 (1-2), 136-140

ED

35. LaFemina, P., Dixon, T.H., Govers, R., Norabuena, E., Turner, H., Saballos, A., Mattioli, G., Protti, M. & Strauch, W. (2009). Fore-arc motion and Cocos Ridge collision in Central America. Geochem. Geophys. Geosyst. 10, Q05S14

CE

PT

36. Lanzo, G., Di Carlo, I., Pichavant, M., Rotolo, S. G., & Scaillet, B. (2016). Origin of primitive ultra-calcic arc melts at crustal conditions—Experimental evidence on the La Sommata basalt, Vulcano, Aeolian Islands. Journal of Volcanology and Geothermal Research 321, 85-101

AC

37. Lloyd, A.S., Plank, T., Ruprecht, P., Hauri, E.H., Rose, W. (2013). Volatile loss from melt inclusions in pyroclasts of differing sizes. Contributions to Mineralogy and Petrology 165, 129-153 38. Longpré, M. A., Stix, J., Costa, F., Espinoza, E., and Muñoz, A. (2014). Magmatic processes and associated timescales leading to the January 1835 eruption of Cosigüina volcano, Nicaragua. Journal of Petrology 55 (6), 1173-1201 39. Marsh, B.D. (1981). On the Crystallinity, Probability of occurrence, and rheology of lava and magma. Contributions to Mineralogy and Petrology 78, 85-98

ACCEPTED MANUSCRIPT 40. Métrich, N., Bertagnini, A., Landi, P., & Rosi, M. (2001). Crystallization driven by decompression and water loss at Stromboli volcano (Aeolian Islands, Italy). Journal of petrology 42 (8), 1471-1490 41. Moretti, R., Papale, P., Ottonello, G. (2003). A model for the saturation of C-O-H-S fluids in silicate melts. Geological Society, London, Special Publications 213 (1), 81-101 42. Navarro, M. C. (1994). Peligro volcánico, volcán Telica, MS, 73 p. INETER Managua

CR

IP

T

43. Novák, Z. & A. Přichystal, (2006). El área volcánica de Telica (Nicaragua) y sus peligros geológicos, in: Krystalinikum, Contribution to the geology and the petrology of crystalline complexes. Moravian Museum Brno, No. 31, 2006, 47-66

US

44. Ochs, F.A. & Lange R.A. (1999). The density of hydrous magmatic liquids. Science 283, 1314-1317

AN

45. Papale, P., Moretti, R. & Barbato, D. (2006). The compositional dependence of the saturation surface of H2O-CO2 fluids in silicate melts. Chemical Geology 29, 78-95

M

46. Patino, L. C., Carr, M. J. & Feigenson, M. D. (2000). Local and regional variations in Central American arc lavas controlled by variations in subducted sediment input. Contributions to Mineralogy and Petrology 138 (3), 265-283

ED

47. Pichavant, M., Di Carlo, I., Rotolo, S. G., Scaillet, B., Burgisser, A., Le Gall, N., Martel, C. (2013). Generation of CO2-rich melts during basalt magma ascent and degassing. Contributions to Mineralogy and Petrology 166, 545-561

CE

PT

48. Portnyagin, M. V., Hoernle, K. & Mironov, N. L. (2014). Contrasting compositional trends of rocks and olivine-hosted melt inclusions from Cerro Negro volcano (Central America): Implications for decompression-driven fractionation of hydrous magmas. International Journal of earth Sciences 103 (7), 1963-1982

AC

49. Putirka, K. (2008). Thermometers and Barometers for Volcanic Systems. In: Putirka K.D., Tepley III G.J. (eds) Minerals, Inclusions and Magmatic Processes. Reviews in Mineralogy and Geochemistry 69, Mineralogic Society of America, 61-120 50. Pyle, D. M. (1989). The thickness, volume and grain size of tephra fall deposits. Bulletin of Volcanology 51 (1), 1-15 51. Roberge, J., Delgado-Granados, H. & Wallace, P.J. (2009). Mafic magma recharge supplies high CO2 and SO2 gas fluxes from Popocatépetl volcano, Mexico. Geology 37, 107-110 52. Robidoux, P., (2016). The nature and source of major magmatic volatiles: open-vent degassing volcanoes in the Central American Volcanic Arc, Ph.D Thesis, Dottorato di

ACCEPTED MANUSCRIPT ricerca in Scienze della Terra – Indirizzo Geochimica Dipartimento di Scienze della Terra e del Mare (DiSTeM) Sez. Chimica e Fisica della Terra, 221 p., Weblink: http://hdl.handle.net/10447/163138 53. Robidoux, P., Aiuppa, A., Rotolo, S., Rizzo, A. L., & Hauri, E. H., (2017). The volatile content of mafic-to-intermediate magmas from San Cristóbal volcano, Nicaragua. Lithos 272-273, 147-163

IP

T

54. Rodgers, M., Roman, D. C., Geirsson, H., LaFemina, P., Muñoz, A. Guzman, C. & Tenorio, V. (2013). Seismicity accompanying the 1999 eruptive episode at Telica Volcano, Nicaragua. Journal of Volcanology and Geothermal Research 265, 39-51

US

CR

55. Rodgers, M., Roman, C. D., Geirsson, H., LaFemina, P., McNutt, S. R., Muñoz, A. & Tenorio, V. (2015). Stable and unstable phases of elevated seismic activity at the persistently restless Telica Volcano, Nicaragua. Journal of Volcanology and Geothermal Research 290, 63-74

AN

56. Roggensack, K., Hervig, R. L., McKnight, S. B., & Williams, S. N. (1997). Explosive basaltic volcanism from Cerro Negro: Influence of volatiles on eruptive style. Science 277, 1639-1642

ED

M

57. Rust, A.C. & Cashman, K.V. (2011). Permeability controls on expansion and size distributions of pyroclasts. J. Geophys. Res.116

PT

58. Sadofsky, S. J., Portnyagin, M. V., Hoernle, K., & van den Bogaard, P. (2008). Subduction cycling of volatile and trace elements through the Central American Volcanic Arc: Evidence from melt inclusions. Contributions to Mineralogy and Petrology 155, 433456

AC

CE

59. Saginor I., Gazel, E., Michael, J., Carr, C., Swisher III, C. C., Turrin, B. (2011). New Pliocene–Pleistocene 40Ar/39Ar ages fill in temporal gaps in the Nicaraguan volcanic record. Journal of Volcanology and Geothermal Research 202, 143-152 60. Scaillet, B. & Pichavant, M. (2003). Experimental constraints on volatile abundances in arc magmas and their implications for degassing processes, in Volcanic Degassing, edited by C. Oppenheimer, D.M. Pyle and J. Barclay, Geol. Soc. Spec. Publ. 213, 23-52 61. Self, S., Wilson, L., and Nairn, I. A. (1979). Vulcanian eruption mechanisms. Nature 277 (5696), 440-443 62. Signorelli, S. & Carroll, M. R. (2000). Solubility and fluid-melt partitioning of Cl in hydrous phonolitic melts. Geochimica et Cosmochimica Acta 64, 2851-2862 63. Sigurdsson, H., Houghton, B., McNutt, S., Rymer, H., & Stix, J. (Eds.). (2015). The encyclopedia of volcanoes, Elsevier

ACCEPTED MANUSCRIPT 64. Simakin, A. G., Salova, T. P., & Babansky, A. D. (2009). Amphibole crystallization from a water-saturated andesite melt: Experimental data at P= 2 kbar. Petrology 17 (6), 591605 65. Spieler, O., Kennedy, B., Kueppers, U., Dingwell, D. B., Scheu, B., & Taddeucci, J. (2004). The fragmentation threshold of pyroclastic rocks. Earth and Planetary Science Letters 226 (1), 139-148

CR

IP

T

66. Spilliaert, N., Allard, P., Métrich, N., and Sobolev, A. V. (2006). Melt inclusion record of the conditions of ascent, degassing, and extrusion of volatile‐rich alkali basalt during the powerful 2002 flank eruption of Mount Etna (Italy). Journal of Geophysical Research: Solid Earth 111(B4)

US

67. Stix J., Zapata, G. J. A., Calvache, V. M., Cortes, J. G. P., Fischer, T. P., Gomez, M. D., Narvaez, M. L., Ordonez, V. M., Ortega, E. A., Torres, C. R. & Williams, S. N. (1993). A model of degassing at Galeras volcano, Colombia, 1988–1993. Geology 21, 963-967

AN

68. Stix, J. (2007). Stability and instability of quiescently active volcanoes: The case of Masaya, Nicaragua. Geology 35 (6), 535-538

ED

M

69. Sugawara, T. (2000). Empirical relationships between temperature, pressure, and MgO content in olivine and pyroxene saturated liquid. Journal of Geophysical Research: Solid Earth 105(B4), 8457-8472

PT

70. Syracuse, E. M., & Abers, G. A. (2006). Global compilation of variations in slab depth beneath arc volcanoes and implications. Geochemistry Geophysics Geosystems 7, Q05017

CE

71. Tenorio, V. (Ed.), 2011d. Boletín Mensual, Sismos y Volcanes de Nicaragua – Junio, 2011, report. INETER, Managua

AC

72. Toplis, M. J. (2005). The thermodynamics of iron and magnesium partitioning between olivine and liquid: Criteria for assessing and predicting equilibrium in natural and experimental systems, Contributions to Mineralogy and Petrology 149, 22-39 73. Venugopal, S., Moune, S., & Williams-Jones, G. (2016). Investigating the subsurface connection beneath Cerro Negro volcano and the El Hoyo Complex, Nicaragua. Journal of Volcanology and Geothermal Research 325, 211-224 74. Vetere, F., Behrens, H., Holtz, F., Vilardo, G. & Ventura, G. (2010). Viscosity of crystalbearing melts and its implication for magma ascent. Journal of Mineralogical and Petrological Sciences 105 (3), 151-163 75. Walker, J. A., Patino, L. C., Michael, J. C. & Feigenson, M. D. (2001). Slab control over HFSE depletions in central Nicaragua. Earth and Planetary Science Letters 192, 533-543

ACCEPTED MANUSCRIPT 76. Wallace, P. J., & Carmichael, I. S. (1994). Petrology of Volcán Tequila, Jalisco, Mexico: disequilibrium phenocryst assemblages and evolution of the subvolcanic magma system. Contributions to Mineralogy and Petrology 117 (4), 345-361 77. Wehrmann, H. (2005). Volatile degassing and plinian eruption dynamics of the mafic Fontana Tephra, Nicaragua. PhD Disertation, Kiel University, Germany, 113p. 78. Wehrmann, H., Hoernle, K., Portnyagin, M., Wiedenbeck, M., Heydolph, K. (2011).

AC

CE

PT

ED

M

AN

US

CR

IP

T

Volcanic CO2 output at the Central American subduction zone inferred from melt inclusions in olivine crystals from mafic tephras. Geochemistry Geophysics Geosystems 12, Q06003

Figure captions

Fig. 1 – Geological map of Telica. The main lithostratigraphy in the Telica area are shown based on the work of the Geological Tcheque Republic Survey (Havlicek et al., 2000). The digital topography effect (1:20m) and the lines from the main Chinandega structures are provided by INETER (1972; 2014). Yellow stars represent rock sampling sites from this study (e.g.

ACCEPTED MANUSCRIPT Robidoux, 2016) and orange stars are sampled outcrops from Havlicek et al. (2000). Isopach approximation for STS fallout area (deposit thickness in meter written with yellow contour bold character for each surface polygon) with maximum area from Havlicek et al. (2000).

CR

IP

T

Fig. 2 – Crystal (olivine, clinopyroxene) - liquid equilibria for phenocryst rims- groudmass glass and phenocryst cores- whole rock pairs. Pale solid lines correspond to constant KdFe-MgOlliq =0.30 calculated at an fO2 = Ni-NiO, dotted lines represent the Kd values of 0.27 and 0.33. Bold dotted lines denote deviation from equilibria of melt inclusions MI crystallized after adding 2 and 5% of olivine for melt inclusion MI* corrected PEC. Both olivine and clinopyroxene hosted inclusions have Mg# calculated and illustrated for equilibrium Kd with Fe2+ (Wallace and Carmichael, 1994).

US

Tree samples of inclusion-olivine pairs show important deviation from equilibria with olivine in comparison with the majority of inclusions (TEL04B5a; TEL04B8b; TEL04Kol6a). Groundmass glass vs olivine rim pairs (blue cross) plot along the Kd value of 0.27.

AN

Matrix glass vs clinopyroxene rims data (red crosses) plot along the red dashed line at Kd value of 0.24.

M

Whole rock vs olivine core compositions are represented by large black squares, while whole rock vs core clinopyroxene data are represented by small black lozenge inside a green frame.

ED

Fig. 3 –

PT

(a, b) Transmitted light microphotographies (crossed polars, 20 x magnification ) a) Sample TEL01 bomb emitted between February 12-20, 1982, note the high vesisularity and the small phenocryst size;. b) Sample TEL04K: scoriae from STS with the cotectic phenocrystic assemblage plg-cpx-ol, with their respective glass inclusions.

AC

CE

(c, f) Backscattered electron detector (BSE) images of c) olivine phenocryst from sample TEL04B2a: crosses refer to olivine composition spots, letters to the melt inclusion MI composition (recasted after PEC), d) Sample TEL03A: an augite phenocryst with inclusions of plagioclase set in a highly vesiculated matrix glass, e) Sample TEL03 phenocryst of plagioclase with augite inclusions, f) Sample TEL02 augite phenocryst with inclusions of plagioclase and magnetite. Fig. 4 – Frequency distribution histogram of Mg/Mg+Fe from mafic phenocrysts in tephra. a) clinopyroxene, b) olivine. Fig. 5 – Quadrilateral classification of pyroxene phenocrysts and microlites a) Pyroxene of post-1970 eruptions, b) Pyroxene of STS deposit. Fig. 6 – Harker diagrams for major elements.

ACCEPTED MANUSCRIPT

IP

T

Symbols: (i) melt inclusions hosted- olivines (MIs)= areas encircled by continuous line (data for north-west Nicaragua limited at Cordillera Los Maribios are from: Roggensack et al., 1997; Wehrman, 2005, Tesis; Atlas, Z. Tesis, 2008; Sadofsky et al.,2008; Wehrman et al., 2011; Portnyagin et al., 2014). (ii) literature whole rock from Telica = colored squares with black frame (data from RU_CAGeochem database of Carr et al., 2014; e.g. including Patino et al., 2000; Bolge et al., 2009; Heydolph et al., 2012; Geirsson et al., 2014). (iii) whole rock data from this study: colored triangles, grouped as, eruption of May 2011 (TEL03A, TEL03B), February 1982 (TEL01, TEL02) and STS deposits (TEL04K, TEL04B, TEL07A, TE07E). (iv) Matrix glasses from scoriae, this study : colored stars.

US

CR

a) K2O vs. SiO2 classification rock diagram from Peccerillo and Taylor (1977). Purple arrow represents the liquid line of decsent for the fractionated assemblage of STEP 1 (see text, 5.2) and red arrow is the liquid line of descent for STEP 2. Crystal fractionation vectors for 10 to 15 % are also reported (modeled by XLFRAC code) , clinopyroxene (CPX), olivine (Ol), magnetite (Mt) and plagioclase (Plg) b) P2O5 vs, c) CaO v. SiO2. d) Al2O3 v. SiO2. e) MgO v. SiO2. f) TiO2 v. SiO2.

ED

M

AN

Fig. 7 – Age-related chemical variations of phenocrysts and melt inclusions trapped in olivine and pyroxene phenocrysts (this study, yellow and green circles, respectively), compared with bulk rock compositions of scoria (this study: blue stars) and lavas from this study (black square : STS pulse #3, Sample TEL07E) and literature (grey and orange squares; from Patino (2000)). a) 100*Mg/Mg+Fe % of olivine and and clinopyroxene phenocrysts

PT

b) CaO/Al2O3, c) S/Cl, d) Ba/La.

AC

CE

Fig. 8 – Volatile contents from melt inclusions at Telica. a) CO2 vs H2O concentrations in melt inclusions from Telica (including micro FT-IR data from Robidoux, 2016 with the serie TEL07A from Bevacqua, 2014) using degassing models from Papale et al. (2006). Cerro Negro data are from Roggensack et al. (1997), Sadofsky et al. (2008), and Wehrman et al. (2011). Nejapa and Granada data are from Sadofsky et al. (2008) and Wehrman et al. (2011). b) H2O vs CO2/S. Telica’s Mg# model (dashed lines) are showing approximate trends for degassing and degassing+crystallization (colored lines with arrow). c) S vs. Cl. Open-path Fourier transform infrared (OP-FTIR) spectroscopy data of in situ volcanic gas compositions are from Conde et al. (2015). Arrows represent approximate trends for degassing (Nicaragua melt inclusions MIs collection; TEL04K samples; TEL04B1 model). Crystallization trend from the TEL04B1 model is pointing out to S/Cl ratios from matrix glasses and the most degassed MIs from Telica. Solid phase may represent EMPA spot on a sulfur-rich mineral. Fig. 9 – Binary diagram of volatiles with degassing/crystallizing path from Telica melt inclusions. a) S vs CaO/Al2O3, b) Cl vs CaO/Al2O3. Dotted lines represent matrix glass

ACCEPTED MANUSCRIPT compositions and straight lines are discriminating glass inclusion samples with solid phases. Solid phases may represent EMPA spot on a sulfur/chlorine-rich mineral.

CR

IP

T

Fig. 10 – Major elements crystallization models. Telica groundmass and bulk rock composition from this study and literature are symbolized in figure 2 and 6. Crystallization vectors for each crystal phase are compared with melt inlcusion data and a two step modeled liquid line of descent represented by purple arrow (STEP 1) and red arrow (STEP 2). Lava whole-rock samples are the same as in figure 6. MELTS calculations indicate that olivine crystallization explains much of the MgO-rich and K2O-poor melt inclusions, while the wide range of their composition is rather explained by the end of the curve tendency with effect of plagioclase±clinopyroxene crystallization.

AN

US

a) K2O vs MgO wt% with the MELTS calculations simulate a possible ascent path (varying both pressure and temperature), using TEL04B1 melt inclusion sample composition with the highest volatile content and pressure values at fO2 = NNO (H2O = 5.2 wt%; CO2 = 1,642 µg/g; 400 MPa). Considering an average melt density of ~2.7 g/cm3 for olivine trapped melts (n = 18; Ochs and Lange, 1999) with mass fractions in figure 10a and average single mineral olivine-pyroxeneplagioclase-magnetite densities taken from Deer et al. (1992) (e.g. Portnyagin et al., 2014), an equivalent volume fraction of 37 % liquid is calculated after STEP 1.

AC

CE

PT

ED

M

b) K2O wt% vs Fo% shows the decreasing temperature evolution of the melt until the most evolved end member composition (matrix glass minimum and maximum liquid-crystal composition) is reached within the critical crystallinity (i.e. eruptability) range (Marsh , 1981). The parental melt (purple star) is the primitive melt inclusion sample TEL01B and the daughter melt (green star) represents the average of evolved inclusions in augite only. Percentage crystallization of the parental melt «F» (red percentage numbers along the curve) is estimated with simplified exponential regression in function of the forsterite content (Fo; see equation with red font). This applies with the supposition of K2O (see equation with black font) being perfectly incompatible during crystallization (e.g. Portnyagin et al., 2014). Grey horizontal bars for critical cristallinity (40 – 50 vol %) means that it is rheologically impossible for the magma to erupt (Marsh, 1981). Fig. 11 – Model of scaled saturations depth for each magma composition at Telica since Holocene. Depth values are based on a crustal geostatic gradient of 27 MPa/km. The shallow zone for volatile separation and possible mixing is illustrated along the orange cylinder (2.4–5.0 km). The horizontal extension and geometry of the plumbing system is unknown.

AC

CE

PT

ED

M

AN

US

CR

IP

T

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

APPENDIX 2:

IP

T

Mineral abundance and chemistry

CE

Mineral chemistry

PT

ED

M

AN

US

CR

Table 1 - Mineral abundances (vol. %). Group Bomb (1982) Pyroclasts post-1970s Scoriae (STS) Period February 1982 2011 - 2012 < 2,300 yrs BP Sample TEL01 TEL02 TEL03A TEL03B TEL04K TEL04B TEL07A Pgl, % 62 61 65 67 50 43 52 Cpx 26 25 15 19 27 26 29 Ol 6 8 8 7 14 17 10 Bt 0 0 tr. tr. 0 0 0 Amph 0 0 0 0 tr. 4 tr. Ox 6 6 12 7 9 9 10 Vesicularity 39 38 55 55 45 44 46 Plg plagioclase, Cpx clinopyroxene, Ol olivine, Bt biotite, Amph Amphibole, Ox oxide. tr. is for trace amounts Analyses are recalculated to 100 %, void–free, after subtracting glass and breakdown corona products with ImageJ software

Olivine phenocrysts are euhedral to subhedral, (Fig. 3c) with the highest Forsterite mol % in

AC

TEL04B sample (core = Fo 83.6 ± 2.9 mol %, rim = Fo = 82.7 ± 2.8 mol %). TEL04K olivines have lower forsterite content (core= 80.5 ± 4.2 rim= . 80.6 ± 3.5). TEL03A has the lowest Fo% (average for both core and rims altogether, 60.2 ± 2.1, Fig. 4a-b). For 7 pairs of core-rim spots in the same crystal, TEL04B and TEL03B have 2/5 samples with slight normal zoning (at least >2 % difference in Fo%) with the other crystal being unzoned. Glass inclusions found in TEL04B have large dimensions (>30 µm), are spherical entirely glassy and many of these contain a bubble (Fig. 3b-c). In TEL04K, glassy spherical inclusions

ACCEPTED MANUSCRIPT are smaller (10–40 µm), but frequently also larger ellipsoidal inclusions are found (20–50 µm)(Fig. 3b). Clinopyroxene phenocrysts are euhedral to subhedral, some having sharp edges against the matrix (Fig. 3d). Crystals have dimensions from 3 to 22 mm and are classified as augites, with some compositional range depending on the layer (Fig. 5a-b). TEL04B shows the largest crystals

T

(>10 mm dimensions), with Mg/Mg+Fe of 0.71–0.74 and Fs15.7-18.0 Wo39.3-41.1. TEL04K

IP

clinopyroxene phenocrysts (10–15 mm) have a Fs13.1-22.5 Wo36.7-42.3 composition with Mg/Mg+Fe

CR

increasing from 0.67 in the core to 0.75 in the rim in several samples (inverse zoning; Fig. 4). TEL01 phenocrysts show Mg numbers of 0.71–0.74 and Fs15.8-18.0 Wo42.2-43.9, and TEL02 phenocrysts show Mg numbers of 0.67–0.70. Three inclusions of augite were found in

US

plagioclase phenocrysts of TEL02, with similar Mg/Mg+Fe of ~0.67 and Fs19.3-20.3 Wo41.1-42.8. Phenocrysts found in TEL03A have 0.61–0.66 and Fs20.3-23.5 Wo41.0-41.4 (Fig. 5a), while TEL03B

AN

has phenocrysts with Mg/Mg+Fe of 0.71 – 0.74 and Fs16.8-17.7 Wo38.1-38.9 (Fig. 5b). It is common to find a complex number of inclusion typologies in the clinopyroxene crystals; we find a

M

concentric series of inclusions from the core to the rim of the crystal, with 5–15 vol.% of bubbles/inclusions and oxide phases (15–25 vol %). Ca-poor plagioclases are found as inclusions

ED

within TEL03A pyroxenes (some low-Mg coupled to An57), but most have normal zoning, from An90-96 (cores), to An61-70 on the rim. Ca enrichment (inverse zoning) frequently occurs with solid

PT

plagioclase inclusions inside clinopyroxene phenocrysts (passing from An82 in the core to An88 closer to the rim; Fig. 3d). This is equally measured with inverse zoning within the clinopyroxene

AC

CE

host (Mg/Mg+Fe passing from 0.70 to 0.74; Fig. 3d).

ACCEPTED MANUSCRIPT

APPENDIX 3 - Mineral chemistry: olivine phenocrysts Sample

Zone

Layer TEL03A

SiO2

FeO

MnO

MgO

CaO Tot. All.

Fo %

(wt%) Rim Core

36.36

36.20

0.76

26.37

0.31

100

36.43

35.75

0.76

26.74

0.33

100

55.99 57.14

tel03Aol1c1

Core

35.65

33.11

0.73

30.20

0.25

100

61.93

tel03Aol1c2

Core

35.35

33.54

0.67

30.14

0.28

100

61.57

Rim Rim

35.94

33.17

0.74

29.77

0.28

35.66

34.02

0.65

29.40

0.28

100

61.54 60.64

tel03Aol2c1

Core

35.63

33.39

0.70

29.96

61.54

tel03Aol2c2

IP 0.27

Core

35.74

32.95

0.68

30.36

0.25

100

62.16

tel03Aol2rim1

Rim

35.56

34.42

0.71

CR

100

29.07

0.24

100

60.09

tel03Aol2rim2

Rim

35.67

34.68

0.69

28.60

0.30

100

59.52

AC

CE

PT

ED

M

AN

100

US

tel03Aol1rim1 tel03Aol1rim2

T

tel03Ai13ol1 tel03Ai13ol2

ACCEPTED MANUSCRIPT Table (continued) Sample

Zone

FeO

MnO

MgO

CaO Tot. All.

Fo %

tel04b2bol

Core

39.12

15.72

0.25

44.68

0.23

100

83.52

tel04b2bolcore tel04b2bolrim

Core Rim

38.38

16.05

0.31

45.05

0.21

100

38.59

16.54

0.21

44.49

0.18

100

83.35 82.75

tel04o3aol

Core

38.27

19.13

0.23

42.14

0.22

100

79.70

tel04o3atry tel04o3olcore

Core Core

38.43

18.35

0.29

42.74

T

100

38.81

19.01

0.33

41.66

0.19

100

80.59 79.62

t04kol4aol t04kol4olrim

Core Rim

40.89

13.91

0.29

44.74

0.17

100

40.51

14.29

0.26

44.75

0.19

100

t04kol5aol t04kol5olrim

Core Rim

39.10

17.51

0.29

42.90

0.19

100

38.42

21.25

0.33

39.77

0.23

100

t04kol6rim t04kol6tim2

Rim Rim

41.10

16.16

0.30

0.16

100

40.50

16.03

0.35

42.90

0.22

100

82.35 82.67

t04kol7aol

Core

38.86

18.47

US

42.29

0.45

42.05

0.17

100

80.23

t04kol7atry1 t04kol7atry2

Core Core

38.69

18.63

0.26

42.18

0.25

100

19.56

0.33

41.70

0.28

100

80.15 79.17

t04kol8arol

Core

39.29

AN

(wt%)

17.26

0.23

43.03

0.20

100

81.64

t04kol8olcore t04kol8olrim

Core Rim

39.34

17.12

0.24

43.11

0.19

100

39.59

18.86

0.33

41.06

0.16

100

81.78 79.51

t4k6oaol

Core

36.63

27.77

0.68

34.72

0.19

100

69.03

t4k6xmi

Core

ED

Layer TEL04K

SiO2

36.39

24.89

0.46

38.07

0.20

100

73.17

39.30

15.44

0.23

44.52

0.29

100

Core Rim

38.46

18.45

0.34

42.47

0.27

100

36.98

23.10

0.47

39.21

0.24

100

80.41 75.16

Core

38.91

16.71

0.25

43.79

0.33

100

82.37

Core

38.80

16.41

0.28

44.25

0.26

100

82.78

tel04ko2aoltry

Core

38.48

16.24

0.23

44.84

0.21

100

83.12

tel04ko2col

Core

39.04

16.38

0.21

44.15

0.21

100

82.77

Rim

38.91

16.15

0.18

44.50

0.27

100

83.09

tel04ko2aol

tel04ko2olrim

M

PT

AC

tel04ko2aolcore

Core

CE

tel04K26aX tel04ko1aolcore tel04ko1aolrim

38.13

CR

IP

0.19

85.15 84.81 81.37 76.94

83.71

ACCEPTED MANUSCRIPT Table (continued) Sample

Zone

FeO

MnO

MgO

CaO Tot. All.

Fo %

(wt%) 39.53

13.85

0.17

46.28

0.18

100

85.63

t04b2bol

Core

39.15

14.20

0.21

46.29

0.14

100

85.32

t04b2col

Core

39.30

14.10

0.26

46.08

0.25

100

85.35

t04b2dol

Core

39.56

14.50

0.22

45.57

0.14

100

84.85

t04b2olcore t04b2olrim

Core Rim

40.36

13.96

0.13

45.36

0.18

100

41.85

14.64

0.22

43.13

0.17

100

85.28 84.01

t04b3aol t04b3olrim

Core Rim

39.66

14.81

0.21

45.14

0.18

100

39.46

14.61

0.26

45.52

0.15

100

84.46 84.74

t04b4aol

Core

43.87

10.22

0.08

45.58

0.25

100

88.83

t04b4olcore t04b4olrim

Core Rim

43.29

9.45

0.12

46.99

0.14

100

44.87

11.45

0.26

43.26

0.17

100

89.86 87.08

t04b5aol

Core

38.06

18.22

0.22

43.30

0.21

100

80.91

t04b5aolcore

Core

38.77

18.01

0.28

42.84

0.10

100

80.92

t04b5aolinter t04b5aolrim

Inter Rim

38.60

17.64

0.33

43.23

0.19

100

33.27

22.44

0.44

43.60

0.24

100

81.38 77.60

t04b8aol

Core

40.19

14.78

0.35

44.45

0.23

100

84.28

t04b8bol t04b8olrim

Core Rim

39.36

14.83

0.25

45.38

0.18

100

38.64

14.85

0.24

46.05

0.21

100

84.51 84.68

t04b9olcore

Core

38.65

18.06

0.23

42.88

0.18

100

80.89

t04b9olinter

Inter

38.59

18.31

0.29

42.66

0.15

100

80.60

t04b9olrim

Rim

39.04

17.72

0.21

42.84

0.18

100

81.17

t04bol6try

Core

39.30

18.15

0.32

42.08

0.15

100

80.52

M

ED

PT

CE AC

CR

IP

T

Core

AN

t04b2aol

US

Layer TEL04B

SiO2

ACCEPTED MANUSCRIPT

APPENDI X4Mineral chemistry : pyroxene phenocry sts SiO2 TiO2T Al2O3A Fe Mn Mg Ca Na2ON Cr2O3C T Wo E Fs J Qu A Mg/(Mg (wt%) iO2 l2O3 O O O O a2O r2O3 ot (∑F n d ad e +Fe) SiO2 . e (wt%) nor m)

T

Spot

0.42

50.64

0.44

50.27

0.42

51.04

0.28

50.51

0.32

50.19

0.45

50.45

0.48

50.60

0.46

CE

50.93

0.41

51.37

0.38

51.39

0.43

50.45

0.40

51.24

0.25

50.88

0.50

51.06

0.45

50.85

0.50

50.19

0.40

50.99

0.33

18. 05 19. 75 18. 84 20. 72 19. 33 20. 21 19. 08 18. 96 19. 85 19. 32 18. 11 18. 64 18. 19 18. 48 20. 89 20. 83 18. 52 19. 04 18. 18

M

ED

0.39 0.30

US

51.35

15. 73 15. 47 15. 62 15. 76 14. 81 15. 11 15. 27 15. 15 16. 81 15. 64 14. 61 16. 10 14. 08 14. 40 16. 17 15. 17 15. 93 15. 68 15. 60

AN

0.46

PT

11. 0.4 2.18 67 6 10. 0.3 2.15 13 9 10. 0.4 3.07 65 2 9.5 0.3 2.60 9 4 11. 0.5 1.73 89 4 10. 0.4 2.29 87 0 11. 0.4 2.40 90 4 11. 0.3 2.34 95 7 9.3 0.1 2.47 7 9 10. 0.4 2.40 55 4 12. 0.5 1.65 87 3 10. 0.3 2.59 18 6 14. 0.4 1.85 23 7 13. 0.4 1.44 38 7 8.4 0.1 2.79 1 1 9.1 0.2 2.77 9 7 10. 0.3 2.91 68 2 11. 0.4 2.56 40 3 12. 0.4 1.89 28 8

51.06

AC

TEL04K5a X Core TEL04K6 X Inter TEL04K8 bX Core TEL04K9a X Inter TEL04K9c xI Inter TEL04K9 dXI Inter TEL04K10 axII Inter TEL04K10 bX Core TEL04K10 cxI Rim TEL04K13 aX Core TEL04K14 aX Core TEL04K15 aX Core TEL04K18 aX1 Core TEL04K18 cX1 Core TEL04K19 aX1 Core TEL04K20 aX Inter TEL04K21 aIXI Inter TEL04K22 aX Rim TEL04K22 bX1 Rim

CR

IP

Sample (EMPA)

0.30 0.26 0.36 0.28 0.27 0.28 0.28 0.24 0.42 0.30 0.32 0.33 0.23 0.23 0.29 0.30 0.25

10 0.00 0 10 0.03 0 10 0.03 0 10 0.04 0 10 0.03 0 10 0.01 0 10 0.00 0 10 0.02 0 10 0.00 0 10 0.03 0 10 0.05 0 10 0.00 0 10 0.00 0 10 0.00 0 10 0.03 0 10 0.03 0 10 0.00 0 10 0.00 0 10 0.00 0

36. 7 39. 9 38. 3 41. 7 39. 3 40. 8 38. 7 38. 5 39. 8 39. 0 37. 1 37. 8 37. 3 37. 7 41. 8 42. 2 37. 6 38. 5 36. 8

44 .5 43 .5 44 .2 44 .1 41 .9 42 .5 43 .1 42 .8 46 .8 43 .9 41 .7 45 .5 40 .2 40 .9 45 .1 42 .8 45 .0 44 .1 43 .9

18 0. .9 0 16 0. .6 0 17 0. .6 2 14 0. .2 0 18 0. .8 0 16 0. .7 0 18 0. .2 0 18 0. .7 0 13 0. .4 0 17 0. .0 0 21 0. .2 0 16 0. .7 3 22 0. .5 0 21 0. .4 0 13 0. .1 0 15 0. .0 2 17 0. .4 2 17 0. .4 0 19 0. .3 0

97. 3. 0 0 97. 2. 7 3 97. 2. 7 1 98. 2. 0 0 97. 2. 3 7 97. 2. 9 1 97. 2. 9 1 97. 2. 8 2 97. 2. 9 1 98. 1. 2 8 96. 3. 8 2 97. 1. 8 9 97. 2. 5 5 97. 2. 5 5 98. 1. 3 7 98. 1. 2 5 97. 2. 8 0 97. 2. 7 3 98. 1. 1 9

0.71 0.73 0.72 0.75 0.69 0.71 0.70 0.69 0.76 0.73 0.67 0.74 0.64 0.66 0.77 0.75 0.73 0.71 0.69

ACCEPTED MANUSCRIPT

51.42

0.42

TEL01aol Core

51.49

0.45

TEL01bol Core TEL04Bp x6cx Core TEL04Bp x6ax Core

51.14

0.44

51.52

0.20

51.10

0.49

Sample (SEMEDS)

SiO2 TiO2 (wt%)

0.36

51.93

0.65

51.62

0.49

51.50

15. 32 16. 06 15. 56 14. 88 15. 05 15. 04

19. 09 18. 87 19. 89 19. 61 19. 31 19. 93

0.46

51.00

0.52

51.42

0.55

51.62

0.56

51.25

0.46

51.26

0.48

10 0.04 0 10 0.00 0 10 n.d. 0 10 n.d. 0 10 n.d. 0 10 n.d. 0

0.41 0.18 0.24 0.31 0.32 0.27

Al2O3 Fe Mn Mg Ca Na2O O O O O

11. 0.4 1.85 56 9 10. 0.5 2.15 44 2

38. 9 38. 1 40. 3 39. 9 39. 3 41. 1

43 .4 45 .1 43 .9 42 .1 42 .7 43 .2

17 0. .7 0 16 0. .8 1 15 0. .8 3 17 0. .9 1 18 0. .0 4 15 0. .7 9

96. 3. 8 2 98. 1. 7 2 98. 1. 2 5 97. 2. 7 2 97. 2. 6 1 98. 1. 0 2

0.72 0.71 0.71 0.74 0.74 0.71 0.71 0.74

13. 06 13. 35

20. 23 20. 24

Cr2O3 T Wo E Fs J Qu A Mg/(Mg ot (∑F n d ad e +Fe) . e nor m)

0.73

10 42. 38 19 1. 95. 2. n.d. 0 3 .0 .7 5 7 8 10 42. 39 18 2. 94. 3. n.d. 0 7 .2 .1 4 6 1

11. 0.4 13. 19. 2.31 92 6 14 47

0.59

10 41. 38 20 2. 95. 2. n.d. 0 1 .6 .4 1 6 3

0.66

11. 0.4 13. 19. 2.55 65 4 05 79

0.55

10 41. 38 19 2. 95. 2. n.d. 0 8 .3 .9 2 8 0

0.67

PT

51.87

AC

TEL02i1c px1 Core TEL02i1c px2 Core Inclus TEL02i2c ion, px1 Plg Inclus TEL02i2c ion, px2 Plg Inclus TEL02i2c ion, px3 Plg TEL02i3c px1 Rim TEL02i3c px2 Rim TEL03Ai1 3cpx1 Core TEL03Ai1 3cpx2 Core

CE

Spot

11. 0.4 2.69 07 1 10. 0.3 2.38 29 8 9.5 0.3 2.43 9 6 10. 0.4 2.31 88 2 10. 0.3 2.29 93 8 9.4 0.3 3.37 4 0

0.31

T

0.55

10 39. 43 17 0. 97. 1. 0.06 0 7 .0 .3 6 8 6 10 38. 43 18 0. 97. 2. 0.00 0 1 .6 .3 1 6 3

0.30

IP

50.42

TEL03B1 X Core TEL03B1a X Core

19. 38 18. 75

CR

0.44

15. 09 15. 44

US

51.07

10. 0.4 2.47 42 0 11. 0.3 2.46 15 9

AN

0.35

M

51.54

ED

TEL04K25 aX2 Core TEL04K25 bX Core

11. 0.2 3.02 43 7 11. 0.3 2.65 51 3 11. 0.4 2.60 23 7 13. 0.4 1.82 83 7 12. 0.3 2.42 08 5

12. 87 13. 34 13. 19 11. 89 13. 19

20. 23 19. 59 19. 80 19. 63 19. 53

0.57

0.66 0.61 0.54 0.64 0.69

10 n.d. 0 10 n.d. 0 10 n.d. 0 10 n.d. 0 10 n.d. 0

42. 8 41. 3 41. 9 41. 5 41. 0

37 .9 39 .2 38 .8 34 .9 38 .6

19 1. .3 9 19 2. .5 1 19 2. .3 6 23 1. .6 2 20 1. .4 3

95. 3. 0 1 95. 2. 4 5 96. 1. 0 4 95. 3. 1 6 94. 3. 8 8

0.67 0.70

0.67 0.67 0.68 0.61 0.66

ACCEPTED MANUSCRIPT

APPENDIX 5 - Original Glass Inclusion Chemistry

TEL04K TEL04K 8a

TEL04K TEL04K 8b

TEL04K TEL04K 9a

TEL04K TEL04K 9cI

TEL04K TEL04K 9d

Average 50.10 0.89 14.27 9.77 0.22 3.83 6.85 2.70 1.66 0.15 0.01 0.03 90.50

8 56.34 0.82 15.29 8.74 0.12 2.58 6.18 2.71 2.32 0.24 0.00 0.03 95.39

59.90 0.63 13.67 5.80 0.22 1.71 4.23 3.16 3.61 0.39 0.00 0.00 93.30

56.08 0.64 14.40 7.51 0.17 2.12 4.96 3.14 3.73 0.29 0.00 0.00 93.03

Average 53.93 0.78 14.73 8.71 0.27 3.39 6.48 3.26 2.03 0.28 0.00 0.03 93.89

19 56.92 0.74 13.30 8.16 0.18 2.75 6.19 2.95 1.98 0.25 0.00 0.11 93.55

57.52 0.68 13.42 9.77 0.34 2.17 5.74 2.71 1.90 0.22 0.00 0.00 94.47

TEL04K TEL04K TEL04K TEL04K1 TEL04K1 TEL04K1 0a 0b 0c Average 56.75 0.85 13.99 8.37 0.21 2.05 5.06 3.29 3.07 0.38 0.01 0.03 94.06

CR

IP

T

TEL04K TEL04K6 a

Average 56.27 0.86 14.10 8.27 0.21 2.41 5.24 3.48 3.00 0.38 0.02 0.04 94.27

53.40 0.99 16.08 9.23 0.20 3.45 6.58 2.70 1.59 0.22 0.00 0.04 94.47

AN

MI Sample No. (#spot/average) SiO2 (wt%) TiO2 Al2O3 FeO* MnO MgO CaO Na2O K2O P2O5 Cr2O3 NiO TOT

STS TEL04K TEL04K20 a

US

Eruption Unit

Table (continued)

M

ED

Average 60.61 1.02 12.72 7.54 0.17 1.69 4.82 3.22 3.39 0.30 0.00 0.02 95.49

May, 2011 TEL03B TEL03B 1

February, 1982 TEL01

February, 1982 TEL01

February, 1982 TEL01

tel01

tel01b

Average 62.74 0.97

Average 60.22 0.96

144 58.89 0.9

PT

Average 58.39 0.68 14.60 7.80 0.19 3.18 5.63 3.40 2.36 0.25 0.02 0.02 96.51

Average 60.68 0.63 12.36 8.33 0.23 2.24 4.44 3.44 2.72 0.24 0.00 0.03 95.32

TEL04K TEL04K2 1aI

Average 56.17 1.02 15.02 9.70 0.21 2.39 6.08 3.80 1.97 0.29 0.02 0.00 96.66

AC

MI Sample No. (#spot/average) SiO2 (wt%) TiO2 Al2O3 FeO* MnO MgO CaO Na2O K2O P2O5 Cr2O3 NiO TOT

STS TEL04K TEL04K TEL04K TEL04K TEL04K1 TEL04K1 TEL04K1 TEL04K1 3a 4b 5a 8c

CE

Eruption Unit

30 55.84 1.08 15.04 10.32 0.27 2.94 5.86 3.43 2.26 0.29 0.00 0.00 97.32

TEL04K TEL04K TEL04K TEL04K TEL04K TEL04K2 TEL04K2 TEL04K2 TEL04K2 TEL04K1 2a 2b 5a 5b 8a Average 56.16 0.72 14.58 8.30 0.20 3.37 5.66 3.02 2.28 0.22 0.02 0.01 94.55

Average 56.74 0.70 14.10 8.72 0.21 3.25 5.77 3.09 2.17 0.26 0.02 0.05 95.09

12 59.01 0.73 14.47 6.69 0.05 2.08 4.84 3.02 3.53 0.34 0.04 0.00 94.79

17 59.02 0.53 17.08 5.86 0.10 1.99 6.04 3.10 2.73 0.31 0.00 0.01 96.78

Average 64.33 0.63 12.76 7.23 0.21 1.44 3.58 3.2 3.73 0.23 0.01 0.03 97.37

Table (continued)

Eruption Unit MI Sample No. (#spot/average) SiO2 (wt%) TiO2

tel01c

STS TEL04K TEL04K2 6a

TEL04K TEL04Ko 1a

149 66.72 0.73

Average 50.07 2.02

126 45.94 0.8

TEL04K TEL04K TEL04K TEL04Kol TEL04Kol TEL04Kol 2a 2b 2c Average 48.17 2.21

133 48.61 2.36

129 48.11 2.03

ACCEPTED MANUSCRIPT 15.06 3.99 0.14 0.97 2.16 4.05 4.79 0.54 0 0.03 95.43

17.24 5.87 0.23 1.56 5.42 3.63 2.4 0.3 0 0 97.83

14.51 9.44 0.18 1.99 4.66 3.6 2.58 0.18 0 0 96.92

15.44 2.48 0.07 0.43 2.85 3.72 2.88 0.26 0 0 95.58

16.26 6.99 0.15 4.55 10.29 3.36 0.87 0.27 0.07 0.01 94.9

16.38 10.1 0.19 6.15 12.43 2.2 0.33 0.05 0 0 94.59

17.52 9.41 0.17 3.63 11.63 2.77 0.51 0.16 0 0 96.17

18.35 8.4 0.16 2.59 11.83 2.6 0.54 0.16 0 0 95.59

18.07 7.72 0.09 2.54 11.76 2.81 0.62 0.19 0 0 93.93

T

Al2O3 FeO* MnO MgO CaO Na2O K2O P2O5 Cr2O3 NiO TOT

TEL04K TEL04Kol8 a

Average 45.49 0.71 18.72 8.5 0.14 6.74 11.41 2.18 0.41 0.13 0 0 94.42

215 44.76 0.83 18.52 9.45 0.21 4.77 12.98 2.38 0.36 0.15 0 0 94.41

Average 46.85 0.76 17.98 9.04 0.19 5.52 10.5 2.29 0.48 0.1 0 0 93.7

Average 43.7 0.73 17.79 9.51 0.19 5.52 12.99 2.04 0.41 0.1 0 0 92.98

Table (continued)

CR

TEL04K TEL04Kol6 a

TEL04B TEL04B 1

M

AN

US

TEL04K TEL04Kol5 a

ED

MI Sample No. (#spot/average) SiO2 (wt%) TiO2 Al2O3 FeO* MnO MgO CaO Na2O K2O P2O5 Cr2O3 NiO TOT

STS TEL04K TEL04Kol4 a

PT

Eruption Unit

IP

Table (continued)

Average 46.4 0.54 18.91 9.31 0.18 5.63 13.17 1.69 0.29 0.11 0 0 96.22

TEL04B TEL04B2 a

TEL04B TEL04B2 b

TEL04B TEL04B2 c

TEL04B TEL04B2 d

Average 48.8 0.53 16.45 7.91 0.13 5.61 12.46 1.92 0.29 0.1 0 0 94.2

Average 48.78 0.7 17.38 7.92 0.18 6.2 12.71 1.87 0.29 0.08 0 0 96.1

Average 46.91 0.66 17.7 7.77 0.16 5.99 12.33 1.76 0.27 0.09 0 0 93.64

185 46.55 0.59 19.76 7.49 0.24 5.91 11.65 1.73 0.29 0.03 0 0 94.23

AC

CE

Eruption STS Unit TEL04B TEL04B TEL04B TEL04B TEL04B TEL04B TEL04B TEL04B TEL04B MI Sample TEL04B3a TEL04B5a TEL04B8a TEL04B8b t4bcpx6oa t4bcpx6ob t4bpx6a t4bpx6a t4bpx6a2 No. (#spot/average) 170 Average Average Average 233 235 Average Average 252 SiO2 (wt%) 46.35 46.24 46.67 46.21 54.12 53.48 55.51 54.38 55.29 TiO2 0.72 0.67 0.57 0.62 0.74 1.05 0.7 0.89 0.77 Al2O3 17.76 17.02 19.05 19.41 13.99 14.89 13.79 13.79 14.34 FeO* 8.45 9.82 8.01 8.05 10.68 9.85 8.71 10.82 7.88 MnO 0.21 0.17 0.12 0.26 0.18 0.23 0.25 0.27 0.11 MgO 6.55 5.95 5.93 6.25 3.33 2.66 2.95 3.01 3.03 CaO 10.89 11.8 11.22 11.37 6.91 7.21 5.86 6.39 6.06 Na2O 2.01 2.02 1.85 1.85 2.85 2.91 3.25 3.58 3.38 K2O 0.27 0.27 0.24 0.23 1.75 1.9 2.54 0.9 2.6 P2O5 0.03 0.08 0.06 0.05 0.22 0.17 0.4 0.18 0.32 Cr2O3 0 0 0 0 0 0 0 0 0 NiO 0 0 0 0 0 0 0 0 0 TOT 93.22 94.04 93.72 94.3 94.76 94.35 93.95 94.21 93.78

ACCEPTED MANUSCRIPT

APPENDIX 6: Standards compositio n Inclusion duplicate (n= 2)

T

International Glass standard

TEL04K2a

GL07-D30-1

PT

ED

M

AN

US

CR

IP

Err. TEL04K2a TEL04K2a Averag S.D. Err. LiteratureaLiterature Duplicat S.D. Rel. 1 2 e % Rel. % a e % % SiO2 (%) SiO2 0.0 48.17 0.04 0.59 49.07 48.83 0.17 (%) 48.14 48.20 0 0.1 2.21 0.03 TiO2 TiO2 2.23 2.19 4.90 0.74 0.87 0.09 7 0.0 17.52 0.06 Al2O3 Al2O3 17.57 17.48 0.95 16.44 16.97 0.37 3 0.0 9.41 0.04 FeO 9.43 9.38 4.33 8.86 8.76 0.07 1 0.5 0.17 0.09 MnO 0.11 0.24 13.58 0.16 0.25 0.07 8 0.0 3.63 0.29 MgO 3.84 3.43 10.03 10.15 9.41 0.52 7 0.0 CaO 11.65 11.61 11.63 0.03 0.09 11.65 12.01 0.25 3 0.1 2.77 0.02 Na2O Na2O 2.76 2.78 4.07 2.13 1.86 0.19 2 0.2 0.51 0.01 K2O K2O 0.51 0.50 6.96 0.07 0.09 0.01 6 P2O5 P2O5

0.18

Cr2O3 Cr2O3

0.00

0.00

NiO

0.00

0.00

Total

0.03

24.21 125.9 0.00 2

CE

Cl

0.16 0.00

0.16

0.29

0.23

0.09 14.76

0.08

0.05

0.07

0.02

96.63

96.27

96.45

0.25

AC

SO3

0.14

0.09

0.03

0.00

0.06

1.0 0.02 0

0.00

0.00

33.10

n.d.

0.00

-

-

n.d.

0.26

-

-

n.d.

0.00

-

-

99.30

99.40

Med. 0.0 (%) 3 «n. d.» is for no data and « - » is for no results for calculations

0.00

7.29

Med. (%)

0.0 9

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

M

AN

US

CR

IP

T

aFine and Stolper, 1986; Hauri et al., 2002, 2006

PT

ED

M

AN

US

CR

IP

T

ACCEPTED MANUSCRIPT

AC

CE

Figure 1

CR

IP

T

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

M

AN

US

Figure 2

US

CR

IP

T

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

M

AN

Figure 3

AN

US

CR

IP

T

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

M

Figure 4

AN

US

CR

IP

T

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

M

Figure 5

CR

IP

T

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

M

AN

US

Figure 6

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

M

AN

US

CR

IP

T

Figure 7

AN

US

CR

IP

T

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

M

Figure 8

AN

US

CR

IP

T

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

M

Figure 9

T

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

M

AN

US

CR

IP

Figure 10

M

AN

US

CR

IP

T

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

Figure 11

ACCEPTED MANUSCRIPT

Table 1 - Bulk analysis of rock samples at Telica. Material

Scoria

Bomb February 2011 - 2012 2011 - 2012 1982

Age Sample

Scoria

Bomb Scoria Scoria Scoria Scoria February < 2,300 yrs < 2,300 yrs < 2,300 yrs < 2,300 yrs 1982 BP BP BP BP

TEL03A TEL03B TEL01 TEL02 TEL04K TEL 07E TEL 07A TEL04B N12°36'26. N12°36'26. N12°36'26. N12°36'26. N12°36'03. N12°01'28. N12°01'28. N12°36'03. 4'' 4'' 3'' 3'' 4'' 2'' 2'' 4'' W86°50'22. W86°50'22. W86°50'13. W86°50'16. W86°53'00. W86°10'32. W86°10'32. W86°53'00. 5'' 5'' 6'' 3'' 8'' 6'' 6'' 8''

SiO2, wt%aSiO2, wt%a TiO2TiO2 Al2O3Al2O3

56.11

56.95

54.90

53.50

51.45

49.58

49.63

53.61

0.66

0.65

0.76

0.74

0.78

1.60

0.87

0.71

CR

Lat.

IP

T

Long.

16.43

17.19

17.65

16.92

16.57

17.74

17.08

9.54

9.46

9.57

9.47

11.01

10.82

11.15

10.19

MnO

0.16

0.16

0.18

0.18

0.17

0.16

0.19

0.18

MgO

2.35

2.50

3.88

4.32

4.78

6.02

5.24

3.89

CaO Na2ONa2O

8.01

8.19

10.06

10.30

10.41

10.08

10.71

9.46

2.62

2.53

2.90

2.69

2.08

2.63

2.17

1.83

1.38

1.36

1.03

0.95

0.93

0.82

0.81

0.82

0.14

0.14

0.11

0.11

0.11

0.18

0.14

0.12

L.O.I.

0.72

0.76

0.78

0.72

0.60

0.00

0.00

2.70

Total Mg# (FeOT)

98.08 0.33

99.13 0.34

101.36 0.45

100.63 0.47

99.24 0.46

98.47 0.52

98.66 0.48

100.59 0.43

AN

M

ED

P2O5P2O5

PT

K2OK2O

US

16.39

Fe2O3bFe2 O3b

AC

CE

aMajor elements analysis performed on the wavelength‐dispersive X‐ray fluorescence (XRF). bLOI values have not been corrected for oxygen uptake upon conversion of FeO to Fe2O3 in the furnace.

ACCEPTED MANUSCRIPT

Table 2 – Electron microprobe analyses of Telica matrix volcanic glass.

TEL0 3B1gl ass 1

TEL0 4Ko2 glass 1 2

5

6

(wt 61 62 62 62 62 62 Si .% .3 .9 .1 .4 .3 .8 4 4 6 5 0 6 O2 )

1. 0. 0. 0. 0. 0. 16 93 87 93 87 99

Al 2O

13 14 13 14 14 14 .9 .0 .5 .2 .0 .1 3 0 9 1 1 3

PT

3

8 6 2 . 9 1 0 . 9 7 1 3 . 3 7 8 . 1 2 0 . 1 7 2 . 0 0 3 . 9 7 3 . 6 4

52. 22

0.3 3

9. 8. 8. 7. 8. 7. 43 13 37 83 24 66

M n O

0. 0. 0. 0. 0. 0. 18 23 13 20 12 15

M gO

0. 0. 0. 1. 0. 0. 53 81 70 05 75 78

Ca O

3. 3. 3. 3. 3. 3. 26 34 42 77 53 70

Na 2O

3. 3. 4. 4. 4. 4. 91 98 25 18 04 42

AC

CE

Fe O

Frag. No. Olivin e Borde r:

TEL04 TEL0 TEL04K6o K10cp 4K26o glass xglass glass 1 1 1

( S w 5 5 i t. 1. 1. O % 6 6 3 4 2 ) T 2. 2. i 0 0 O 2 0 2

52. 06

0.90

1.19

1.3 1

15.94

16.7 2

12. 96

9.91

8.69

10. 82

9.4 6

F e O

0.2 9

M n O

0. 0. 1 1 7 7

0.28

0.25

0.2 6

14. 70

M g O

4. 3. 3 9 2 9

3.44

2.65

3.6 4

18. 36

C a O

8. 8. 0 2 4 3

8.15

8.35

7.0 1

0.3 4

N a2 O

2. 2. 9 9 6 8

2.69

3.44

2.6 1

2.5 7

A l2 O 3

1 5. 5 4 1 0. 3 7

52.84

54.9 5

1 5. 9 0 1 0. 6 7

ED

Ti O2

7 6 0 . 5 1 0 . 9 7 1 4 . 0 7 8 . 6 2 0 . 1 6 1 . 0 9 4 . 1 8 4 . 1 5

US

4

T

TEL03 Abulk 1 2 3

Frag Frag. No. . No. Olivine Olivin Border: e TEL04K6o Border glass :

IP

No. Frag. Spot No. Olivi ne Borde r:

CR

Bulk matrix spots:

Frag. No. Olivi ne Bord er:

AN

Frag. No.

M

No. Spot

ACCEPTED MANUSCRIPT

150

3 4 64 86 n. 1 4 0 50 0 d. 0 0

0

0

350

(µ g/g 14 20 16 0 0 4 S )

0

0

8

0

(µ g/g 21 16 16 17 21 17 80 30 50 50 90 20 Cl )

0 1 4 9 0 9 9 . 6 5

749

250

98. 31

-

-

C ( r2 µ O g/ 3 g) ( N µ i g/ O g) ( µ g/ S g) ( µ C g/ l g) ( T w o t. t % al ) S / C l

3.0

1.75

2.0 0

0. 0. 3 3 1 3

0.27

0.23

0.3 7

n. n. d. d.

n.d.

320

0

n. n. d. d.

n.d.

0

100

1 6 4

2 9 6

485

8

216

7 3 0

7 6 0

1620

1430

153 0

9 7. 4 6

9 6. 6 6

96.09

98.2 2

93. 02

0. 0. 2 3 2 9

0.30

0.01

0.1 4

CE

0. 0. 0. 0. 06 12 10 - 00 «n. d.» is for no data and « - » is for no results Sulphur contents (μg/g) are converted from original SO3 (wt.%).

5

1.69

AC

S/ Cl

PT

ED

(wt 97 99 97 98 97 98 To .% .7 .1 .5 .9 .7 .6 2 3 2 9 4 2 tal )

1 8 8 1 7 2 0 9 7 . 4 2 0 . 1 1

O

AN

(µ Ni g/g O )

4 32 8 0 0 0

21 34 56 0 0 0

0

2

0.0 3

M

Cr (µ 2O g/g ) 3

O P

1. 1. 4 4 5 2

T

0. 0. 0. 0. 0. 0. 54 48 47 42 37 48

2

IP

P2 O5

K 0.0 1

CR

3. 4. 3. 3. 3. 3. 45 28 57 95 52 45

4 . 0 9 0 . 4 1

US

K2 O

3 . 2 8 0 . 4 0

No. Spot

SiO2 (wt.%)

Frag. No. Olivine Border: TEL04Ko2glass 1 2 51.63

51.64

Frag. No. Frag. No. Frag. No. Olivine Border: Olivine Border: Olivine Border: TEL04K6oglass TEL04K10cpxglass TEL04K26oglass 1 1 1 52.84

54.95

52.06

ACCEPTED MANUSCRIPT 2.02

2.00

0.90

1.19

1.31

Al2O3 FeO MnO MgO CaO

15.90 10.67 0.17 4.32 8.04

15.54 10.37 0.17 3.99 8.23

15.94 9.91 0.28 3.44 8.15

16.72 8.69 0.25 2.65 8.35

12.96 10.82 0.26 3.64 7.01

Na2O

2.96

2.98

2.69

3.44

2.61

K2O

1.45

1.42

1.69

1.75

2.00

P2O5

0.31

0.33

0.27

0.23

n.d. n.d. 164 730 97.46 0.22

n.d. n.d. 296 760 96.66 0.39

n.d. n.d. 485 1620 96.09 0.30

320 0 8 1430 98.22 0.01

CE

PT

ED

M

IP

0.37

CR US

AN

(µg/g) (µg/g) (µg/g) (µg/g) (wt.%)

AC

Cr2O3 NiO S Cl Total S/Cl

T

TiO2

0 100 216 1530 93.02 0.14

ACCEPTED MANUSCRIPT Table 3 – Electron microprobe analyses of Telica glass inclusions. Inclusion Phenocryst Phenocryst chemistry (Fo) Cpx (Ens) Inclusion shape and size Type Bubble (vol %) SiO2, wt.%

TEL04K20a _ 75 s50X45 Fully enclosed -

TEL04K6a 73 i65X53 Fully enclosed -

TEL04K8a 72 e60X35 Fully enclosed -

TEL04K8b 72 e50X30 Fully enclosed -

TEL04K9a 75 e107X50 Fully enclosed -

TEL04K9cI 69 e90X45 Fully enclosed 60.8

58.9

64.2

60.3

57.4

1

0.9

0.7

0.7

0.8

0.8

Al2O3 FeO* MnO MgO CaO

15.8 10.8 0.2 4.2 7.6

15.7 9.2 0.1 3 6.8

14.6 6.2 0.2 1.8 4.5

15.5 8 0.2 2.3 5.3

15.7 9.3 0.3 3.6 6.9

14.2 8.7 0.2 2.9 6.6

Na2O

3

2.8

3.4

3.4

3.5

3.2

K2O

1.8

2.4

3.9

4

2.2

2.1

0.2

0.014 0.038 100 1122

n.d. 0.028 100 1093

H2O (wt.%) nSIMS CO2 (µg/g) nSIMS S (µg/g) EMPA Cl (µg/g) EMPA Sat. Press. (MPa) by PPL

1.37 n.d. 519 1785 14

1.68 n.d. 300 1960 2.6

IP CR

0.4

0.3

0.3

0.3

n.d. n.d. 100 1077

n.d. n.d. 100 1085

0.004 0.032 100 1110

n.d. 0.114 100 1098

0.62 n.d. 256 2900 16.7

0 n.d. 92 2320 -

0.98 n.d. 356 2220 1.2

0 n.d. 409 2070 -

ED

M

AN

0.2

Cr2O3 NiO TOT T (°C) by KP

US

P2O5

T

55.4

TiO2

AC

CE

PT

Sulphur contents (μg/g) converted from original SO3 (wt%). Saturation pressures (MPa) , based on H2O-CO2 concentrations in MI, were calculated with solubility model of Papale et al. (2006).

Inclusion Phenocryst Phenocryst chemistry (Fo) Cpx (Ens) Inclusion shape and size

TEL04K9d 71 e80X40

TEL04K10a 70 e100X50

TEL04K10b 69 e90X45

Fully enclosed -

Fully enclosed -

SiO2, wt.%

60.2

TiO2

0.7

Al2O3 FeO* MnO MgO CaO

13.5 10.5 0.4 3 6.9

Type Bubble (vol %)

Fully enclosed -

TEL04K10c 76 e85X35 Fully enclosed -

TEL04K13a 73 e30X15 Fully enclosed -

TEL04K14b 67 e50X25 Fully enclosed -

60.2

59.7

56.5

57.7

63.5

0.9

0.9

1

1

1.1

14.7 8.9 0.2 2.4 5.6

15 8.8 0.2 2.6 5.6

16.9 9.8 0.2 3.7 7

14.9 10.1 0.2 3.1 6.9

13.3 7.9 0.2 1.8 5.1

ACCEPTED MANUSCRIPT 2.7

3.5

3.7

2.8

3.8

3.4

K2O

1.9

3.2

3.2

1.7

1.9

3.5

P2O5

0.2

0.4

0.4

0.2

0.3

0.3

Cr2O3 NiO TOT T (°C) by KP

n.d. n.d. 100 1086

0.013 0.032 100 1083

0.016 0.037 100 1090

n.d. 0.038 100 1111

0.018 n.d. 100 1089

n.d. 0.021 100 1076

H2O (wt.%) nSIMS CO2 (µg/g) nSIMS S (µg/g) EMPA Cl (µg/g) EMPA Sat. Press. (MPa) by PPL

0 n.d. 445 2360 -

0.91 n.d. 46 2490 1.1

1.4 309* 176 2740 17.9

0 n.d. 705 1950 -

0 n.d. 102 2020 -

1.16 n.d. n.d. 4135 -

CR

IP

T

Na2O

Type Bubble (vol %)

TEL04K18c

TEL04K21aI

TEL04K22a

TEL04K22b

TEL04K25a

74 e60X35 Fully enclosed -

66 e35X10 Fully enclosed -

73 e35X15 Fully enclosed -

71 i330X130 Fully enclosed -

69 i110X50 Fully enclosed -

72 e110X15 Fully enclosed -

63.7

57.3

59.4

59.7

62.3

0.7

1.1

0.8

0.7

0.8

15.1 8.1 0.2 3.3 5.8

13 8.7 0.2 2.4 4.7

15.2 10.6 0.3 3.2 6.2

15.4 8.8 0.2 3.6 6

14.8 9.2 0.2 3.4 6.1

15.3 7.1 0 2.2 5.1

3.5

3.6

3.5

3.2

3.3

3.2

2.4

2.8

2.3

2.4

2.3

3.7

0.3

0.3

0.3

0.2

0.3

0.4

0.016 0.021 100 1104

n.d. 0.026 100 1087

n.d. n.d. 100 1099

0.024 0.011 100 1109

0.023 0.057 100 1106

0.044 n.d. 100 1084

1.72 n.d. 344 2595 16.6

0 n.d. 110 1925 -

0 n.d. 421 2410 -

0.95 n.d. 340 2175 1.3

0 n.d. 344 2130 -

1.09 n.d. 0 2580 1.6

60.5

TiO2

0.7

K2O P2O5 Cr2O3 NiO TOT T (°C) by KP

CE

Na2O

AC

Al2O3 FeO* MnO MgO CaO

PT

SiO2, wt.%

M

TEL04K15a

ED

Inclusion Phenocryst Phenocryst chemistry (Fo) Cpx (Ens) Inclusion shape and size

H2O (wt.%) nSIMS CO2 (µg/g) nSIMS S (µg/g) EMPA Cl (µg/g) EMPA Sat. Press. (MPa) by PPL

AN

US

Inclusion shapes are resumed as ‘’s’’, spherical shape; ‘’e’’, elliptical shape; ‘’I’’, irregular shape. ‘’A’’ is for textural assemblage order among other inclusions assemblage found in the same crystal.

ACCEPTED MANUSCRIPT

IP

59.13

68.8

0.94

0.85

0.75

16.42 6.35 0.24 2.68 6.65

12.94 9.93 0.21 4.19 7.18

15.41 2.99 0.08 1.26 3.89

4.2

3.43

3.15

3.7

AN

T

Inclusion Phenocryst TEL04K25b TEL04K18a TEL03B1 TEL01 TEL01b TEL01c Phenocryst chemistry (Fo) Cpx (Ens) 71 64 74 74.3 70.9 70.9 Inclusion shape and size e80X20 e42X15 i90X50 i130X30 i«?» i«?» Type Fully enclosed Fully enclosed Fully enclosed Fully enclosed reentrant reentrant Bubble (vol %) -

2.26

2.23

2.85

66.1

65.6

0.5

0.6

1

Al2O3 FeO* MnO MgO CaO

17.7 6 0.1 2.1 6.2

13.1 7.4 0.2 1.5 3.7

15.7 4.2 0.1 1.1 2.4

Na2O

3.2

3.3

K2O

2.8

3.8

P2O5

0.2

0.004 0.013 100 1084

0.007 0.034 100 1070

H2O (wt.%) nSIMS CO2 (µg/g) nSIMS S (µg/g) EMPA Cl (µg/g) EMPA Sat. Press. (MPa) by PPL

0 n.d. 0 2210 -

1.73 102* 72 2840 27.9

5

0.5

0.3

0.18

0.26

n.d. 0.026 100 1054

n.d. n.d. 100 1139

n.d. n.d. 100 1131

n.d. n.d. 100 1130

0 n.d. 695 4280 -

0.25 65 332 621 2.9

0.46 17 2079 2215 13.2

n.d. n.d. 136 -

AC

CE

PT

ED

M

0.3

Cr2O3 NiO TOT T (°C) by KP

60.72

CR

61

TiO2

US

SiO2, wt.%

Inclusion Phenocryst Phenocryst chemistry (Fo) Cpx (Ens) Inclusion shape and size

TEL04K26a

TEL04Ko1a

TEL04Kol2a

TEL04Kol2b

TEL04Kol2c

TEL04Kol4a

84 i45X20 Fully enclosed -

80.1 e150X120 Fully enclosed 5, 12, 12

82.1 s35X30 Fully enclosed -

82.5 s30X25 Fully enclosed -

82.6 s25X20 Fully enclosed 10

84.9 e30X25 Fully enclosed -

SiO2, wt.%

52.7

47.96

50.81

52.17

52.21

47.12

TiO2

2.1

0.82

2.37

2.62

2.25

0.69

Al2O3

17

16.69

18.76

20.38

20.04

18.26

Type Bubble (vol %)

ACCEPTED MANUSCRIPT 11.5 0.2 5 10.8

12.1 0.21 6.92 12.68

7.19 0.16 4.58 12.45

4.94 0.13 2.94 13.14

5.03 0.07 3.35 13.04

11.23 0.17 8.71 11.17

Na2O

3.5

2.24

2.97

2.89

3.12

2.13

K2O

0.9

0.34

0.55

0.6

0.69

0.4

P2O5

0.3

0.04

0.17

0.18

0.21

0.12

0.1 0.013 100 1132

n.d. n.d. 100 1165

n.d. n.d. 100 1155

n.d. n.d. 100 1090

n.d. n.d. 100 1101

n.d. n.d. 100 1267

1 n.d. 941 560 1.4

4.1 36 2368 1844 103

1.73 1584 1223 468 166

1.87 1642 1245 484 130

1.72 1359 1254 489 111

810 1540 -

TEL04B1

TEL04B2a

TEL04B2b

85.5 s130X125 Fully enclosed 9

85.1 s130X125 Fully enclosed -

50.91

49.61

SiO2, wt.%

47.5

K2O

AC

Na2O

TEL04Kol8 a

IP

68.5 e35X30 Fully enclosed -

81.4 i210X140 Reentrant -

85.2 s40X40 Fully enclosed 9

50.55

46.72

47.22

0.87

0.86

0.77

0.5

0.54

0.67

19.65 9.35 0.22 5.56 13.78

20.44 9.65 0.12 3.19 11.92

18.73 10.71 0.21 6.48 13.69

17.79 10.96 0.21 8.9 12.45

16.39 9.3 0.15 7.96 12.45

16.7 10.06 0.2 8.32 12.27

2.53

2.61

2.15

1.59

1.91

1.8

CE

TiO2 Al2O3 FeO* MnO MgO CaO

81.1 e50X35 Fully enclosed -

TEL04Kol6a

PT

Type Bubble (vol %)

TEL04Kol5a

ED

Inclusion Phenocryst Phenocryst chemistry (Fo) Cpx (Ens) Inclusion shape and size

M

AN

US

H2O (wt.%) nSIMS CO2 (µg/g) nSIMS S (µg/g) EMPA Cl (µg/g) EMPA Sat. Press. (MPa) by PPL

CR

Cr2O3 NiO TOT T (°C) by KP

T

FeO* MnO MgO CaO

0.38

0.55

0.43

0.28

0.29

0.28

P2O5

0.16

0.11

0.1

0.1

0.09

0.08

Cr2O3 NiO TOT T (°C) by KP

n.d. n.d. 100 1187

n.d. n.d. 100 1117

n.d. n.d. 100 1212

n.d. n.d. 100 1231

n.d. n.d. 100 1270

n.d. n.d. 100 1279

H2O (wt.%) nSIMS CO2 (µg/g) nSIMS S (µg/g) EMPA Cl (µg/g) EMPA

1060 1906

700 1234

1050 1825

5.22 360 1020 1450

865 1542

865 1133

ACCEPTED MANUSCRIPT -

-

198

-

TEL04B3a

TEL04B5a

85.1

84.7

s135X130 Fully enclosed -

s130X125 Fully enclosed 11

84.3 s70X70 Fully enclosed ?

80.7 s90X90 Fully enclosed ?

84 e50X40 Fully enclosed ?

84.3 e45X40 Fully enclosed 10

SiO2, wt.%

48.95

48.27

48.4

48.9

48.94

48.05

TiO2

0.64

0.57

0.7

0.71

0.58

0.61

Al2O3 FeO* MnO MgO CaO

17.37 10.21 0.19 8.4 12.15

19.27 10.14 0.27 8.07 11.42

17.32 11.51 0.24 8.91 10.68

17.82 10.9 0.18 6.66 12.36

19.15 10.05 0.15 7.65 11.32

19.18 10.44 0.29 8.06 11.28

Na2O

1.73

1.69

1.96

2.11

1.86

1.83

K2O

0.27

0.28

0.26

0.28

0.24

0.23

P2O5

0.09

0.03

Cr2O3 NiO TOT T (°C) by KP

n.d. n.d. 100 1281 895 875 -

CR

US

AN

M

0.08

0.06

0.05

n.d. n.d. 100 1297

n.d. n.d. 100 1243

n.d. n.d. 100 1264

n.d. n.d. 100 1274

780 1754 -

950 1622 -

990 1426 -

985 1332 -

870 1039 -

ED

0.03

n.d. n.d. 100 1273

AC

CE

H2O (wt.%) nSIMS CO2 (µg/g) nSIMS S (µg/g) EMPA Cl (µg/g) EMPA Sat. Press. (MPa) by PPL

TEL04B8b

IP

TEL04B2d

Type Bubble (vol %)

TEL04B8a

-

TEL04B2c

PT

Inclusion Phenocryst Phenocryst chemistry (Fo) Cpx (Ens) Inclusion shape and size

-

T

Sat. Press. (MPa) by PPL

Inclusion Phenocryst t4bcpx6oa t4bcpx6ob t4bpx6a t4bpx6a t4bpx6a2 Phenocryst chemistry (Fo) Cpx (Ens) 71.1 71.1 74 74 74 Inclusion shape and size e60X45 e55X40 e45X35 e45X35 e50X40 Type Fully enclosed Fully enclosed Fully enclosed Fully enclosed Fully enclosed

ACCEPTED MANUSCRIPT Bubble (vol %)

-

-

-

-

SiO2, wt.%

56.82

56.19

58.55

56.72

58.78

TiO2

0.75

1.03

0.73

0.89

0.81

Al2O3 FeO* MnO MgO CaO

14.18 11.31 0.2 4.05 7.84

14.68 10.59 0.26 3.79 8.57

13.97 9.35 0.26 3.87 7.07

13.13 11.43 0.28 4.73 8.48

15.02 8.45 0.12 3.5 6.78

Na2O

2.88

2.86

3.26

3.33

3.53

K2O

1.76

1.85

2.54

0.83

2.71

P2O5

0.21

0.17

0.39

0.18

Cr2O3 NiO TOT T (°C) by KP

n.d. n.d. 100 1141

n.d. n.d. 100 1138

n.d. n.d. 100 1141

n.d. n.d. 100 1141

H2O (wt.%) nSIMS CO2 (µg/g) nSIMS S (µg/g) EMPA Cl (µg/g) EMPA Sat. Press. (MPa) by PPL

292 1700 -

425 1760 -

621 2220 -

IP

0.32 n.d. n.d. 100 1141

977 1285 -

196 2440 -

CR

US AN

M ED PT CE AC

T

-

ACCEPTED MANUSCRIPT

Table 4 – NanoSIMS 50L data for CO2– H2O – S –Cl in Telica glass inclusions. Inclusion (NanoSIMS #)

CO2 (µg/g) H2O (wt%) S (µg/g) Cl (µg/g)

65 17

0.25 0.46

621 2215

3540 2677

TEL04K10b

309

1.4

185

3180

TEL04K18a

102

T

TEL01 (Cpx) TEL01a TEL01b

1.73

82

2626

TEL04K9I*

1211*

2.01

305

2139

TEL04K9II

80

293

2116

1.16

102

2576

21

0.67

45

1634

5

0.25

64

192

36

4.07

2368

1844

1584

1.64

1223

468

1642

1.66

1245

484

1359

1.58

1254

489

334

1.8

1985

1375

98

1.26

82

2535

56

0.95

59

2781

360

5.22

1513

1442

329

4.8

1354

1281

353

4.96

1464

1379

STD 519

12.9

7

10.9

13.9

STD D30

14.8

7.3

11.9

15.6

Total STD DEV.

13.8

7.1

11.4

14.8

CR

1.88

24

TEL04K14try

US

TEL04K14tryb TEL04K18z TEL04K (Ol)

AN

TEL04Kol1 TEL04Kol2a TEL04Kol2b

M

TEL04Kol2c TEL04Kol3a

ED

TEL07 (Cpx) TEL07PXB TEL07APXiG2

PT

TEL04B (Ol) TEL04B1

AC

CE

TEL04B2a TEL04B2b

IP

TEL04K (Cpx)

No data are determined for « - ». TEL04K9I* is discarded, because of visible contamination of graphite from carbon quoting find along microfractures interpreted with NanoSIMS12C− mapping; 1211 ppm CO2

ACCEPTED MANUSCRIPT Highlights

CE

PT

ED

M

AN

US

CR

IP

T

The CO2 content in Telica olivine melt inclusions are the second highest in Nicaragua From 15 km to surface, magma evolves in two crystallizing depth zones below Telica Augite-hosted melt inclusions record shallow residual Cl-rich magma Magma mixing at ~2.4–5.0 km preceded VEI > 3 Scoria Telica Superior eruptions Magma viscosity increase at <2.4 km depth likely governed more recent (VEI 1-2) Telica magmatic explosive eruptions

AC

    