To better estimate the extrusive flux of the Central American Arc, from 2002-2008, we obtained sixty one high precision 40Ar/39Ar ages on geographically well-situated lavas and tephra from Costa Rica and Nicaragua. Here, we describe a number of observations encountered during this study using four examples that well document the precision, accuracy and general reliability of the 40Ar/39Ar ages. First, low K2O values, particularly in samples from Nicaragua, is a major limitation in or attempts to obtain reliable dates on samples under 1 My. Second, extensive weathering of samples due to the tropical climate of Central America has resulted in various levels of argon loss even when the hand sample appeared unaltered. Third, our field and geochronological data lead us to conclude that eruptive rates have not been constant over the past 15 to 20 My, but rather appears punctuated by gaps of up to several million years. We attempted to address the temporal gaps in several ways. First, geochemical analyses were used to identify samples that may have erupted during time periods without known volcanism. For example, U/Th values in the active Central American arc are significantly higher than those obtained from the Miocene Coyol Group except for four samples with intermediate values that were dated to determine if their ages were intermediate as well. However, all of these samples were found to be from a period with known volcanism. Second, we sought to locate the oldest sections of the active arc and the youngest sections of the Coyol Group in order to better constrain the timing and duration of the apparent gap in volcanic productivity. This approach also failed to locate samples from periods without known volcanism. When these methods proved largely unsuccessful, our focus shifted to dating regions of minor volcanism between the active and Coyol volcanic fronts as well as between Cosigüina and San Cristóbal, the longest stretch of the Central American Volcanic Front without active volcanism. This effort yielded ages on samples ranging from 1.1 to 3.6 Ma and, thus, substantially reduced the apparent volcanic gap in Nicaragua.
Keywords: Central American Volcanic Front, Geocrhonology, 40Ar/39Ar
Resumen
Con el principal objetivo de realizar mejores cálculos de la producción volcánica, en el Arco Volcánico de América Central, realizamos 61 dataciones 40Ar/39Ar en tefras y lavas en localidades de frente volcánico de Costa Rica y Nicaragua por medio de varias campañas de campo y análisis de laboratorio (preparación de muestras, envió al reactor nuclear, espectroscopía de masas de las muestras radiactivas, al laboratorio de gases nobles del Departamento de Ciencias de la Tierra y Planetarias de la Universidad Rutgers) del 2002-2008. En este artículo presentamos diferentes observaciones que resultaron de este estudio donde se describe la precisión, exactitud y la confiabilidad de las edades 40Ar/39Ar. El primer resultado muestra las limitaciones de esta técnica en muestras baja en K2O con edades < 1 Ma, especialmente en Nicaragua. El segundo resultado muestra los efectos de la pérdida de Ar por los procesos de meteorización típicos del trópico, inclusive afectando las muestras sanas a nivel críptico. El tercer resultado sugiere que las razones eruptivas del frente volcánico no han sido constantes durante los últimos 15-20 Ma, sino más bien representan ciclos eruptivos. Con el propósito de evaluar los hitos entre diferentes ciclos eruptivos, evaluamos los datos geoquímicos a lo largo del frente volcánico, en especial las relaciones U/Th, cuyos valores son inferiores (Grupo Coyol) en Nicaragua, comparados con el frente volcánico activo. Sin embargo, las muestras seleccionadas con base en geoquímica, resultaron ser de edades conocidas y no produjeron datos que llenan los hiatos entre ciclos de actividad volcánica. La siguiente estrategia que manejamos fue extender el muestreo en el Grupo Coyol, inmediatamente detrás del frente volcánico activo, sin embargo, esta estrategia no fue exitosa para encontrar muestras que llenaran los hiatos de actividad. No obstante, obtuvimos edades de 1.1-3,6 Ma entre los volcanes Cosiguina and San Cristóbal, los cuales limitan el hiato de actividad y mejoran nuestra percepción de los ciclos de actividad volcánica en América Central.
Palabras Clave: Frente Volcánico de América Central, Geocronología, 40Ar/39Ar
Introduction
The development of 40Ar/39Ar
dating methods made it possible to obtain high precision ages from a wide variety of igneous material. Radiometric dates appear simple to interpret, but this accessibility often leads to misunderstandings about how these dates are obtained and even what geologic events in the rock’s history the dates refer to. It is therefore important to be clear about what argon ages mean, how they are derived, and what affects their reliability.
]]>
The basis of the 40Ar/39Ar dating technique is a modification of the 40K/40Ar dating method based on the decay of 40K to 40Ar with a half-life of 1.26 Ga (Beckinsale and Gale, 1969).
The40Ar/39Ar technique differs from the 40K/40Ar method in that the K of a sample, is measured by conversion of 39K to 39Ar
by neutron bombardment in nuclear reactor. Since the ratio of 39K to 40K is known, the derived 39Ar can be used as proxy for 40K
measurement (Merrihue and Turner, 1966). This technique has an advantage over conventional K-Ar dating, because both the radioactive parent 39Ar and the daughter 40Ar can be measured by the same method, mass spectrometry, on the same sample split at the same time. This permits the measurement of smaller sample sizes (since splits are not needed for K and Ar measurement) and ]]>
]]>
The 40Ar/39Ar technique can be applied to any K-bearing igneous rock or mineral phase that has not experienced extensive weathering or reheating since its formation. Within each rock, 40K is perpetually decaying, however 40Ar will not begin to accumulate ]]>
40Ar is a relatively large atom and cannot escape the typical silicate mineral crystal lattice unless the sample is reheated or physically abraded. Once the mineral has cooled beneath its blocking temperature, the “clock” begins and the daughter 40Ar
accumulates. Blocking temperatures vary for different minerals, with amphibole ~500ºC (Hanson and Gast, 1967), biotite ~400ºC, and plagioclase ~200ºC (Berger and York, 1981). It is important to note that when we refer to a 40Ar/39Ar “age”, we are referring to this moment in time and not necessarily the formation age. For example, plutonic rocks can remain above their blocking temperature for millions of years and metamorphic rocks can have their radiometric clocks reset by subsequent ]]>
While 40Ar/39Ar dating is an extremely useful tool, it is not without its drawbacks, particularly in Central America and obtaining reliable ]]>
Radiogenic Argon
Prior to the eruption of a lava or tephra, the magma is generally well above the blocking temperature of most K-bearing minerals, and normally ]]>
40Ar/36AR ratio of the atmosphere (295.5, Steiger and Jager, 1977). All extrusive volcanic rocks therefore contain some 40Ar at the time of formation, but 40Ar of radiogenic origin (abbreviated 40Ar*) only begins to accumulate after they cool below their blocking temperature. A reliable date can be obtained if ]]>
40Ar* accumulates to distinguish from 40Ar derived from equilibration with the atmosphere.
The accumulation of 40Ar
* is controlled by two main factors: The initial potassium content of the sample and the elapsed time since it dropped below the blocking temperature. A high potassium sample can accumulate sufficient 40Ar* to yield a reliable age in less time than a low potassium sample. A problem arises however, when trying to date young rocks with low potassium content, because there is very little 40Ar
* to measure.
While potassium contents of Central American arc basalts are comparable to basalts from other arcs (generally between 0.5 and 2 wt.%, Luff, 1982; Marsh, 1982; Smith et al., 1980; Wills, 1974), there is significant internal variability with central Costa Rican values almost doubling those in northwest Nicaragua (Figure 1, Carr et al., ]]>
Volcanic material as young as a few thousand years has been successfully dated using 40Ar/39Ar dating techniques (Renne et ]]>
2O contents and selecting only those samples with the highest levels can help solve this problem.
Weathering and hydrothermal alteration can also be problematic for argon geochronology of lavas, because both of these processes can cause the release of 40Ar* that has been building up ]]>
Figure 2; Ching-Hua et al., 1994). Under certain conditions, weathering and hydrothermal alteration can cause the release of all pre-eruption 40Ar
*, even in samples less than 100 ka. For example, on the southern flank of San Cristóbal, one smooth, lobe-shaped lava was found (12.7315ºN/87.0540ºW) to be so hydrothermally altered that it could be easily broken by hand. This sample was not collected, but is a reminder of how quickly alteration can occur. Kis often lost from glass phases during weathering and hydration.
Another factor that can ]]>
40Ar to 36Ar will be 295.5. However, if preeruption 40Ar is not allowed to degass, the sample is said to have “excess” argon. This can be seen in an isochron diagram of sample VE-071605-2 (Figure ]]>
), which is described later in this paper.
Methods
Samples for 40Ar/39Ar dating discussed here, were obtained from the cores of large boulders or lava blocks collected from the study areas. Samples were observed under binocular and petrographic microscopes and ]]>
]]>
It is worth noting that all ages are based on either matrix or mineral separates as opposed to whole rock analysis. The main difference is that with the whole rock method, the entire sample is analyzed and may include radiogenic argon from weathered phases, pre eruption phenocrysts, or xenolithic contamination. By analyzing only the fine-grained matrix, we minimize these factors and limit the possibility of obtaining anomalously old ages. Even with this precaution, it can be difficult to eliminate pre-eruption phases entirely, particularly if they are small and of the same minerals that ]]>
For these reasons, samples were incrementally heated with up to 15 increasing temperature steps and the released gases measured each time. Each temperature step is based on an increase in laser power wattage (typically between 2 and 32 watts) and those power limits have been empirically determined to vary the temperature of the sample between ]]>
39Ar would form an age plateau at the 95% confidence level. Typically, the plateau age (if there is one) is reported along with the integrated age, which is simply the weighted average of each individual temperature step and is essentially the same as a total fusion age.
The incremental-heating method is also beneficial because multiple ]]>
Figure 2). In this type of diagram, 36Ar/40Ar is plotted against 39Ar/40Ar In an ideal sample, the individual temperature steps plot along a single ]]>
36Ar/40Ar value.
Case Studies
The first sample case study is RO-4 (Figure 2A), a lava collected from ]]>
39Ar, yet the apparent age is below 0. The low 40Ar* yield is most likely due to ]]>
In the absence of a reliable plateau, how should this data be interpreted? The isochron yields a negative age and an atmospheric initial 40Ar/36Ar, so we can conclude that there ]]>
The next sample case study is CR-TE-04- 8b (Figure 2B), a lava ]]>
Figure 3, Carr et al., 2007). All temperature steps form a 0.369 ± 0.012 Ma plateau at the 95% confidence level and the 40Ar* yields range from 4 to 16%. The isochron age agrees with the plateau age within the margin of error at 0.39 ± 0.03 Ma and the initial 40Ar/36Ar ratio is within the range of ]]>
40Ar* yield (relative to Nicaraguan volcanic front samples) is consistent with the majority of other Costa Rican samples. For all of these reasons,this is considered to be an excellent example of a reliable age in contrast to the previous sample RO-4.
The third case study is a pillow basalt from Isla Venado off Costa Rica’s Nicoya Peninsula (VE-07165-2; Figure 2C). This sample is overlain by a unit of reworked volcaniclastic sediments that was dated using biostratigraphy to 100 m.y.. This sample had an initial 40Ar/36Ar ]]>
Figure 2C) show the individual temperature steps falling from 105 Ma to 45 Ma throughout the analysis. Not only does this sample lack a plateau, but the isochron is also problematic with a mean squared weighted deviation of 14, well above the variation accounted for by the errors. One could attempt to refit the isochron with a different line by including only the high temperature steps that ]]>
40Ar/36Ar ratio, but we believe that would be over interpreting very low-quality data. Although a reliable age cannot be obtained from this sample, the high initial 40Ar/36Ar ratio suggests that it the age is probably much younger than the >100 Ma estimate based on field data. This may simply mean that the contact between the ]]>
The last sample case study is C-51 (Figure 2D), a lava collected from a ]]>
39Ar, the first temperature step is not included and is almost 2 Ma younger than ]]>
Volcanic History
Previous geochronological work in Nicaragua (Ehrenborg, 1996; Elming et ]]>
]]>
First, U/Th values in Nicaraguan volcanics experienced a significant increase between the Miocene and present arc (Figure 3, Plank et al., 2002), however there were three samples thought to be Miocene that had U/Th values between those typical for the Miocene and active arcs. These samples were selected for dating, because their intermediate U/Th values raised the possibility that their ages were intermediate as well. Sample LL-4 is from the Las Lajas volcano just east of the Nicaraguan depression and samples Bal-8 and Bal-10 are from the Balsamo ]]>
Second, an attempt was made to constrain both the timing and duration of this apparent hiatus in Nicaraguan volcanism by refining estimates for the youngest material within the Coyol Group as well as the oldest ]]>
40Ar/39Ar ages obtained using this method were reported in Carr et al. (2007), with a maximum age of 330 ka. Plank et al. (2002) suggested that Miocene volcanism in Nicaragua ceased ~7 Ma after the Coyol ]]>
Figure 4, Carr et al., 2007).
This approach extended the geographic distribution of the Coyol Group ~10 km into the Nicaraguan Depression, although did not extend its temporal range. Two samples were found in this area with ages of 1.13 ]]>
We also collected samples between Cosigüina and San Cristóbal, the longest stretch of the volcanic front without an active volcano. This effort revealed two previously unknown volcanic units: The Tinajas (2.5-1.3 Ma) and the Encanto (3.6-3.2 Ma). Both of these are described in Saginor et al. (2011). These two units also decreased the duration of the temporal gap in Nicaragua, leaving only ]]>
Acknowledgements
The authors acknowledge revisions and comments by Dr Guillermo Alvarado and the editorial work of Dr. Percy Denyer. This work was partially supported through the NSF Margins Program, grants EAR0203388 and NSF ]]>
Conclusions
Samples from the Nicaraguan volcanic front are generally lower in K2O than samples from the Costa Rican front. In addition, the front may be slightly younger in Nicaragua than Costa Rica. These two factors make obtaining reliable Nicaraguan
40Ar/49Ar ages difficult because it leads to lower 40Ar* yields.
The volcanic history in Nicaragua is incomplete and several attempts have been made to locate samples that fill in these temporal gaps. ]]>
References
BECKINSALE, R.D., & GALE, N.H. 1969: A reappraisal of the decay constants and branching ratio of 40K.- Earth Plan. Sci. Lett. 6: 289-294. [ Links ]
BERGER, G.W., & YORK, D. 1981: Geothermometry from 40Ar/39Ar dating experiments.- Geochim. Cosmochim. Acta. 45: 795-811. [ Links ]
CARR, M.J., FEIGENSON, M.D., PATINO, L.C., & WALKER, J.A. 2003: Volcanism and geochemistry in Central America: progress and problem.- In Eiler, J. & G. Abers (Eds): Inside the Subduction Factory.- AGU Geophysical Monograph 138: 153-179. [ Links ]
CARR, M.J., SAGINOR, I., ALVARADO, G.E., BOLGE, L.L., LINDSAY, F.N., MILLIDAKIS, K., TURRIN, B.D., FEIGENSON, M.D., & SWISHER III, C.C. 2007: Element fluxes from the volcanic front of Nicaragua and Costa Rica.- Geochem. Geophysics Geosystems. 8(6): 1525-2027. [ Links ]
CHING-HUA, L., ONSTOTT, T.C., CHIANGHWA C., & TPHOON, L. 1994: An assessment of 40Ar/39Ar dating for the whole- rock volcanic samples from the Luzon Arc near Taiwan.- Chem. Geol. 114(1-2): 157-178. [ Links ]
CONDIE, C. 2005: Continuity (Ba) and change (U) in Central American geochemistry: New evidence from the Miocene Balsamo Formation in El Salvador.- Rutgers University [Tesis Maestría]. [ Links ]
EHRENBORG, J. 1996: A new stratigraphy for the Tertiary volcanic rocks of the Nicaraguan highland.- Geol. Soc. Am. Bull. 108: 830- 842. [ Links ]
ELMING, S.A., LAYER, P., & UBIETA, K. 2001: A paleomagnetic study of Tertiary rocks in Nicaragua, Central America.- Geophys. J. Int. 147: 294-309. [ Links ]
LANPHERE, M. 2000: Comparison of conventional K–Ar and 40Ar/39Ar dating of young mafic volcanic rocks.- Quat. Res. 53(3): 294-301. [ Links ]
LUFF, I.W. 1982: Petrogenesis of the island arc tholeiite series of the South Sandwich Islands.- University of Leeds, U.K. Tesis Doc. [ Links ]
MARSH, B.D. 1982: The Aleutians. In Andesites: Orogenic andesites and related rocks.-R.S. Thorpe (ed.), Chichester, Wiley: 99-114. [ Links ]
MERRIHUE, C., & TURNER, G. 1966: Potassium-Argon dating by activation with fast neutrons.- J. Geophys. Res. 71: 2852-2857. [ Links ]
PECERILLO, R., & TAYLOR, S.R. 1976: Geochemistry of Eocene calc-alkaline rocks from the Kastamonu Area, northern Turkey.- Contrib. Min. Pet. 58: 63-81. [ Links ]
PLANK, T., BALZER, V., & CARR, M.J. 2002: Nicaraguan volcanoes record paleoceanographic changes accompanying closure of the Panama Gateway.- Geol. 30: 1087-1090. [ Links ]
RENNE, P.R., SHARP, W.D., DEINO, A.L., ORSI, G., & CIVETTA, L. 1997: 40Ar/39Ar dating into the historical realm: Calibration against Pliny the Younger.- Science, 277: 1279-1280. [ Links ]
SAGINOR, I, GAZEL, E., CARR, M., SWISHER, C., & TURRIN, B. 2011: New Pliocene– Pleistocene 40Ar/39Ar ages fill in temporal gaps in the Nicaraguan volcanic record.- J. Volc. Geothermal Res. 202: 143-152. [ Links ]
STEIGER, R.H., & JAGER, E. 1977: Subcommision on geochronology: Convention on the use of decay constants.-Earth Plan. Sci. Lett. 36: 359-362. [ Links ]
SMITH, A.L., ROOBAL, M.J., & GUNN, B.M. 1980: The Lesser Antilles – A discussion of the Island arc magmatism.- Bull. Volc. 43: 287-302. [ Links ]
WILLS, J.K. 1974: The geological history of southern Dominica, and plutonic nodules from the Lesser Antilles.- University of Durham, U.K. Tesis Doc. [ Links ]
*Correspondencia: Ian Saginor: Keystone College, Natural Sciences and Mathematics, One College Green, La Plume, PA 18440. Autor para contacto: ian.saginor@keystone.edu Esteban Gazel: Virginia Tech, Department of Geological Sciences, 4044 Derring Hall (0420), Blacksburg, VA 24061 ]]>
Michael J. Carr: Rutgers University, Department of Geological Sciences, 610 Taylor Rd., Piscataway NJ 08854 Carl C. Swisher III: Rutgers University, Department of Geological Sciences, 610 Taylor Rd., Piscataway NJ 08854 Brent Turrin: Rutgers University, Department of Geological Sciences, 610 Taylor Rd., Piscataway NJ 08854
1Keystone College, Natural Sciences and Mathematics, One College Green, La Plume, PA 18440 2Virginia Tech, Department of Geological Sciences, 4044 Derring Hall (0420), Blacksburg, VA 24061 3Rutgers University, Department of Geological Sciences, 610 Taylor Rd., Piscataway NJ 08854 *Autor para contacto: ian.saginor@keystone.edu
![Figure ]]> ), which is described later in this paper.](img/revistas/rgac/n45/a03i2.jpg){kind=link}
