Geological Society of America Special Paper 340 1999
Jurassic continental rift magmatism in northeast Mexico: Allogenic metaigneous blocks in the El Papalote evaporite diapir, La Popa basin, Nuevo L e h , Mexico Jennifer M. Garrison* and Nancy J. McMillan Department of Geological Sciences, Department 3AB, New Mexico State University, LQS Cruces, New Mexico 88003, United States; E-mails:
[email protected],
[email protected]
ABSTRACT Large allogenic blocks of greenschist facies metavolcanic and metaplutonic rocks in the El paPaiote evaporite diapir, located in the Lararnide La Popa foreland basin of northeastern Mexico, are interpreted as local basement entrained in the evaporite during diapirism. On the basis of rock textures and geochemistry, the metaigneous rocks can be divided into four groups: metaplutonic, high-Nb metavolcanic, intermediate-Nb metavolcanic, and low-Nb metavolcanic rocks. The metavolcanic rocks have been thoroughly recrystallized to chlorite-actinolite-epidote-albite, but remnant volcanic textures remain. The rocks exhibit no metamorphic fabrics. The 40Ar/39Arplateau ages on biotite in metaplutonic blocks indicate that the blocks last cooled through 350 "C in the latest Jurassic at 145-146 Ma. Relatively high Nb and Ta concentrations and low La/Ta ratios, typical of continental rift magmas, suggest that both plutonic and volcanic protoliths were produced during continental rifting and preclude their association with either upper Paleozoic arc volcanic rocks or Triassic arc plutons of the nearby Coahuila block. The absence of foliated textures suggests that the blocks are not part of the middle Paleozoic or Precambrian crystalline basement. These relations suggest that the metavolcanic blocks represent continental rift basalts erupted on extended basement in the Sabinas basin during the Jurassic. Metaplutonic blocks may have been rift plutons exposed on horsts and eventually covered by the Callovian Minas Viejas evaporite, which was deposited in the subsiding rift basin. Because diapir initiation is inhibited by thick overburden (Schulz et al., 1993), it is likely that diapirism began by loading and/or extension in the Sabinas basin in the Early Cretaceous, after the metamorphic event. Allogenic blocks were entrained at the salt-basement interface and transported by the salt to the surface during subsequent diapirism.
*Present address: Department of Earth and Space Sciences, University of California, Los Angeles, CA 90024. Garrison, J. M., and McMillan, N. J., 1999, Jurassic continental rift magmatism in northeast Mexico: Allogenic metaigneous blocks in the El Papalote evaporite diapir, La Popa basin, Nuevo Le6n, Mexico, in Bartolini, C., Wilson, J. L., and Lawton, T. F., eds., Mesozoic Sedimentary and Tectonic History of North-CentralMexico: Boulder, Colorado, Geological Society of America Special Paper 340.
J. M. Garrison and N. J. McMillan
320
INTRODUCTION
metaigneous allogenic blocks in the El Papalote diapir, which developed in the Laramide La Popa foreland basin in Nuevo The presence of exotic blocks in salt diapirs has long been Le6n, northeastern Mexico (Fig. l), and propose a mechanism for recognized (e.g., Roberts, 1924). Diapirs associated with large salt block entrainment. El Papalote is elliptical in shape, has a surface deposits often contain blocks of various lithologies, representative area of 4-5 km2, and is exposed on the north-dipping limb of a of surrounding country rock and indicative of geologic history large anticline (Fig. 2). The diapir pierced the Jurassic and Creta(Roberts, 1924; Wilson, 1975; Weinburg, 1993). For example, ceous sedimentary section within the basin, and is currently Wilson (1975) used the presence of Precambrian basement blocks exposed between members of the Upper Cretaceous Potrerillos in a diapir of the Hormuz salt on the Arabian Peninsula to infer a Formation (Laudon, 1984). The evaporite is the Callovian Minas regonal igneous event that took place during the late Precam- Viejas Formation (Winker and Buffler, 1988), which is also brian. In the Hormuz salt, only units of the same age or older than exposed in adjacent nearby diapirs and the salt-cored anticlines the source bed are entrained in the salt (Wilson, 1975); rocks from Potrero Garcia and Potrero Chico (Wall et al., 1961). Varieties of lithologies are found as allogenic blocks in the overlying strata are not found as exotic, or allogenic, blocks. In this chapter we use major and trace element geochemistry, diapu, including metavolcanic, metaplutonic, and carbonate age determinations, and mineralogy to determine the origin of rocks. Blocks range in diameter from 1 to 200 m; the largest
100
200 km
Jurassic features:
Figure 1. Map of northeastern ~Mexicoillustrating spatial relations between Jurassic extensional features, Cretaceous-Tertiary Laramide compressional features, and La Popa Laramide foreland basin. Base map is from Sedlock et al. (1993).Jurassic basement highs and lows are from Winker and Buffler (1988),inferred from distribution and ages of Jurassic sedimentary rocks, labeled as follows: CTB = Chihuahua trough-Bisbee basin; SB = Sabinas basin; ML = Magicatzin low; CB = Coahuila block; BPH = Burro-Picacho high; TH = Tamaulipas high. Positions of Laramide Parras and La Popa basins and Sierra Madre deformational f~ontare from McBride et al. (1974). Other symbols: star M = Monterrey;SM = San Marcos Jurassic normal fault (from Sedlock et al., 1993).
,
Jurassic continental rift magmatism, northeast Mexico: Allogenic metaigneous blocks, La Popa basin, Mexico
evaporite carbonate lentil
321
Laboratory at the New Mexico Institute of Mining and Technology. Biotite grains were extracted by heavy liquid and magnetic separation techniques, encapsulated with a Fish Canyon Tuff standard, irradiated, and analyzed using standard techniques. The decay constant and isotopic abundances are those suggested by Steiger and Jager (1977), and all errors are reported at the 2 G confidence level, including the uncertainty in the J-factor.
PETROLOGY OF THE METAIGNEOUS ALLOGENIC BLOCKS Mineralogy Figure 2. Map showing location of El Papalote evaporite diapir in La Popa basin (see Fig. I), modified from Laudon (1984); Tertiary ages of units are from Vega-Vera and Penilliat (1989). Abbreviations: Kpa = Parras Shale; Km = Muerto Formation; TKpo = Potrerillos Formation; Ta = Adjuntas Formation; Tv = Viento Formation.
blocks are concentrated near the edges of the diapir. Although the diapir pierced a thick Jurassic-Cretaceous sedimentary section, very few of the carbonate blocks correspond to Mesozoic formations (Garrison, 1998). Laudon (1984) identified one of the larger carbonate blocks as the Upper Jurassic Zuloaga limestone, which conformably overlies the Minas Viejas evaporite in undisturbed stratigraphic sections (Winker and Buffler, 1988).
METHODS Samples were collected from the interiors of the blocks to avoid the chemical effects of interaction with the host evaporite (see following discussion of reaction aureoles). Major elements and the trace elements Sc, V, Cr, Ni, Zn, Nb, Be, Sr, Ba, Zr, andY were determined by inductively coupled plasma atomic emission spectrometry (ICP-AES) in the Department of Geosciences at Texas Tech University using U.S. Geological Survey standards on sample splits ground in a tungsten carbide mill. Precision for the major oxides is +2%, except for MnO at +I%. The rare earth elements (REE), As, U, Co, Cs, Hf, Sb, Ta, and Th were determined by instrumental neutron activation analysis (INAA) at the Research Reactor Center in Columbia, Missouri, on splits ground in a corundum mill. Analytical precision is 23% for La, Sm, Ce, Eu, Hf, Ta, and Th; *3%-5% forYb, Lu, and Cs; and +5%-10% for Nd, Tb, As, U, and Sb. Major element analyses are reported as analyzed and normalized to a volatile-free basis for graphing and interpretation. Electron microprobe analyses to determine amphibole and plagioclase compositions in the metavolcanic blocks were performed in the Department of Geology at Louisiana State University using a JEOL 733 Superprobe. Well-characterizedsilicate standards were used. Analytical accuracy based on counting statistics is 20.1 wt%. The 40Ar/39Ar age determinations were made on biotite grains from two samples at the New Mexico Geochronological Research
Laudon (1984) described the metaigneous blocks as highly epidotized diorites. Our more detailed study of the blocks shows that both metaplutonic and metavolcanic rocks are present, all thermally metamorphosed to greenschist facies. The blocks exhibit neither chilled margins nor pillow structures as if the magmas had cooled against the evapoilte, nor foliated textures, indicative of dynamothermal metamorphism. Metaplutonic blocks have equigranular phaneritic textures and comprise monzonite (orthoclase-plagioclase-quartz-biotite) and biotite diorite (plagioclase-biotite-hornblende). Igneous minerals have been overprinted by the metamorphic mineral assemblage chlorite-epidote-actinolite; plagioclase is replaced by sericite. Metavolcanic rocks have aphanitic and aphanitic porphyritic textures; several of the blocks are amygdaloidal, with vesicles filled with secondary calcite and chlorite. The metavolcanic blocks have been thoroughly recrystallized to greenschist facies minerals actinolite-chlorite-epidote-albite. Microprobe analyses of plagioclase grains show that their composition is nearly pure end-member albite (An,-2); the amphiboles are low-alkali, low-A1 metamorphic actinolites rather than hornblende (Tables 1 and 2). Sparse olivine phenocrysts have been pseudomorphed by serpentine and brucite. Olivine and plagioclase phenocrysts range in size from several millimeters to 1 cm in length. Reaction Aureoles Many of the metaigneous blocks are surrounded by reaction aureoles, which represent a transition zone of interaction between the blocks and the evaporite. This zone is characterized by a pale green mixture of evaporite and highly weathered brecciated rock. The amount of this alteration material decreases away from the block, as do concentrations of SO2, A1203,Fe203,TiO,, Na20, MnO, and MgO (Garrison and McMillan, 1997), while CaO increases. Because the aureoles have not been disrupted by flow within the diapir, the reaction appears to post-date emplacement. This raises the question of whether geochemical analysis of the allogenic blocks can yield useful information about their preentrainment igneous and metamorphic history. Two samples of the same metavolcanic block at different positions from the block-transition-zone interface were analyzed to assess the mobility of elements within the blocks as a result of
TABLE 1. MICROPROBE ANALYSIS OF ACTlNOLlTE GRAINS IN METAOLCANIC BLOCKS
Sample
87
87
91
97.46
98.34
98.23
SiO, TiO,
MnO MgO CaO Na,O
K20 F CI Total
Stoichiometric proportions based on 15 cations (except K) Tetrahedral site Si 7.769 7.901 7.909 Ti 0.033 0.025 0.010 Al 0.213 0.099 0.091 Fe3+ 0.018 0.000 0.000 Ml+M2+M3 site 0.113 0.092 Al 0.000 0.000 0.063 Fe3+ 0.200 Fez+ 1.429 1.599 1.444 0.035 0.054 0.031 Mn2+ 3.265 3.330 Mg 3.31 5 M4 site Excess Ml+M2+M3 0.007 0.038 0.000 Ca 1.913 1.951 1.925 Na 0.042 0.037 0.087 A site 0.000 0.008 Na 0.000 0.005 0.013 K 0.008 0.000 0.000 FI 0.000 , 0.001 0.000 CI 0.000
Total
15.016
15.043
15.014
91
98
98
53
53
64
64
Jurassic continental rzj? magmatism, northeast Mexico: Allogenic metaigneous blocks, La Popa basin, Mexico
I
I
323
TABLE 2. MICROPROBE ANALYSES OF PLAGIOCLASE GRAINS IN METAVOLCANIC BLOCKS Sample
87
87
64
98
98
91
91
53
53
86
86
Mineral
An,
A",
A,'
4
An,.
An,
An,
A",
A",
An,
A",
99.92
100.06
100.22
99.87
101.96
100.05
101.88
100.75
99.63
98.49
SiO, TiO,
Fe203
MnO MgO CaO N%O K2O Total
99.47
Stoichiometric proportions based on 8 oxygens Si Ti Al Fe Mn Mg Ca Na K Total
2.966 0.000 1.043
2.955 0.000 1.049
2.931 0.000 1.038
2.958 0.001 1.056
0.002 0.000
0.003 0.000
0.025 0.003
0.004 0.000
0.000 0.020
0.000 0.022
0.036 0.016
0.000 0.001
0.956 0.004
0.974 0.004
0.969
0.976
0.002
0.006
4.990
5.008
5.020
5.001
4.988
interaction with the host evaporite. Sample 65 was collected in the center of a 50-m-diameter block as far from the evaporite as possible, and represents the least altered composition. Sample 66 was collected from the block near the block-transition-zone interface and records changes in composition due to interaction with the evaporite. The large ion lithophile elements (LILE) Ba, Rb, K, and Sr have different concentrations in the two samples (Fig. 3). These are elements that form weak bonds and are easily mobilized by alteration and weathering processes. In contrast, Th and the high field strength elements (HFSE) Nb, Zr, Hf, Ti, Y, and the heavy REEs Tb and Yb, which form tight bonds in minerals due to their high valence and small radius, have almost identical concentrations in the two samples. This indicates that these elements have not been mobilized by reaction processes in the aureole, and that they maintain their preentrainment concentrations. Other elements, such as P, La, Ce, Nd, and Sm, exhibit some loss in the altered sample, although not as much as the LILEs. There are three lines of evidence that suggest that these elements can be used in conjunction with the HFSEs and the heavy REEs as petrogenetic indicators, and that the data set represents preentrainment compositions. First, the samples, with the exception of block edge sample 66, are on Si0,-MgO trends that can be easily interpreted as having an igneous origin (Fig. 4). Second, the samples are in well-defined groups when plotted on chon-
5.010
4.990
4.992
4.997
4.991
5.017
drite-normalized incompatible trace element diagrams (Fig. 5); none of the samples exhibit the light REE-depleted characteristics of alteration. There is wide scatter in the mobile elements Ba, Rb, K, and Sr in Figure 5, but the HFSEs and REEs exhibit the smooth pattern typical of continental basalts (Thompson et al., 1983). Third, the overall patterns of metavolcanic samples strongly resemble (with the exception of the mobile elements) the mean composition of fresh mafic lavas younger than 5 Ma from the western U.S. Basin and Range (Fitton et al., 1991).
Metaplutonic Rocks Three metaplutonic rocks have been analyzed for major and trace element concentrations (Tables 3, 4, and 5). Two diorites with low Si0, (49.3%-50.0%) and high MgO (5.8%-6.2%) are broadly similar in composition to the most mafic metavolcanic blocks; a monzonite (SiO, = 60.7%; MgO = 0.8%) has much higher SiO, than any analyzed metavolcanic rock (Fig. 4). The chondrite-normalized incompatible trace element pattern of the monzonite (Fig. 6) exhibits marked depletions in Ti and P, which indicate fractionation of Fe-Ti oxides and apatite. Compared to the metavolcanic suite (Fig. 5), the mobile elements Ba, Rb, K, and Sr do not exhibit extreme variation in the metaplutonic rocks despite the wide range of igneous protolith compositions. This
J. M. Garrison a)
324
Figure 3. Chondrite-normalized incompatible trace element diagram (normalized to values of Thompson et al., 1984) of two samples taken at different distances from block-evaporite transition zone.
--I0
Plutonic rocks
A
High-Nb metavolcanicrocks Intermediate-Nb rnetavolcanicrocks
@
Low-Nb metavolcanicrocks
0
Samdes ss and 66
56'
66'
(block edge)
(block interior)
Figure 4. Variation diagram of Si02vs. MgO, showing plutonic blocks plotted with three groups of metavolcanic blocks.
may indicate that some of the element mobility in the metavolcanic rocks occurred prior to entrainment rather than during interaction with the evaporite, and that the plutonic protoliths underwent less thorough metamorphism than the volcanic protoliths. High Nb and Ta concentrations, light REE enrichment, and LaN/TaN< l(i.e., no trough at Ta in Fig. 6) suggest that the plutons were produced during continental riftmg. Metavolcanic Rocks Metavolcanic blocks can be divided into high-Nb, intermediate-Nb, and low-Nb chemical subgroups on the basis of concentrations of the immobile elements Nb, Ta, Zr, P, and Y on chondrite-normalized incompatible trace element diagrams (Fig. 5). Disregarding the mobile elements Ba, Rb, K, and Sr, which exhibit large variation, each subgroup has a distinct trace
element pattern. For example, samples in the high-Nb subgroup (Nb > 55 ppm, normalized values of 190-230) have consistently higher NbN/ZrN(6.2-7.2) and P,/Y, (3.1-5.4) ratios than samples in the low-Nb subgroup, (Nb = 25-52 ppm, Nb, = 71-149, NbN/ZrN= 2.8-4.2, and PN/YN= 1.7-2.3). The intermediate-Nb group has intermediate Nb concentrations (43-46 ppm, Nb, = 123-131) as well as intermediate elemental ratios (NbN/ZrN= 4.4-5.9, and PN/YN= 2.3-2.7). The major elements of the subgroups also plot in distinct groups (Fig. 4). The high-Nb subgroup has the lowest SiO, (49.4%-51.2%) of the groups and exhibits consistent MgO (5.9%-7.1%) over that range. In contrast, the low-Nb group exhibits a larger range in SiOz (50.3%-54.0%) and MgO (3.5%7.8%), which are negatively correlated. The intermediate-Nb group has the widest variation in major element composition (SiO, = 49.1%-55.0%; MgO = 4.4%-7.0%). These differences in major and trace element compositions preclude a simple genetic relation between the subgroups. For example, the low-Nb subgroup could be considered parental to the high-Nb subgroup because incompatible elements increase in concentration during fractional crystallization; however, the lower MgO concentrations in the low-Nb subgroup would indicate the opposite relation, i.e., that they are more evolved than the high-Nb subgroup. Similarly, it is impossible to produce the subgroups by different degrees of partial melting of the same mantle source. For example, the high-Nb subgroup has similar Zr concentrations, lower Y concentrations, and higher Nb and P concentrations than the low Nb samples. If the high-Nb group had been produced by lower degrees of partial melting of the same mantle, the residual mantle would have to contain minerals that retained much of the Y, some of the Zr, and none of the Nb and P. Although garnet in the source could buffer the Y concentrations, melting of typical spinel or garnet peridotite cannot produce an increase in Nb and P without concomitant increases in Zr, unless an unusual residual mineral assemblage were present. A more reasonable process for generation of the subgroups is partial melting of a heterogeneous mantle source, possibly in two or more events. Age Determinations Age determinations and the compositions of metamorphic minerals constrain the age of greenschist facies metamorphism. The 40Ar/39Ar age determinations for biotite grains in plutonic samples 35 and 67 yield plateau ages of 146.5'2 1.6 Ma and 145.6 + 1.0 Ma, respectively (Fig. 7). These data indicate that the rocks most recently passed through the blocking temperature of biotite (-350 "C; Dickin, 1995) during latest Jurassic time. The presence of nearly pure albite and low-Al actinolite indicates that the rocks were thoroughly recrystallized at temperatures within the greenschist facies but below the 400 OC peristerite gap in the plagioclase solid solution series (Zoltai and Stout, 1984). These rnineralogic indicators support the con-
-
hirassic continental rift magmatism, northeast Mexico: Allogenic metaigneous blocks, La Popa basin, Mexico
0
.+ -
~
~
325
3
rmediate-Nb group
b
u 2
0 1c? L
0 2 \
3
2
0'
rn
*
2
Figure 5. Chondrite-normalized (Thompson et al., 1984) incompatible trace element diagramsfor low-, intermediate-, and high-Nb samples. Mean composition of mafic (
