3. Description of the deformation and fabric gradient
Spatially oriented samples were collected at closely-spaced intervals along a transect, starting at the contact between the Cerro de Costilla tonalite and the orthogneiss wall-rocks, and progressing 28.3 m into the tonalite until a point was reached where the rock showed no field evidence of deformation (Fig. 2b). Sample collection was most closely spaced at the edge of the tonalite, where the strongest fabric gradient and mineralogical changes were noted. Two foliation-perpendicular thin sections were prepared from each sample: one parallel to and the other perpendicular to the lineation. Below we describe the progression from the least to most deformed rocks.
3.1. Least deformed tonalite
The least deformed tonalite observed shows a weak magmatic foliation delineated mainly by biotite aggregates and aligned plagioclase grains. The plagioclase commonly has oscillatory zoning contained within core-to-rim compositional zoning, as discussed in Section 4. It also has well developed growth twinning and a few crystal faces, though most boundaries are irregular. Commonly it contains a few rounded quartz inclusions, as well as local rounded to euhedral and irregularly shaped inclusions of hornblende. Some plagioclase grains have been bent, producing local concentrations of lenticular deformation twins. Local evidence of minor strain-induced grain boundary migration is observed (Fig. 3). Interstitial quartz shows subgrains parallel to the c-axis, but has no chess-board subgrain patterns indicative of c-slip. It shows evidence of subgrain rotation recrystallization, as well as some grain boundary migration (Fig. 4). Brown-green hornblende occurs as irregularly shaped to subhedral grains and aggregates, some with inclusions of quartz and biotite, together with inclusions of plagioclase that vary from rounded to euhedral with rounded corners. Some of the hornblende has simple or multiple twinning, and some is optically zoned, with brown-green cores and blue-green rims.
Biotite occurs as greenish brown, subhedral grains with {001} faces, as well as ‘decussate’ aggregates, a few of which locally pass into incipient foliae (Section 4), suggesting that some deformation and recrystallization occurred in the least deformed rocks. Locally biotite is intergrown with hornblende. Some biotite grains are bent, some have slightly misoriented, irregular to ragged ‘subgrains’ that appear to have been formed by fracturing (Fig. 5), and rare grains have been broken into fragments with ragged edges, the resulting spaces having been filled with quartz, fine-grained titanite and small rotated biotite fragments (Fig. 6). Locally biotite adjacent to plagioclase has been partly replaced by fine-grained, symplectic aggregates of titanite and plagioclase, the plagioclase being commonly in optical continuity with the plagioclase grain on which it nucleated (Fig. 7).
K-feldspar occurs as small interstitial grains marginally replaced by fine-grained myrmekite (Fig. 8), the formation of myrmekite probably being related to the deformation (e.g., Simpson, 1985; Simpson and Wintsch, 1989; Vernon, 1991).
Titanite occurs locally as scattered large grains, but mainly as small grains, aligned aggregates and veinlets, especially in biotite. Also present are rare pink to blue-black tourmaline, apatite and brown allanite.
3.2. Initiation of foliae
Quartz recrystallization increases with increasing fabric intensity, though the recrystallized grainsize remains relatively large. The recrystallized aggregates become elongated in more deformed rocks, but are highly elongated only in the most strongly deformed rocks, where they help to delineate the foliation. Quartz also forms incipient foliae in aggregates with biotite (Fig. 9), with or without fine-grained titanite or apatite.
Plagioclase locally shows small new marginal grains and aggregates of grains that, on the basis of optical evidence, could have been formed by either primary recrystallization or ‘sintering’ of small fragments. Though apparent subgrains occur locally, deformation temperatures were probably too low for subgrain rotation recrystallization to have played an important role (see sections 3.5 and 4.3). Therefore, local minute subgrains and new grains along internal fractures and grain boundaries (Fig. 10) probably originated mainly as fragments (e.g., Vernon et al., 2004; Johnson et al., 2004), although some genuine subgrains may be present (e.g., Fig. 10a). These fine-grained aggregates may contribute to foliae by providing pathways for linking biotite-rich foliae. Locally plagioclase grains have been fractured and pulled apart, some with fillings of undeformed quartz, others with fillings of recrystallized quartz, others with fillings of quartz and aligned biotite in a crude ‘beard’ structure, and still others with hornblende and plagioclase fragments that broke away from matrix grains as they flowed in to fill the gap (Fig. 11). No fillings involving feldspar were observed, confirming that melt was absent during the deformation (cf. Vernon et al., 2004).
Biotite is the main contributor to foliation initiation, in the form of (1) small biotite grains that grow into fractures on the edges of primary plagioclase grains (Fig. 12a), and are eventually smeared out into long foliae as they separate the plagioclase into pieces (Fig. 12b); (2) aligned biotite aggregates cutting through hornblende grains, some following straight fractures, and others apparently partly replacing hornblende; (3) lenticular biotite grains and aggregates splintering during ‘necking down’ at their ends to form incipient foliae, owing to sliding of cleavage fragments on one another, followed in some instances by apparent recrystallization (Fig. 13); (4) development of sub-parallel splinters at bends in biotite grains; (5) deformed biotite grains bending into an alignment, accompanied by the development of subgrains and recrystallized aggregates (Fig. 14) (6) biotite-rich foliae leading away from hornblende that was partly replaced by biotite, the foliae commonly carrying small fragments of hornblende (Fig. 15), (7) thin veinlets of very fine-grained biotite and quartz, locally with titanite, epidote or apatite, mostly in primary grain boundaries, though locally cutting through grains, suggesting initiation by fracturing (Fig. 16), (8) ‘beards’ of biotite and biotite ± quartz on primary plagioclase grains (Figs 11a, 17), and (9) rare biotite splays or fans cut by foliae, to which they contribute (Fig. 18).
Figure 16. Continuous foliae of biotite and quartz with small fragments of titanite, epidote and hornblende.
Figure 18. Biotite grain cut through by a foliae to which it contributes by recrystallizing into aligned new grains.
Many of the new biotite-rich foliae follow grain boundaries of strong minerals, especially plagioclase and hornblende, constituting the beginnings of an S/C-type foliation pattern (Fig. 19). This is consistent with preferential fracturing along grain boundaries, rather than across grains, though some biotite-rich foliae, apparently occupying former fracture sites, do cut plagioclase and hornblende grains.
Hornblende shows rare internal deformation in the form of slightly misoriented ‘subgrains’ that probably formed by fracturing. In moderately deformed rocks, ‘beards’ of aligned blue-green hornblende (optically similar to the rims of zoned primary grains), with biotite and quartz, locally develop between drawn apart hornblende fragments (Fig. 20). Where hornblende is extensively replaced by biotite, the resulting hornblende relics can be incorporated into biotite-rich foliae (Fig. 15). Hornblende splinters have locally been smeared, with biotite, along some foliae. Uncommonly, hornblende grains neck down at their ends, releasing small fragments into incipient foliae (Fig. 21). In more strongly deformed rocks, some hornblende grains have been broken into cleavage fragments that have been separated and drawn out into crude foliae, linked by biotite and embedded in recrystallized quartz.
Figure 20. Blue-green hornblende, biotite and quartz form a ‘beard’ between pulled apart fragments of green-brown hornblende.
Titanite makes a minor contribution to the initiation of foliae, in the form of small lozenges and irregular grains in aligned biotite grain boundaries and rare foliae consisting mainly of fragments dislodged from originally large primary titanite grains (Fig. 22). Titanite-plagioclase symplectites replacing biotite (Fig. 7) form a fine-grained reaction product that may help to initiate foliation development, forming recrystallized, fine-grained aggregates (Fig. 23).
Figure 23. Biotite cut by a folium containing titanite and disaggregated plagioclase-titanite symplectite.
Myrmekite partly replacing the minor amount of interstitial K-feldspar (Fig. 8) can also contribute to the initiation of foliae, because the fine-grained quartz and plagioclase recrystallize to very fine-grained aggregates (Fig. 8) that promote grain-boundary deformation processes (e.g., Vernon et al., 1983; LaTour, 1987).
3.3. More continuous foliae
Many relatively continuous foliae consist of recrystallized biotite ± quartz, with or without irregularly distributed titanite, plagioclase, epidote and hornblende (Fig. 16). In some foliae the biotite is locally ‘decussate’ (Fig. 16), whereas in others it is well aligned (Fig. 15a). Some biotite-rich foliae splay into several foliae. Most foliae pass between primary grains and so tend to anastomose around large, strong grains of plagioclase and hornblende in an S/C-type relationship (Fig. 24).
Initially, the biotite-rich foliae are discontinuous, with insufficient linkages to make a through-going foliation (e.g., Johnson et al., 2004). Links or bridges consist mainly of fine-grained aggregates, with varying proportions of biotite, quartz, titanite and plagioclase (Fig. 25). Some are very fine-grained plagioclase-quartz aggregates representing recrystallized myrmekite (Fig. 8), whereas others appear to be small plagioclase grains, many of which may have originated as minute fragments along fractures (Fig. 10). After the links have been established, further strain can accumulate by preferential deformation of continuous, weak, biotite-rich foliae, producing mylonite, as discussed by Johnson et al. (2004).
3.4. Strongly deformed rocks
Foliae in the most strongly deformed rocks consist of: (1) continuous aggregates rich in biotite, with varying proportions of quartz, minor ilmenite and rutile, and local fine-grained aggregates derived from the symplectites replacing biotite; (2) elongate aggregates of recrystallized quartz (Fig. 26); and (3) finer-grained aggregates rich in biotite and quartz forming ‘beards’ on and between plagioclase clasts, all anastomosing around the larger plagioclase grains (Fig. 27). Biotite and, less commonly, quartz appear to have replaced plagioclase at the edges of the ‘beards’ (Figs 17, 28), although in some instances these minerals may simply be growing into jagged embayments where fragments of the plagioclase have been detached. In many places, these detached plagioclase fragments of various sizes are incorporated into the ‘beards’ (Fig. 27). We infer that the internal foliation in the ‘beards’ delineates the stretching direction.
As shortening normal to the foliation increases, the plagioclase clasts separate, with the result that the ‘beards’ become extended into fine-grained, foliated aggregates and continuous fine-grained foliae (Figs 27, 29), eventually constituting a foliated matrix between residual plagioclase clasts. At this stage of foliation development, the S/C-type character of the less deformed rocks is replaced locally by a more uniform, anastomosing foliation as the biotite and quartz foliae collapse into zones between separating plagioclase clasts. The plagioclase clasts are locally broken and can be extensively fragmented, the fragments being cemented by recrystallized quartz (Fig. 30).
In one layer (about 2 cm thick) of one of the more strongly deformed rocks, the plagioclase shows marginal replacement by lobes and irregular patches of symplectic, myrmekite-like aggregates (Fig. 31), some plagioclase grains having been extensively to completely replaced; locally the replacing plagioclase is devoid of quartz blebs (Fig. 31). The rocks show an abrupt change from layers without plagioclase replacement to the layer with extensive plagioclase replacement. We infer that this layer was preferentially infiltrated by fluid, in order to produce the symplectite (discussed below).
The most deformed rocks (sample A) contain quartz-tourmaline patches and veins, in which the tourmaline occurs as angular fragments dispersed in recrystallized quartz. Some of the tourmaline consists of close fragments that could be fitted back together, separated by single-crystal quartz fillings, suggesting deposition of quartz from solution.
In the mylonitic rock immediately at the edge of the pluton (sample A), the grainsize of the biotite-quartz aggregates is much smaller than in samples B and C, and the amount of matrix between the plagioclase clasts appears to have increased (Fig. 32). Extensive recrystallization of the matrix produced fine-grained decussate aggregates of biotite and aggregates of quartz and biotite with low-energy grain shapes.
3.5. Microstructural evidence of deformation conditions
We infer from microstructural observations that the deformation temperatures for the Cerro de Costilla rocks were 400-550°C, because (1) quartz shows no evidence of c-slip, which is generally inferred to require temperatures above 600°C (e.g., Kruhl, 1996, 1998); (2) quartz shows widespread evidence of subgrain rotation recrystallization, and some evidence for grain-boundary migration recrystallization, suggesting a temperature range of 400-500°C (Hirth and Tullis, 1992; Tullis et al., 2000; Stipp et al., 2002a, b); and (3) plagioclase shows no clear evidence of dislocation creep or subgrain rotation recrystallization, which require temperatures above 500-550°C (Tullis et al., 2000; Rosenberg and Stünitz, 2003); instead it appears to have undergone minor grain-boundary migration and fracturing, though some recrystallization (‘sintering’) of fragmental aggregates may have occurred — a process that can occur at less than 400-500°C (Fitz Gerald and Stünitz, 1993). Given the local mobility of plagioclase grain boundaries, and that we cannot completely rule out very rare subgrain development through rotation recrystallization, we place an upper limit on temperature of 550°C. In Section 4.3 we present hornblende-plagioclase thermometry that suggests a deformation temperature less than approximately 540°C. In view of the microstructural evidence and thermometry results, we suggest that the deformation temperature was 475 ± 50°C.