Thermo�tectonic History of the Marlborough Region, South Island, New Zealand

Apatite fission track analysis has been used to study the thermal and tectonic history of the Marlborough Region, South Island, New Zealand. The very young ages ( <10 Ma) of apatite in the vicinity of the Alpine Fault bend and Seaward Kaikoura Range, are consistent with the recent rapid uplift/erosion in these areas. Most of the apatite ages are younger than depositional ages, indicating that the host rocks in Marlborough have ex­ perienced exposure to temperatures in the zone of partial annealing for apatite. In addition, apatite ages obtained along the Wairau Fault are al­ ways younger than those of other areas. Apatite fission track ages and mean lengths show that there are two major cooling events, one occurring from the early Miocene (-20 Ma) and the other in the mid-Cretaceous (-100 Ma). Modeled thermal histories of eleven samples selected from Marlborough samples are consistent with the stratigraphic record and re­ flect that in the Wairau block the timing of the main Neogene uplift/erosion event is earlier (mid to late Miocene) than to the southeast in the Seaward Kaikoura Range (late Pliocene-Pleistocene). (

Apatite fission track data provide not only information about numerical ages but also estimation of the thermal history of host rocks.Fission tracks in U-bearing crystals such as apatite result from the spontaneous fission of 238U, and can be applied to thermal and tectonic studies.The annealing of fission tracks is an important aspect of the fission track thermochronometer.Because of a kinetic understanding of annealing in apatite (Green et al. 1986(Green et al. , 1989b;;Laslett et al. 1987;Duddy et al. 1988;Crowley et al. 1991), thermal histories can be reconstructed from forward modeling of time-temperature histories and comparison of pre dicted and measured fission track ages and lengths.Different minerals have different "closure temperatures"." Closure temperature " is a concept that links the observed age to the tempera ture at which fission track age starts to accumulate (Dodson 1973;Hodges 1991).For example, the closure temperature for apatite ranges from 110°C to 125°C depending on apatite compo sition (Gleadow and Duddy 1981;Green et al. 1989b).
In this study, apatite fission track analysis is applied to assess the thermal and tectonic history of basement in Marlborough.The results of apatite fission track data will be discussed in four sub-regions: the Wairau, Inland Kaikoura, Seaward Kaikoura, and Kahutara.Additionally, the software of fission track thermal history modeling (Monte Trax), developed by Gallagher (1995), is used to reconstruct the thermal histories of selected samples with good track length data.

Geologic Setting
The Torlesse Supergroup constitutes the basement of the Marlborough region.Bradshaw et al. (1981) explained that the angular unconformity separating the Torlesse Supergroup from younger units, represents the stratigraphic expressions of the end of the early Cretaceous Rangitata Orogeny.The age of basement rocks ranges from the Late Jurassic to Early Cretaceous, with Triassic successions in the far west.The Alpine Schist crops out adjacent to the Alpine Fault.Its exposure is a result partly of Cretaceous denudation (Suggate 1978) but mainly Neogene denudation, as is shown here.The depositional ages of cover strata range from Cretaceous to Quaternary (Fig. 1).Early Cenozoic sequences accumulated during a tectonically quiet period when Marlborough was part of a passive margin environment.This was followed by the Kaikoura Orogeny, dating from the early Miocene, which reflects devel opment of the modern Australia-Pacific plate boundary in the region (Browne 1995).In Marlborough, magmatism and extension occurred at about 100 Ma.After the initiation of extension and magmatism (-100 Ma), marine sedimentary sequences (greensands/limestone) accumulated and subsided through thermally controlled processes (Lensen 1962).This tec tonic quiescence lasted from 90 to 25 Ma (Baker and Seward 1996).From the early Miocene onwards, crustal shortening and strike-slip faulting are considered to have become increas ingly important in the Marlborough region (Carter and Norris 1976;Suggate 1978;Baker and Seward 1996).

Tectonic Setting
The Marlborough region lies within the Australian-Pacific plate boundary zone (Fig. 1) at a critical position between the southern end of the Hikurangi margin (where the oceanic Pa cific plate subducts beneath the continental Australia plate) and the Alpine Fault section (where the continental Pacific plate collides with continental Australia plate).The Marlborough faults

Sampling
Eighty-eight samples (94 14-1 to -88) were collected from outcrops along roads through out Marlborough, or by helicopter from the Seaward Kaikoura Range.Most of the samples were taken from within the Marlborough Fault System and the region of the Alpine Fault bend.These samples can be divided into four groups, located at four sub-regions respectively (Fig. 1).The four sub-regions are: the Wairau, Inland Kaikoura, Seaward Kaikoura, and Kahutara.

Experimental Procedures
The experiment procedures applied in this study were followed using the methods de scribed in the papers of Green (1986), Kamp et al. (1989), andTippett andKamp (1993).Apatite concentrates separated from rock samples (-3 to 4 kg) were obtained by various methods, including standard magnetic and heavy liquid techniques.The apatite concentrates were mounted in Petropoxy™ resin at -140°C on glass slides and ground with silicon carbide paper to disclose internal crystal surfaces.After polishing with a slurry of alumina powder, the crystals were then etched in 5 M HN03 for 20s at -24°C.
The following procedures were followed: (a) all mounts were cut to 1 x 1.5 cm and cleaned with detergent and alcohol, (b) low-uranium mica external detectors were sealed directly in contact with the mounts by using envelopes of heat-shrink plastic, (c) pinpricks were made at the corners of each mount-mica sandwich for subsequent location, (d) all mounts were stacked vertically with dosimeter glass standards (SRM 612 for apatites) placed at the top and bottom of each stack for irradiation.Each dosimeter was also mounted with a mica detector.Afterwards, all stacks were packed into canisters and irradiated in the X-7 facility of the HIFAR reactor, New South Wales, Australia.The nominal fluences of thermal neutrons were 1x 1016 to 5 x 1016 neutrons cm•2 for apatites.
In this study, the external detector method described by Gleadow (1981) was applied in the dating.The fission track ages were determined by using the zeta calibration method (Hurford and Green 1982;Green 1985).The measurement of fission track lengths was followed by using the recommendations of Laslett et al. (1982).A chi-square statistic was used to deter mine the probability of grains counted in a sample belonging to a single population of ages (Galbraith 198 1).The results of weighted mean zetas are reported in Table 1.The apatite weighted mean ( is 348.4 ± 5.8 (SRM 612).The results of calibration of horizontally con fined track length determinations on apatite are listed in Table 2.

FISSION TRACK RESULTS AND MODELED THERMAL HISTORY
Apatite fission track results reveal that the fission track data are strongly influenced by the regional tectonics.The fission track data for Marlborough samples are shown in Table 3.The distributions of apatite fission track ages and track lengths are shown in Figs. 2 and 3. Uncer- ' value for v degrees of freedom where v =(Number of crystals-I) [Galbraith 1981]; pooled p/p, ratio used to calculate /',; and uncertainty where P(X')>5%; mean p/p, ratio used to calculate /',; and uncertainty where P(X2)>5% [Green 1981].Standard ages used are Fish Canyon Tuff 27.8 ± 0.7 Ma, Tardree Rhyolite 58.7 ± LI Ma [Hurford and Green. 1983]; Durango apatite 31.4 ± 0.5, Mount Dromedary Igneous Complex 98.7 ± 0.6 Ma, Lak e Mountain Rhyodacite 367.6 ± 1.5 Ma, Mount Warning Complex 22.8 ± 0.5 Ma [Green 1985]; Buluk Member tuff 16.2 ± 0.2 Ma [Hurford and Watk ins 1987].An uncertainty component from the independent age is included in the error on each 1',; value; apatite mean /',; and its error weighted according to uncertainties on indi vidual /',; values .Apatite s determination fulfils the requi rements proposed by Hurford[ 1990].Figure 4 is a plot of mean track length versus apatite age for samples originating in Marlborough from south of the Wairau Fault.It is useful to examine the data together in this plot before considering the data in transects, because it should reveal broad patterns about the occurrences of annealing zones and the timing of significant cooling events (e.g., Green 1986).
A general boomerang trend is shown in the data, although there are complexities in the pattern for samples with 90 million years or more of age.Samples with very young ( < 10 Ma) ages have long lengths (>15 µm), reflecting very recent and rapid cooling of the host rocks from temperatures exceeding the closure temperature of fission tracks in apatite (taken as 110 0C).The decrease in mean track length with increasing age from 10 Ma to around 67 Ma is due to a change of the proportion of shorter tracks, annealed during burial (heating) of the basement leading up to a late Cenozoic regional cooling event, versus longer tracks formed during the late Cenozoic cooling phases and contributing to the total mean length.This com ponent of the boomerang originates through different samples experiencing increasing levels of partial annealing as the age decreases and length increases.There is a trend in the plot (Fig. 4) for some samples to then increase in mean length with an increase in apatite fission track age from 67 to about 100 Ma.This corresponds to decreasing levels of partial annealing for those samples that have more apparent age.The samples with around 100 Ma of ages and long lengths cooled rapidly at around that time, and have remained at low temperatures since then in order to have retained the long mean lengths.
There are eight samples with ages of 90 Ma or more, but lengths of 13 µ m or less.These samples and their host rocks have probably experienced two or more phases of partial annealing, Easting and northing refer to New Zealand Ma p Series 260.Track densities (p) are xlO' tracks cm•'.All analyses are by external detector method using 0.5 for the 4rr127t geometry correction factor.Apatite ages calculated using dosimeter glass SRM 612 and zeta-612 = 348.4± 5.8 (lcr); P (X') is probability of obtaining x ' value for v degrees of freedom (where v is number of crystals -1) [Galbraith 1981]; pooled p/p, ration is used to calculate age and uncertainty where P(X,2)>5%; mean p/p, ration is used to calculate age and uncertainty where P(X,2)<5% [Green 1981].

Apatite Results and Interpretation
In summary, the plot can be broken up into four parts: (i) Samples-33, -60 and -81 are   easternmost end of the Awatere Valley.The host rocks of Sample 94 14-45 are of mid-Creta ceous age.The modeled thermal history of this sample (Fig. 5b) indicates that apatites cooled rapidly in the source area during the mid-Cretaceous, were deposited at about 100 Ma and reached maximum temperature (65°C) prior to Neogene cooling.

Seaward Kaikoura sub-region
Samples 9414-83 was collected from the Seaward Kaikoura block (Fig. 6.a).This sample was difficult to model, that is, to get all predicted and observed parameters to match.However, the modeling was sufficient to estimate the maximum temperatures prior to Neogene erosion.Sample 9414-83 is a Motuan (Albian) sandstone.Modeling shows that this sample has prob ably been heated to a maximum temperature of about 82°C (Fig. 6a).The apatites seem to have cooled during the early Cretaceous, were deposited in the basin with about 20 m.y. of inherited age, experienced some cooling during the early-late Cretaceous, and heating through burial during the late Cretaceous-Miocene.

Kahutara sub-region
Samples 9414-61 comes from the site south of the Hope Fault (Fig. 1), and has Jurassic depositional age.In Fig. 4 this sample has long mean lengths and comparatively old age.The modeling result is moderately successful (Fig. 6b).The sample shows a strong mid-Creta ceous cooling phase that brought the host rocks up to near the surface.This was followed by late Cretaceous-Cenozoic heating through burial and accommodation of cover strata, followed by late Pliocene-Pleistocene uplift and erosion of that cover succession.Maximum tempera ture achieved prior to the latest phase of uplift and erosion is -45°C.

4.3.S Interpretation of modeled thermal history
In this section samples with good length data have been selected for modeling of the thermal histories.The samples were selected to sample all parts of the length-age plot (Fig. 4).What has emerged is that all samples modeled retain same evidence for a mid-Cretaceous cooling event, even where there has been major late Cenozoic cooling.Samples with very young ages and long lengths have not been modeled because of the inadequacy of the length data, but these would not see back to the mid-Cretaceous event because they were totally reset by the recent uplift and denudation.The results of modeled Jurassic rocks, far western parts of Marlborough, tend to show the cooling event as having occurred during the Motuan (Albian), whereas the annealed Motuan rocks show a slightly later phase of uplift (94-90 Ma) and retain a provenance signal that "sees" back to the early Cretaceous.
In all cases the modeling is consistent with late Cretaceous-Oligocene burial/heating, but this can be minor in the western parts of Marlborough where these rocks are not present.Meanwhile, the modeled results show that the timing of the main Neogene uplift/erosion event varies, being earlier (mid to late Miocene) in the north (Wairau block) and later (late Pliocene Pleistocene) in the southeast (Seaward Kaikoura Range).The samples in Fig. 4 that have older ages (>90 Ma) but intermediate lengths ( 11-13 µ m) appear to be from host rocks that were uplifted during the mid-Cretaceous into lower parts of a partial annealing zone, and experienced some subsequent late Cretaceous-Cenozoic burial.The consequence of spending considerable time in the zone of partial annealing, and only recent uplift, has been retention of age relative to length.

CONCLUSIONS
The very young ages (<10 Ma) of apatite in the vicinity of the Alpine Fault bend and Seaward Kaikoura Range, can be correlated with the recent rapid uplift/erosion in these areas.
Most of the apatite ages are younger than depositional ages, revealing that the host rocks in Marlborough have experienced exposure to temperatures in the zone of partial annealing for apatite.In addition, apatite ages obtained along the Wairau Fault are always younger than those of other areas.Apatite fission track ages and mean lengths indicate that there are two major cooling events, one occurring from the early Miocene (-20 Ma) and the other in the mid-Cretaceous (-100 Ma).Modeled thermal histories of selected samples with good length data are consistent with the stratigraphic record and reflect that in the Wairau block the timing of the main Neogene uplift/erosion event is earlier (mid to late Miocene) than to the southeast in the Seaward Kaikoura Range (late Pliocene-Pleistocene).

Fig. 1 .
Fig. 1.Tectonic and geological map of the Marlborough region, South Island, New Zealand.
487 have usually been explained as secondary transforms connecting the Hikurangi subduction margin with the main Alpine Fault oblique-slip boundary (Wellman 1971; Christoffel 1971).Stock and Molnar (1982) stated that a change in the direction of migration of the Australian Pacific pole of rotation occurred about 9.8 Ma ago.The subduction of the Pacific plate has probably propagated southwards during the Late Neogene, causing continuous activation of the Marlborough Faults (Carter and Carter 1982).Several palaeomagnetic studies while docu menting the occurrence of rotations of blocks between the faults also support this concept of successive propagation of faults during the Neogene (Walcott 1978; Lamb 1988; Mumme et al. 1989; Roberts 1992; Vickery and Lamb 1995).

Fig. 2 .
Fig. 2. Distribution of apatite fission track ages (in Ma) of Marlborough.Each sample number, completed by prefix 9414-, is shown in a parenthesis .The main faults (Fig. 1.) are indicated by solid lines .
reset.(ii) Samples on the plot between 9414-1 and -53 are heavily partially annealed and are inferred to have experienced the lower levels of a partial annealing zone prior to late Cenozoic cooling.(iii) Samples between 9414-5 and -6 1, and possibly -63, experienced the upper part of a partial annealing zone prior to late Cenozoic cooling.(iv) Samples 9414-39, -42, -43, -44, -45, -47, -70 and -84 have the most complicated thermal history, and either retained a prov enance record, or experienced partial annealing in Marlborough during two intervals.

Fig. 4 .
Fig. 4. Plot of mean track length versus apatite age for Marlborough samples.Uncertainties are at 1 cr level.Each sample number is completed by prefix 9414-.

Table 1 .
Results of calibration of fission track age determinations by the zeta approach.
Apatite Mean 1',; 348.4 ± 5.8Analyses of apatite age standards are by external detector method; track densities (p) are (xl06 cm-2); N is number of track s counted .P(X') is the probability of obtaining x

Table 2 .
Results of calibration of horizontally confined track length determina tions on apatite.
Standard Sample Number of Tracks .Mean length ± 1 cr, µm Standard Deviation cr, µm tainties of fission track ages are reported at the 1 a level.Modeled thermal histories of four samples selected from different sub-regions are shown in Figs. 5 and 6. 4.1 Apatite Fission Track Age versus Mean Track Length

Table 3 .
Fission track data for Marlborough samples.