The stratigraphic sequence recovered at Site 1122 consists of Miocene through Pleistocene strata, consisting of lower and middle Miocene sediments (~130 m), overlain by a thin and stratigraphically highly condensed and incomplete upper Miocene and Pliocene section (<50 m) and a 450-m-thick Pleistocene section. The uppermost few meters at the site include the Holocene. Calcareous and siliceous microfossils are present throughout the whole section. The preservation of microfossils is good or very good in the upper part of the sequence but deteriorates downward.
The main paleontological results are summarized in Figure F16. The sequence of biohorizons indicates an apparently continuous section at the top, ranging from the upper Quaternary to the upper Pliocene. Calcareous nannofossil data indicate a slight unconformity at ~454 mbsf, which separates the Pleistocene-upper Pliocene sediments from a short lower Pliocene interval. A more pronounced hiatus is detected at ~490 mbsf. Below this unconformity, the sediments are middle Miocene in age, ranging from ~10.9 Ma (based on the last occurrence [LO] of Coccolithus miopelagicus, which is still present in the section) down to 16.7 Ma (last common occurrence [LCO] of Globorotalia zealandica).
In the Pliocene-Pleistocene interval, the planktonic foraminiferal assemblages are characteristic of an oceanic environment, and benthic assemblages are a mixture of deep-water and redeposited shelf forms. The upper part of the Miocene sequence is characterized by almost barren sediment or by rare deep-water benthic species, while the lowermost part of the section has a normal oceanic plankton fauna and mixed mid-shelf and deep-water benthic assemblages.
In the late Pleistocene interval, diatoms are represented by a mixture of autochthonous pelagic species and neritic-nearshore species (probably derived from the inner part of the Bounty Trough). The early Pleistocene and Pliocene sediments are rich in reworked older Pliocene and Miocene Antarctic/subantarctic species as well as some Paleogene diatom valves. The Miocene sediments yield again higher numbers of autochthonous Antarctic and subantarctic species, whereas in the lowermost part of the section only reworked Paleogene diatoms (Eocene and Oligocene in age) occur.
The micropaleontological biostratigraphy of Site 1122 is mostly based on the onboard study of core-catcher samples. Hole 1122A samples were used for the uppermost part of the section and Hole 1122C samples for the lower part. Additional samples were taken from within selected cores to address specific age and paleoenvironmental questions. The absolute ages assigned to biostratigraphic datums follow the references listed in Tables T2, T3, T4, and T5, all in the "Explanatory Notes" chapter).
Nannofossils obtained from Cores 181-1122A-1H to 13X (0-119.97 mbsf) and from core catchers of Cores 181-1122C-15X to 68X (111.28-617.80 mbsf) were analyzed. There is an 8.6-m overlap between the two holes in our analysis. Additional samples were selected and investigated strategically to increase biostratigraphic resolution. The occurrence of taxa are presented in Table T4. Because of the paucity of major marker species, especially during the Miocene, we did not attempt to correlate the sequence with Martini's (1971) standard zonation.
Calcareous nannofossils are common and well preserved in the upper part of the sequence (0-400 mbsf), except for several short turbidite intervals where only a few nannofossils were found. The preservation of nannofossils deteriorates downward from Core 181-1122C-46X (400-617.8 mbsf) at the same time as the concentration of nannofossils in sediments increases. Reworked nannofossils occur frequently throughout the sequence. The common presence of reworked species makes it hard to use last occurrence datum levels, as, for instance, with Helicosphaera sellii and Cyclicargolithus floridanus.
Thirteen nannofossil biohorizons were recognized and used for dating the drilled sequence. The upper 420 m represents apparently a continuous sequence of Pleistocene-late Pliocene age, if small diastems were negligible. The absence of marker species as well as the frequent occurrence of reworked nannofossils makes it difficult to date the lower part of the sequence (420-617.8 mbsf). Nevertheless, two hiatuses were detected at ~454 mbsf and 490 mbsf. A short early Pliocene interval (~3.5 Ma in age) is sandwiched between these two hiatuses (Fig. F16). Beneath the lower hiatus, a middle Miocene sequence (between 11 to 17.4 Ma as constrained by nannofossils) was recovered.
Samples 181-1122A-1H-CC to 7H-CC contain abundant, moderately preserved nannofossils characterized by the presence of Emiliania huxleyi. This interval is correlated with Zone NN21 (Martini, 1971), with its bottom estimated at 0.24 Ma (Naish et al., 1998). Because of the dilution effect of terrigenous influx caused by turbidites, it was difficult to assess the dominance of E. huxleyi, and, therefore, the acme zone of this species was not recognized within Zone NN21. The base of NN20, the LO of Pseudoemiliania lacunosa, estimated at 0.42 Ma (Sato and Kameo, 1996; Naish et al., 1998), was recognized at Sample 181-1122C-22X-CC.
From Samples 181-1122C-22X-CC to 44X (176.9-391.5 mbsf), the sequence contains assemblages diagnostic of the early to late Pleistocene (e.g., Matsuoka and Okada, 1989; Takayama, 1993; Sato and Kameo, 1996). Sato and Kameo's (1996) recent revision of the estimated ages of Pleistocene nannofossil datum levels was adopted for dating this core section. The LO of Helicosphaera sellii was assessed to be at Sample 181-1122C-44X-3, 13 cm, which is inconsistent with other biohorizons, and, therefore, was not included in the age-depth model.
A biostratigraphic break occurs between Samples 181-1122C-50X-CC and 51X-2, 58 cm. Below this break, well-preserved Sphenolithus neoabies and Dictyococcites antarcticus occur commonly and persistently. The occurrence of Sphenolithus neoabies (mean LO estimated to be slightly younger than 3.54 Ma; Spencer-Cervato et al., 1994) and Pseudoemiliania lacunosa (FO at 3.54 Ma; Spencer-Cervato et al., 1994) in Sample 181-1122C-51X-2, 58 cm, to 51X-CC (452.2-463.6 mbsf) suggests a correlation of this interval to the NN15 Zone of the late early Pliocene. The break itself might represent a small hiatus separating the late Pliocene from early Pliocene. Sample 181-1122C-52X-CC yields a few Helicosphaera sellii, whose first appearance was reported to be at the basal boundary of Zone NN12 (de Kaenel and Villa, 1996), indicating an age younger than 5.5 Ma (Berggren et al., 1995).
Samples taken from Core 181-1122C-54X are almost barren of nannofossils and, therefore, no age assessment can be made for this core. Below this barren interval, Sample 181-1122C-55X-CC yields assemblages characteristic of middle Miocene age. We surmise that a major hiatus separating the lower Pliocene from the middle Miocene occurs between Samples 181-1122C-55X-CC (499.7 mbsf) and 53X-CC (477.8 mbsf), at around 490 mbsf. The samples immediately below the unconformity are characterized by the occurrence of a moderately preserved assemblage containing Calcidiscus premacintyrei, Coccolithus miopelagicus, and medium-sized Reticulofenestra spp. Although these species occur also in samples above this unconformity as a result of reworking, they occur in large quantities with better preservation below this level. The top of the sequence below the unconformity is estimated to be older than 10.9 Ma, based upon the appearance of Coccolithus miopelagicus (Gartner, 1992).
The middle Miocene section (499.7-617.8 mbsf) contains a few badly preserved Discoaster spp. The general paucity of major marker species hampers our effort to correlate the sequence with Martini's (1971) zonation. Nevertheless, several constraints can be given: (1) the bottom of the deepest core contains Calcidiscus premacintyrei, which indicates an age younger than 17.4 Ma (Gartner, 1992); (2) Sample 181-1122C-60X-CC (541.1 mbsf) marks the FO of Calcidiscus macintyrei, which indicates a middle Miocene age of 12.34 Ma (Raffi and Flores, 1995); and (3) the LO of Sphenolithus heteromorphus (dated 13.52 Ma, Raffi and Flores, 1995) in Sample 181-1122C-62X-CC (563.3 mbsf) gives another constraint to bracket the interval into the middle Miocene. In summary, nannofossil assemblages suggest that the lower quarter of the core (499.7-617.8 mbsf) was deposited between 17.4 and 10.9 Ma.
The uppermost part of Site 1122 is late Pleistocene to Holocene in age (0-0.45 Ma), based on the sporadic presence of Globorotalia hirsuta (Samples 181-1122B-1H-CC, 181-1122A-7H-CC, and 181-1122A-11X-CC). Many high spiro-conical globorotaliids, with features similar to those in Globorotalia hirsuta and G. praehirsuta, occur in Samples 181-1122C-30X-CC and 31X-CC, but they are here considered to be precursors of true G. hirsuta populations and are not indicators of a similar late Pleistocene age (Tables T5, T6).
Globorotalia inflata is the dominant unkeeled globorotaliid from the top of the section down to Sample 181-1122C-34X-CC, where it is replaced in dominance by Globorotalia puncticuloides. In other parts of the region, this datum has been estimated to be ~0.9 Ma (Hornibrook and Jenkins, 1994). At Site 1122, this datum lies ~10 m above the Brunhes/Matuyama boundary (0.78 Ma), indicating that the previously estimated level may be a little too old. As in other stratigraphic sections in this region (e.g., DSDP Site 594 and Site 1119), Globorotalia truncatulinoides is sporadic and small when it first appears close to the FCO of Globorotalia inflata (in Sample 181-1122C-26X-2, 7-8 cm, well above the Brunhes/Matuyama boundary). Similarly, the age of the FO G. truncatulinoides, previously estimated for this region at ~0.9 Ma, is possibly also a little early.
Assemblages with unkeeled globorotaliids dominated by G. puncticuloides (3.6-0.7 Ma) and with common Globorotalia crassula (2.6-0 Ma) extend down to Sample 181-1122C-31X-CC and indicate a late Pliocene to early Pleistocene age for this section. Unaccompanied G. puncticuloides extends down to Sample 181-1122C-51X-CC and indicates that this interval is no older than ~3.6 Ma (middle Pliocene). Sample 181-1122C-53X-CC contains the LO Globorotalia puncticulata (3.7 Ma).
No planktonic foraminifers or age-specific benthic forms are present in Samples from 181-1122C-54X-CC to 57X-CC, nor from 1122C-59X-CC to 61X-CC, with the exception of a 7-cm whitish marl interval in otherwise brown-green clay at 181-1122C-55X-4, 22-24 cm. The assemblage in the latter sample is a fine silt fraction (~60 µm) foraminifer ooze, exclusively composed of cold water-mass taxa, including Globigerina quinqueloba and Neogloboquadrina pachyderma (FO ~11.3 Ma). The assemblage is undoubtedly winnowed, either from current erosion or suspension separation in a gravity flow. A small assemblage in Sample 181-1122C-58X-CC contains several Globorotalia miotumida (FO 13.2 Ma), numerous Neogloboquadrina continuosa (LO ~11 Ma), and Paragloborotalia mayeri (LO 10.8 Ma), indicating a late middle Miocene age (11-13.2 Ma, late Lillburnian-Waiauan Stages).
Sample 181-1122C-62X-CC heralds a return of planktonic foraminifers and contains Globorotalia praemenardii (LO ~13.2 Ma), G. conica (LO ~13 Ma), and one Orbulina suturalis (FO 15.1 Ma). These indicate a middle Miocene age (15-13.2 Ma, Lillburnian Stage). The evolutionary transition from Globorotalia miozea to G. praemenardii occurs in Sample 181-1122C-63X-CC, indicating an early middle Miocene age (~15.8 Ma, Clifdenian Stage). The lower 40 m of the hole extends back across the early to middle Miocene boundary, with the lowest Sample (181-1122C-68X-CC) containing numerous five-chambered Globorotalia miozea (FO 16.7 Ma), one Globigerinoides bisphericus (FO 17 Ma), and the first good specimens of Globorotalia zealandica (LCO 16.7 Ma), indicating a late early Miocene age (~16.7 Ma, mid- to late Altonian Stage) for the bottom of Hole 1122C.
Diatoms are present nearly throughout the whole recovered sequence except for the lowermost 34 m, which is characterized by silica diagenesis that led to dissolution of diatoms and formation of authigenic zeolites. In the cores above, preservation of diatoms varies (Table T7).
The upper Pleistocene sequence of turbidites is dominated by autochthonous, planktonic diatoms of the Bounty Trough surface waters, whereas the early Pleistocene turbidite and Pliocene contourite deposits below are dominated by reworked older diatoms from the Pliocene, Miocene, and Paleogene.
In the upper Pleistocene turbiditic sequence, down to the Brunhes/Matuyama boundary (i.e., 0-300 mbsf), autochthonous, planktonic diatoms provide three datums: the well-dated late Pleistocene shifts in abundance of Hemidiscus karstenii and Actinocyclus ingens and also the last occurrence of Thalassiosira elliptipora (Figs. F16, F17; Tables T7, T8; and Table T4 in the "Explanatory Notes" chapter).
For the deposits below, a comparison with the biostratigraphic results from calcareous nannofossils, foraminifers, and the paleomagnetic reversal record showed that in the interval from Sample 181-1122B-52X-CC to 62X-CC (470 to 570 mbsf), the occurrences of Denticulopsis dimorpha (10.1-10.7 to 12.2 Ma) and Simonseniella barboi (1.2-1.8 to 12.5 Ma) fit with the middle Miocene age determined by these other methods. Here the planktonic foraminifers are also better preserved and more diverse, which suggests that this is a period of higher sedimentation rates, resulting in the preservation of at least the more robust autochthonous planktonic diatoms. In the contourite sediments above and below (300-470 mbsf and 570 mbsf to core base), age-diagnostic, autochthonous diatoms cannot be found and dissolution-resistant reworked diatoms from Eocene and Oligocene strata are present (see discussion under paleoenvironment).
The radiolarian biostratigraphy at Site 1122 is based on the examination of 81 core-catcher samples (Table T9). Throughout the turbidite and contourite units, radiolarians are sporadic and occur very rarely. However, radiolarian faunas in twenty samples are moderately well-preserved and provide useful age and paleoenvironmental information. Most samples between 0 and 402 mbsf yield common to abundant Antarctissa denticulata, Antarctissa strelkovi, Lithelius nautiloides, and Saccospyris antarctica, which represent subantarctic waters. The lower sandy units (>452 mbsf) are almost barren, except for Samples 181-1122B-62X-CC and 63X-CC.
The uppermost interval between Samples 181-1122C-1H-CC and 17X-CC (0-133 mbsf) is of late Pleistocene age (younger than 0.46 Ma), based on the LO of Stylatractus universus (0.46 Ma) in Sample 181-1121C-18X-CC (140.5 mbsf). This datum is well known and well established worldwide; however, in this section, Stylatractus universus occurs only sporadically. Rare, reworked forms of middle to upper Miocene species like Cyrtocapsella japonica, C. tetrapera, Stichocorys peregrina, and Eucyrtidium calvertense were observed in the upper part of the section.
Samples 181-1122C-48X-CC and 49X-CC yield common Eucyrtidium calvertense (LO 1.92 Ma) and 49X-CC also contains Lithelius nautiloides (FO 1.93 Ma). Therefore, the interval between 181-1122C-18X-CC and 48X-CC (140.5-432.4 mbsf) is dated as latest Pliocene to late Pleistocene (0.46-1.9 Ma). The FO datum of Lithelius nautiloides is apparently too high, resulting from reworking, because samples below 444 mbsf were barren or contained only rare radiolarians. Samples 181-1122C-62X-CC and 63X-CC yield common radiolarians including Calocycletta sp., Lychnocanoma sp., and Cyrtocapsella tetrapera of Miocene age.
In the upper Pliocene-Quaternary section (Cores 181-1122A-1H to 181-1122C-53X; 0-478 mbsf), lithostratigraphic Subunits IA through IC consist of turbidite beds of graded, very fine sand to silt ("coarse beds"), deposited on the Bounty Channel levee by periodic turbidity currents, which are interbedded with thin beds of hemipelagic to pelagic mud ("fine beds") deposited from suspension during the intervening periods (Fig. F18).
Samples from the "fine beds" (e.g., Samples 181-1122A-3H-CC, 181-1122C-25X-CC, and 181-1122C-26X-2, 7-8 cm) have planktonic forms composing ~80%-98% of the foraminifers. The benthic foraminiferal assemblage (Table T6) is dominated by typical upper abyssal calcareous assemblages containing varying mixtures of fairly small Trifarina angulosa, Epistominella exigua, Melonis barleeanum, Cassidulina carinata, Cibicides pachyderma, Oridorsalis umbonatus, Pyrgo murrhina, Quinqueloculina venusta, and Globocassidulina subglobosa. The fine beds also contain a few, small, shallow-water, benthic forms derived from "inner and mid-"shelf depths (e.g., Elphidium advenum and E. charlottense).
Samples from the "coarse beds" (e.g., Samples 181-1122A-7H-CC, 9X-CC, and 12X-CC) have planktonic forms comprising 75%-90% of the foraminifers. The benthic foraminiferal assemblage is dominated by small (mostly <0.2 mm) Elphidium advenum, E. charlottense, Uvigerina dirupta, Miliolinella subrotundata, Nonionellina flemingi, Notorotalia, and Pileolina spp. apparently displaced in with the turbiditic sand from "shelf" depths, and mixed with other deeper water bathyal or abyssal forms (e.g., Trifarina angulosa, Melonis spp., Pullenia bulloides, large Globocassidulina spp., and Nuttallides umbonifera) that may have either been picked up en route or have lived in the turbiditic sediment after it was deposited on the levee. The benthic foraminifers in these turbidites thus exhibit a mixed origin from inner shelf depths (e.g., Pileolina spp. and Zeaflorilus parri) through to abyssal.
The Quaternary planktonic foraminiferal assemblages in lithostratigraphic Unit I are dominated by a mix of small Neogloboquadrina pachyderma, Globigerina quinqueloba, larger Globigerina bulloides, Globorotalia inflata, and lower in the interval by Globorotalia puncticuloides and G. crassula. Some samples have a cooler water aspect with a greater dominance of small N. pachyderma and G. quinqueloba (e.g., Samples 181-1122C-26X-CC and 34X-CC); others have a slightly warmer aspect with the addition of a moderate number of large Orbulina universa, Globorotalia crassula, and G. hirsuta, and a greater percentage of large Globorotalia inflata or G. puncticuloides (e.g., Samples 181-1122A-7H-CC, 181-1122C-16X-CC, and 181-1122C-20X-CC).
In the lower part (lithostratigraphic Subunit IIA) of the Pliocene-Quaternary section (from Samples 181-1122C-47X-CC to 53X-CC), most of the planktonic assemblages are sparse and mainly composed of bigger, thicker-walled specimens with only a few, small, thin-walled forms present. It would seem that the planktonic assemblage is being affected by dissolution, with only the thicker-walled specimens and a few rapidly buried thinner-walled ones not being dissolved. The assemblages suggest that the Carbonate Compensation Depth (CCD) was possibly at a slightly higher level during this period than it was later in the Quaternary. Alternatively, corrosive cold-water flow may have been active.
The Miocene section (Cores 181-1122C-54X to 68X, 488 to 618 mbsf) contained foraminiferal assemblages that are predominantly partly recrystallized and poorly to moderately preserved. In the upper half of the interval (lithostratigraphic Subunit IIB), many of the fine sands are barren of foraminifers (e.g., Samples 181-1122C-54X-CC, 56X-CC, 59X-CC, and 61X-CC).
A light mud (Sample 181-1122C-55X-4, 22-24 cm) from this largely barren interval contains a rich assemblage of extremely small planktonic and benthic foraminifers, the result of winnowing from currents or suspension separation in a gravity flow. Its low-diversity planktonic assemblage with Neogloboquadrina pachyderma and Globigerina quinqueloba indicates cool water above. This sample also contains evidence (two minute specimens of Chiloguembelina) of reworking from Oligocene oceanic strata. A similar assemblage of dominantly small planktonic forms occurs in Sample 181-1122C-55X-CC, together with a sparse assemblage of deep-water benthic foraminifers, which is dominated by large Nodosaria longiscata and frequent Oridorsalis umbonatus.
Sample 181-1122C-55X-CC contains no planktonic foraminifers and a sparse deep-water benthic assemblage (e.g., Bolivinopsis, Cibicidoides pachyderma, Eggerella bradyi, Globocassidulina subglobosa, Laticarinina pauperata, Melonis pompilioides, and Oridorsalis umbonatus). We conclude that the site was below the CCD and that all planktonic foraminifers dissolved before they settled on the bottom, whereas the benthic forms were preserved by rapid burial.
Foraminiferal assemblages are present throughout the lower half (lithostratigraphic Unit III) of this Miocene interval and contain rich, planktonic assemblages with a wide range of sizes. Temperate overhead water is indicated by the abundance of forms from the Globorotalia miozea and Globorotalia zealandica lineages. The almost total absence of orbulines argues against any subtropical influence.
Entirely deep-water benthic assemblages are present in Samples 181-1122C-62X-CC and 63X-CC (lithostratigraphic Subunit IIIA), but below this in Subunit IIIB there is a substantial component of "mid-shelf" (~50-100 m depth) benthic foraminifers. These benthic forms appear to be of similar age as the deep-water assemblage to which they have been added, although a late Oligocene or early Miocene age cannot be entirely ruled out. The unusual aspect of these allochthonous shallow-water benthic assemblages is the large size of many specimens (>0.25 mm) and their abundance (composing up to 90% of the benthic assemblage in Samples 181-1122C-66X-CC and 67X-CC). Although abrasion and breakage is evident in some specimens, others still have crisply preserved ornament and show little rounding. This suggests downslope transport by mass-flow mechanisms and no long-distance transport by strong bottom currents. The shallow-water assemblage is dominated by varying mixtures of Notorotalia spinosa and Nonionella novozealandica, a well-known New Zealand, lower Miocene association from middle shelf depths (~50-100 m). Other shallow-water components include Anomalinoides fasciatus, A. macraglabra, Cibicides perforatus, C. notocenicus, Discorotalia tenuissima, Elphidium advenum, Kolesnikovella australis, Melonis maorica, and Siphonina australis.
Thus the lower half (lithostratigraphic Subunit IIIB) of the Miocene interval has an allochthonous, planktonic-dominated assemblage that accumulated at similar abyssal depths as today and a large component of mostly benthic foraminifers derived from middle shelf depths.
In the upper Pleistocene section, the diatom assemblages are a mix of subantarctic and cosmopolitan, planktonic species with an additional input of local neritic and shallow-water species from the Bounty Trough head. The diatom valves are well preserved because of the high sedimentation rates of fine-grained, clastic terrigenous material. The sediment facies is mainly turbidites interbedded with thin hemipelagic sediments. The characteristic occurrence of diatom assemblages within these turbidites is that no diatoms are found in the quartz sand at the turbidite bases, whereas higher in the clayey part of the turbidite, autochthonous, planktonic diatoms occur together with species typical for more nearshore areas and with benthic diatoms, which must be derived from shallow, coastal areas. With the transition to nonturbidite sediments, these dislocated shallower water diatoms decrease in abundance.
The provenance of the diatoms is different in the underlying early Pleistocene turbiditic sequence and the Miocene-Pliocene contourite sediments (below 300 mbsf; Fig. F19), the latter being shaped by Antarctic Bottom Water (AABW). The average sedimentation rates determined for these sediments are considerably lower than for the upper Pleistocene turbidite sequence above: approximately one-quarter lower for the lower Pleistocene turbiditic sequence and one-eighth to one-quarter lower for the contourite deposits. Here, only the relatively dissolution-resistant diatom valves are preserved. In these sediments, subantarctic/Antarctic diatoms of Pliocene and Miocene age are found as well as reworked Eocene-Oligocene species. Except for the middle Miocene section mentioned above, where the ages derived from diatoms agree with those derived from the calcareous microfossils, a large part of the assemblage has to be considered allochthonous or reworked, brought in by AABW.
The changing sedimentation rates and the changing provenance of the diatoms indicate that the dominating influence of the AABW, which is characteristic through most of the Neogene sediments at this site, is overcome by turbidite sedimentation during the late Pleistocene. This change in sedimentation rate and provenance of diatoms may reflect an increased late Pleistocene uplift of South Island New Zealand, which has been reported to reach 6 mm/yr for the Marlborough ranges (e.g., van Dissen and Yeats, 1991).
Upper Pliocene to Pleistocene radiolarian faunas obtained from the section (Cores 181-1122A-1H to 181-1122C-49X) are characterized by the consistent occurrences of abundant Antarctissa denticulata, Antarctissa strelkovi, Cycladophora davisiana davisiana, Lithelius nautiloides, and Saccospyris antarctica. These species are representative of Antarctic/subantarctic affinity. In addition, there is common to abundant occurrence of Cycladophora davisiana davisiana in Samples 181-1122A-1H-CC, 181-1122A-6H-CC, 181-1122B-1H-CC, 181-1122C-14X-CC, and 181-1122C-35X-CC. It is well known that the relative abundance curve of Cycladophora davisiana corresponds well with oxygen isotope curve (Morley and Hays, 1979).