Early Cretaceous striate tricolpate pollen from the Borehole Mersa Matruh 1, North West Desert, Egypt

Striate tricolpate pollen has been recovered from Early Cretaceous samples of the borehole Mersa Matruh 1 in the N.W. Desert of Egypt. Scanning Electron Microscope (SEM) study has revealed fine details of their exine scupture, on the basis of which four new taxa, STRIOTRI-OVAL, STRIOTRI-SMOOTHMUR, STRIOTRI-SEGMUR and STRIORET-SMOOTH, are distinguished. The stratigraphic ranges of these are discussed and they are compared with other published species. They are among the oldest striate tricolpate pollen yet described, appearing in sediments of Early Aptian age, slightly predating the first reliable records of reticulate tricolpates in the sequence studied. The lack of earlier reticulate grains is attributed to possible sample failure. There is evidence that the true first appearance of tricoplate pollen in Egypt may be late Barrenian.

were of limited quantity, usually two to ten grams, because of the small size of the original samples. Standard palynological extraction methods were used for sample preparation, involving HCI, HF, centrifuging and the use of zinc bromide (S.G. = 2.0) as a heavy liquid.
Unwanted organic debris was removed by treatment with concentrated nitric acid for periods of between five and ten minutes, followed by brief alkali clearance using 5 % ammonium hydroxide solution. Where particularly large quantities of organic debris were present in a sample, a saturated solution of chromium trioxide in concentrated nitric acid was found to be a more suitable oxidant ; treatments with this solution were limited to 1-3 minutes. Some samples were also improved by cautious short centrifuge treatment to remove very fine debris.
Strew mount preparations were made of the organic residues on stubs equipped with Mark 3 Cambridge Geology SEM grids (cf. Hughes et al., 1979). After coating with gold the stubs were traversed systematically and a record made of the palynomorphs encountered. Specimens were recorded by photography on 70mm Ilford FP4 film and may be relocated on the stubs by their grid coordinates.
All material relating to this study is deposited in the Department of Earth Sciences, Cambridge University, U.K.

SYSTEMATIC DESCRIPTIONS
Records were made using the biorecordkomparison record system of Hughes (1976) within the format suggested by Hughes (1986). This will enable future workers to recover individual records which can then be assessed on their own merit. The records are stored on floppy discs for easy transmission.   Exine s e m i t e c t a t e , s t r i a t e .
Muri rounded i n c r o s s s e c t i o n , w i d t h 0.15 (0.28) 0.39 um, height 0.15 (0.28) 0.35 urn, s c u l p t e d w i t h f a i n t t r a n s v e r s e r i d g e s , supported on inconspicuous columellae up t o 0.1 um t a l l which are v e r y hard t o observe; muri meander f r e e l y over g r a i n s u r f a c e , overlapping, branching and anastomosing r e g u l a r l y but running p a r a l l e l along a p e r t u r e margins. Nexine smooth, v i s i b l e only occasionally through t h e mat of muri.
Apertures p a r a l l e l t o t h e long a x i s , with margins s l i g h t l y infolded and poorly-defined; one a p e r t u r e is longer than t h e o t h e r two, g i v i n g t h e g r a i n a s l i g h t l y asymmetrical o u t l i n e when viewed from some angles.

Occurrence of STRIOTRI-OVAL
The youngest record of STRIOTRI-OVAL is in core sample MMX-1 7875 (Late Aptian). CfA records range down to the earliest reliable record in core sample MMX-1 9508 (Early Aptian). CfA records also It is most abundant is in the two deepest cuttings samples mentioned above, where it accounts respectively for 10.9 and 12.5% of the total angiosperms present. In the cores its highest abundance is in sample MMX-1 9560 (6.9%). Although the possibility of caving in the two deepest cutting samples cannot be ignored, it is interesting to note the high abundance of STRIOTRI-OVAL there relative to the other angiosperms present (which could themselves have been subject to caving), especially as the next highest aundance is in the deepest core 9182/3, 9508/2, 11620*/1. sample. This suggests that this pollen type may have Exine s e m i t e c t a t e , s t r i a t e .

Descr iptf on M (All observations made with SEM).
Muri rounded i n c r o s s s e c t i o n , w i d t h 0.15 (0.25) 0.31um, h e i g h t 0.15 (0.2) 0.23 um, smooth, supported on small columellae which a r e a t b e s t only p a r t i a l l y v i s i b l e through t h e mat of muri; muri meander f r e e l y over g r a i n s u r f a c e , branching, anastomosing and overlapping f r e q u e n t l y . There are occasional transverse connections between the muri, but a t r u e s t r i o r e t i c u l a t e condition is never a t t a i n e d .
Muri run p a r a l l e l along t h e a p e r t u r e margins.

DISCUSSION AND COMPARISON a. Recognition of taxa
The four striate taxa recorded above are distinguished from one another on the structure of the tectum and on aperture configuration.

Two of them, STRIORET-SMOOTH and STRIOT-R I -S M O O T H M U R , h a v e unsculpted m u r i .
STRIORET-SMOOTH is easily distinguished because it has a strioreticulate exine in the central part of the grain, while STRIOTRI-SMOOTHMUR has a purely striate sexine.
For the other two, S T R I O T R I -O V A L and STRIOTRI-SEGMUR, the only easily recognisable difference is in the apertures. STRIOTRI-OVAL has two short colpi, the third being slightly longer, while STRIOTRI-SEGMUR has longer colpi which are of equal length. In addition, STRIOTRI-SEGMUR has a well-developed sculture of irregular exinous spherules on the aperture membrane, a feature which is absent in STRIOTRI-OVAL. This feature is unfortunately of limited use as it can only be seen on the more expanded specimens. The sexine is also slightly different. In both the muri meander freely over the grain surface, branching, anastamosing and overlapping. However in STRIOTRI-SEGMUR there are frequent blind projections and sharp loops of the muri which project up through the mural network. This form also has a tendency to have a double layer of muri in the sexine and sometimes appears to have a pseudoreticulate base layer as a result. b. Comparison with existing taxa Comparison of the Egyptian pollen with published species has been difficult because most published data are based on light microscope observations. This makes it impossible to make clear comparisons, particularly as the Egyptian forms are, with the possible exception of STRIORET-SMOOTH, themselves indistinguishable without the use of SEM. Nevertheless, comparable forms do exist which are discussed below.
Grains with a strioreticulate sexine similar to that of STRIORET-SMOOTH include Striatopollis paraneus (Norris) Singh, 1971, Retitricopites vermimurus Brenner, 1963 and Striatricopites reticulatus Regali et al., 1974. S. paruneus is known from sediments of Albian to Cenomanian age in Canada (Norris, 1967;Singh, 1971;Playford, 1971), U.S.A. (Doyle & Robbins, 1977;Hedlund & Norris, 1968;Srivastava, 1975) and from Australia (Dettmann, 1973). Dettmann's (1973) SEM details reveal secondary transverse striations on the muri (PI. 2, figs. 21 and 22) which she points out as being invisible at magnificatons of less than x3000. This clearly refutes any relation between S. paraneus and STRIORET-SMOOTH. S . paraneus is, however, quite similar to STRIOTRI-SEGMUR in its possession of transversely-striate muri and the irregular projec-tions which were interpreted by Dettman (1973) as free standing columellae on the aperture membrane. In addition, there is good agreement between the sizes of the two forms. Also of interest is the secondary layer in the sexine of STRIOTRI-SEGMUR which sometimes forms a pseudoreticulum. This might agree with the observation of the pseudoreticulate sexine of S . paraneus, particularly apparent in the light microscope figures of Doyle & Robbins (1977). However, STRIOTRI-SEGMUR and S. paraneus differ in minor details, especially in the dimensions of the muri which are larger in S . paraneus, with a much more solid appearance. STRIOTRI-SEGMUR occurs in Aptian strata leading to the conclusion that it might represent a precursor of S. paraneus.
Striatricolpites reticulatus Regali et al., 1974, which is of possible Aptian age, has a stratigraphic range similar to that observed for STRIOTRI-SMOOTHMUR. However, it is entirely different, being much larger (52pm) and therefore cannot be compared.
Retitricolpites vermimurus was described from the Potomac by Brenner (1963) and was also illustrated by Doyle & Robbins (1977). It is similar to STRIORET-SMOOTH, both in the size and in the possession of a pseudoreticulum (especially clear in P1. 4, fig. 30 of Doyle & Robbins, 1977). However, the two cannot be compared further without SEM details for the North American examples.
Laing (1975) illustrated the species Striatopollis sarstedtensis Krutzsch, 1959 from Albian to Cenomanian strata of S. England and N. France, providing useful SEM illustrations. Although he mentions a polar microreticulum, this feature is not obvious from the SEM picture, suggesting that this is again the result of the existence of a secondary layer in the sexine. The muri show a clear orientation parallel to the colpi, a feature which is also seen in the specimen illustrated from Portugal (Groot & Groot, 1962) and the Eastern North Atlantic (Kotova, 1978). Laing also points out the presence of "small cones" on the muri which are equivalent to the transverse striae on the muri of STRIOTRI-OVAL and STRIOTRI-SEGMUR. S. sarstedtensis is not, however, sufficiently similar to either of these types to invoke any specific connection.
Other species with comparable exine structure include Striatopollis dubius Jardine & Magloire, 1965 and Striopollenties terasmei Rouse, 1962. Unfortunately SEM detail is not yet available for either. However, both bear a resemblance to STRIOTRI-OVAL, although S. terasmei is less likely to compare as it occurs only in the Late Cretaceous and is also larger.
S . dubius is Albian to Cenomanian in stratigraphic distribution and is known from offshore N. America (Hochuli & Kelts, 1980), S. America (Regali et al., 1974) and from offshore and mainland W. Africa (Jardine & Magloire, 1965;Morgan, 1976;Kotova, 1978). It is very similar to STRIOTRI-OVAL in the way the muri run down into the colpi and then turn to follow the colpus margins (see PI. 12, figs. 12 a,b,c of Kotova, 1978). However, it differs from STRIOTRI-OVAL in being larger.
S. dubius is also quite similar to STRIOTRI-SMOOTHMUR but is again larger. In addition, STRIOTRI-SMOOTHMUR has a more random arrangement of its muri, which do not show such a clear tendency to run perpendicularly into the colpi as is the case in STROTRI-OVAL.
There have been several unidentified striate tricolpates in the literature. Of particular interest are the possibly Aptian examples of  and Gubeli et al. (1985, no illustration), because they coincide better with the ranges observed for the Egyptian species. Unfortunately these forms are not yet sufficiently well documented to allow comparison.

DISCUSSION
Statigraphic Setting. The almost exlusively terrestrial character of the palynoflora has presented problems with the dating of the sequence in Mersa Matruh 1, mainly due to the paucity of dinocysts for independent control. Thus the chronology of the sequence is based entirely on observations of the spores and pollen which are present.
The earliest angiosperms in Mersa Matruh 1 occur in sediments which are of probable Late Barremian age (Penny, 1986a) and compare well with those of other Late Barrenian floras from W. Africa , N. America (Doyle et al., 1975;Hickey & Doyle, 1977) and S. England (Hughes et al., 1979), particularly with respect to the Stellatopollis forms which are present.
The strata immediately underlying the first tricolpates have been dated as Late Barremian to Earliest Aptian and are marked by the first occurrences of Afropollis Doyle et al., 1982. Of these, the earliest known specimens are possibly Late Barremian in age (Penny, 1986b). Also occurring for the first time at this level is the Retimonocolpites-reticulatus-peroreticulatus group of Schrank (1982), which has a ?late Barremian earliest appearance (Penny, 1988). It is just above these strata that the first striate tricolpates occur, close to the lower limit of definitely Aptian strata. These are also the earliest tricolpates to occur in the well. In other areas the first tricolpates to appear are reticulates (e.g. Herngreen, 1975;Gebeli et al., 1985). The apparent reversal of the situation in Egypt is almost certainly a false impression due to the extreme rarity of reticulates in older strata. This possibility is reinforced by the appearance of a single reticulate tricolpate grain in the Barremian core sample MMX-1 11832. The upper range of the striates coincides with the ranges of Retimonocolpites mawhoubensis Schrank, 1983 and Reyrea polymorphus Herngreen, 1973. R. mawhoubensis was described from the Late Aptian to possibly Early Albian strata penetrated by the Mawhoub West 2 well (Schrank, 1983) which is to the south of Mersa Matruh 1. R. polymorphus has a similar range (Herngreen, 1973;Thusu & Van der Eem, 1985;Regali et al., 1974). The age inferred by the presence of these two species in Mersa Matruh 1 is therefore Late Aptian to Early Albian. The striates disappear just after the last occurrence of R. mawhoubensis in strata of presumed early Albian age. Early Tricoplate Succession. The possible appearance of striates before reticulates is probably due to the rarity of reticulates in earlier strata rather than to a reversal of the recognised evolutionary succession. Nevertheless, it is interesting that in the early part of their range the striates account for the great majority of the tricolpate grains which were recovered. This may reflect a local predominance of striate producers in the flora. In terms of the actual numbers of taxa which are present there is greater parity, although the early reticulates are all sporadic in occurrence. This situation is reversed in the Late Aptian-Early Albian succession as the striates decline and reticulates become more common, both in the number of taxa and in the number of specimens recovered.
Doyle et al. (1977) indicate the possibility that the appearance of tricolpates in the W. African sequence could be used as a stratigraphic reference to identify the lower limit of the Aptian. However, it is possible that the tricolpates appear earlier in Eastern Gondwana. This is indicated by the observations of Brenner (1974Brenner ( , 1976 who found tricolpates in possible Barremian strata from boreholes in the Northern Negev of Israel. Support for this contention is provided by the single occurrence of a reticulate tricolpate in the probable Barremian strata of Mersa Matruh 1. Unfortunately both the observations of Brenner (1974Brenner ( , 1976 and those in the present study are subject to a certain degree of uncertainty. In the Negev case caving in the well may have led to downhole contamination and in Mersa Matruh (although the grain in question did come from a core sample) the number of specimens is too small.
For the present the earliest record of tricolpates in Eastern Gondwana must be viewed as possible earliest Aptian semu . This is in spite of the tantalising suggestion that this early stage of angiosperm pollen evolution might yet be pushed into the Barremian succession. Mr. D. Newling and Mr. R. Lee were of invaluable assistance with the photography. I a m very grateful to the Department of Earth Sciences, Cambridge University, for the use of their facilities during the preparation of this paper.

Manuscript received August 1987
Revised manuscript received February 1988