Capel Horeb Quarry

Highlights

Capel Horeb Quarry has yielded the earliest examples of plant axes with vascular tissue found anywhere in the world (Figure 3.13). This is thus the oldest unequivocal evidence of a land vascular plant.

Introduction

This site is a disused quarry near Llandovery, Powys [SN 844 323], and has yielded plant fossils of Ludlow and also possibly Přídolí age. The fossils are usually very fragmentary, and some early descriptions of the site make little or no mention of them (Straw, 1930; Potter and Price, 1965).

Better preserved material has been found, as described by Heard (1939), Edwards (1970b, 1982), Edwards and Davies (1976) and Edwards and Rogerson (1979).

Description

Stratigraphy

The geology of this site has been described by Straw (1930), Potter and Price (1965), Edwards and Richardson (in Friend and Williams, 1978) and Edwards and Rogerson (1979). The lower part of the sequence consists of shallow marine siltstones and sandstones with a restricted shelly fauna, and belongs to the Upper Roman Camp Formation. Faunal evidence discussed by Potter and Price (1965) suggests a late Ludlow (mid-Ludfordian) age for these strata, a conclusion supported by the microfossils (Doming in Edwards and Davies, 1976). Lying unconformably above these beds is the Long Quarry Formation (Figure 3.14), which has been interpreted as either upper Ludlow (Richardson and Lister, 1969) or lower Přídolí (Potter and Price, 1965). They are probably littoral siltstones and sandstones. The topmost part of the sequence belongs to the Red Marls Formation.

Palaeobotany

The following plant fossils have been reported from the Upper Roman Camp Formation here:

Phaeophycophyta(?):

Nematothallus sp.

Nematoplexus sp.

Rhyniophytoids:

Cooksonia hemisphaerica Lang

cf. C. caledonica Edwards

Steganotheca striata Edwards

The Long Quarry Formation has yielded the following rhyniophytoids:

Cooksonia cf. pertoni Lang

S. striata Edwards

cf. Renalia sp.

The thalloid-like structure described by Heard and Jones (1931a) as Eohepatica dyfriensis, and by Heard and Jones (1931b) as Thallomia Ilandyfriensis, is now believed to be part of a dictyocarid arthropod (Rolfe, 1969).

Interpretation

The Upper Roman Camp flora appears to be dominated by what are probably non-vascular land plants, such as Nematothallus. Bulk maceration has yielded cuticles showing a reticulate pattern on their inner surface, characteristic of the genus (Edwards, 1982). There is considerable variation in this patterning, but it is not yet clear whether more than one species is present. The evidence from Capel Horeb suggests that Nematothallus was a thalloid, encrusting plant, rather than a leaflike structure of a Prototaxites, as suggested by Lang (1937) and Jonker (1979).

Edwards (1982) also found a variety of fine tubes in her bulk maceration samples. The majority were isolated tubes of uncertain affinity. There were, however, some clusters of tubes that resemble those found in Prototaxites axes. These were named Nematoplexus by Edwards, following the nomenclature of Lyon (1962).

Rhyniophytoid plants are represented here by a number of different taxa recognized on the basis of their reproductive structures. The most abundant belong to Cooksonia. The specimens from the Upper Roman Camp Formation have mostly globose sporangia, which can be identified as C. hemisphaerica (Edwards and Rogerson, 1979), although one was assigned to cf. C. caledonica (Edwards and Rogerson, 1979, pl. 1 fig. d). Other specimens from these strata had more elongate sporangia, which lie outside the circumscription of Cooksonia as defined by Lang (1937), and were not identified by Edwards and Rogerson. C. hemisphaerica is known to have a variety of sporangial shapes (Edwards, 1979a) and the possibility that these slightly more elongate forms belong there cannot be excluded.

Only one fertile specimen of Cooksonia has been reported from the Long Quarry Formation here (Edwards and Rogerson, 1979, pl. 1 fig. h). This showed squatter sporangia than those from the Upper Roman Camp Formation and was identified as C. cf. pertoni Lang.

This is the type locality for another type of rhy-niophytoid plant: Steganotheca striata Edwards, 1970b (fig. 8b and (Figure 3.15)). It occurs in both the Upper Roman Camp and Long Quarry formations. Like Cooksonia, it has slender, dichotomous axes with terminal sporangia. The sporangia are, however, elongate and less well individualized, and usually show a heavily carbonized, lenticular structure at the apex. It is at present unclear whether this structure is simply due to compression of the sporangial tip, or is evidence of a dehiscence structure. Because of the elongate shape of the sporangia and isotomous branching, Edwards (1970b) initially placed Steganotheca in the Rhyniaceae, but it has subsequently been described as rhyniophytoid (Pratt et al. , 1978; Edwards and Edwards, 1986).

The holotype of Cooksonia downtonensis Heard, which originated from Capel Horeb (Heard, 1939), was transferred to S. striata by Edwards (1970b). It is arguable that Heard's name for this species should take priority, but this is not the place to propose a new combination.

A single specimen from the Long Quarry Formation (Edwards and Rogerson, 1979, pl. 1 fig. i) shows a more complex branching pattern than the other rhyniophytoid species found at Capel Horeb. Edwards and Rogerson compared it with Cooksonia hemisphaerica Ananiev and Stepanov, 1969, non Lang, which Gensel (1976) has in turn compared with Renalia. The latter shows characters intermediate between the Rhyniaceae and Zosterophyllaceae and there has been some disagreement as to its taxonomic position. However, Edwards and Edwards (1986) have recently argued that it belongs to the Rhyniaceae, and that its zosterophyll-like characters are due to evolutionary convergence.

Far commoner than the fertile rhyniophytoid specimens discussed above are unbranched and dichotomous axes without sporangia. Being sterile, it is impossible to identify them beyond Hostinella sp. However, examples from the Upper Roman Camp Beds have been shown to have in situ tracheids (Edwards and Davies, 1976), the earliest known examples of axes with such tissue. Prior to Edwards and Davies' (1976) study, the oldest known axes with in situ tracheids were from the basal Devonian at Targrove Quarry (Lang, 1937). As with the Capel Horeb specimens, none of the Targrove axes with tracheids bore sporangia and so it has been impossible to confirm their identification. However, they were found in association with fertile specimens of Cooksonia hemisphaerica, and it has been widely asserted that they belonged to the same plant (e.g. Taylor, 1981). Dispersed tracheids have been reported from strata as old as the Cambrian (Gosh and Bose, 1952; Jacob et al., 1953) but Banks (1975a) has argued that these cannot be relied on as evidence of the presence of the Tracheophyta, because of possible reworking or contamination. The Capel Horeb specimens are thus the oldest indisputable evidence of land vascular plants.

Conclusion

Capel Horeb has yielded the oldest unequivocal evidence of plants with vascular conducting tissue (xylem) from anywhere in the world; the fossils are c. 415 million years old. The development of this tissue was one of the key steps that helped plants overcome the hydraulic problems inherent with living on the land, and thus paved the way for the evolution of land vegetation as we see it today.

References