82 



J.A. ALLEN, H.L. SANDERS AND F. HANNAH 



40-W\L 

 30- 

 80-|H\L 

 70- 



•• • 



:* * 



60- 

 50- 



PL\TL 



• • • • • 



Length (mm) 



Fig. 134 Portlandia abyssorum. Variation in ratios of height H/L 

 width W/L and postero-umbonal length PL/TL to length against 

 length of specimens from Sta. 195, Angola Basin (closed 

 triangles) and Sta. 8528', Cape Verde Basin (closed circles). 



Fig. 135 Portlandia abyssorum. Lateral view from the right side of 

 the internal morphology of a specimen from Sta. 8528 \ Cape 

 Verde Basin. (Scale = 1.0 mm). For identification of the parts see 

 Fig. 34. 



DISCUSSION 



Of all the protobranch bivalves of the Atlantic, the yoldiellids 

 are by far the most difficult nuculanid subgroup in which to 

 discern evolutionary pathways. Despite the large number of 

 species we believe that these are closely related within a 

 subfamily. With possibly one exception, of 28 species of 

 Yolidella described here, there is a fine gradation in morpho- 

 logical features that both combines them within a single genus 

 and distinguishes them as a separate group. It must be 



emphasized that as in all deep-sea protobranch species, the 

 shell proportion changes with increasing size. In general the 

 post-umbonal length increases at a rate greater than the other 

 dimensions. In some cases this also applies to the height of 

 the shell. As a result the small shells of a species may be 

 mistaken as being of a different species. This also creates 

 difficulties in making comparisons between species. 



To the experienced eye, species and subspecies can be 

 separated and defined. Of particular diagnostic importance is 

 the form and course taken by the hind gut. Although there 

 may be some variation within a species, the course and 

 diameter of the gut alone is usually diagnostic. Nevertheless, 

 it is often difficult to define and only becomes clear after close 

 comparison of a range of species. We can distinguish eight 

 broad configurations of the hindgut of which the simplest, a 

 single loop to the right of the viscera occurs in 17 of 32 species 

 and subspecies of Yoldiella (Table 3) (Allen, 1992). The most 

 complex configuration occurs in only one species (see below). 



The species of Portlandia form a compact group in which 

 the shells are relatively robust, elongate, with the posterior 

 margin approximately central to the horizontal midline. The 

 hind gut has a single loop to the right and has a large 

 diameter. The adductor muscles are large and oval and more 

 or less equal in size. The hinge plates are long and stout, 

 continuous with the amphidetic internal ligament. The lunule 

 and escutcheon are usually well-defined. Three of the four 

 species are from upper to mid-slope depths and one P. 

 abyssorum, is from abyssal depths. The latter, apart from 

 having larger palps, smaller adductor muscles and more 

 dorsal anterior and posterior limits to the shell outline than 

 the other species described, its morphology is basically the 

 same. 



One rare species of Yoldiella, Y. veletta, has many of the 

 characteristics of Portlandia described above. However, it is a 

 fragile shell without lunule or escutcheon. Furthermore, the 

 hinge plates form in a narrow bridge below the umbo with the 

 amphidetic ligament ventral to it. With only three specimens 

 at hand we defer categorical judgement, but hypothesize that 

 the primitive form of Yoldiella must have had similar charac- 

 teristics. 



On the premise that the simplest form of hind gut is likely 

 to reflect the primitive condition we derive an evolutionary 

 pattern that originates in species with this character but 

 among others, for it is unwise to base evolutionary conclu- 

 sions on one character alone. In passing, it should be said that 

 it is a sad fact that for many malacologists, it is the shells 

 rather than the viscera that are all-important. 



It seems reasonable to assume that deep-water species in 

 the Atlantic originated either from shallow water, possibly 

 tethyan and arctic seas, or by migration at depth from the 

 Southern Ocean. For reasons that we will describe elsewhere 

 (Allen and Sanders in m/s), we think it unlikely that the 

 major colonization of the deep Atlantic was from high 

 southern latitudes. Yet there are only a limited number of 

 yoldiellids present in shelf sediments, but all of these species 

 have a relatively short hindgut with a simple single loop to the 

 right side of the viscera. This character is also found in other 

 shallow-water nuculanids and in shallow-water neilonellids 

 and tindariids, however, species of Yoldiella differ from these 

 latter three groups in not having heavy, concentrically orna- 

 mented shells. When concentric ornamentation is present in a 

 yoldiellid it is always of a fine, delicate nature, and more 

 often than not confined to the periphery of the shell. 



We identify two species from shallow water in the North 





