Colossendeis macerrima, Wilson, 1881

Bamber, Roger N., 2002, Bathypelagic pycnogonids (Arthropoda, Pycnogonida) from the Discovery deep-sea cruises, Journal of Natural History 36 (6), pp. 715-727 : 720-723

The Colossendeis macerrima complex

The complex of species which have been variously attributed to Colossendeis macerrima Wilson, 1881 is of taxa with a long, slender proboscis (1.5 to 3 times as long as the trunk), distinct cuticular-reinforce d anterolatera l corners on the cephalon, and a palp article 3 about half the length of 5.

Calman (1923) mentioned forms (of `Colossendeis macerrima ’) with a distally narrow, upturned proboscis in some Arabian Sea material; Stock (1975), referring to these and considering C. minor Schimkewitsch, 1893 and C. gardineri Carpenter, 1907 , proposed to`follow Calman, and consider these diOEerent forms to belong to one, variable, species’.

Attempts to unravel the Colossendeis macerrima complex were undertaken by Stock (1978b) who reinstated C. leptorhynchus Hoek, 1881 , subsequently (1984) reinstating C. minor and C. gardineri , in the latter case after re-examination of the type and (then) only known specimen; unfortunately, he did not present data on the leg article lengths of C. gardineri , and the type has been re-examined herein to con®rm the data presented by Carpenter (1907) ( table 1).

Stock’s conclusions may be summarized as follows:

Colossendeis minor and C. gardineri both have a signi®cantly tapering proboscis, with the distal half about half the diameter of that proximal to the mid-proboscis swelling, and the tip markedly upturned, the mouth pointing upwards. Colossendeis minor has been collected over a range of sizes comparable to the known range of C. macerrima s. str.: its single consistent distinction is the relative lengths of the distal palp articles, with the 10th article signi®cantly longer than the 9th (ca 1.5 times as long), whereas it is subequal to each of articles 8 and 9 in the other three taxa.

Colossendeis gardineri has the C. minor proboscis morphology, but not the elongate distal palp article; in addition, the sole spines of both tarsus and propodus are conspicuously stronger than those of any of the other three taxa, wherein they appear as setae.

Colossendeis macerrima and C. leptorhynchus have a proboscis distally tapering slightly or not at all, straight or slightly upturned; the distal three palp articles are subequal; propodal sole spines (setae) are inconspicuous. Stock (1978b) redescribed the types of Colossendeis leptorhynchus , a distinct species characterized by article 7 of the palp being longer than the 6th (1.2 times as long, compared with <1.5 times as long in all other species of the complex) and the tarsus being more than twice the length of the propodus (2.3 times as long, compared with <1.9 times as long in all other species of the complex). Otherwise, it closely resembles C. macerrima s. str.

Fry and Hedgpeth (1969) undertook morphometric analyses of a number of (predominantly Antarctic) species of Colossendeis , but including C. macerrima , and indicated heterogonic growth, usually with ranges of various meristic ratios in

Coxa1 0.70 P3 3.425

Coxa2 0.83 P4 0.49

Coxa3 0.83 P5 4.815

Femur 13.93 P6 0.915

Tibia 1 13.89 P7 0.925

Tibia 2 11.83 P8 0.41

Tarsus 3.46 P9 0.38

Propodus 1.88 P10 0.48

Claw 0.59

relation to size (growth). They also presented formulaic interpretations of the compound spine ®elds of the distal oviger articles. While they present data for ranges of article sizes and ratios by species, unfortunately their data for the percentage lengths of leg articles for C. macerrima are incorrect (presumably by typographical error). Fortunately, their ratios of leg article lengths to each other and to other body parts are correct. What they did conclude for the genus was that, owing to diOEerential rates of growth between diOEerent structures, relative proportions vary continuously through size ranges within species.

I have examined the spine ®elds of the distal oviger articles of the Colossendeis gardineri holotype, and found them to conform with the de®nition given for C. macerrima by Fry and Hedgpeth (1969) (not subchelate; a distinct endal row of`needle’ spines and a ®eld of four or ®ve further rows of needle spines).

It is against this backgroun d that members of the Colossendeis macerrima complex need to be interpreted. Ratios of leg articles of these taxa generally overlap, owing to their diOEerential growth. Colossendeis gardineri Carpenter, 1907 has only been collected as small specimensÐCarpenter’s type at 6.5 mm trunk length, and Stock’s (1986) specimen of which he claimed to`hardly see any diOEerences at all between’ it and the type. While the second tibia was reported as considerably shorter in relation to the other leg articles in C. gardineri than in C. macerrima s. str. (tarsus plus propodus plus claw 0.5 times the length of tibia 2 in C. gardineri , compared with up to 0.3 times in C. macerrima ), this may well be a size-related feature, and leg article ratios (including tarsus to propodus ratio) may not be justi®able for splitting this taxon. Similar arguments may apply to the ratio of third to ®fth palp articles.

A number of specimens of the Colossendeis macerrima complex were present in the Discovery material, all small (and none resembling C. leptorhynchus ). In order to determine their identity, morphometric analysis was undertaken of this material, together with the type of C. gardineri , eight specimens of C. macerrima and two of C. minor , plus an undetermined specimen collected by Mauchline (1984), all from the north-east Atlantic (NHM collections: see Bamber and Thurston, 1995) together with four specimens of C. minor from waters oOE Fiji (courtesy of the MuseÂum National d’ Histoire Naturelle, Paris).

For each specimen, the 3rd to 10th articles of the palp, the distal four articles of the 2nd or 3rd leg (where available), and the maximum and distal diameters of the proboscis were measured. In addition, the lengths of propodal spines/setae were measured in comparison with the propodal diameter. Attempts to quantify the`upcurve’ of the proboscis proved impractical.

The lengths of the larger propodal sole spines of the Colossendeis gardineri holotype were approximately half the propodal diameter. Two specimens of the present material also had such large propodal spines and were tentatively attributed to this species; the remaining material bore propodal setae of length 0.03 times or less the propodal diameter, other than one C. macerrima (0.13 times) and the largest C. minor on which no propodal setae or spines were observed.

Some other consistent diOEerences were apparent ( table 2). Thus, the material attributed to Colossendeis minor showed a 10th to 9th palp article ratio (P10/P9) greater than 1.7, the ratio for the other taxa ranging between 1.2 and 1.6. The P5/P3 ratio distinguished all three taxa, being less than 1.5 for C. gardineri specimens (the type plus those of the present material with large propodal spines), greater than 1.9 for C. minor , and ranging from 1.6 to 1.93 for C. macerrima (thus slight overlap and standard deviations of parameters distinguishing Colossendeis between the last two). Ratios of distal to maximum diameter of the proboscis were less distinctive, being <0.65 for C. gardineri , between 0.4 and 0.8 for C. minor , and between 0.7 and 1 for C. macerrima . The distinctions of the analysed material on these three parameters are shown in ®gure 3.

Ratios of distal leg articles were not distinctive. No characters were related to size.

On the basis of these analyses, the Discovery material was attributed to Colossendeis gardineri (two specimens, representing the third occurrence of this species) and C. macerrima . Further, the specimen of Colossendeis sp. of Mauchline (1984) is now attributed to C. macerrima .