A description of two new species of the genus Erenna (Siphonophora: Physonectae: Erennidae), with notes on recently collected specimens of other Erenna species Author Pugh, P. R. Author Haddock, S. H. D. text Zootaxa 2016 4189 3 401 446 journal article 10.11646/zootaxa.4189.3.1 4ecc0f25-80ae-4624-8062-b316874b3c92 1175-5326 166117 0F0EE4C3-05B9-4248-944D-B41DC8361CB2 Erenna cornuta Pugh, 2001 Diagnosis. Nectophores with apico-laterals not dividing close to ostium and very weakly expressed in distal half. Thrust block small, with no median indentation or conical protuberances. Lateral radial canals typically have 'horn' canals branching off at apico-lateral margins of nectosac. Two types of bract both with weak transverse ridge and asymmetric pair of lateral cusps. One type with characteristic elongated ectodermal process at distal end of bract, on upper surface. Three types of tentilla; the most abundant and largest one with nematocysts, on the long cnidoband, separated into bundles, and a spherical terminal process containing a pair of photophores and a redpigmented spot. Material examined. Pugh's (2001) description of Erenna cornuta was based on a single specimen that, as will be shown, appeared to be immature. Eight further colonies, together with a fragment, have been captured by the MBARI ROVs and we have been able to study six of these, as listed below.

ROV	Dive	Date	Position	Depth (m)	T°C
Ventana	1402	8 May 1998	36°41.83'N 122°01.87'W	833	4.15
Tiburon	412-D3	23 March 2002	36°36.20'N 122°22.62'W	1673	2.47
Tiburon	964-D4	7 April 2006	36°19.78'N 122°54.02'W	1392	2.92
Tiburon	1039-D3	30 September 2006	34°17.65'N 124°03.09'W	1328	3..25
Tiburon	1039-SS11	30 September 2006	34°17.66'N 124°03.94'W	1325	3.28
Tiburon	1155-SS2	1 December 2007	35°42.24'N 122° 34.75'W	1629

Description. Colonies of Erenna cornuta can instantly be recognized in situ by the long streak of ectodermal cells on the distal ends of the bracts ( Figure 14 ). Pneumatophore : c. 1.5 x 1 mm, with no pigmentation. Nectophores : A close-up of the in situ nectosome ( Figure 15 A) showed that the white lateral ostial processes are a prominent feature as was the black pigmentation in the ostial ring canal. It was also present in the radial canals, but this was not apparent from the photograph. It may be that the prominent white lateral ostial processes are another character that distinguishes this species in situ , as for the other species they were generally less obvious, except for Erenna laciniata (see Figure 9 B) where the proximal part is black, but with a distinct white distal portion. The type material included two developing and 13 larger nectophores that measured up to 16 mm in length, 14 mm in width and 4 mm in depth. This, as with other characters discussed below, seems to indicate that the colony was quite young as the mature nectophores from the Tiburon Dive 1155 specimen averaged 32 mm in length and width, with a maximum of 35 by 33 mm , respectively. Four very young nectophores, four developing and 27 mature ones from that specimen were preserved. FIGURE 14. In situ photograph of Erenna cornuta taken during Tiburon Dive 1155 at 1628m. MBARI. FIGURE 15. Erenna cornuta Pugh, 2001 . A . Detail of nectosome MBARI.; B . young nectophore from Tiburon Dive 1155-SS2 specimen. Scale bar 1 mm. The arrangement of the ridges on the youngest nectophores ( Figure 15 B) was somewhat different from the fully mature ones. Neither the upper or lower lateral ridges reached to the ostium, and the upper ones closely approached each other distally, then diverged laterally and petered out. Although they appear to do so in the photograph, the vertical lateral ridges did not join with the upper laterals at this stage. The lateral ostial processes were very obvious and were packed with nematocysts, which measured c. 110 x 35 µm. As the discharged ones did not appear to have a shaft it was presumed that they were isorhizas. The processes were penetrated by a canal that arose from the ostial ring canal at the same level as the insertion of the lateral radial canals. The thrust block formed a small mesogloeal protuberance, slightly emarginated in the mid-line. There was no sign of a muscle-free zone on the nectosac, while the lateral "horn" canals were greatly enlarged and included a broad canal. All the canals were obviously pigmented. The patch of ectodermal cells on the sides of the nectophore just distal to the vertical lateral ridges had yet to appear. FIGURE 16. Upper (left) and lower (right) views of young nectophore of Erenna cornuta Pugh, 2001 from Tiburon Dive 1155- SS2. Scale bar 5 mm. For the slightly larger nectophore shown in Figure 16 , the axial wings had greatly increased in size as had the mesogloeal forming the thrust block. The arrangement of the ridges remained much the same, although there was now a weak connection between the vertical and upper lateral ridges. The upper lateral ridges often overlapped each other before they diverged close to the ostium, but that was probably a result of preservation. Unlike Erenna richardi these ridges did not divide. The nematocysts on the lateral ostial processes had by now been lost, by abrasion or usage, and so the central canal became more apparent. At this stage the lateral parts of the nectosac extended more proximally than the mid-region, but again this probably was the result of preservation. The lateral "horn" canals had narrowed to a similar diameter as that of the lateral radial canals, but there was still no sign of a muscle-free zone. The lateral photophores also had yet to appear. The nectophores of Erenna cornuta illustrated by Pugh (2001, Figure 12 ) showed a slightly more advanced stage of development. The proximal halves of the upper lateral ridges now paralleled the lower ones, and formed the upper lateral margins of the nectophores. However, distal to the vertical lateral ridges, they still curved in toward the mid-line, and then down toward the ostium, while rapidly petering out. Often they could only be traced by staining the material. The lower lateral ridges extended down to the lower lateral margins of the ostium. The proximal end of the nectosac was almost flat, and the muscle free zone in that region was quite distinct. The lateral ostial processes remained quite prominent and the pair of lateral photophores appeared just distal to the vertical lateral ridges. In their preserved state they were brown in colour, but would have been black in life. The new material showed that that was not the final developmental stage. For the largest nectophores ( Figure 17 ) the upper and lower lateral ridges then almost paralleled each other throughout their length and formed the lateral margins of the nectophore. The thrust block had considerably increased in size now formed a very characteristic swollen mesogloeal process on the upper side of the nectophore. It was only attached to the main body of the nectophore by a thin strip of tissue and, thereby, formed a sort of a thickened flap, with a broad but shallow proximal emargination. The nectosac was T-shaped and had a distinct muscle-free area on either side of the proximal parts of the lateral radial canals. The thickened ascending mantle canal arose on the upper side of the nectophore, below the thrust block, and passed through a proximal groove onto the lower side where it gave rise to the narrower pedicular canal. This ran straight to the nectosac and gave rise directly to the lateral radial canals. For the type specimen no pigmentation was noted in the radial canals, but for the Tiburon Dive 1155 nectophores dark brown granulations could be seen in all the canals, including the "horn" canals, and particularly in the ostial ring canal. This pigmentation was probably black in life. No nectophores were observed where the "horn" canals were absent. They usually curved inward proximally, and their walls were often irregular, occasionally with small diverticula. FIGURE 17. Lower view of mature nectophore of Erenna cornuta Pugh, 2001 . Scale bar 1 cm. One strange feature of some of the nectophores was the presence, at least in the preserved material, of thickened pieces of tissue that hung off the upper or lower lateral ridges anywhere proximal to the vertical lateral ridges or just distal to them. They appeared to be formed by a thickening of the tissue along the ridge. One end of this thickening then broke free, and gradually the free part lengthened until finally the tissue projected out from the ridge as a sort of digitate process. The size of these processes varied greatly but they could be up to 5 mm in length ( Figure 17 ). At first they were thought to be preservation artefacts but as they were found at all stages of development, and were present on more than half of the mature nectophores of the Tiburon Dive 1155 specimen, it was thought that this was something that may have occurred in life, although the reason for doing so remains obscure. FIGURE 18. Detail of siphosome of Erenna cornuta Pugh, 2001 from Tiburon Dive 1155. MBARI. Siphosome : The siphosome of the Tiburon Dive 1155 specimen in situ is shown in Figure 18 and clearly shows the long, white patches of cells on the upper surfaces of the bracts; the elongated gastrozooids, with their tentacles, and the gonodendral clusters of male gonophores. Because the stems were so highly contracted during preservation, it was not possible to study the organisation of the individual cormidia in detail. FIGURE 19. Erenna cornuta Pugh, 2001 . Type A bracts. Scale bar 5 mm. Bracts : As was described for the type material ( Pugh, 2001 ), two types of bract were found with the specimens of Erenna cornuta . The longer, Type A , bracts of the type specimen measured up to 16 mm in length but, as with the nectophores, the bracts found with the present specimens were considerably larger, reaching more than 40 mm in length ( Figure 19 ). They were also about twice as common as the other type . With this further increase in size the transverse ridge, which had been incomplete in the smaller examples, now became complete. It formed a distinct flap across the bract, and on the inner side it extended distally as a process thickened with mesogloea. The more proximal of the two asymmetrically placed lateral cusps, which were associated with it, disappeared, while the distal one became more prominent, as did the more distal lateral cusp, on the inner side of the bract. The central rounded spot on the upper side just distal to the transverse ridge usually could not be discerned on the largest bracts, but was quite evident on the smaller ones ( Pugh, 2001, Figure 15 ). The bracteal canal arose on the upper side of the bract and on passing over onto the lower side it gradually increased in diameter up to about one third the length of the bract. It then maintained that diameter until close to its distal end, where it thinned considerably as it penetrated into the mesogloea and ran at first obliquely and then perpendicularly to end below the proximal part of the elongate ectodermal process on the upper side of the bract. The canal might have extended further along beneath this process, but that could not be confirmed with certainty. The elongate distal process gave the bracts a very distinctive appearance that made it easy to identify the specimens in situ . Numerous nematocysts, which were anisorhizas measuring c. 130 x 38 µm before discharge but slightly smaller after, were present there. FIGURE 20. Erenna cornuta Pugh, 2001 . Type B bracts. Scale bar 5 mm. The shorter Type B bracts ( Figure 20 ) were also considerably longer those of the type material. The Tiburon Dive 1155 ones measured up to 30 mm in length and 15 mm in width. In basic design they were very similar to the Type A bracts, except that the transverse ridge lay in the distal half of the bract instead of the proximal one. The changes, as the bracts enlarged, in this ridge and the associated lateral cusps were very similar to those described for the Type A bracts, in that the ridge became complete and one of the cusps, the outer, was lost, while the inner one, and again the more distal lateral cusp, became more distinctive. The arrangement of the bracteal canal was exactly the same. The main change that occurred as the bracts grew was that the distal ectodermal process on the upper side of the bract became elongate, as in the Type A bracts. The process, as with the other Erenna species, was packed with nematocysts. Similarly, the median photophore, just distal to the transverse ridge, was usually not apparent on the largest bracts. Gastrozooid : The gastrozooids can clearly be identified on the in situ picture ( Figure 18 ), being swollen proximally and a long narrower stomach region, with a translucent distal proboscis region showing numerous longitudinal hepatic stripes. In the preserved state the basigaster was quite prominent and had a yellowish-orange colour. FIGURE 21. Erenna cornuta Pugh, 2001 . A . Tentillum of first type. dc : diverticular canal; gvc : gastrovascular cavity. Scale bar 1 mm. B . Distal end of mature tentillum of second type, from Tiburon Dive 964-D4 specimen. Scale bar 1 mm. C . Distal part of cnidoband from same specimen. Scale bar 2.5 mm. Tentilla : On the original holotype specimen Pugh (2001) described what he considered to be two different types of tentillum. However, as noted above, there are several reasons to consider the type specimen as a small, not fully developed colony, and it seems that what he considered to be the young stage of the second type of tentillum (his Figure 8 A) actually belongs to a third type. Pugh (2001) gave a detailed description of the first type of tentillum, which was relatively small, and had the nematocysts concentrated into four sub-terminal patches. This appeared to have been the commoner of the two types that he considered to be present. However, for the six further specimens presently available for examination, this type of tentillum was only found on one specimen, from Tiburon Dive 1039, and only three were found. These tentilla ( Figure 21 A) differed slightly from that originally described in that the tentillum was divided from its pedicle by a distinct annulation, distal to which the gastrovascular cavity at first ran along one side of the tentillum before abruptly moving to the mid-line. In addition there were five, not four, patches of nematocysts, three arranged in a row at the distal end of the tentillum, while the two others formed a lateral pair borne on short stalks. The nematocyst complement was not investigated further. All three of the present tentilla were detached from the tentacle, but Pugh (2001) noted both his types of tentillum were attached to the same tentacle. For the present material both the second and third types of tentillum were similarly attached to the same tentacle. It is not clear, at present, how to reconcile this apparent anomaly as to the general absence of the first type of tentillum from most of the specimens examined, but it is clear that this arrangement is highly unusual. It is unlikely that during development one transforms into another, as is the case for the two types of tentilla of Resomia species (Pugh, 2006), and a close examination of a tentacle with several young tentilla still attached showed that the order in which each type of tentillum was developed appeared to have no pattern. However, this phenomenon needs to be investigated further. FIGURE 22. Various stages in the development of the Type 2 tentillum of Erenna cornuta Pugh, 2001 . Scale bars 5 mm. The holotype specimen only possessed a few mature tentilla of what we here designate as the second form. In these the broad elongate cnidoband was largely filled with large, vacuolated endodermal cells that gave it a spongelike appearance. Many such tentilla were found with more recent specimens ( Figure 21 B, C) and were generally much larger. The cnidoband was penetrated by a narrow gastrovascular cavity that distally opened out into a spherical vesicle. Within this vesicle there were two photophores and a red-pigmented sphere ( Figure 21 B). The nematocysts were grouped into distinct domed patches along one side of the cnidoband, and as each patch was more transparent centrally, it took on the appearance of a sucker on a cephalopod tentacle. Pugh (2001) noted only a few "mature" tentilla of this type that possessed up to 17 of these nematocyst patches. However, during an initial examination of the specimen collected during Doc Ricketts Dive 32, before preservation, it was noted that one tentillum had 172 of these patches arranged along the cnidoband that extended to 15 cm in length. Again this exemplified that the holotype specimen was not fully mature. Figure 22 shows some stages in the development of the second type of tentillum of Erenna cornuta . At early stages the tentillum was basically a narrow tube overlain on one side, and for the most part, by the undifferentiated cnidoband. As the latter differentiated the nematocysts became grouped into distinct rounded patches, although they were tightly packed together to begin with. Towards maturity the main body of the tentillum expanded as it was filled by the spongy material surrounding the gastrovascular cavity, and the two photophores became apparent in the spherical terminal process. Unlike the other types of tentilla, there was no sign of a distal diverticular canal. FIGURE 23. Various stages in the development of the Type 3 tentillum of Erenna cornuta Pugh, 2001 . Scale bars 5 mm. Stages in the development of the third type the nematocysts are shown in Figure 23 . For this type , once the cnidoband has differentiated, it always formed a single patch along most of the length of the tentillum and, at its most extensive, had undulating lateral margins ( Figure 23 , top). The large photophores were apparent from a very early stage in the short, pointed terminal process. In life this process was observed to flick from side to side at regular intervals. The main body of the tentillum was filled with spongy material with the gastrovascular cavity running through it. This ran to almost the tip of the terminal process and gave off a diverticular canal that then ran proximally below the cnidoband to about half its length ( Figure 23 , top). The cnidoband of the fully developed third type of tentillum ( Figure 23 , bottom) appeared to be greatly reduced but, in actuality and probably during preservation, the cnidoband itself had become stripped from the tentillum leaving just a small ridge of black material. Concomitant with this stripping off, most of the nematocysts of one type in the cnidoband discharged so that the cnidoband then appeared as a fuzzy mat and these were frequently found loose in the preserved material. The nematocysts from one of these mats of tissue were examined and found to include the three types of nematocyst that had been found in the cnidobands of other Erenna species. The largest measured up to 172 x 42 µm ( Figure 24 A), and so were somewhat larger than those found with the type specimen, namely c. 130 x 35 µm. They were the only ones that were found to have discharged and were presumed to be microbasic mastigophores. The nematocysts shown in Figure 24 B were by far the most numerous and measured c. 45 x 20 µm, while the third type ( Figure 24 C), measuring c. 38 x 12 µm, was relatively rare. These sizes were similar to those found for the holotype specimen. Although no discharged ones of these types were found they were presumed to be the same pear-shaped and cylindrical ones observed in other species. FIGURE 24. Erenna cornuta Pugh, 2001 . A–C . Nematocysts from the cnidoband of the type 3 tentillum. D . Three types of nematocysts found on the proboscis region of the palpon. Scale bars A. & D. 50 µm; B, C, 25 µm. FIGURE 25. A . In situ frame grab showing some tentacles of Erenna cornuta Pugh, 2001 from Tiburon Dive 333. See text for details. B . In situ frame grab of posterior end of Tiburon Dive 1155 colony, showing the extremely long type 2 tentilla. MBARI. One of the in situ frame grabs taken during Tiburon Dive 333 of a colony of Erenna cornuta is shown in Figure 25 (left). The characteristic long distal patches on the upper sides of the bracts can clearly be seen as can, with close inspection, the three types of tentilla on the tentacles. The small Type 1 ones appeared to be more common on the posterior-most tentacles, being supplemented by the long irregular type 2 ones more anteriorly and then by the type 3 ones with the distal "ocelli" clearly showing up as white spots. On the other hand the Tiburon Dive 1155 colony showed the total dominance of the type 2 tentilla that reached extraordinary lengths ( Figure 25 right). Palpon : The palpons of the Tiburon Dive 1155 specimen were also much larger than those of the type specimen, measuring up to 19 mm long and 7 mm in diameter. They consisted of a very short pedicle to which a short palpacle was attached; an inflated stomach region the internal walls of which were covered with irregular patches of gastrodermal cells; and a short proboscis. The ring of pigmented spots of gastrodermal cells found in the smaller palpons of the type specimen at the proximal end of the proboscis had coalesced into a solid darkly pigmented region . No nematocysts were found on the palpacles, but at least three different types ( Figure 24 D) were found to be present in the proboscis region. The two larger types were by far the most common, but as well as the smallest, whose contents appeared to be completely transparent, a few very small nematocyst, 14 x 10 µm, were occasionally observed. What type any of them were was not ascertained as no discharged ones were found. Gonophore : For all the specimens examined no mature gonodendra were found, although some minute, very immature ones were found with the Tiburon Dive 1155-SS 2 specimen that were too young to sex. Distribution. Including the material examined, the following specimens have been either identified from in situ frame grabs or collected (emboldened).

ROV	Dive	Date	Position	Depth (m)	T°C
Ventana	1402	8 May 1998	36°41.83'N 122°01.87'W	833	4.15
Tiburon	333	16 July 2001	36°19.49'N 122°54.04'W	1830	2.28
Tiburon	412-D3	23 March 2002	36°36.20'N 122°22.62'W	1673	2.47
Tiburon	479	23 September 2002	36°20.19'N 122°54.47'W	1483	2.81
Tiburon	480	24 September 2002	36°34.53'N 122°31.38'W	464	6.323
Tiburon	514	23 November 2002	36°34.29'N 122°31.36'W	1381	3.09
Ventana	2394	1 July 2003	36°41.91'N 122°02.93'W	995	3.87
Ventana	2572	15 September 2004	36°42.47'N 122°04.11'W	1060	3.91
Tiburon	543	26 March 2005	23°37.73'N 108°46.99'W	1340	3.17
Tiburon	964-D4	7 April 2006	36°19.78'N 122°54.02'W	1392	2.92
Tiburon	964	7 April 2006	36°19.78'N 122°54.02'W	1320	3.02
Tiburon	965	8 April 2006	36°34.35'N 122°31.25'W	1526	2.77
Tiburon	1039-D3	30 September 2006	34°17.65'N 124°03.09'W	1334	3.25
Tiburon	1039-SS11	30 September 2006	34°17.66'N 124°03.94'W	1324	3.28
Tiburon	1106-D1	28 July 2007	36°07.00'N 123°38.00'W	1608	2.61
Tiburon	1109-D3	31 July 2007	35°50.00'N 122°40.00'W	1766	2.44
Tiburon	1115	9 August 2007	36°19.87'N 122°56.02'W	1567	2.76
Tiburon	1152	6 November 2007	36°38.09'N 122°08.98'W	1282	3.07
Tiburon	1152	6 November 2007	36°38.09'N 122°08.98'W	1366	2.91
Tiburon	1155	1 December 2007	35°42.24'N 122° 34.75'W	1542	2.63
Tiburon	1155-SS2	1 December 2007	35°42.24'N 122° 34.75'W	1629	2.55
Doc Ricketts	32-SS9	31 May 2009	36°15.00'N 123°10.00'W	1715	2.56
Doc Ricketts	310	2 November 2011	36°33.01'N 122°30.10'W	1407	2.85
Doc Ricketts	326	3 December 2011	36°04.27'N 122°16.98'W	1396	2.85
Doc Ricketts	343	25 February 2012	24°21.71'N 109°15.34'W	1418	3.32
Doc Ricketts	725	11March 2015	23°37'N 108°45'W	1590	2.87
Doc Ricketts	725	11March 2015	23°37'N 108°45'W	1323	3.27
Doc Ricketts	728	14 March 2015	23°41.5'N 108°49'W	1419	3.20
Doc Ricketts	728	14 March 2015	23°41.5'N 108°49'W	1404	3.22

All but six of these specimens of Erenna cornuta were seen or collected in a very small area in Monterey Bay, California and, as with other Erenna species, over a relatively narrow depth range of 833 to 1830 m , mean 1426 ± 207.3 m ; ignoring the one anomalously shallow value. The mean temperature was 2.9 ± 0.16°C. In the 1000–1500 m depth range E. cornuta co-occurred with both E. richardi and E. laciniata , but it was easily distinguished from them by the long patch of cells at the distal end of the bracts. The six specimens from the Gulf of California came from similar depths, mean 1438 ± 105.8 m , with a mean temperature of 3.17 ± 0.16°C. The type specimen came from off the Dry Tortugas , in the Gulf of Mexico at a depth of 896 m . As previously reported ( Pugh, 2001 ) four specimens have been found in the Discovery collections, two at 30°N, 23°W between 1250–1500 and 1500–2000 m , and two from off Bermuda (c. 31°45'N , 63°45'W ) at depths of 1250–1500 m . In addition, Totton (1965) mentioned a specimen with horns collected at Discovery St. 4255 from a depth of 1000 m (between 36°31'N , 11°24'W and 36°22'N , 11°15'30"W , to the west of the Straits of Gibraltar ). Thus, it is probable that this rare species has a worldwide distribution, but it has not yet been found off Japan although its sister species E. laciniata has (Dhugal Lindsay, personal communication).