Plumularia lagenifera (sensu Millard, 1975)

Schuchert, Peter, 2013, The status of Plumularia lagenifera Allman, 1885 (Cnidaria, Hydrozoa) and related species, Zootaxa 3613 (2), pp. 101-124 : 106-112

publication ID

publication LSID

persistent identifier

treatment provided by


scientific name

Plumularia lagenifera


Plumularia lagenifera

Most of the type material of P. lagenifera (BMNH 1959.10.21.1) is a dry herbarium preparation as already described by Allman (1885). Only a small part is kept in alcohol and it was only this material that was available for study. It consists of four stem fragments lacking soft tissues or gonothecae. The partially shrunken appearance of the material strongly suggests that it was also once part of the dry material and put into alcohol either by Allman himself or a later investigator. There are only a few hydrocladial segments left, but all of them have hydrothecae with a distinctly curved outer wall ( Fig. 8B–D View FIGURE 8 ). The internal ridges of the hydrocladial segments are distinct, but not more pronounced than those seen in P. setacea ( Fig. 4 View FIGURE 4 ). One peculiarity is the slight notch on the outside of the proximal annular thickenings, this in both the main and intersegment. However, this could be an artefact due to the drying out of the material. Also the strong curvature of the hydrotheca could in fact be pronounced by some inevitable shrinking during the initial dry preservation.

Other material of P. lagenifera collected in the region of Vancouver Island was identified as such if several of their hydrothecae had a convex outer wall ( Fig. 9 View FIGURE 9 , lowest row). However, the curvature was rarely as pronounced as in the type material (e. g. Fig. 9D View FIGURE 9 3 View FIGURE 3 , comp. with Fig. 8C–D View FIGURE 8 ). Usually a number of hydrothecae within a single plume had almost straight walls ( Fig. 9B View FIGURE 9 5 View FIGURE 5 ), whilst others were more or less curved ( Fig. 9B View FIGURE 9 3–B View FIGURE 3 4 View FIGURE 4 ). There is thus a considerable variability of this character, although concave walls were never seen. The variability of the curvature could be used as an argument to consider those Pacific colonies with a straight hydrotheca, here attributed to P. setacea ( Fig. 6 View FIGURE 6 ), as merely extreme variants of the same population. The only, but weak, counter-argument is the occasional presence of two apophysal nematothecae ( Fig. 7A View FIGURE 7 ), a trait that was never seen in the P. lagenifera samples examined for this study.

(A) MHNG-INVE-54566, Plymouth, GB, 12– 18 m. (B) MHNG-INVE-36290, Roscoff, 0 m. (C) MHNG-INVE-29393, Bay of Morlaix , 20 m. (D) MHNG-INVE-70533, Roscoff, 70 m. (E) MHNG-INVE-37537, Spain, 20 m. (F) MHNG-INVE-25346, Iceland, 206 m.

(A) MHNG-INVE-36290, Roscoff , 0 m. (B) MHNG-INVE-54566, Plymouth, GB, 12- 18 m. (C) MHNG-INVE-37537, Spain, 20 m. (D) MHNG-INVE-25346, Iceland, 206 m. (E) MHNG-INVE-29393, Bay of Morlaix. (F) MHNG-INVE-70533, Roscoff, 70 m. (D) MHNG-INVE-78076, Vancouver Island , Canada, 22 m.

Comparing the figures of the Atlantic P. setacea ( Fig. 4 View FIGURE 4 ) and P. lagenifera ( Fig. 8 View FIGURE 8 ), the latter seem to have distinctly larger stem segments, but more samples could not confirm this (table 1). For all variables there is an overlap of their sizes and a single character variable would not allow the two populations to be distinguished reliably. Principal components analysis ( Dunn & Everitt 1982) can analyse numerous variables simultaneously by producing a smaller number of abstract variables (linear combinations of the original variables, principal components) and it provides a graphical summary which visualizes the data by placing similar samples close together, while dissimilar samples will result in points which are more distant. If this graph shows that the data result in distinct, spatially separated clusters, then independent characters can be used to identify these groups (in this case the convex hydrotheca or the geographic origin). The variables of table 1 can be combined so that two new variables (PC1 and PC2) represent about 54 % of the original variation ( Fig. 11 View FIGURE 11 ). The broken stick criterion ( Frontier 1976; Jackson, 1993) showed that only components 1 and 2 are usable (43.9 and 20.5 %, see figure 11). Exclusion of the variable "plume height" which is highly dependent on environmental conditions did not yield in significantly different clustering (not shown).

(A) MHNG-INVE-78965, Puget Sound, 118 m. (B) MHNG-INVE-78076, Vancouver Island , 22 m (C) MHNG-INVE-78077, Vancouver Island, 0.5 m .

(A) BMNH 1915.3 .6.44, Vancouver Island , identified by C. M. Fraser. (B) MHNG-INVE-25120, floating docks of Friday Harbor Laboratories. (C) MHNG-INVE-78974, floating docks of Friday Harbor Laboratories. (D) MHNG-INVE-78079, Juan Fuca Strait, 22 m.

(A) MHNG-INVE-78964, floating docks of Friday Harbor Laboratories. (B) MHNG-INVE-78966, floating docks of Friday Harbor Laboratories. (C) MHNG-INVE-25120, floating docks of Friday Harbor Laboratories. (D) MHNG-INVE-78079, Juan Fuca Strait, 22 m.

The clusters of Atlantic P. setacea a nd Pacific P. lagenifera show a more or less good separation ( Fig. 11 View FIGURE 11 ), but it is not complete. Moreover, the samples of P. setacea of the NE Pacific are intermediate to both populations and make a distinction even more difficult using numeric values alone.

The measurements thus give us no additional, solid argument for a separation of P. lagenifera from P. setacea . We are thus still left with the curvature of the outer hydrothecal wall as the only trait that enables us to separate the two nominal species. Unless future population genetic analyses indicate otherwise, it seems preferable to regard P. lagenifera as distinct from P. setacea .

The status of the South African P. lagenifera sensu Millard, 1975 , here attributed to the name P. gaimardi , is discussed in the following section.