By Dr. Steve O'Shea
Auckland University of Technology, Earth and Oceanic Sciences Research Institute, Private Bag 92 006, Auckland, New Zealand.
Note: Steve welcomes discussion in TONMO's Architeuthidae forum.
Architeuthis is a leviathan amongst cephalopods, with mantle lengths reputed to, but in reality unlikely to ever approach 4−6 metres (Clarke 1966, Roper & Boss 1982). We stress the unlikely nature of such exaggerated mantle lengths because a marked contrast exists between ‘4−6 metres’ and measurements based on numerous stranded specimens, sperm whale gut contents, and fisheries bycatch, where mantle length characteristically ranges ~ 0.5−2.4 m (Aldrich 1991, Roeleveld & Lipiński 1991, Jackson et al. 1991, Gauldie et al. 1994, Fernandéz-Nunez & Hernandez-Gonzàléz 1995, Norman & Lu 1997, Lordan et al. 1998, Förch 1998, and personal observation).
As nobody has been able to rear deep-sea species of squid in captivity (yet), especially Architeuthis, we really have no idea how old this species is, and how fast it grows. Published estimates for growth rate in Architeuthis vary from 5.06−2.62 mm mantle length (ML)/day (d-1), with larger individuals appearing to grow faster than smaller individuals (Tables 1−3). These estimates are based on examination of two tiny bones (statoliths; Figs 1, 2) removed from the ventral surface of the squid’s head, that in cross-section reveal numerous rings, much like those in sectioned tree trunks; by counting the rings and comparing this to the mantle length of the squid we come up with an estimate of both growth rate and age. The problem is that these estimates have not been validated for deep-sea species — we really don’t know if an individual ring on the statolith is deposited on a daily basis. Although daily growth increments within the cephalopod statolith have been validated for Todarodes (Nakamura & Sakurai 1991), Illex illecebrosus (Dawe et al. 1985, Hurley et al. 1985), Sepioteuthis lessoniana (Jackson 1990), Idiosepius pygmaeus (Jackson 1989), and Alloteuthis subulata (Lipinski 1986), they have not for Architeuthis, nor for any other deep-sea/cold water species of cephalopod (although the growth rings are similar in appearance to daily growth rings of other squids (Dunning & Lu 1998)).
A wide variety of growth curves have been reported for squids, e.g., Boyle (1990) for Loligo sp., wherein it is argued that cephalopod growth is either slowed drastically or becomes erratic at the onset of sexual maturity. For species with a short-term breeding season, this point of slow growth might be reached simultaneously at the same age and at the same body size by the bulk of the population. In other species, sexual maturity occurs over a very wide range of body sizes. Growth also appears to vary widely from season to season (e.g., Hatfield (1991) for Loligo gahi), and from individual to individual of the same brood when ample food is available (Hurley 1976) and Yang et al. 1986, for Loligo opalescens).
Young squid are capable of fast, exponential growth when food is not limiting (Yang et al. 1986). Asymptotic mantle size as a function of time has been observed in wild populations, e.g., Natsukari et al. (1988) for Photololigo edulis, Spratt (1978) for Loligo opalescens, Patterson (1988) for Loligo gahi (although no evidence for asymptotic growth was found for L. gahi by Hatfield (1991), and for Illex argentinus (Hatanaka et al. (1985)). Linear, logarithmic, or exponential growth has been reported by Hurley (1976), Turk et al. (1986), Yang et al. (1986) and Hatfield (1991) for Loligo opalescens, for L. vulgaris, and L. opalescens respectively — all hatchery-reared animals where presumably neither food limitation nor temperature fluctuation were major issues.
New Zealand Architeuthis statoliths, when viewed from posterior and anterior views (Figs. 1, 2), are similar to those of Roeleveld & Lipiński (1991), Gauldie et al. (1994) and Lipinski (1997). They have a strong wing with regularly indented dorsal and lateral lobes, cavernous anterior “proto-sulcal” groove with an in-folding lobe edge, and a prominent rostrum.
While Lordan et al. (1998) present an image of a ground statolith, showing rings near the nucleus, neither they nor others have precisely defined its’ position in the statolith. Analysis of the dorso-lateral lobes under SEM indicated that they emanated from the capella region; intuitively, the nucleus should be contained therein. Thus, in this study, transverse thin sections of the Architeuthis statolith were orientated orthogonal to each other through the capella, dorso-ventrally or medial-laterally using the left and right statoliths of one specimen (see Fig. 2). Resultant thin sections demonstrated clearly that the nucleus did lie within the capella region, close to its anterior edge (Figs 3, 9).
At higher magnifications (Fig. 4), axial complexities or interruptions are evident. These could be attributable to incorrect alignment of the section, where the grinding plane has effectively crossed over one of the dorso-lateral lobes leading into the capella; the ability to count increments is consequently reduced. Broad mid-axis increments are a common feature at this magnification, with widths of ~ 25 µm comparing well to an average value of ~ 26 µm for similar broad bands, described as ‘dark check-rings’ caused by the fusion of micro-increments, by Gauldie et al. (1994). What periodicity, if any, these broad check rings have is open to speculation, but they do form a regularly repeating construct
Micro-increment pattern and width within the statolith is site specific and depends on the plane of optical focus. Mid-axis increments in the dorso-ventral orientation can be counted as large rings of width ~ 2.22 µm. Defocusing across the axis shows these larger rings to be composed of finer increments with widths of ~ 0.89 µm (Fig. 5). We propose that the larger rings are an artefact of the sub-surface crystallography, caused by an optical summation as light enters and exits the thickness of the preparation. The finer increments may then represent a surface-focused effect, minimising presumed interference or diffraction patterns in the sub-surface layers. This raises the possibility that ring counts within the published literature may be underestimating age since reported counts typically involve rings of width ~ 3 µm in Architeuthis. That lateral axes are smaller than the dorso-ventral ones infers that any available rings will be forced into a tighter agglomeration. Lateral thin-section data shows similar widths of the order of 1−1.5 µm, but also bunches of increments are discernible having broad dimensions of ~ 10 µm (Fig. 6).
The regularity of increment appearance can be striking as is seen in a post-nuclear image, antero-laterally aligned with ring widths between 1.25−1.5 µm (Fig. 7). These rings resemble those described for several teleosts (Pannella 1971, Brothers et al. 1976, Jones 1986, Campana & Neilson 1985, Campana 1992).
Postero-medial ring patterns are more complex in dorso-ventral section, with post-nuclear rings reducing in width to ~ 1.2 µm before disappearing to be become broad bands (~ 9 µm) composed of finer (~ 1.85 µm) rings (Fig. 8). By contrast, the orthogonal transverse section in the medial-lateral plane is approximately three times shorter than the corresponding dorso-ventrally aligned section (see Fig. 2). Temporal (age) information across this shorter axis should be condensed in space and time, with very narrow increments exceedingly closely spaced, but with major check marks easier to define. In section, through this axis from an 1125 mm ML specimen, 15−17 sharply defined large rings are apparent (Fig. 9). If these rings were of a lunar/monthly periodicity then this specimen would have an estimated age of ~ 476 days, which falls within the calculated age range of 431 and 1437 days from Table 3.
Defocusing through the preparation reveals finer micro-increments that become unreadable toward the margin (Fig. 10). Upon magnification, the finest rings yet discovered in Architeuthis statoliths are visible just post the presumptive larval hatching check; here their widths are ~ 0.87 µm (Fig. 11), and they are similar to those described previously for the mid-dorsal position. Indeed, average increment widths derived by dividing mid-capella (i.e., nuclear) to dorsal edge distances by the number of counted rings shows a remarkable uniformity with an average of ~ 2.9 µm, compared with those in the published literature (see Table 3). Thus interpretation of total ring number is extremely difficult, especially when placed within the context of validated cephalopod ring widths. For example Nakamura & Sakurai (1991) report ring widths of ~ 3.05 µm from a near adult Todarodes pacificus, and Hurley et al. (1985) report validated width range of 2.8 to 1.7 µm for Illex illecebrosus statoliths.
To add to the difficulties, it is known from the examination of microincrements within teleost otoliths that feeding can affect increment width (Victor 1982) and that differences in growth rate can alter the total number of countable rings (Volk et al. 1995). Raya et al. (1992) report non-daily increments from the statoliths of the sepiod Sepia hieredda. The statolith of Architeuthis is small in comparison to overall mantle length (Lipinski 1997), adding to the complexity and importance of looking for very small rings to substantiate age estimates. We consider that current statolith-based age estimates are to be treated with suspicion as none to date describe consistently fine accretions of increments of the order of 0.87 µm that we have described.
By using the average observed increment widths calculated from the literature for Architeuthis, we derive a value of ~ 2.9 µm (see Table 3, excluding some anomalous counts from alternate axes by Jackson et al. (1991) and Lipinski (1991)). This is close to the mean increment width data observed in a broad range of cephalopod statoliths listed in Table 2. We also propose a lower bound to the mean increment width to be = 0.87 µm. Thus, upper and lower bounds to age, derived using a method similar to Ralston (1985), can be obtained by dividing the statolith nucleus-to-dorsal growth edge distance by both values, giving age estimates differing ultimately by a factor ~ 3 (2.9/0.87). Using this method maximum age is estimated to be 1747 days (~ 4.8 years) in a 1680 mm ML, which is within the maximum likely age speculated to be ~ 5 by Forsythe & van Heukelem (1987). The same specimen has a lower age bound (SFD/2.9 µm) of ~ 1.5 years.
By converting putative age ranges into mean-mantle growth rates, and summing across all specimen lengths, we can suggest two possibilities. First, Architeuthis dux is a fast-growing squid, with a growth rate upper bound of ~ 3.6 mm ML d-1. Second, A. dux is a slow-growing squid, with a growth rate lower bound of ~ 0.97 mm ML d-1. There is, however, no current agreement regarding which model best describes squid growth. Early growth of larvae is exponential (Forsythe & van Heukelem 1987, Bigelow & Landgraf 1993, Bower 1996), while post larval growth is thought to be logarithmic. Therefore, Architeuthis growth rate is likely to fall somewhere between the upper and lower bounds calculated above, and differ at different stages of its life cycle.
It proves difficult to reconcile known adult sizes and weights with historic (1880’s) records of 20 m long squids weighing one tonne — such animals would have to exceed an unprecedented 3 m ML. The estimate of 6 metres maximum mantle length proposed for this animal (Roper & Boss 1982), and the lesser maximum of 4 metres (Clarke 1966) should not be perpetuated — they are both likely to be overestimates of the true maximum mantle length this squid achieves. Of approximately 110 recently examined (over 8 years) specimens the largest we have seen was a fully mature female of ML 2.3 m. Remains of specimens collected in the 1880’s held at the Museum of New Zealand indicate such reported lengths are quite exaggerated, with remains consistent in size with the largest individuals now known, recently collected and examined during this study from local waters.
Table 3. Estimates of growth rate in Architeuthis based on 22 specimens based on statolith microsculpture, and the assumption that micro-increments are deposited daily (as validated in different species of laboratory-reared squids).
Estimating age and growth rate in Architeuthis dux
Dr. Steve O'Shea's paper is a guide for marine biologists studying giant squid