NOTARBAKTOI.O-DI SCIARA: NATl'HAI. HISToHY OF MOBVLA 



winter diet of M.Japanica was lacking, as four of 

 the five specimens collected then had empty stom- 

 aches; the fifth, a large female, contained a small 

 fragment of a partially digested fish carcass. 

 Since quantifiable stomach contents were found 

 only in large rays between April and July, no size 

 or seasonal differences in the diet of M.japanica 

 could be detected. 



Mohiila Japanica was often found carrying 

 Remora remora , usually seen clinging to the out- 

 side of body, but found once inside a spiracle. Six 

 specimens of R. remora (range 109-217 mm SL) 

 were collected from M. japanica . Only one speci- 

 men of/?, albescens (97 mm SL) was found, in the 

 mouth cavity of a M.japanica. A pilot fish, Nau- 

 crates ductor (Carangidae) also associated with 

 M. japanica, swam alongside a harpooned ray 

 that was being towed inshore and remained for 

 some time at the water's edge, where the ray was 

 beached. Mobula japanica was parasitized by the 

 following crustaceans: Nerocila acuminata 

 (Isopoda: Cymothoidae), Pupulina brevicauda 

 and P . minor (Copepoda: Caligidae) on the skin; 

 Eudactylina oliveri (Copepoda: Eudactylinae) in 

 the gills; and Kroeyerina sp. (Copepoda: Kroyeri- 

 idae) among the olfactory lamellae. Unidentified 

 trypanorhynch cestodes were occasionally found 

 within the pleuroperitoneal cavity. 



Habitat preference of M . japanica did not ap- 

 pear to differ from that of M. thurstoni . However, 

 the use of the habitat differed seasonally: in April 

 and May, when M. thurstoni was abundant at the 

 surface, M. japanica was never seen, and few 

 specimens were bottom gillnetted during those 

 months. Conversely, M.japanica, in June and 

 July, was seen in the late morning hours at the 

 surface in groups of several individuals swim- 

 ming parallel to the shore. Occasionally speci- 

 mens were seen in water <1 m deep. Mobula 

 japanica is not known to school, and I never ob- 

 served schooling. 



Coles (1910) reported that M . olfersi (= M. hy- 

 postoma ) utters a "musical, bell-like bark" when 

 dying. A similar account was given by Risso 

 (1810) of Cephalopterus massena {= M . mobular). 

 This information led subsequent authors (Nor- 

 man and Fraser 1937; Bigelow and Schroeder 

 1953) to wonder whether mobulid rays are capa- 

 ble of producing sounds while in the water. Sound 

 production is a fairly widespread phenomenon 

 among bony fishes (Fish and Mowbray 1970; 

 Tavolga 1971); however, elasmobranchs lack the 

 traditional structures used by teleosts to generate 

 sound, i.e., the swim bladder and bony skeletal 



parts (Marshall 1962), and recognizable sounds 

 have not been recorded from these animals 

 (Backus 1963). Sound production among elasmo- 

 branchs has been reported only for the Atlantic 

 cownose ray, Rhinoptera bonasus (Myrberg 

 1981); in that case clicks and scraping sounds 

 were presumably produced with the dental plates, 

 elicited when strongly prodding three rays which 

 were confined in a tank (Fish and Mowbray 

 1970). Mobula japanica, when beached alive, 

 often emitted a distinctive noise which could have 

 been the equivalent of Coles' "bark". This noise, 

 however, was apparently caused by the periodic, 

 spasmodic contractions of the mandibular, pha- 

 ryngeal, and hypobranchial musculature of the 

 asphyxiating ray, which forced air from the 

 mouth cavity out of the gill openings through the 

 meshlike branchial filter plates. Although under- 

 water sonic recordings have never been made, it 

 seems unlikely that under normal circumstances 

 any audible sound could be produced in this fash- 

 ion by submerged mobulids. 



This area served as a spring and summer feed- 

 ing and mating ground for adult M. japanica, 

 rather than as a pupping or nursery ground, as 

 indicated by the lack of small-sized specimens. 

 Seasonal abundance of M .japanica in the surface 

 waters was indicated by the catch data (Table 5) 

 and is comparable to the seasonal abundance of 

 M . thurstoni (Fig. 6). No M . japanica were ob- 

 served in March; in April and May they occurred 

 occasionally. By mid-June large numbers ap- 

 peared in the nearshore surface waters near 

 Punta Arena de la Ventana, and were easily har- 

 pooned. Most of the July mobulid catch consisted 

 of M .japanica, when the numbers of M. thurstoni 

 had declined. Data are lacking for the August- 

 October period, therefore it was impossible to tell 

 whether the peak of abundance occurred in July 

 or later. Fishermen's reports were not clear, al- 

 though there was agreement on an overall decline 

 of mobulid abundance in late summer. Mobula 

 japanica fed exclusively on the euphausiid A^yc- 

 tiphanes simplex, and its numbers apparently de- 

 clined concomitant with the late summer decline 



Table 5. — Mean number of daily captures of Mob- 

 ula japanica (symbols as in Table 1). 



59 



