Sparidae: L III D6

Rhabdosargus holubi (Steindachner, 1881).

Silverbream or Cape stumpnose


Egg diameter in µm

Number of oil globules

Diameter of oil globule in µm

Yolk texture

Perivitelline space

Position of oil globule at hatch

Gut length   at eye- pigment stage


ca 850






40% of NL



Egg: The larva in the egg has more, but paler yellow pigment than LIIIC5, The specimen in Plate G was reared from a batch of eggs listed as LIIIC5, and some Pagellus (LIIIC5) were reared from the same batch. However, to add to the confusion, a specimen was reared with a batch of R. sarba eggs (LIIID9, and see below), which are quite different in appearance, to Pagellus eggs.

Larva: The NH larva is also unrecorded, but the comment above may apply. All the larvae depicted in Plates B-E are tentative, reared from a larva which, at 3 days, had a red blotch on the posterior edge of the anus (B). The juvenile shown in Plate F is a wild-collected fish, taken in a fine-meshed net during an evening spring flood tide into the Lovu estuary. Plate F1 shows the tri-cusped teeth by which these recruits can be identified.  B: 3, C: 15 days, D: 21 days, E: 24 days, F: circa 30 days, G: 70 days (23-25°C). Plate H shows another reared specimen, at 17 months, in which the tri-cusped teeth can still be seen, though rather worn compared with Plate F1.

Juveniles with tricusped teeth have been reared from five batches of eggs, from Park Rynie, 18 September 1993, 7 August 1994, 5 August 1995 (all from offshore samples), and from 2km offshore of Amanzimtoti on 16 July and 19 September 1997. In the first three cases, Pagellus was reared from the same batch, while in the last two instances, Rhabdosargus sarba was also reared. Undoubtedly others have been missed and catalogued with Pagellus (LIIIC5) or R. sarba (LIIID9, but see also comments under LIIIC5, LIIID5 and LIIID7). Five larvae hatched from eggs, have matched the barcode of R. holubi (BOLD), but all were submitted under the code LIIID9. The barcode of the only egg sequenced under the code LIIID6 has matched 5 locally collected adult Polysteganus coeruleopunctatus, locally known as blueskin, or trawl soldier. Clearly the egg is being confused with R. sarba.. The first three images above are of an egg collected on 5 June 2011, the 2day larva in ventral view, and the 3 day larva from that egg, which has barcoded as R. holubi.

R. holubi has a wider spawning season than Pagellus, Cheimerius and Chrysoblephus, possibly as a strategy to assist its larvae to recruit to estuaries that are not always open to the sea (Whitfield 1990). Note however that the coastal marine sparid Leiognathus mormyrus (LIIID1) also displays year-round spawning, but is quite rare in estuaries. Harrison (2003) recorded over 30 000 juvenile R. holubi in South African estuaries during his surveys, and only 2 L. mormyrus. And the estuarine dependent sparid Acanthopagrus vagus has a rather restricted spawning season (see sheet LIIIE3A).


During a 2005 study of spring flow-tide recruitment of juvenile fishes to the Lovu and Mkomazi estuaries, run concurrently with this study, recruitment of R. holubi was found to be continuous through the year, with a peak in April/May and another from August to December (both graphs). In all catches of recruits, it was evident that larger larvae, which might have missed the previous spring tide recruitment, were not collected on the following spring tide, two weeks later. This despite the capture in the net of large (ca 25-30mm SL) juvenile Trachinotus, which appeared to rule out net avoidance by bigger R. holubi juveniles. All the recruiting juvenile R. holubi measured 9.9-11.9mm SL (n=901), and were estimated to be about 30 days old. Growth rates of small juveniles of similarly sized sparids (Sarpa salpa, Crenidens crenidens, Diplodus sargus, Pagellus natalensis), reared under laboratory conditions during this study, all indicate a growth rate of at least 0.46mm/day, between day 30 and day 45. If these growth rates occur in the wild, then after an additional 14 days, the same hatchlings could be expected to measure at least 16.3mm. No R. holubi larvae of that size were collected, suggesting the previous spring tide recruits either die, or move out of the surf zone before the next spring tide.

Some of the recruiting larvae had distended stomachs, from feeding on the small coastal marine cladoceran Evadne, just prior to entering the estuary.