Cappo et a\: A new approach to validation of periodicity and timing of opaque zone formation 



477 



Figure 2 



Photomicrographs of transverse sections of the sagittae of (AlL. erythropterun si 799 (21+yr) by trans- 

 mitted white Ught showing the ventral LI and sulcal L3 axes used for counting opaque zones; (B) the 

 outer LI axis of L nvuhtus (s 1996 1 showing the position of an OTC mark iTi by reflected ultraviolet 

 light; (C) the same outer LI axis of L rivulatiis (sl996) showing narrow opaque zones by transmitted 

 white light. 



zones along these axes we recognized sharply defined 

 "structural check rings" (sensu Gauldie, 1988) that could 

 be followed from the ventral leading edge of the otolith 

 around to the sulcus. These check rings were evident on the 

 outermost edges of the opaque zones and were counted and 

 measured after adjustment for aspect ratios. The position 

 of the OTC mark was measured from the primordium to 

 the outside edge of the mark and from there to the edge of 

 the otolith along the LI and L3 axes. The distance between 

 twin OTC marks was measured from the distal edge of the 

 first mark to the proximal edge of the second. 



Data analyses 



Analyses of the results for periodicity and timing of 

 increment formation each comprised both an informal 

 approach analogous to those used in earlier validation 

 studies and a model of otolith growth. The informal 

 validation of periodicity assumed a calendar date of 1 

 October for completion of opaque zones, based on previous 



Australian studies (Fowler, 1995), to report frequency 

 distributions of the observed number of zones minus the 

 expected number of zones formed outside OTC marks. 

 The informal, "method of best fit" approach to estimate 

 timing of completion of opaque zones was to assume that 

 formation was completed annually on a particular day 

 of the year. For each otolith section, and for every single 

 day of the year, the number of opaque zones expected to 

 be completed during time at liberty was calculated from 

 the dates elapsed between marking and sacrifice. Only 

 those dates for which the expected number of opaque zones 

 formed past the OTC mark equalled the number actually 

 observed were retained for analysis. The mean dates (MDi 

 from frequency distributions of these possible calendar 

 dates were selected as best informal estimates of time of 

 completion of opaque zones. 



We defined an "increment cycle" as the complete formation 

 of adjacent translucent and opaque zones visible on an 

 otolith section (Fig. 3). Two models were developed in a 

 "direct method" to estimate the periodicity and timing 



