712 



Effect of analytical conditions in wavelength 

 dispersive electron microprobe analysis on the 

 measurement of strontium-to-calcium (Sr/Ca) ratios 

 in otoliths of anadromous salmonids 



Christian E. Zimmerman 



Oregon State University 



Department of Fisheries and Wildlife 



Corvallis, Oregon 97331 



Present address: US. Geological Survey 

 Alaska Science Center 

 101 1 E Tudor Road 

 Anchorage, Alaska 99503 



E-mail address: czimmermanfSusgs gov 



Roger L. Nielsen 



College of Oceanic and Atmospheric Sciences 

 104 Ocean Administration 

 Oregon State University 

 Corvallis, Oregon 97331 



The use of strontium-to-calcium (Sr/Ca) 

 ratios in otoliths is becoming a standard 

 method to describe life history type and 

 the chronology of migrations between 

 freshwater and seawater habitats in 

 teleosts (e.g. Kalish, 1990; Radtke et al., 

 1990; Secor, 1992; Rieman et al., 1994; 

 Radtke, 1995; Limburg, 1995; Tzeng et 

 al. 1997; Volk et al., 2000; Zimmerman, 

 2000; Zimmerman and Reeves, 2000, 

 2002). This method provides critical 

 information concerning the relation- 

 ship and ecology of species exhibiting 

 phenotypic variation in migratory 

 behavior (Kalish, 1990; Secor, 1999). 

 Methods and procedures, however, vary 

 among laboratories because a standard 

 method or protocol for measurement of 

 Sr in otoliths does not exist. In this 

 note, we examine the variations in 

 analytical conditions in an effort to 

 increase precision of Sr/Ca measure- 

 ments. From these findings we argue 

 that precision can be maximized with 

 higher beam current (although there 

 is specimen damage) than previously 

 recommended by Gunn et al. ( 1992). 



Wavelength dispersive electron mi- 

 croprobe analysis (WD-EM) has been 

 used by most researchers, although 

 other methods such as ()rot()n-induced 

 x-ray emission (PlXExBabaluk et al.. 



1997; Markowitz et al, 2000 ) have been 

 used. WD-EM remains a common and 

 relatively inexpensive method. The 

 conceptual approach among research- 

 ers using WD-EM is similar but the 

 methodological approach or analytical 

 (operating) conditions vary. In a com- 

 parison of laboratories using common 

 otoliths, Campana et al. (1997) found 

 among-laboratory variation in mean 

 Sr concentrations that could not be 

 described by otolith variability. Al- 

 though the laboratories were internally 

 consistent in applying their methods, 

 comparisons between laboratories dif- 

 fered. Campana et al. suggested that 

 the sensitivity of WD-EM to operating 

 conditions might have led to this varia- 

 tion between laboratories. 



Development of analytical techniques 

 for measuring Sr/Ca ratios has been re- 

 viewed to validate techniques in specific 

 studies (Kalish, 1990; Secor, 1992; Toole 

 and Nielsen, 1992; Limburg, 1995). 

 Gunn et al. (1992) analyzed effects of 

 counting times, beam current, accel- 

 erating voltage, and beam diameter 

 on measures of Sr and other elements 

 and they warned that beam powers 

 required for WD-EM were sufficient 

 to cause specimen damage including 

 pitting and chemical change. As a re- 



sult, Gunn et al. (1992) recommended 

 limiting beam power densities to 

 < 3pW/pm''. This recommendation has 

 been followed in most studies using Sr/ 

 Ca ratios to reconstruct the chronology 

 of migrations between the freshwater 

 and marine environments (Table 1). 

 Toole and Nielsen ( 1992), however, con- 

 cluded that Sr/Ca precision could be in- 

 creased, with no loss of accuracy, by using 

 analytic conditions that lead to a beam 

 power density of just over 15pW/pm- 

 (5-pm beam diameter; accelerating 

 voltage=15 loA; beam current=25 kV). 

 The inherent beam damage was not 

 critical because of the similar behavior 

 of Sr and Ca during progressive beam 

 damage. 



In published studies using WD-EM 

 to measure Sr/Ca ratios in otoliths, the 

 operating conditions, including beam 

 power densities, have varied greatly 

 (Table 1). Establishing a microprobe 

 protocol for measurement of Sr/Ca 

 ratios in otoliths involves a balancing 

 act of counting times, beam current, 

 and beam diameter. The selection of 

 optimum conditions is constrained 

 by financial resources, allocation of 

 time for use of instruments, and the 

 required resolution of Sr/Ca ratios 

 for any specific application. Each re- 

 searcher must weigh the benefits and 

 costs to best answer the question at 

 hand. Generally, these parameters are 

 manipulated to optimize precision and 

 accuracy of analyses in relation to vari- 

 ability within the otolith and implica- 

 tions of the results. 



For Sr/Ca ratios to remain an ac- 

 cepted and accurate means of describ- 

 ing migration histories and other life 

 history events, continued analytic and 

 technical refinement and validation 

 are required. We examined the effects 

 of crystal choice, beam diameter, beam 

 current, and beam power densities on 

 Sr/Ca measurements (expressed as 

 atomic ratios) in salmonid otoliths: 1) 

 we measured Sr using both the TAP 

 and PET crystals in regions with high 

 Sr/Ca 00.003) and low Sr/Ca XO.OOl) 



Manuscript approved for publication 

 12 February 2003 by Scientific Editor 



Manu.script received 4 April 2003 at NMFS 

 Scientific Publications Office. 



Fish Bull. 101:712-718 (2003). 



