68 Allan E. Houston and Michael R. Pelton 



METHODS 



This study was conducted in the upper headwater basin of the 

 North Fork of the Wolf River in west Tennessee, largely on the Ames 

 Plantation. From November 1984 through May 1985, beaver jaws 

 from 169 beavers were collected from legal trapping efforts on 

 approximately 1,619 ha of the watershed. 



Molars and premolars from lower mandibles were extracted and 

 cleaned. If tooth basal openings indicated a specimen to be >4 years 

 old, age was determined by grinding the lingual surface of premolars 

 or molars with a 120-grit stone to expose longitudinal cross-sections. 

 Ground surfaces were polished with a 400-grit emery cloth and cementum 

 layers were carefully counted using hand-held magnification (Larson 

 and van Nostrand 1968). Twenty-eight beavers, >4 years old as aged 

 in this manner (Table 1), provided the samples (i.e., dependent variables) 

 for our study. 



Premolar cementum length and noncementum length were measured 

 to the nearest millimeter along the estimated maximum longitudinal 

 cross-sectional centerline of one premolar per specimen (Fig. 1). An 

 exponential model predicting age in years was developed by combining 

 a growth curve for cementum length and a decay curve for noncementum, 

 using nonlinear regression (PROC NLIN) and the multivariate secant 

 method (DUD) to set initial parameters (SAS Institute, Inc. 1985). 

 Intercept was specified at zero to prevent predicted ages from 

 dipping below zero. An "r 2 like" statistic was calculated by taking 

 [1 - (residual SS/corrected total SS)]. 



RESULTS AND DISCUSSION 



The following model was developed using cementum length and 

 noncementum length of premolars as independent variables: 



AGE = + 6 1784 * e (-o.io37*noncemen) + 2.6513 * e (° ul9 * CEMENTUM ) t 



The "r 2 like" statistic was 0.93. The predicted curves are combined 

 to produce a response surface (Fig. 2). 



Our data were obtained from specimens on the headwaters of a 

 single watershed. Although beavers can travel considerable distances, 

 we assumed our sample represented only a small region. The study 

 area may not fully represent variation existing range-wide or within 

 adjacent watersheds. It is possible that genetic differences and dietary 

 regimes will yield different tooth size, wear, and cementum accretion 



