We examined the relationship between the abundance of alewives Alosa pseudoharengus and the recruitment of yellow perch Perca flavescens to determine if alewives were potentially responsible for yellow perch recruitment failures in southern Lake Michigan after 1988. We used annual June-August bottom trawl data from two 5-m-depth index sites near Michigan City, Indiana, during 1984Indiana, during -1998 to index alewife abundance as the catch per unit effort (CPUE) of age-1 and older fish and yellow perch recruitment as the CPUE of age-2 fish. The relationship between alewife abundance and yellow perch recruitment was modeled as log e R tϩ2 ϭ 11.7 Ϫ (2.12) log e A t , where log e R tϩ2 is the natural log of the CPUE of age-2 yellow perch in year t ϩ 2 and log e A t is the natural log of the CPUE of alewives age 1 and older in year t. The model explained greater than 70% of the variability in recruitment of the 1984-1996 yellow perch year-classes. We conclude that yellow perch year-class failures in Indiana waters of Lake Michigan were largely explained by an increase of mean alewife abundance after 1988 and that the yellow perch fishery is unlikely to fully recover unless mean alewife abundance returns to an extremely low level, similar to the early 1980s.

This study was undertaken to quantify the length-fecundity and length-egg size relationships for yellow perch Perca flavescens in the Indiana waters of Lake Michigan. Data were pooled from gill-net collections made in 1985, 1986, and 1999, resulting in a wide length range of mature female yellow perch (172-332 mm total length [TL]). The length-fecundity relationship was log 10 F ϭ Ϫ3.220 ϩ 3.223·log 10 TL (r 2 ϭ 0.89), where F is fecundity. The mean preserved egg volume (V; mL) increased with yellow perch TL and was represented by the following equation: log 10 V ϭ Ϫ2.06 ϩ 1.10·log 10 TL (r 2 ϭ 0.48). These results reveal that larger females produced both more and larger eggs than did smaller females. Therefore, the intense harvest targeting large yellow perch (primarily females) in the 1980s and 1990s may have had an effect on the quantity and quality of eggs spawned by the population, possibly resulting in reduced recruitment.

In freshwater, the optimal choice between radiotelemetry or acoustic telemetry has often been unclear because the combined effects of tag power, conductivity, tag depth, and antenna type on radio tag detection distances have not been quantified. To enable more informed decisions regarding the best telemetry methods to use at particular study sites, we measured maximum detection distances of an acoustic tag and two different 48-49-MHz radio tags over a wide range of conductivities, depths, and acoustic conditions in Minnesota lakes and rivers. Radio tag detection distances increased with increasing tag power, decreased with increasing conductivity, and generally decreased with increasing depth. Detection distances measured with a Yagi antenna were typically about double those measured with a loop antenna. Maximum detection distances of the radio tags were predicted by the equation log e R ¼ 13.69 À 0.005771 C À 0.006575 D þ 0.1044 P þ 0.7275 A À 0.001302 C D À 0.00008208 C P (where R ¼ maximum detection distance, m; C ¼ surface conductivity, lS/cm, at ambient temperature; D ¼ tag depth, m; P ¼ tag output power, dB relative to 1 mW; and A ¼ antenna type [1 ¼ Yagi, 0 ¼ loop]). Acoustic detection distances were less variable overall than radio detection distances but still differed substantially between and within water bodies. We saw no consistent relationship between acoustic detection distance and tag depth, but detection distance was negatively affected by ambient noise. Our model indicates that a 48-49-MHz radio tag can be detected with a boat-mounted Yagi antenna at a depth of 50 m or more if conductivity is low enough, but the maximum detection depth decreases drastically as conductivity increases. The equation above can be used to predict whether detection distances of radio tags similar to the ones we tested will be adequate for a particular study.

Because the yellow perch Perca flavescens is a valuable sport and commercial species in southern Lake Michigan, forecasts for abundance of quality‐size fish would be beneficial. We analyzed a time series of annual index trawl catch per unit effort (CPUE) in order to identify a model that could be used in forecasting. Relative abundance of stock‐size (≥130 mm) and quality‐size (≥200 mm) fish in Indiana waters of Lake Michigan varied by about two orders of magnitude from 1975 to 1996 and declined to a low level during the 1990s. Cross‐correlation was used to identify a strong positive relation between CPUE of stock‐size fish (S) in year t and quality‐size fish (Q) in year t + 2. This relation was described by the linear model, √Qt+2 = 2.68 + 0.00572·St, and was due to survival and growth of sub‐quality–stock‐size fish from t to t + 2. The CPUE of quality fish predicted by the model closely approximated the trend in observed values. The model predicted that relative abundance of quality‐size yellow perch in Indiana waters of Lake Michigan would remain extremely low in 1997 and 1998. This method of forecasting may be applicable to other populations for which adequate data are available.

The temporal signature technique can be used to assign age to fish that show an incomplete or indistinct growth history at the margins of their scales. The temporal signature technique matches part of an individual's "environmental" growth history to characteristic patterns found in a master chronology that was developed from reliably aged specimens of a species in a particular environment. An error sum of squares measures the concordance between an individual's environmental growth history and the master chronology. Ages assigned to walleye (Stizostedion vitreum) by the temporal signature technique and by examination of scales were compared to assess the performance of the new technique. Scale-age agreed with one of the three most likely signature-ages in 67–77% of the comparisons using ail observed increments. These results are purposely conservative because of the methods employed and the nature of the example. All observed growth increments should be used in applying the temporal signature technique, but age may still be accurately assigned if as few as three increments are available. The temporal signature technique will perform best for species that exhibit high interannual variation in growth.