Aging yellow bullhead using pectoral-spine sections was relatively precise, as long as sections distal to the basal recess were used for age assignment to prevent under-aging of individuals. Levels of aging precision obtained for yellow bullheads while using pectoral spines (CV = 8.6%, APE = 6.1%) was comparable to that obtained in a majority of studies that aged fish using otoliths, where the median value for the CV was 7.6% and for D (equivalent to average percent error when otoliths are aged twice) was 5.5% (Campana 2001). In addition, our study provided the first validation of pectoral-spine aging in yellow bullhead, demonstrating that one complete annulus (one opaque zone and one translucent zone) was deposited each year. Although not conclusive for all yellow bullheads injected with OTC because of the small sample size, the validation study was successful in demonstrating that a 7-year old yellow bullhead deposited one complete annulus in a 12-month period (Fig. 7). It would be beneficial to conduct further validation studies that encompass the range of ages observed in yellow bullhead from south Florida.

Precision in aging yellow bullheads using sections from otoliths was unacceptably low (CV = 23.7%, APE = 16.8%) relative to using pectoral spines. Otoliths in ictalurids can be problematic for aging because the sagittae are not the largest otolith pair as they are in the majority of other fishes. In yellow bullheads, as in channel catfish (Barbour and Kollmar 2003) and armoured catfish Hoplosternum littorale (Ponton et al. 2001), the lapilli are larger than the asterici and wider than the sagittae, which are thin and fragile compared to the lapilli (pers. obs.). Most aging of teleosts using otoliths, however, has developed around using the sagittal otolith for aging because of its usually larger size and relatively clearly demarcated growth zones. In yellow bullheads, difficulties in enumerating annuli were exacerbated by the morphology of the lapilli. Unlike the sagittae and the asterisci, the lapilli are orientated horizontally in the utriculus, are dorso-ventrally flattened (instead of laterally flattened as in the sagittae and asterisci), and rest with the ventral surface in contact with the macular bed (Assis 2005). The lapilli therefore have a distinguishing, relatively large, macular hump (gibbus maculae) on their ventral surface (Assis 2005), and it is this area of the lapillus that was unreadable in cross-section. The macular hump created a discontinuity in the ventral part of the otolith section, making it impossible to trace annuli around the entire section. This undoubtedly led to problems in recognizing annuli versus false annuli or "checks" in the cross-sections. The South American pimelodid catfish Hypophthalmus edentatus have been successfully aged using asterisci, with CVs ranging from 3.5-18.4% depending on the age class (Ambrósio et al. 2003). Crumpton et al. (1984) were unsuccessful using otoliths to age brown bullhead, channel catfish, and white catfish. However, more recently, Nash and Irwin (1999) successfully aged flathead catfish using sagittal otoliths with 85% between-reader agreement. Testing a variety of modifications in the aging method, Buckmeier et al. (2002) aged 1-4 year old channel catfish using sectioned sagittae also with 85% between-reader agreement. Their method could be considered to have been validated for 2-4 year old catfish since the fish used in their study were initially obtained from an aquaculture facility, grown-out for a period of years, and were therefore of known age. Although yellow bullhead lapilli used in our study were small (2-4 mm length) but similar in shape to the lapillar (sic sagittal) otolith illustrated in Buckmeier et al. (2002), the majority of cross-sections produced for aging yellow bullhead using the same methods were not adequate. Without further development of otolith aging in yellow bullhead, pectoral spines would continue to be the aging structure of preference. An added advantage to using a pectoral spine for aging is that, unlike the use of otoliths, the fish would not have to be sacrificed to obtain a spine if it was desired to return the fish alive to the water.

Yellow bullheads in south Florida grew relatively slower and were smaller at age compared to other populations throughout the U.S. (Fig. 12). Growth in yellow bullheads was similar in Florida, Oklahoma (Jenkins et al. 1952; Finnell et al. 1956), and Washington (Wydoski and Whitney 1979) for their first 3 years, after which fish from Oklahoma and Washington achieved much larger size with increasing age. Although based on very limited data (n = 5 fish aged), only yellow bullheads specifically from Rowland Lake, Washington, were observed to have similarly slow growth in ages 4-6 (Jackson and Caromile 2000) as we observed in southern Florida. Yellow bullheads have been previously documented as having variable growth, with fish from a tributary of the Mississippi River also being much smaller at ages 2-3 (<200 mm SL, or ~ 244 mm TL) compared to yellow bullhead from Reelfoot Lake in Tennessee (Ross 2001). The growth rate reported for yellow bullhead populating Reelfoot Lake was much greater than that of yellow bullheads studied in other regions, including south Florida. For example, Reelfoot Lake bullheads that were 2-5 yr old were more than double the size of yellow bullhead of the same age from other populations (Fig. 12), and they also attained unusually large sizes, reaching a maximum of 470 mm TL (corresponding to ~1.0-1.4 kg) (Carlander 1969). Yellow bullhead in regions other than Tennessee, however, have also been documented to attain relatively larger sizes compared to south Florida. According to the International Game Fish Association, the world all-tackle angling record for yellow bullhead is 1.92 kg (4.25 lb) for a fish caught in Mormon Lake, Arizona, in 1984, although Sternberg (1987) reported an Illinois state record catch of 2.4 kg (5 lb 4 oz). In addition to their slow growth and small size, yellow bullheads in south Florida also attained a much older maximum age (12 years) (Fig. 12) than the previously reported 7 years (Scott and Crossman 1973). Given that only a moderate number of yellow bullhead from south Florida were aged in our study (n =144), it is probable that even older fish might be detected with increased sampling in south Florida.

Figure 12.- Mean total length at observed age for yellow bullhead collected from south Florida (This Study), Illinois River and Tenkiller Reservoir, Oklahoma (Jenkins et al. 1952), Reelfoot Lake, Tennessee (Schoffman 1955), Little River, Oklahoma (Finnell et al. 1956), Washington (Wydoski and Whitney 1979) and Rowland Lake, Washington (Jackson and Carolmile 2000). Vertical bars denote the range in size of yellow bullhead at age for south Florida, Illinois River, and Reelfoot Lake and are offset from one another for clarity. Solid line represents estimated von Bertalanffy growth curve for yellow bullheads from south Florida. [larger image]

Assuming yellow bullheads from Tennessee do not differ genetically from yellow bullheads in south Florida, then we would expect that fish from south Florida should have a similar potential for growth as fish from Tennessee. The fact that they are smaller indicates that they are limited in their growth by biotic and abiotic environmental influences. This variability in growth might be partially explained by density-dependent processes, such as relative abundance and availability of prey on a per-capita basis. Ross (2001) has observed that yellow bullhead exhibit stunted growth when living in overpopulated ponds, which would support a food limitation hypothesis. Low densities and small maximum sizes are common among aquatic animals in southern Florida, and are thought to be related to the oligotrophic nature of aquatic systems there (Turner et al. 1999; Rice et al. 2005). Limitations in available prey resources could also result from intra- and interspecific competition or interference with other predators, such as wading birds in the area. Loftus (2000) found that feeding by many fishes in the Everglades is reduced during the spring dry season, when bullheads and other fishes are concentrated in dry-season refuges like alligator holes. Densities of large fishes there are very high, prey are depleted, and water quality poor (Loftus and Kushlan 1987; Nelson and Loftus 1996). Growth of fishes would be slow under those conditions.

In general, any factors that influence the density of yellow bullhead may contribute to density-dependent growth, including habitat quality and quantity. Yellow bullhead have been observed to be most common in shallow, relatively clear-water portions of lakes, ponds and streams with heavy vegetation (Scott and Crossman 1973; Trautman 1981; Laerm and Freeman 1986; Loftus and Kushlan 1987). They also occur in relatively slow-moving waters rather than in faster-flowing rivers (Scott and Crossman 1973; Laerm and Freeman 1986), and Jackson (1996) observed that yellow bullhead were, while common in oxbow lakes, rare in the main channel of the Yockanookany River, Mississippi. Similarly, in south Florida, few yellow bullhead were collected in the deeper, box-cut canals with patchy vegetation compared to the densely vegetated, shallow, marshes.

While the tolerance of yellow bullhead to general habitat perturbation and pollution may be debatable, it is apparent that they have a negative response to loss of vegetation. Trautman (1957) observed a decrease in the abundance of yellow bullhead with aquatic vegetation loss and increased turbidity in an Ohio lake, as well as a general inverse relationship between the abundance of yellow, as well as black Ameiurus melas and brown bullheads, with habitat degradation (Trautman 1981). Scott and Crossman (1973) have also documented that removal of suitable habitat, such as vegetation and logs, leads to a decrease in the number of yellow bullheads, although less so than black or brown bullheads. Koonce et al. (1996) ascribed the change in yellow bullhead occurrence in Lake Erie from common in the pre-1800 period to rare in this decade as being due to habitat degradation. Simon (1991), however, considered yellow bullhead to be highly tolerant to declines in stream quality. In south Florida, this species is one of the most tolerant fishes able to survive crowding and poor water quality in isolated ponds in the dry season, when water temperatures are high, ammonia levels are high, and dissolved oxygen levels are very low (Kushlan 1974; Loftus and Kushlan 1987).