Assessments of body condition in dipperscinclus cinclus:potential pitfalls in the derivation and use of condition indices based on body proportions

Factors influencing the body condition of Welsh Dippers were assessed from 1153 measurements of mass and wing length, and 789 measurements of tarsal length, made on over 1000 individuals. Condition indices were derived by comparing actual masses with those predicted from regressions of mass on wing length and tarsal length after log : log transformation. The indices gave an indication of the mass of a bird for a given body size. Different regression models (least squares and Model II) gave condition values which were almost identical. Age (within sexes), sex (within ages), time of year, time of day and habitat quality all had significant effects on condition indices derived from measurements of birds of all ages and sexes combined. Effects due to sex probably resulted from systematic differences in body proportion between males and females, though condition indices within age and sex groups could not be derived accurately because of other sources of variability. Apparent effects due to age differed markedly depending on whether condition indices were calculated from tarsal length or wing length. Using wing length, first years had higher condition scores than older birds, with the reverse true using tarsus. This pattern reflected significantly shorter wings in first year birds. Effects due to moult and feather growth also affected condition indices based on wing length. Measurements like these, which vary during different stages of a bird's life, have limited value in condition indices by comparison with more constant features such as tarsal length. Effects on condition due to time of year followed known events in the annual cycle such as breeding, moult and moderate winter fattening. Circadian changes were probably related to overnight energy use and daytime gain, with peaks in condition indices occurring within a few hours after dawn, and prior to dusk. Birds of both sexes had significantly lower condition indices on acidic streams than on non‐acidic streams. Although many of these effects are expected, they highlight the need for condition indices which account for all known sources of variability before further studies are carried out on condition values based on body proportions. Assessments of fat scores or muscle condition could provide alternative methods. It might also be difficult to assess whether high or low condition scores are adaptive under the changing circumstances of a bird's life.