Chick growth and food provisioning of northern fulmars Fulmarus glacialis were studied at Fair Isle, Shetland, in 2 years of contrasting environmental conditions. Low sandeel Ammodytes marinus availability resulted in very low breeding success in many seabirds in Shetland in 1998, whereas breeding success in 1997 indicated a normal season. Fulmars at Fair Isle experienced lower chick survival in 1998 than in 1997. There were also significant differences in diet and food provisioning in the 2 years; sandeels were the most abundant prey items in 1997, but were completely absent from the diet in 1998. High nestling mortality occurred shortly after hatching in 1998. Those chicks that survived this episode had previously been heavier for their age, and remained heavy throughout development in comparison to chicks in 1997, despite a continued absence of sandeels from the diet. Surviving chicks in 1998 were subsequently fed comparatively small meals at frequent intervals, but with no overall difference in the amount of food delivered. We discuss the implications of differential mortality and resulting growth patterns for the use of seabird breeding parameters to monitor marine environments.

This is a review of previously published data from experiments designed to measure the chicks' requirement for an essential amino acid in diets containing surplus protein (generally in the range 220–300 g crude protein (CP)/kg diet). The evidence shows that, within this range, the requirement for a first limiting amino acid increases nearly in direct proportion to the CP content of the diet. To explain this, the following possibilities are considered: (1) that energy supply was limiting the response to a critical amino acid, (2) that the experiments were conducted in environments which limited heat disposal by the chicks, (3) that results from the trials have been misinterpreted because growth rate rather than protein deposition was used as a response measure, (4) that the availability of amino acids in the diets was lower than had been assumed, (5) that nutrients other than the amino acid under study were limiting performance, and (6) that there was an imbalance of amino acids in the diet. It is concluded that only the last explanation fits the facts. The imbalances reported have led to a decline in the efficiency of the utilisation of the first limiting amino acid in some instances. The depressed growth observed with a high-protein diet limiting in one essential amino acid could not be attributed to lowered feed intake in most cases. The imbalance is, in these respects, different from classical imbalances which result from loading the diet with a mixture of essential amino acids. From the evidence reviewed, some doubt is cast on the validity of the dilution method for estimating amino acid requirements; nevertheless it its concluded that the dilution method remains the most trustworthy procedure for making estimates to be used in practical diet formulation. It is recommended that, in future, programmes for diet formulation should be modified to prevent surplus protein being permitted in the solution, unless a proportional adjustment is made to the minima prescribed for amino acids likely to be present in limiting proportions.