A stochastic model to predict annual egg production of a flock of laying hens.
Ovulation rate in laying hens is determined by the interaction of two biological systems; namely, a circadian rhythm that restricts the release of luteinising hormone to an eight- to ten-hour period of the day, and the process of follicle maturation. Etches and Schoch (l984) used a two-compartmental model to represent the circadian rhythm and a Gompertz equation for follicle maturation. In doing so, they were able to predict ovulation times for two- to nine-egg sequences. This model has been improved by replacing their table of values with continuous functions that predict the values for each parameter in the ovulatory model, for any ovulation rate. Consequently, ovulation times may be predicted for any sequence length. A population model that simulates annual egg production has been developed in Visual Basic. Each parameter in the model is allocated a mean and standard deviation, so that variation is introduced into the flock. Mean age at first egg is predicted from the age at photostimulation and the lengths of the photoperiods applied during rearing. Quadratic-by-linear functions are used to predict changes in the hen's internal cycle length over time, which in turn determine changes in the ovulation rate and rate of lay. Short egg sequences, frequently observed at onset of lay in experimental flocks, are simulated initially, followed by the prime (or longest) sequences, which are produced at the time of peak rate of lay, before gradual increases in the internal cycle length cause the egg sequences to become shorter once more. In view of the fact that the interval between oviposition and the subsequent ovulation is about 30 minutes, time of lay may be predicted from ovulation time for all eggs other than the last egg of a sequence, because in this case there is no associated ovulation. A curvilinear function is used to predict the value of the last interval from the ovulation rate, because experimental data show that short sequences have longer intervals between the last two eggs than long sequences. The circadian rhythm of LH release is linked to the onset of darkness, so that mean time of lay occurs 13 to 14 hours after sunset. The distribution of oviposition times is unimodal for young flocks and bimodal for older flocks. Yolk weight is predicted from hen age using a function appropriate for the genotype. Allometric functions are used to predict albumen weight from yolk weight and shell weight from the weight of the egg contents. Egg.weight is given by the sum of the three components. With advancing hen age, the proportion of yolk in the egg increases at the expense of both albumen and shell. Random events, such as internal ovulations, and the production of soft-shelled and double-yolked eggs, are accounted for in the model. Their incidence is linked to the genotype and to the age of the hens and their occurrence is restricted to a proportion of the flock. Internal ovulations cause interruptions to egg sequences, thereby reducing overall mean sequence length. This model could be of benefit to a producer wanting to know how a change to the lighting programme would affect the laying performance of the strain, or to a nutritionist desiring to determine changes in voluntary feed intake and to the nutrient requirements of the birds over the laying period. It may also be used as a teaching aid, so that students gain a thorough understanding of the process of egg production and are able to test the response of layers to different environmental stimuli. The user has control over a number of inputs, thereby making it a generalised model that can be used for different strains. With a few modifications, the model may be used to simulate the erratic and variable laying behaviour of broiler breeders.