Sleep length and sleep quality scores were collected on board ships over periods of up to two weeks from 38 watchkeepers working a '4-on/8-off routine' and 29 dayworkers. All watchkeepers exhibited fragmented sleeping patterns, which indicated a lack of adaptation of the sleep/wakefulness cycle to the hours of work. There were only slight differences in total sleep length between watchkeepers and dayworkers, however, both groups did not obtain an adequate amount of sleep. Within the watchkeeping crews the 3rd Officers had by far the shortest sleep length. Concerning sleep quality, daytime sleep was generally given the lowest ratings, whereas sleep starting before midnight was on average evaluated as the best, both by watchkeepers and dayworkers. Watchkeeping personnel do not normally have any "days off" during a voyage so that missed sleep might even amount to a sleep deficit. A solution for this problem could perhaps be a new, stabilized system that allows a single uninterrupted sleep, which is required for full recuperation, to be taken each day.

The safety of a ship depends substantially on its bridge watchkeepers, whose alertness and efficiency must be maintained at all hours of the day and night. Fatigue, circadian rhythms, and sleep disruption occasioned by the unusual working hours of these personnel may all affect their performance. A methodology for assessing the magnitude of this problem is proposed. The application of this methodology in a large-scale shipboard study of merchant mariners on extended voyages is then described, and details given of the techniques used to measure sleep and activity, and temporal variations in a range of physiological und psychological parameters. A summary of the data collected in the study is provided as a reference point for the reports on the different aspects of the results that follow in subsequent articles.

Oral and rectal temperature, urinary excretion of adrenaline and noradrenaline, and heart rate were measured in 28 watchkeepers working a "4-on/8-off" routine, and in 25 dayworkers, on board ships. Readings of oral temperature were taken over 4-hourly periods of up to two weeks; the rectal temperature, urine, and heart rate data were collected on selected days within these periods. Analysis of watchkeepers' temperatures and unconjugated catecholamine excretions showed slight signs of an interactive adaptation to time of day and hours of work, but it was clear that full phase adjustment of the circadian rhythms to shifted hours of work did not occur. The "split" pattern of sleep imposed by the watch system may be a major factor in preventing complete adaptation of physiological rhythms to shift work in the shipboard situation; this problem could be overcome by devising a system that allows sleep to be taken in a single uninterrupted block each day.

Performance scores on both a simple detection task and a complex visual judgment task, together with subjective alertness ratings, were collected 4-hourly from 19 watchkeepers working a "4-on/8-off" routine, and from 20 dayworkers, over consecutive days of voyages on merchant ships. Adaptation of the circadian rhythms in the measures to the shifted hours of work of the three sections of the watchkeeping crews was at best only partial. These rhythms resulted in a depression of the levels of most measures during the night hours, which was exacerbated by the effects of recent awakening from sleep. It is suggested that these unwanted variations in operational effectiveness could be largely removed by replacing the 4-on/8-off watchkeeping system, with its associated "split" sleep pattern, by one which allows a single full length sleep each day. Such a system would encourage better adaptation of the rhythms; effects of recent awakening could be avoided by the provision of an adequate "waking up" period before duty begins.

Daily diary records of sleep and activity, and 4-h measurements of body temperature, performance and subjective alertness were collected on board ship from 15 watchkeepers on the 4-on/8-off system, and from 28 dayworkers, on both westward and eastward transatlantic voyages. The data from a balanced sample of the subjects were analysed over selected 8-d periods of the voyages where four or five time zones were crossed. During these periods the average amount of daily sleep obtained by dayworkers on the eastward voyage was more than 1 h less than that on the westward voyage, and its quality was rated lower. Watchkeepers' main sleep was also shorter when travelling eastward, but this reduction was partially compensated for by a slightly longer secondary sleep. With the exception of subjective alertness on the eastward voyage, the basic phase of the circadian rhythms in the measured variables adjusted appropriately to the clock changes associated with the time zone crossings. The normal shape of the average daily curves was, however, altered differentially in the two directions of travel; as a result, morning levels of all variables were lower on the eastward voyage than on the westward, but evening levels were higher. These distortions of rhythm waveforms, which probably arose from a combination of endogenous and exogenous factors, add another dimension to the basic problem caused by the effects of circadian rhythms on operational efficiency in the shipboard situation. This problem can only be solved by the development of alternative watchkeeping systems which take full account of these rhythms.