We examined the results of a natural 'experiment' to determine the effects of fishing on Caribbean reef fish communities. We repeated a trap survey in 1986 of reef fish populations at exposed. offshore sites on southeast (SE) and southwest (SW) Pedro Bank and protected, nearshore sites around the Port Royal Cays (PRC), Jamaica, that had been surveyed in 1969-73. The 'control' area -SW Pedro Bank -was virtually unexploited in 1969-73 and was lightly-to-moderately exploited in 1986. As 'treatment' areas, exploitation of SE Pedro Bank increased from moderate to heavy between surveys, and from heavy to very heavy around PRC. Overall catch rates declined significantly in heavily exploited areas -82 % over SE Pedro Bank and 33 % around PRC -but did not decline over SW Pedro Bank. The largest fishes commonly caught in the traps (lutjanids and large serranids and scarids), f a m~e s particularly vulnerable to trap fish~ng (acanthurids and bahstids), and other commerc~ally preferred species groups (small serraruds, haemulids, mullids) dechned most consistently. Increases in catch rate were generally found only in commercially less desirable reef fishes, such as chaetodontids, holocentrids, and tetraodontiformes. The catch (i.e. species) composition changed significantly between surveys in h e a d y exploited areas (SE Pedro Bank and PRC) but not over SW Pedro Bank. Similar changes in 2 different reef environments (SE Pedro Bank and PRC) suggest that quantitative relationships of catch rate and composition with fishing effort may be useful in fishery management.

1. Four experiments designed to investigate the effect of fertilizers on the rate of bulking of Majestic potatoes are described and the results discussed.2. Irrespective of fertilizer application yield increased until the haulm died and then decreased slightly.3. Sulphate of ammonia increased the yield by increasing the number of tubers and increasing the yield in all size-grades. Thus its effect was seen throughout the season, no matter what grades yield was based on. It also increased dry-matter percentage.

3-year ley).In treatments including 1-year ley, broad red clover was compared with Italian rye-grass, and, in those including 2-and 3-year leys, pure lucerne leys were compared with pure cocksfoot leys and a lucerne/cocksfoot mixture, all the leys being underlst year Superphosphate (19-5% P a O ) (cwt./acre) Muriate of potash (60 % K a O) (cwt./acre) 0 sown in the preceding barley crop. Thus there were nine combinations of cropping sequence and composition of leys and these formed the main treatments of a split-plot design. All leys were compared at two levels of management (M), namely, leys always mown and the produce removed, and leys always grazed by sheep. Defoliation was carried out at the same time, three times each year on all treatments, so that management should cause little difference in botanical composition. Of the allarable plots half were undersown with trefoil in the barley crop immediately preceding the potatoes, and this was ploughed in without cutting or grazing in the autumn; the other half of these plots had no legume ploughed in. Other subtreatments were (a) three levels of nitrogenous manuring (N) applied to all crops as follows:cwt.

An experiment comparing six-course rotations which included 1-, 2-and 3-year leys of pure grass, pure legume and grass/legume mixture is described, and the results from the fifth and sixth phases of the first cycle and first and second phases of the second cycle are discussed.Compared with all-arable cropping, the inclusion of leys increased the yield of barley grown as the second and third crops after the leys were ploughed up, although the difference was less when 58 kg N/ha had been applied. Neither length nor type of ley affected yield in the second crop after ploughing the leys, but in the third crop 2-and 3-year leys led to greater yield than 1-year leys.Only rotations containing 1-year leys were compared with all-arable cropping in the fourth and fifth crops after ley, but some effects of these were still apparent in that both Italian ryegrass and broad red clover leys led to an increased sugar percentage of sugar beet grown as the fourth crop when no nitrogen was applied, and broad red clover leys led to a greater response to nitrogen in barley grown as the fifth crop after ley. INTRODUCTIONHanley, Ridgman & Jarvis (1964a) discussed the results of an experiment, conducted on light gravelly soil at Cambridge, which compared the effects of different sequences of cropping, some of which contained leya of various types and duration, different levels of nitrogenous manuring and the application of farmyard manure on the yield of potatoes. The experiment has continued as a rotation experiment and the present paper reports the results of the four arable crops which followed potatoes in the first cycle.