A total of 285 samples representing 51 feedstuffs used in Belgian feed-mills were quantitatively analysed for phytase activity, phytate-P and total P. It was concluded that the number of feedstuffs showing significant phytase activity (more than 100 units kg(-1)) is rather limited (n = 13). Of the cereals analysed, only rye (5130 units kg(-1)), triticale (1688 units kg(-1)), wheat (1193 units kg(-1)) and barley (582 units kg(-1)) were rich in phytase. Wheat by-products, such as fine bran meal (4601 units kg(-1)) or pellets (2573 units kg(-1)) I, middlings (4381 units kg(-1)), feed flour (3350 units kg(-1)) and bran (2957 units kg(-1)) were very phytase rich. Pelleted wheat fine bran samples were, on average, only 56% as active as non-pelleted wheat fine bran. Malt sprouts pellets (877 units kg(-1)) as well as corn distillers (385 units kg(-1)) also showed a moderate phytase activity. All other feedstuffs analysed showed zero or extremely low phytase activity. Phytase activity was not related to total P content or phytate-P content. A significant linear relationship was found between phytate-P and total P, for two feedstuff classes-wheat plus wheat by-products (R(2)=0.953; RSD=0.060) and maize plus maize by-products (R(2)=0.928; RSD=0.042)-but not for cereals or oil meals. Phytate-P content, as a percentage of total P content, was higher in cereals and in wheat by-products than in oil-seed meals and legume seeds. In malt sprouts, corn distillers and moist ensiled maize, phytate-P seems to be totally or partly hydrolysed as a result of processing or ensiling. In feedstuffs derived from roots and tubers, as well as in citrus pulp, cocoa shells, soyabean hulls, flax chaff, maize cob meal, dehydrated alfalfa and a dried mycelium sample, phytate-P was not detectable.
