Abstract: (1) The distribution of excreta on areas of reseeded grassland in the New Forest used by free-ranging cattle, ponies and fallow deer was shown to be non-random. Distinct latrine areas were recognized where the faeces of all three herbivore species were concentrated, and where the majority of urinations occurred. The mosaic of latrine and non-latrine areas can be detected in aerial photographs in which non-latrine areas appear as light-grey patches set in a matrix of the dark grey latrine areas. During the 3 years of the study the position of the mosaic proved to be relatively stable. (2) The latrine areas were characterized by an uneven sward about 50 mm tall with abundant thistles (Cirsium spp.) and ragwort (Senecio jacobaea). Non-latrine areas had an even and very closely cropped sward between 10 and 20 mm tall. Soil chemical analysis of the two kinds of area revealed significantly higher levels of exchangeable potassium in latrine areas, and on one site significantly higher levels of magnesium and organic matter. No significant differences were detected in soil reaction, nor in phosphorus or calcium levels. (3) Observations of grazing animals revealed a tendency, at all times of year, for ponies to avoid grazing in latrine areas. In winter and spring this tendency was very slight, but from midsummer until late autumn a substantial majority of grazing ponies were to be found in non-latrine areas. In contrast, only 2% of the cattle observations made over a period of 20 months were of animals grazing in non-latrine areas. (4) The standing crop of dung and the rate of dung production on the two kinds of area were monitored for 12 months on one lawn. The amount of pony dung produced on non-latrine areas was only 16.5% of that in latrine areas, while for cattle the corresponding value was 28.7%. It is argued that the observed pattern has been created by selective grazing and eliminatory behaviour of the ponies, and that the excreta of cattle and deer are largely confined to pony latrine areas because these animals are unable to graze the very short herbage of non-latrine areas.

Abstract: Herring gulls drop hard-shelled mollusks and hermit crab-inhabited molluskan prey in order to break the shells and gain access to the edible interior. A field study of predatory shell dropping on Cape Cod, Massachusetts, U.S.A. showed that the gulls usually drop the same shell repeatedly, orient directly to dropping sites that are invisible from the point at which the mollusks are captured, drop preferentially on hard surfaces, adjust dropping heights to suit the area and elasticity of the substrate, orient directly into the wind while dropping, sever the large defensive cheliped of hermit crabs before consumption, and rinse prey that is difficult to swallow. Proficiency in prey dropping is acquired through dropping objects in play, trial-and-error learning, and perhaps, observation learning.
Observable attributes of predatory shell-dropping support inferences that the gulls are capable of extended concentration, purposefulness, mental representation of spatially and temporally displaced environmental features, cognitive mapping, cognitive modeling, selectivity, and strategy formation. Identical cognitive processes have been inferred to underlie the most sophisticated forms of chimpanzee tool-use.
Advanced cognitive capacities are not restricted to chimpanzees and other pongids, and are not associated uniquely with tool use. The chimpocentric bias should be abandoned, and reconstructions of the evolution of intelligence should be modified accordingly.