The question as to why some communities are more invasible than others has pro‐found implications for conservation biology and land management. The theoretical issues involved go right to the heart of our understanding of species coexistence and community assembly. The experiment reported here indicates that for productive, small‐scale grassland plots, species identity matters more than species richness in determining both the number of invading species and the total biomass of invasives.

The Park Grass Experiment at Rothamsted in southeast England was started in 1856, making it the longest-running experiment in plant ecology anywhere in the world. Experimental inputs include a range of fertilizers (nitrogen, phosphorus, potassium, and organic manures) applied annually, with lime applied occasionally, and these have led to an increase in biomass and, where nitrogen was applied in the form of ammonium sulfate, to substantial decreases in soil pH. The number of species per plot varies from three to 44 per 200 m 2 , affording a unique opportunity to study the determinants of plant species richness and to estimate the effect sizes attributable to different factors. The response of species richness to biomass depends on the amount and type of nitrogen applied; richness declined monotonically with increasing biomass on plots receiving no nitrogen or receiving nitrogen in the form of sodium nitrate, but there was no relationship between species richness and biomass on plots acidified by ammonium sulfate application. The response to lime also depended on the type of nitrogen applied; there was no relationship between lime treatment and species richness, except in plots receiving nitrogen in the form of ammonium sulfate, where species richness increased sharply with increasing soil pH. The addition of phosphorus reduced species richness, and application of potassium along with phosphorus reduced species richness further, but the biggest negative effects were when nitrogen and phosphorus were applied together. The analysis demonstrates how multiple factors contribute to the observed diversity patterns and how environmental regulation of species pools can operate at the same spatial and temporal scale as biomass effects.

For insects exploiting spatially structured arrays of resource patches (host plants, fungi, carrion, etc.), the distribution of individuals among patches can have important consequences for the coexistence of competitors. In general, intraspecific aggregation of consumer individuals over the landscape of patches stabilizes competition. Oviposition behavior of individual females can generate aggregation of larvae across patches and, therefore, strongly influences the outcome of competition between co-occurring species. We used simulation models to evaluate the consequences (for the coexistence of competitors) of different movement behaviors by females before and between oviposition events. Coexistence times increase when females are more likely to travel among neighboring patches than among distant ones. Coexistence times are also longer when females begin egg laying near the site of their emergence. Preoviposition dispersal is, therefore, destabilizing. We also considered responses by females to edges of resource arrays. Edge effects are generally stabilizing, delaying competitive exclusion by increasing larval aggregation, but different responses to edges have dramatically different effects on coexistence. The longest coexistence times occur when edges are "sticky", such that females encountering an edge tend to remain there.

The preparations of cobalt(II1) complexes of a series of tripeptides are reported in which the peptides are coordinated as quadridentate chelates through the terminal NH,, two peptide N-and terminal CO, groups. The 'H and I3C N M R spectra are given for the free peptides in the anionic, zwitterionic, and cationic forms and for the cobalt(II1) complexes. The chemical shift data are analyzed in terms of the effects of protonation and coordination, and the coupling constants for the a-CH-P-CH2 fragments of the leucyl, phenylalanyl, and tyrosyl residues are analyzed in terms of the possible rotamers for the side groups. In [Co(NH3),(g1y-gly-~-phe)], the phenyl group adopts a conformation that results in an adjacent NH3 ligand resonating at the very unusual position of 6 0.77. IntroductionAs part of an investigation into the reactions of metal peptide complexes, a series of diammine(tripeptide)cobalt(III) complexes has been prepared with the peptide coordinated as a quadridentate chelate (structure 1). Although the prepa-

No abstract