3. COMPETITION AND CONSTANCY. 



Significance. — Although a number of studies have been made with 

 special reference to the constancy of a particular species of pollinator to 

 one species of flower, there has been little or no experimental work in this 

 field. Competition between different species or genera of plants for the 

 visits of insects has received practically no attention and no experiments 

 are known that bear directly upon this problem. 1 The early view that the 

 bees in particular are constant or nearly so has been shown to be only partly 

 true by Bulman, Plateau, and others, and the whole subject of habit in 

 pollinators has been found to be much more complex than supposed. The 

 idea that constancy was essential to the effective cross-pollination of a 

 species, as well as to the prevention of hybridization, has necessarily been 

 modified, and is now rather to be regarded as a matter for investigation 

 in terms of habit and efficiency. Moreover, there have been several dif- 

 ferent concepts of constancy, as discussed later. In the present treatment, 

 the experimental results of competition studies are first considered. This 

 is followed by an account of the composition and weight of the pollen loads 

 of various species, and the chapter is concluded with a resume of the work 

 of other investigators and a summary of the general principles involved. 



COMPETITION. 



General plan. — The methods employed in competition experiments 

 have been discussed in detail in the introduction, the one most used in the 

 present study being that of reciprocal bouquets. The primary objectives 

 have been to throw light on the degree of habit fixation on the one hand 

 and to furnish evidence as to the relative attractiveness of different flowers 

 on the other. At the same time, questions of efficiency have been taken into 

 account. For the most part, the species with regular corollas were used in 

 order that visitors might not be excluded or handicapped by structural 

 features, but a few zygomorphic flowers were employed in order to disclose 

 the effect of structure. 



The bouquet method was the one regularly used, though occasionally 

 normal plants grew sufficiently close to each other that equal areas or equal 

 numbers of flowers could be marked off for comparison. In all cases the 

 natural plant was regarded as the standard, and the comparison was made 

 by means of a bouquet placed in its midst or attached to the branches. 

 The bouquet consisted in some cases of a mass of flowers or a group of 

 branches and in others of single flowers or branches put in vials and attached 

 to the stems. In practically all cases where this method was used, the 

 records were made in the normal habitat of the standard plant. The 

 bouquets were arranged among the stems or branches so that the insects 

 would necessarily pass them in moving about. In a few instances the 

 bouquet consisted of several species from other localities, in order to deter- 

 mine the reaction to species probably not seen previously but similar to 

 those regularly visited. 



1 Knoll has made an incidental study of competition, which has just come to hand (1922: 

 215); he employed Pelargonium zonale or Satureia nepeta in competition with Linaria vulgaris 

 in experiments with Macroglossa stellatarum. 



94 



