54 BASES AND CRITERIA. 



upon the kind and thoroughness of the experimental test. The planting of a 

 trial crop by a settler will give some idea of the indicator meaning of the native 

 vegetation that has been removed. In such a case the evidence is slight and 

 its value tentative. If the planting is repeated for several years or is extended 

 to other farms or localities, its value increases accordingly. As this is the 

 usual course for a crop in a new region, it is obvious that ordinary agricultural 

 practice must suggest indicator correlations with crop plants. This is well 

 known to be the case, but the actual utilization of indicators by farmers seems 

 always to have been inconsiderable. This is largely due to a lack of knowledge 

 of native plants, especially in a new region, but also to the fact that this 

 knowledge was needed most in selecting land and choosing crops, at a time 

 when it was still to be acquired. Thus, while the aggregate experience of a 

 neighborhood might possess real value, there has rarely been any method of 

 formulating it and making it effective. 



The extension of experiment stations and substations throughout the West 

 initiated the period of scientific study of agricultural problems. The investi- 

 gations were directed chiefly to the selection of the best varieties for different 

 regions and soils and to the improvement of yields. Unfortunately, the bota- 

 nist was not interested in the problems of field crops and the agronomist was 

 little or not at all concerned with native vegetation. The result was that a 

 great mass of experimental data remained unavailable because it lacked corre- 

 lation. It was possible to give this only through ecological studies, and then 

 only after quantitative methods had been devised for the analysis of habitat 

 and conununity. As a consequence, exact and purposeful studies on indica- 

 tors date from the present decade for each of the three great fields, agriculture 

 (Shantz, 1911), forestry (Clements, 1910), and grazing (Clements, 1916 : 102; 

 1917 : 303; 1918 : 296; 1919). In spite of this late beginning, the recognition 

 and utilization of indicators are destined to undergo rapid development. This is 

 especially true of forestry and grazing, owing to the fact that the corresponding 

 experiment stations and reserves are organized upon the basis of exact ecology. 



Essentials. It has already been insisted that experiment affords the only 

 decisive test of an indicator. A single experiment may do this if properly 

 checked, but repetition is regularly necessary to cover the range of conditions 

 in space and in time. The experiment itself must be made with the fullest 

 knowledge of the factors concerned as well as the vegetation to be correlated. 

 As already pointed out, this involves quadrat study of the community and its 

 successional relations, and instrumental study of the habitat and its variation 

 through the climatic cycle. The thoroughgoing appUcation of this method 

 makes it possible to take advantage of countless natural happenings to convert 

 them into experiments. The number of such possibilities furnished by denu- 

 dation, lumbering, fire, cultivation, grazing, etc., is countless. If adequately 

 utilized, they will not only greatly reduce the number of set experiments 

 necessary, but will also make the latter possible on a scale otherwise out of the 

 question. The natural experiment has the advantage in economy of time and 

 effort, and in repetition of examples. The checked experiment permits of a 

 definite choice as to time and place, and allows greater control. It is the essen- 

 tial task of experimental ecology to combine these into a complete method, 

 which will give quantitative results throughout the field of ecology as well as in 

 connection with indicators. This is one of the primary objects of the present 

 treatment, though the indicator relations are necessarily given first place. 



