358 FOREST INDICATORS. 



tion are relict survivors, or trunks and stumps. Less obvious but equally 

 conclusive are charred fragments or pieces of charcoal in the soil. In those 

 cases where there is no direct evidence of the original forest, the desired clues 

 are readily afforded by indicator communities which bear a definite relation 

 to the forest. Such are serai and especially subclimax communities which 

 exhibit a successional relation to the forest cUmax, and societies of shrubs or 

 herbs which formed layers in it. WhiJe the latter are frequent in burns and 

 clearings, they are usually accompanied by tree relicts which furnish more 

 direct evidence. In some cases, however, they are the sole indicators of the 

 former existence of forest in a particular spot. Subclimaxes are by all odds 

 the best indicator communities of forest cHmaxes, since they show that the 

 habitat has reached the condition in which the chmax dominants can thrive. 

 The earlier communities of a subsere have nearly the same value, since the 

 habitat undergoes relatively slight change. In the case of a prisere, only the 

 grass and scrub stages indicate that the slow reaction upon the originally bare 

 area has reached a point in which remaining changes may be compensated by 

 planting operations. Afforestation indicators are savannah, chaparral, or 

 grassland of tall-grasses, in which the water requirements are sufficiently near 

 those of trees that the gap may be bridged by planting methods, and espe- 

 cially by making use of the increased rainfall of the wet phase of the climatic 

 cycle. 



Furthermore, the indicators of sites for planting or sowing serve also to 

 indicate the preferred species. In the case of reforestation, the general rule 

 is that these are the climax trees that were in possession, but reasons of manage- 

 ment may make it desirable to employ a subclimax dominant, such as lodge- 

 pole pine. Similarly, the growth-form best adapted for planting in a region 

 is the one developed by that region, as the Forest Service has repeatedly 

 demonstrated at its experiment stations. In the case of afforestation, the 

 indications as to species must be derived from tree communities somewhere in 

 contact with the grassland or scrub, as from pine in the case of the sandhills of 

 Nebraska, from the indications of an intermediate conamunity, such as scrub, 

 or from the comparative study of habitats. 



Prerequisites for planting and sowing. The critical part played by rodents 

 and by competition in natural reproduction was recognized more than a 

 decade ago (Clements, 1910). Extensive tests of sowing in many national 

 forests by the Forest Service has shown that destruction or control of the 

 rodents is imperative (Tillotson, 1917:50). In fact, it seems evident that 

 for practically all regions rodents are the most serious enemies of both natural 

 and artificial reproduction, and that they should be systematically and 

 permanently cleared out of all areas in which reproduction is important. Com- 

 petition is a process which is less readily controlled on a large scale. Com- 

 petition for water is much more decisive as a rule than for light, the latter 

 usually becoming critical only in dense scrub or similar communities. The 

 disturbance of the soil involved in planting seedUngs or in sowing by the seed- 

 spot method usually suffices to reduce water competition sufficiently, except in 

 a grass sod. The latter is usually encountered in clearings and in grassland 

 associations in which afforestation is the method to be employed. In climax 

 grassland, where the annual rainfall is less than 25 inches, the grasses use all 

 of the water-content during the drier portions of the season. As a consequence 



