dance from the newer apical growth to the older basal 

 foliage. If the width of the branch at mid-length is 

 measured, then the length of the foliated part times 

 the width gives the foliage surface. The total foliage 

 surface of representative trees is determined from 

 felled individuals, and the total foliage surface per 

 unit area may be computed from the known density 

 of trees. If the arthropods vary in abundance at dif- 

 ferent levels in the tree, representative sampling must 

 be taken at each level. Considerable variation in ani- 

 mal density also occurs from tree to tree so that sam- 

 pling must be well distributed over the area under 

 investigation (Morris 1960). 



Soil animals 



One must resort to a variety of methods to 

 census the difTerent kinds of animals in the soil be- 

 cause of great differences in their size, physical char- 

 acteristics, and behavior (Fenton 1947, Van der Drift 

 1950, Kevan 1955). The uiegafaitna consists of the 

 larger millipedes, centipedes, snails, amphibians, rep- 

 tiles, and small mammals. Mammals must usually he 

 trapped. For the other forms mentioned, if there are 

 a half-dozen workers available, a plot 10 meters on 

 a side (100 m-) may be marked out and the ob- 

 servers, forming a line at one side, may gradually 

 work over the plot, turning over all the leaves and 

 sticks. This gives a good count but must be repeated 

 in various parts of the community. 



FIG. 4 4 Tullgren modification of 

 a Berlese funnel for quantitative 

 sampling of soil animals. 



For quantitative sampling the fauna of fallen logs 

 and decaying stumps, it is convenient to mark out an 

 area 50 meters on a side (0.25 hectare) and then 

 measure the length of all logs and the height of 

 stumps. A medium-sized log and stump are com- 

 pletely torn apart and all animals counted. The total 

 population for the whole area may then be calcu- 

 lated. 



The inacrofauiia, consisting of the larger insects 

 and spiders, the smaller millipedes, centipedes, and 

 snails, and the earthworms may be censused by means 

 of a steel ring, 7.5 cm wide, having a sharpened edge, 

 and covering 0.1 m-, which is pressed into the ground 

 until it is flush with the surface. The litter and top 

 10-12 cm of the soil, which contain most of the 

 ground animals, may then be sorted by hand, either 

 in the field or in the laboratory. Samples brought in 

 from the field should be transported in paper or 

 plastic sacks or tight containers and sorted as soon 

 as possible before the predatory animals in the sam- 

 ple have consumed prey species. 



Lumbricid earthworms commonly penetrate well 

 below the topmost few centimeters of the soil. Hand- 

 sorting of considerable amounts of soil is both la- 

 borious and time consuming, but gives the most de- 

 pendable results. In one study, small sample plots 

 were thoroughly soaked with a potassium permanga- 

 nate solution, and a later check by hand sorting indi- 

 cated that 80 per cent of the adult and 100 per cent 

 of the immature worms were forced to the surface 

 (Evans and Guild 1947). Attempts at earthworm 

 censusing with other chemicals have been less satis- 

 factory (Svendson 1955). Driving worms to the sur- 

 face by a discharge of alternating current from a 

 probe thrust into the ground has also been tried 

 (Kevan 1955). 



Enchytraeid or pot-worms belong to the meso- 

 fauna as do the smaller arthropods, such as spring- 

 tails, symphylans, pauropods, proturans, mites, and 

 various insect larvae. Enchytraeids may be extracted 

 quickly and efficiently by putting a layer of sand on 

 top of a soil sample in a special container, and ap- 

 plying heat and water from below. This forces the 

 animals to accumulate in the sand, from which they 

 may be easily separated (Kevan 1955). 



A common method of extracting small arthropods 

 from litter and soil is by means of the Tullgren modi- 

 fication of the Berlese funnel. Where possible, the 

 soil sample should be kept intact as a block and in- 

 verted into the funnel, bottom side up with an electric 

 light bulb placed above the sample. Forced to retreat 

 from the light and heat and the gradual drying of the 

 soil from the top to the bottom, the animals move 

 down the funnel and fall into the bottle of alcohol 

 below. A week or ten days is usually required to ob- 

 tain all possible animals from the sample. The pro- 

 cedure is subject to a number of faults, however, and 



38 Background 



