Popul5 4 



(Anscombe 19^0). These factors determine the size of the bulk- 

 sample, and the greater this is, the nearer its mean concentration 

 approaches the true mean of the parcel of land under investigation. 

 Local variability is created by the cultivation of host plants, both 

 laterally and at depth, especially where wide row spacing is used. 

 Cultivations and ploughing in subsequent years tend to even up these 

 local variations and render the surface soil more uniform. Gross 

 variability is encountered, especially in the early stages of infes- 

 tation, and tends to become less as the field becomes more uniformly 

 infested. 



Recently Fenwich (work about to be published) has attempted to get 

 estimates of within-field variability and to deduce from this the 

 influence of (a) size of count and (b) number of borings taken. His 

 results are very briefly summarized in Table U. The picture given 

 is more favourable than would be obtained in the field for the follow- 

 ing reasons: (1) the plots used were experimental plots and have had 

 more unifonn treatment than fields; (2) their eelworm populations were 

 all relatively high; (3) their size was small so that the sampling 

 rate per vinit area (even at low levels) was higher than is usually 

 possible in fields; and (h) the assumption that within spot variability 

 is distributed according to the Poisson distribution. This is not so. 



The success of sampling for advice to farmers on the desirability of 

 growing susceptible crops on infested land, which has been practiced 

 in Europe for many years, may be attributed to the fact that such 

 popiilation estimates are required mainly for fields in the heavily 

 infested class and also to the high absolute level of the 'economic 

 zeros' for cultivated crops. For potato root eelworm, this level has 

 been estimated by Johnson & Thompson (19U5) at 0.5 cysts with con- 

 tents/g. soil or ^00 x 10° cysts with contents/acre. Similar levels 

 apply also for the beet and cereal root eelworms. 



In advisory work, spatial variability has to be accepted as one of the 

 hazards of soil sampling. In research, control of spatial variability 

 is highly desirable. A measure of control can be obtained by use of 

 "fixed plot," "fixed grid" sampling, or by employment of "microplots" 

 or pots. In the last two, infested soil can first be homogenised by 

 mixing. Small pots, however, are unnatural in their water, temperature, 

 and aeration relationships; and it is difficult to translate results 

 obtained in them in terms of field conditions. Larger glazed pots, as 

 used by Peters (1952), are more satisfactory; but a closer approach to 

 field conditions can be got with 'microplots' (Jones 1956). 



Assay of Cyst Contents 



Morgan (1925) and other early workers in this field attempted to use 

 cysts as a measure of the soil population, but it soon became clear 

 that cyst coixnts were invalid for this purpose. Figure 2 (Jones 19U5) 

 shows the lack of correlation between cyst and egg counts for the beet 

 eelworm. A better measure is the number of cysts with contents 



