The Harvest Method
This method involves taking samples, usually as soil cores, dry-sorting the organic material or rinsing it free by use of water or other flotation media, then sieving, sorting, and obtaining dry mass values.
For sorting and categorizing roots, three factors need to be considered: root diameter, spatial distribution, and also temporal distribution (Fogel, 1985). Much of the existing data have been derived from thousands of cores that have been washed, sorted, and analyzed by legions of weary researchers. Some of these data have been truly informative and worth the effort. Other efforts, perhaps a majority of the published papers, have limited value. In the course of measuring root production by the harvest method, scientists often use what is known as the "peak-trough" calculation, in which the peaks and valleys of root-standing crops through the course of a growing season as represented on a graph are successively added or subtracted about some general mean level. Unfortunately, there can be a fairly frequent occurrence of no net changes in root biomass, perhaps as often as 30% of the time in grasslands studies (Singh et al., 1984); these are known as zero-sum years, which have no net production because the increases in production are canceled out by those periods which show decreases. These problems were reviewed by Singh et al. (1984) (Fig. 2.1). They extensively analyzed a grassland root production data set, looking for effects of sample (replicate number) size and sampling frequency, and coming to the conclusion that fairly frequently (perhaps in 3 years out of 10) one could expect to measure no significant increments to growth when using the peak-trough harvest method. In addition, they compared the amount of NPP that one would expect from the peak-trough harvest method with a multiple-year-based computer simulation model of root production and turnover. They found that the peak-trough method at times overestimated either the "true" or the simulated root production by as much as 150% because of widely varying means; this led to spuriously high "production" values. The simulated production was not more "real" than the data, of course, but the researchers raised the question that perhaps the peak-trough method, as applied usually, may often lead to some significant overestimates of root production rates.
Considerable information is available on fine root production (FRP) in forested ecosystems. Nadelhoffer and Raich (1992) compiled 59 published estimates of annual net FRP from 43 forest sites worldwide. They compared four techniques used by investigators: (1) sequential core method (calculated as differences in means of fine root biomass between sampling periods and measured across growing seasons); (2) maximum-minimum method (simpler than the first method in that it uses only the difference between annual minimum and maximum fine root biomass to estimate FRP; (3) ingrowth core method (similar to the method of Steen [1984, 1991], cited later in this chapter); and (4) the nitrogen budget method (based on annual measures of net nitrogen mineralization in soil and net nitrogen flux into aboveground tissues. Annual nitrogen allocation to fine roots is calculated from the difference
Was this article helpful?