A clay loam, a silt loam and a sand soil were gradually dried from field moisture content to air-dryness at 25�C in the laboratory. Microbial C measured by substrate-induced respiration (SIR), fumigation-incubation (FI) and fumigation-extraction (FE), microbial N-flush measured by FI and FE, microbial ATP content, and soil phosphatase and sulfatase activities were monitored throughout a drying period of approx. 60 h achieved over 16 days. All the microbial and enzyme variables declined as the gravimetric soil water content ( W) decreased to air-dryness. In general, the relationship between microbial C or N-flush and W was linear, but was exponential between ATP or phosphatase and W. Soil texture appeared to affect the rates of decline and also the amounts of the microbial and enzyme variables remaining in air-dry soil; e.g., the lowest rate of microbial C decline and the largest amount remaining at air-dryness occurred in the clay loam soil. Sulfatase activity was not significantly affected by soil drying. Agreement between the SIR and FE estimates of microbial C was good (r = 0.92***). These two methods were applicable over a wide range of water contents. Microbial N-flush, estimated by the FE method, also showed a consistent trend and correlated highly with microbial C estimated by SIR or FE. In contrast, microbial C and N-flush estimated by the FI method were not significantly correlated with W or any of the other variables. ATP and phosphatase activity appeared to relate more closely to microbial activity (CO2 respiration/microbial C) than microbial mass. The reliability of the methods to measure the biomass and the influence of soil texture, water and carbon contents on microbial survival are discussed.

Microbial respiration, mass and activity (respiration/mass) in three silt loam soils with different moisture regimes (1000-2700 mm annual rainfall) were measured as the field-moist soils were gradually air-dried at 25�C in the laboratory. Microbial CO2 and O2 respiration and mass (estimated from substrate-induced respiration), and oxidizable organic carbon, extractable in 0.5 M K2SO4, were monitored throughout the 60 h drying period and also in samples of these gradually dried soils which were rewetted with water (2 ml g-1 soil). Water availability was the major factor controlling microbial respiration and activity as the soils dried. Both respiration and activity declined continuously as gravimetric water content (W) decreased. Microbial mass also declined, but only after W fell below 0.1-0.3. The microbial mass of the lowest-rainfall soil resisted desiccation better than the higher-rainfall soils. Extractable carbon levels increased after considerable soil drying (when W < 0.1-0.2). This increase in extractable carbon could be accounted for by the carbon released from micro-organisms killed by drying. Microbial respiration and activity increased sharply within 30 min of rewetting gradually dried soils. These increases appeared to be largely due to the carbon released from micro-organisms, killed by drying, being metabolized by the surviving micro-organisms when water was present. The ability of the surviving micro-organisms to metabolize this (extractable) carbon was remarkably similar for all three soils, suggesting that the carbon released from killed cells is of a very similar qualitative nature, and/or that the surviving populations were also similar between the soils.

Temporal fluctuations in rates of nitrogen mineralization and the activities of six enzymes were measured in topsoil, predominantly a Typic Haplaquoll, from two plots that contained pastures of different age in the Wairarapa area. Samples were taken at c. 4-weekly intervals for over one year. Organic carbon contents averaged 6.7 and 3.6% in soil from the older and younger pastures respectively. Net mineral-nitrogen production at 25�C was higher in soil at a standardized water potential (-3 and -4 kPa for soil from the older and younger pastures respectively) than at field moisture content. It was initially higher in soil from the younger than from the older pasture. Generally, distinct seasonal patterns were not apparent. All of the enzyme activities showed significant temporal fluctuations. Amylase and cellulase activities fluctuated more than invertase activity, but all three carbohydrase activities were generally high in wet spring samples. When plot effects were removed, only the fluctuations in amylase activity were related positively and significantly to soil moisture content. Fluctuations in cellulase, urease, phosphatase and sulphatase activities were correlated negatively with soil moisture content. The temporal fluctuations in enzyme activities were, when plot effects were removed, mainly independent of the small variations that occurred in soil organic carbon and total nitrogen contents. Interrelationships of these biochemical properties, and relationships with rates of CO2 production and indices of microbial biomass, are discussed.

Three silt loam soils from a climosequence (1000-2700 mm annual rainfall) were gradually dried from field moisture content to air-dryness at 25�C in the laboratory. Microbial C measured by substrate-induced respiration (SIR), fumigation-incubation (FI) or fumigation-extraction (FE), microbial N-flush measured by FI and FE, microbial ATP content and soil phosphatase and sulfatase activities were monitored throughout the drying period (approx. 60 h). All indices declined as the gravimetric soil water content (W) decreased until reaching air-dryness. Significant declines in the biomass sometimes occurred only following a large decrease in W, dependent on the soil. In general, when microbial C and N-flush declined, the rates of decline were linearly correlated with W. However, ATP and soil phosphatase were exponentially related to W. When expressed as a ratio of the total change in microbial indices against the total change in W for the whole drying period, the ratios were consistent between the soils. Agreement between the SIR and FE estimates of microbial C, whilst significant (r = 0 58***), was poor, especially for the low rainfall soil, although the FE C- and N-flushes correlated well (r = 0-76***). In contrast, the FI C- and N-flushes correlated very poorly (r = 0.30**) and were not significantly correlated with W or the other indices. ATP and soil phosphatase activity were strongly correlated (r = 0.89***). The reliability of the methods and the influence of soil moisture regimes on microbial survival are discussed.

Grazing and fertilizer management practices are of prime importance for maintaining summer-moist hill pastures of introduced grasses and clovers in New Zealand for sheep and cattle production. The influence of withholding grazing (a pastoral fallow) from spring to late summer on microbial biomass, C and N mineralization, and enzyme activities was investigated in a Typic Dystrochrept soil from unfertilized and fertilized (rock phosphate and elemental S) low-fertility pastures at a temperate hill site. The fallow increased pasture but not legume growth in the following year in the unfertilized treatment, but had no effect on pasture or legume growth in fertilized plots. High background levels of the biochemical propel-ties examined, and very variable rates of N mineralization, complicated data interpretation. Extractable-C concentration and CO2-C production were enhanced at the completion of the fallow. Increases in net N mineralization (14-56 days incubation), following initial immobilization, after the fallow were clearly indicated in the unfertilized treatment, but were less distinct, in the fertilized treatment. The fallow had no detectable influence on the concentrations of total C and N or microbial C and P, or on invertase, phosphodiesterase and sulfatase activities. Some small changes in microbial N and an increased proportion of bacteria in the microbial population were, however, suggested. Results are consistent with the concept of fallowing giving a short-term increase in pools of readily decomposable soil organic matter. Generally, the changes that did occur in these soil biochemical properties are, with the partial exception of increased N availability, unlikely to have had any pronounced impact on subsequent pasture performance.