Wood ant nests as potential hot spots for carbon and nitrogen mineralisation

Red wood ants (the Formica rufa group) build large nests which can last for many years. The ants often bring large quantities of conifer litter to their mounds. In this study we compared chemical properties of the forest floor and ant-nest materials of red wood ants at two different sites in Sweden. We also did an incubation study in the laboratory to determine C and N mineralisation rates of soil and nest materials at two moisture levels. Some chemical properties, e.g. pH and C/N ratio, differed between sites. Nests were always drier, had a higher C/N ratio and often had higher NH₄⁺ concentrations than the surrounding forest floor. This indicates that the nests increase spatial heterogeneity of the forest floor not only because they contain large amounts of organic matter but also because of qualitative differences in nest composition. In the incubation study, experimentally increased moisture levels in the nest material resulted in net N immobilisation, while surrounding litter and humus materials showed net N mineralisation. The CO₂ evolution was clearly increased by increased moisture. Dry conditions in the nest material prevented it from being decomposed and, since there were no plants to take up N, inorganic N accumulated. However, based on our results we hypothesise that ant nests do not produce large amounts of inorganic N as long as the ants bring fresh litter with high C/N ratio to the nests and keep the nests dry. However, the nests can be considered as potential "compost heaps" which probably decompose faster and will be a source of N leaching when the ants no longer maintain the specific dry conditions.     

Impact of soil nitrogen concentration on Glomus spp.-Sinorhizobium interactions as affecting growth, nitrate reductase activity and protein content of Medicago sativa

Our objective was to evaluate how increasing levels of N in the medium (0, 4, 8 and 16 mmol N added kg^-1 soil) affect the interaction between Sinorhizobium and arbuscular mycorrhiza (AM) fungi in the tripartite symbiosis with Medicago sativa. Growth response, nutrient acquisition, protein content, and nitrate reductase (NR) activity were measured both in plant shoots and roots. Results showed that N levels in soil did not affect mycorrhizal colonization but they strongly influenced nodulation, particularly of mycorrhizal plants. Mycorrhizal colonization was required for a proper nodulation when no N was applied to soil. In contrast, the addition of 4 mmol N kg^-1 soil reduced nodulation only in mycorrhizal plants and 8 mmol N added kg^-1 soil allowed nodule formation only in non-mycorrhizal plants. Nodulation was totally inhibited in all treatments with the addition of 16 mmol N added kg^-1 soil. N addition enhanced NR activity in all the treatments, while AM colonization increased the proportion of NR allocated to roots. This effect was more pronounced under the lowest N levels in the medium. The two AM fungal species showed different distribution pattern of enzymatic activities in plant tissues indicating specific physiological traits. Protein content as well as the relative proportion of protein in roots were greatly increased after mycorrhizal colonization. Glomus intraradices-colonized plants had the highest protein content in shoot and root. Mycorrhizal effects on growth, N acquisition and biochemical variables cannot be interpreted as an indirect P-mediated effect since P content was lower in mycorrhizal plants than in those which were P fertilized. Mycorrhizal colonization increased the N content in plants irrespective of the N level, but the effectiveness of AM fungi on plant N acquisition depended on the AM fungus involved, G. intraradices being the most effective, particularly at the highest N rate. N₂ fixation, enhanced by AM colonization, contributed to N acquisition when a moderate N quantity was available in the soil. Nevertheless, under a high N amount the nodulating process and/or fixing capacity by Sinorhizobium was reduced in AM plants. In contrast, the AM fungal mycelium from a particular mycorrhizal fungus may continue to contribute efficiently to the N uptake from the soil even at high N levels. These results demonstrate the particular sensitivity of AM fungal species in terms of their growth and/or function to increasing N amounts in the medium. A selection of AM fungi used to address specific environmental conditions, such as N fertilization regimes comparable to those used in agronomic practices, is required for a better use of N applied to soil.     

Arbuscular-mycorrhizal inoculation of five tropical fodder crops and inoculum production in marginal soil amended with organic matter

Five fodder crops, Zea mays, Medicago sativa, Trifolium alexandrinum, Avena sativa, and Sorghum vulgare were inoculated with a consortia of indigenous arbuscular mycorrhizal (AM) fungi in non-sterile PO₄^3- deficient sandy loam soil amended with organic matter under field conditions. Shoot and root dry weights and total uptake of P and N of all the test plants were significantly increased by AM inoculation. Mycorrhizal inoculation increased yield in terms of shoot dry weight by 257% in T. alexandrinum followed by 50% in A. sativa, 28% in Z. mays, 20% in M. sativa and 6% in S. vulgare. Variation in dependence on mycorrhiza was observed among the fodder crops. T. alexandrinum showed a maximum dependence of 72% in contrast to 5.7% dependency in S. vulgare. Plant species showed differences in percentage AM colonization, with a high root infection recorded in Z. mays (76%). Spore production and infectious propagules (IP) were as high as 78 spores/IP g^-1 and 103 spores/IP g^-1 in S. vulgare. This study clearly indicates the potential of using indigenous AM inoculations in fodder crops grown in marginal soils along with in situ large-scale production of AM inocula.     

Variability in soils and vegetation associated with harvester ant (Pogonomyrmex rugosus) nests on a Chihuahuan Desert watershed

The effects of harvester ant (Pogonomyrmex rugosus) nests on the density and cover of spring annual plants and on soil characteristics were measured at three locations characterized by different soils and dominant vegetation on a desert watershed. There were few differences in vegetation and soils associated with harvester ant nests at locations at the base of the watershed where brief periods of flooding and sediment deposition occur at periodic intervals. At mid-slope locations, there were significant increases in total nitrogen, inorganic phosphorus, and cover (biomass) of four species of spring annuals at the edges of nest disks when compared with reference sites. The spring annuals that exhibited increased cover were species that increase biomass as a function of available nitrogen. At a clay-loam, Scleropogon-Hilaria, grassland site, there were significant reductions in the concentrations of Ca²⁺ and Mg²⁺, significant increases in nitrate and total nitrogen, but a significant increase in cover in only one species of annual plant. The data demonstrate that the effects of ants on soil properties and vegetation vary with site location and soil type.     

Nitrous oxide emissions from the wheat-growing season in eighteen Chinese paddy soils: an outdoor pot experiment

To identify the key soil parameters influencing N₂O emission from the wheat-growing season, an outdoor pot experiment with a total of 18 fertilized Chinese soils planted with wheat was conducted in Nanjing, China during the 2000/2001 wheat-growing season. Average seasonal N₂O-N emission for all 18 soils was 610 mg m^-2, ranging from 193 to 1,204 mg m^-2, approximately a 6.2-fold difference between the maximum and the minimum. Correlation analysis indicated that the seasonal N₂O emission was negatively correlated with soil organic C (r²=0.5567, P<0.001), soil total N (r²=0.4684, P<0.01) and the C:N ratio (r²=0.4530, P<0.01), respectively. A positive dependence of N₂O emission on the soil pH (r²=0.3525, P<0.01) was also observed. No clear relationships existed between N₂O emission and soil texture, soil trace elements of Fe, Cu and Mg, and above-ground biomass of the wheat crop at harvest. A further investigation suggested that the seasonal N₂O-N emission (E, mg m^-2) can be quantitatively explained by E=1005-34.2SOC+4.1S_a (R²=0.7703, n=18, P=0.0000). SOC and S_a represent the soil organic C (g kg^-1) and available S (mg kg^-1), respectively.     

The impact of harvester ants on decomposition, N mineralization, litter quality, and the availability of N to plants in the Mojave Desert

In arid areas of North America, nests of the seed-harvesting ant Pogonomyrmex rugosus tend to be elevated in mineral nitrogen and other soil nutrients relative to other microhabitats. We investigated the roles of decomposition, N mineralization, and plant nutrient uptake in maintaining high standing stocks of nutrients in P. rugosus ant nests. Decomposition rates of standard cellulose substrates placed on the surface of ant nests and other desert microhabitats suggest that conditions found in ant nests and bare areas are conducive to higher rates of decomposition than conditions under shrubs. In laboratory incubations of moist soil, net N mineralization rates were significantly higher in soil from ant nests than from bare areas and under two of three plant species. Net N mineralization rates measured in situ were much lower than those measured in laboratory incubations, but ant nest soil still exhibited higher rates at one of two sites. Litter collected from ant mounds, composed chiefly of seed chaff, was similar in N content to litter collected from underneath the dominant plant species, but had a significantly higher mean δ¹⁵N. Using this distinctive isotope signature as a tracer, we found no evidence that large perennial shrubs tap ant nests as a source of N. An invasive, annual grass species was significantly enriched in ¹⁵N, had higher leaf %N, and produced more seeds when growing on the mound than when growing several meters away; however P. rugosus nest surfaces are typically free of such annuals. We conclude that both high rates of nutrient cycling relative to other Mojave Desert microhabitats and low N utilization by the surrounding vegetation contribute to high standing stocks of mineral N in P. rugosus nests.     

Dehydrogenase activity, redox potential, and emissions of carbon dioxide and nitrous oxide from Cambisols under flooding conditions

Samples from topsoils (0-10 cm) of 16 Polish arable Cambisols developed from different parent materials (sand, silt, sandy gravel, loess, loam and clay), were incubated under flooded conditions with NO₃^-. Dehydrogenase activity, redox potential (Eh), and emissions of CO₂ and N₂O were measured. According to dehydrogenase activity, the soils were divided into two groups: those of low activity (I), where the final dehydrogenase activity was <0.03 nmol triphenylformazan (TPF) g^-1 min^-1, and those with high final dehydrogenase activity (II), >0.03 nmol TPF g^-1 min^-1. Generation of CO₂ and of N₂O under flooded conditions was shown to be significantly related to dehydrogenase activity. Soil dehydrogenase activity increased curvilinearly with organic matter content, showed a maximum at pH 7.1, and decreased curvilinearly with Eh. The final cumulative CO₂ production increased linearly with soil organic matter content and curvilinearly with dehydrogenase activity and decreased linearly with Eh. The most significant relationship was found with dehydrogenase activity (R²=0.74, P<0.001). The final cumulative N₂O production decreased linearly with Eh and increased curvilinearly with pH and dehydrogenase activity but linearly with organic matter content; the most significant relation being found with dehydrogenase activity (R²=0.69, P<0.001). The CO₂:N₂O ratio in the gases evolved increased curvilinearly with Eh and decreased with dehydrogenase activity and N₂O and CO₂ production.     

Correlations between pesticide transformation rate and microbial respiration activity in soil of different ecosystems

Cecil sandy loam soils (ultisol) from forest (coniferous and deciduous), pasture, and arable ecosystems were sampled (0-10 cm) in the vicinity of Athens, Georgia, USA. Soil from each site was subdivided into three portions, consisting of untreated soil (control) as well as live and sterile samples treated with the fungicide metalaxyl and the herbicide propachlor at 10 mg kg^-1 soil. Pesticide transformation rate, basal respiration (basal) and substrate-induced respiration (SIR) rates, and microbial metabolic quotient (qCO₂) were measured for the initial application of metalaxyl [methyl-N-(2,6-dimethylphenyl)-N-(metoxyacetyl)-DL-alaninate] or propachlor (2-chloro-N-isopropyl-acetanilide) at 22°C and 60% water holding capacity. Positive correlations were found for the following: metalaxyl transformation rate constant (K_met) and basal (r=0.73); K_met and SIR (r=0.83); propachlor transformation rate constant (K_pr) and basal (r=0.89); and K_pr and SIR (r=0.91). Regression analysis of pesticide transformation rate and soil respiration activity, coupled with specific soil properties (pH, C_org, and clay content), revealed a positive correlation between K and SIR for C_org (r=0.88 and 0.98, for metalaxyl and propachlor, respectively). qCO₂s were not significantly different (P=0.05) in propachlor-amended and pesticide-free soils. Metalaxyl amendment resulted in a change in the ecophysiological status of the soil microbial community as expressed by qCO₂. The qCO₂ values in metalaxyl-amended soils were significantly greater (P=0.05) in pine forest (by 25%) and arable and pasture (by 20%) soils compared to unamended soils. Differences in qCO₂ values may represent the magnitude of pesticide-induced disturbance. The duration of this disturbance was greater in the pine forest soil (48 days) compared to arable and pasture soils (21 and 15 days, respectively).     

Unique soil microbial assemblages associated with grassland ant species with different nesting and foraging strategies

Ants are important ecosystem engineers and can be abundant in extensively managed grassland ecosystems. Different ant species create nests varying in structure and size, and tend to have a variety of feeding strategies. Differences in food imported to the nest and contrasting nesting behaviour may control soil microbial community structure in nest soil, with cascading effects on nutrient cycling, but this has not been tested in grassland ants. Soil and ants were sampled from nests of three ant species: two formicines; Lasius flavus (aphid farmer/scavenger, mound builder) and Formica lemani (scavenger/hunter, non-mound builder), and a myrmicine; Myrmica sabuleti (hunter/scavenger, non-mound builder), in an extensively grazed temperate grassland and compared to similar soils without ants. Microbial assemblages were determined using molecular approaches (terminal restriction length polymorphism and automated ribosomal intergenic spacer analysis). Both aboveground (vegetation diversity) and belowground (soil physico-chemical properties) components were measured to assess the potential of the different ant species to modify the environment. Stable isotope ratios (δ¹³C and δ¹⁵N) of ant tissues and nest soil organic matter confirmed differences in trophic distances. Significant changes in soil pH, moisture content, total C and total N, and in vegetation composition, demonstrated ant ecosystem engineering effects. In turn, nests of L. flavus, M. sabuleti and F. lemani had different microbial activities and harboured significantly different microbial assemblages (total bacteria, total fungi, ammonia-oxidising bacteria and nitrogen-fixing bacteria), but the diversity was similar. These findings suggest that grassland ants can control microbial assemblages via changes in physical and biological soil characteristics in their nests, and as such, different ant species harbour unique microbial assemblages in nests.     

Naidid worms (Oligochaeta, Naididae) in paddy soils as affected by the application of legume mulch and/or tillage practice

Reproduction, intrinsic rate of natural increase and population density of naidid worms were investigated in submerged paddy fields and the laboratory. No tillage plus legume-mulching increased the population density of naidid worms. Soil treatments with neither tillage nor legume mulch, and tillage practice alone, did not increase the number of worms. Dero dorsalis Ferronnière was dominant in soil of the no-tillage treatment. In laboratory experiments, legume-mulching with the proper amount of dissolved O₂ accelerated asexual reproduction of D. dorsalis through zooid budding. For the legume and aeration treatment, (N_i+1-N_i) N_i^-1 values (where N_i and N_i+1 are the populations at times t=i and t=i+1) were plotted against N_i+1. Utilizing this linear relation, this data fitted the logistic curve (r²=0.885, P<0.05). Based on the linear relation, the intrinsic rate of natural increase (r), carrying capacity (K), and doubling time (T) were calculated as 0.2125 day^-1, 12,666 m^-2, and 3.26 days, respectively. The amounts of legumes applied were highly correlated with the population of D. dorsalis, indicating that the weight of legume is a limiting factor with respect to carrying capacity. A literature review indicated a significant correlation (P<0.01) between intrinsic rate of natural increase and maximum body length of naidids with temperature conversion of the growth rate. Sexually mature worms were rarely found in submerged paddy fields. Sexual reproduction seems to be adopted in response to soil desiccation after paddy field drainage.     

Effects of nitrogen fertilization on soil nitrogen pools and microbial properties in a hoop pine (Araucaria cunninghamii) plantation in southeast Queensland, Australia

A field study was conducted to investigate the effects of N fertilization on soil N pools and associated microbial properties in a 13-year-old hoop pine (Araucaria cunninghamii) plantation of southeast Queensland, Australia. The treatments included: (1) control (without N application); (2) 300 kg N ha^-1 applied as NH₄NO₃; and (3) 600 kg N ha^-1 as NH₄NO₃. The experiment employed a randomized complete block design with four replicates. Soil samples were taken approximately 5 years after the N application. The results showed that application of 600 kg N ha^-1 significantly increased concentrations of NH₄⁺-N in 0-10 cm soil compared with the control and application of 300 kg N ha^-1. Concentrations of NO₃^--N in soil (both 0-10 cm and 10-20 cm) with an application rate of 600 kg N ha^-1 were significantly higher compared with the control. Application of 600 kg N ha^-1 significantly increased gross N mineralization and immobilization rates (0-10 cm soil) determined by ¹⁵N isotope dilution techniques under anaerobic incubation, compared with the control. However, N application did not significantly affect the concentrations of soil total C and total N. N application appeared to decrease microbial biomass C and N and respiration, and to increase the metabolic quotient (qCO₂) in 0-10 cm soil, but these effects were not statistically significant. The lack of statistical significance in these microbial properties between the treatments might have been associated with large spatial variability between the replicate plots at this experimental site. Spatial variability in soil microbial biomass C and N was found to relate to soil moisture, total C and total N.     

Earthworm excreta (mucus and urine) affect the distribution of springtails in forest soils

The collembolan species Heteromurus nitidus, exclusively located in soils at pH>5, can be cultured in acidic humus. As this species is attracted to the excreta of earthworms from calcic mull, its distribution is supposed to be determined only indirectly by soil pH through the distribution of earthworms. Higher densities and biomasses of Lumbricidae were observed in a calcic mull (pH 7.8) than in an acidic mull (pH 4.8) and a moder humus (pH 4.2). Choice experiments were performed to compare the attraction of H. nitidus to the mucus-urine mix of five earthworm species from the calcic mull and the acidic mull. H. nitidus was attracted to the excreta of the five species, whatever their ecological category and the humus form from which they originated. The collembolan Heteromurus major, which was indifferent to soil pH, was not attracted to earthworm excreta, which emphasizes the significance of this phenomenon for the distribution of H. nitidus over a pH range. The attraction of H. nitidus to earthworm excreta tended to be weaker and more variable when earthworms originated from acidic mull compared to calcic mull, particularly in the case of Lumbricus terrestris. Increased earthworm density reinforced by better mucus quality and quantity could determine the distribution of H. nitidus according to soil acidity. The only urine compound capable of attracting H. nitidus was NH₃ at a low concentration (0.03 g l^-1). The NH₃ content of the mucus-urine mix which attracted H. nitidus was 0.037 g l^-1, and was therefore responsible, at least partly, for the attraction.     

Short-term patterns of carbon and nitrogen mineralisation in a fallow field amended with green manures from agroforestry trees

The mineralisation of green manure from agroforestry trees was monitored with the objective to compare the temporal dynamics of mineralisation of litter from different species. Green manures from five agroforestry tree species were used on a fallow field during the long rainy season of 1997 (March-August) and from two species in the following short rainy season (September-January) in western Kenya. Different methods, i.e. measurements of isotopic ratios of C in respired CO₂ and of soil organic matter (SOM) fractions, soil inorganic N and mass loss from litterbags, were used in the field to study decomposition and C and N mineralisation. Soil respiration, with the separation of added C from old soil C by using the isotopic ratio of ¹³C/¹²C in the respired CO₂, correlated well with extractable NH₄⁺ in the soil. Mineralisation was high and very rapid from residues of Sesbania sesban of high quality [e.g. low ratio of (polyphenol+lignin)/N] and low and slow from low quality residues of Grevillea robusta. Ten days after application, 37% and 8% of the added C had been respired from Sesbania and Grevillea, respectively. Apparently, as much as 70-90% of the added C was respired in 40 days from high quality green manure. Weight losses of around 80%, from high quality residues in litterbags, also indicate substantial C losses and that a build-up of SOM is unlikely. For immediate effects on soil fertility, application of high quality green manure may, however, be a viable management option. To achieve synchrony with crop demand, caution is needed in management as large amounts of N are mineralised within a few days after application.     

Effect of long-term fertiliser use on acid and alkaline phosphomonoesterase and phosphodiesterase activities in managed grassland

Soil phosphatase activities play an important role in the mineralisation of soil phosphorus (P). In this study acid and alkaline phosphomonoesterase and phosphodiesterase activities of soils under long-term fertiliser management (ca. 100 years) were measured to determine the effects of fertiliser inputs on the cycling and availability of P. Enzyme activities were compared with microbial biomass P, determined by fumigation-extraction, and with extractable P using NH₄F-HCl. Experimental plots were divided into three groups: those receiving farm-yard manure (FYM), those receiving mineral P and those receiving no P amendment. Plots receiving FYM had the highest extractable P values and the greatest enzyme activities. There was no obvious relationship between extractable P and microbial biomass P except in those plots where no P was added (r²=0.778), emphasising the importance of fertiliser management in P dynamics in soils. Acid phosphomonoesterase activity was high in all plots, including those where microbial biomass P levels were low. This supports the findings of previous studies suggesting that acid phosphomonoesterase activity in soils is primarily of root origin. All phosphatase enzyme activities were significantly correlated with extractable P in plots receiving mineral P. This relationship was negative for acid phosphomonoesterase activity (r²=-0.947), suggesting that acid phosphomonoesterase activity is suppressed by extractable P in managed grasslands receiving mineral P fertilisers.     

Influence of soil parameters on the effect of 3,4-dimethylpyrazole-phosphate as a nitrification inhibitor

Nitrification inhibitors specifically retard the oxidation of NH₄⁺ to NO₂^- during the nitrification process in soil. In this study, the influence of soil properties on the nitrification-inhibiting effect of 3,4-dimethylpyrazole-phosphate (DMPP), a newly developed nitrification inhibitor, has been investigated. Based on short-term incubation experiments, where the degradation of DMPP could be largely disregarded, the oxidation of the applied NH₄⁺ was more inhibited in sandy soils compared with loamy soils. The influence of soil parameters on the relative NO₂^- formation could be described by a multiple regression model including the sand fraction, soil H⁺ concentration and soil catalase activity (R²=0.62). Adsorption studies showed that the binding behaviour of DMPP was influenced markedly by soil textural properties, viz. the clay fraction (r²=0.61). The adsorption of DMPP was found to be an important factor for the inhibitory effect on NH₄⁺ oxidation in a short-term incubation (r²=0.57). It is concluded that the evaluated soil properties can be used to predict the short-term inhibitory effect of DMPP in different soils. The significance of these results for long-term experiments under laboratory and field conditions needs further investigation.     

Effects of harvester ant (<Emphasis Type="Italic">Messor</Emphasis> spp.) activity on soil properties and microbial communities in a Negev Desert ecosystem

Harvester ants (Messor spp.) function as an essential link between aboveground resources and below-ground biota such as the microbial community. We examined changes in soil microbial biomass and functional diversity resulting from harvester ant (Messor spp.) activity in the Negev Desert, Israel. Abiotic and biotic soil parameters were recorded during two seasons—wet and dry—also representing food availability periods for the ants (low and high seed availability, respectively). Soil samples were collected monthly from the 0- to 10- and 10- to 20-cm soil layers: (1) near the nest entrance, (2) under chaff piles, and (3) at a 2-m radius from the nest entrance (control). Harvester ant activity increased the percentage of organic matter, total soluble nitrogen, and microbial activity in nest-modified soils in comparison to the control soils. Higher CO₂ evolution was recorded in the low-seed season in ant nest soils than in the control soils. During the high-seed season, higher carbon dioxide evolution was recorded only at the nest entrance locations. There were no differences in microbial biomass between the low- and high-seed seasons, but highest microbial biomass was found under chaff in low-seed season and in nest soils in high-seed season. Microbial functional diversity was higher in nest-modified soils than in the control soils. This study suggests that the effect of harvester ant nests on soil fertility is due to increased microbial biomass and microbial activity in ant nest-modified soils.     

Do long-lived ants affect soil microbial communities?

This study was designed to test the hypothesis that desert ant species that build nests that remain viable at a particular point in space for more than a decade produce soil conditions that enhance microbial biomass and functional diversity. We studied the effects of a seed-harvester ant, Pogonomyrmex rugosus, and two generalist ant species, Aphaenogaster cockerelli and Myrmecocystus depilis, on soil microbial communities. Microbial biomass was higher in P. rugosus-modified soils than in reference soils when soil water content was higher than 3%. Microbial biomass was either higher in reference soils or exhibited no difference in reference soils and nest-modified soils of A. cockerelli and M. depilis. There were differences in microbial functional diversity and microbial community level physiological profiles (MicroResp method) between ant-nest-modified and reference soils of the three ant species on some sampling dates. Temporal patterns of soil microbial communities associated with the ant species resulted from differences in soil moisture, density, and species composition of the annual plant communities associated with the ant nests and in reference areas. Differences in annual plant communities associated with ant nests and surrounding areas resulted in different chemical inputs into the soil organic-matter pools. This study shows that generalizations about the effects of long-lived ant nests on soil biota in arid regions must consider feeding behaviors of the ant species and temporal patterns of rainfall.     

Carbon partitioning in plant and soil,carbon dioxide fluxes and enzyme activities as affected by cutting ryegrass

The effect of a single cut (simulated grazing) and regrowth of Lolium perenne on CO₂ efflux from soil (loamy Haplic Luvisol), on below-ground C translocation and on the distribution of plant C among different soil particle size fractions was investigated under controlled conditions with and without N fertilization by pulse labelling with ¹⁴C 7 times (four before and three after the cutting). The amount of ¹⁴C respired from the rhizosphere of Lolium decreased by a factor of about 3 during 1 month of growth. At the same time the amount of ¹⁴C stored in soil increased. Cut and non-fertilized plants respired less C in the rhizosphere compared to the uncut plants and cut fertilized plants. About 80% of the root-derived CO₂ efflux originated from the C assimilated after defoliation, and 20% originated from the C assimilated before cutting. N fertilization decreased the below-ground C losses (root respiration and exudation) during regrowth. The shoot is the main sink of assimilated C before and after the defoliation. N fertilization led to higher C incorporation into the shoot parts growing after defoliation compared to unfertilized plants. A lower incorporation of ¹⁴C was observed in the roots of N fertilized plants. The relative growth rates (expressed as ¹⁴C specific activity) of roots and stubble were minimal and that of shoot parts growing after defoliation was maximal. Twelve percent of ¹⁴C was found in the newly grown leaves after regrowth; nevertheless, 4.7% and 2.4% of ¹⁴C in the new shoot parts were translocated from the root and shoot reserves of unfertilized and fertilized plants, respectively. Most of the C retranslocated into the new Lolium leaves originates from the stubble and not from the roots. Between 0.5% and 1.7% of ¹⁴C recovered in shoots and below-ground C pools was found in the soil microbial biomass. Cutting and fertilization did not change ¹⁴C incorporation into the microbial biomass and did not affect xylanase, invertase, and protease activities. Tracing the assimilated C in particle size fractions revealed maximal incorporation for the sand and clay fraction.     

Cultivable heterotrophic N<Subscript>2</Subscript>-fixing bacterial diversity in rice fields in the Yangtze River Plain

Heterotrophic N₂-fixing bacteria are a potentially important source of N₂ fixation in rice fields due to the moist soil conditions. This study was conducted at eight sites along a geographic gradient of the Yangtze River Plain in central China. A nitrogen-free solid malate-sucrose medium was used to isolate heterotrophic N₂-fixing bacteria. Numbers of the culturable N₂-fixing bacteria expressed as CFU (colony forming units) ranged between 1.41ǂ.42᎒⁶ and 1.24ǂ.23᎒⁸ in the sampled paddy field sites along the plain. Thirty strains with high ARA (acetylene reduction activity) were isolated and purified; ARA of the strains varied from 0.9 to 537.8 nmol C₂H₄ culture^-1 h^-1, and amounts of ¹⁵N fixed ranged between 0.008 and 0.4866 mg·culture^-1·day^-1. According to morphological and biochemical characteristics, 14 strains were identified as the genus Bacillus, 2 as Burkholderia, 1 as Agrobacterium, 4 as Pseudomonas, 2 as Derxia, 1 as Alcaligenes, 1 as Aeromonas, 2 as Citrobacter, and 3 strains belonged to the corynebacter-form group.     

Effects of earthworms on Zn fractionation in soils

Laboratory incubation experiments were conducted to examine the effect of earthworm (Pheretima sp.) activity on soil pH, zinc (Zn) fractionation and N mineralization in three soils. No Zn uptake by earthworms was observed. Zinc addition decreased pH of red soil (soil 1) and hydragric paddy soil (soil 3) by 0.5 and 0.2 unit, respectively, but had no effect on alluvial soil (soil 2). The effect of Zn on soil pH was possibly due to a specific adsorption mechanism between Zn and oxides. Earthworm activity significantly decreased the pH of the red soil, a key factor affecting Zn solubility, but not of the other two soils. Earthworm activity significantly increased DTPA-Zn (DTPA-extractable) and OxFe-Zn (NH₂OH-HCl-extractable) in the red soil, but had little effect on other fractions. In the alluvial soil, earthworm activity significantly increased OxFe-Zn but decreased organic-Zn (organic-associated Zn). In the hydragric paddy soil, earthworm activity significantly increased MgCl₂-Zn (MgCl₂-extractable) and organic-Zn. The level of CaCl₂-extractable Zn in all three soils was not affected by earthworm activity. Nitrogen mineralized as a result of earthworm activity was equivalent to 110, 120 and 30 kg N ha^-1 in soils 1, 2 and 3, respectively. Zinc added at rates less than 400 mg Zn kg^-1 did not seem to affect the activity of N-mineralizing microorganisms. The present results indicated the possibility of increasing the metal bioavailability of relatively low level metal-contaminated soils, with a higher organic matter content, by earthworm inoculation.