Abstract:
This project used an electronic nose (E-nose) system composed of an array of 14 nonspecific
conducting polymer sensors for soil and water diagnostics, based on
qualitative microbial volatile production patterns. It tested the feasibility of using soil
microbial volatile fingerprints for detecting and monitoring changes in microbial
activity in three soils, as a response to key environmental factors such as temperature
(16, 25, 37°C), water potential (-0.7, -2.8 MPa), and nutrient (glucose and wheat
straw) inputs. It also investigated their potential use for atrazine detection when
applied to soil at usual field application rates (2.5 ppm) as well as for monitoring its
bioremediation using the white-rot fungus Trametes versicolor (R26), for up to 24
weeks. Furthermore, statistical correlations were investigated between soil volatile
profiles and traditional microbial parameters for characterising microbial communities
and their metabolic activities such as respiration, dehydrogenase (DHA) and laccase
(LAC) activities, bacterial and fungal colony counts and fungal community structure
under different soil conditions. Finally, this study explored the potential of microbial
volatile production patterns for monitoring the activity and differentiation of two
Streptomyces species (S. aureofaciens A253 and S. griseus A26) in potable water and
in soil, as well as the production of geosmin in both environments.
Data in this research has demonstrated that the production of volatile organic
compounds (VOC) in soil is likely to arise from microbial metabolism. The E-nose
was able to detect variations in the patterns of volatile production from soil according
to treatments, functioning as indicators of shifts in microbial activity and community
structure. The potential for discrimination between soil types in relation to
environmental factors and nutrient addition has been demonstrated for the first time
using principle component analysis (PCA). Significant (p<0.05) correlations were
also found between soil volatile patterns (through PC1) and traditional soil microbial
parameters. The close relationship (r>0.80) between PC1 and soil respiration was
particularly relevant, since it indicates that microbial volatile fingerprints, similarly to
respiration, respond quickly to changes in soil conditions. The sensor array was also
able to detect Streptomyces activity and differentiation as well as discriminate between bacterial species at different concentrations in potable water and in soil.
Using this approach, the presence of geosmin was detected in water at 0.5 ppb (below
its human odour threshold detection, OTD) and in soil at 100 ppb (OTD not
established).
This study has, therefore, demonstrated that an E-nose can be employed as a rapid,
sensitive, reproducible and non-invasive tool for characterising changes in soil
environmental conditions, as well as for monitoring key soil processes such as organic
matter decomposition and atrazine degradation. It also suggests that this approach can
complement, and perhaps replace, some of these methods for a quick and routine
evaluation of the impact of environmental factors on soil microbial communities.
Furthermore, this study showed that an E-nose can also be employed for assessing
Streptomyces activity and detecting geosmin production at an early stage in water and
soil.