UC Berkeley

In this dissertation I use gene expression by indigenous soil microbial communities to better understand microbial resuscitation, the biogeochemical process of nitrification, and the consequent ramifications to global nitrous oxide emissions. The first rainfall following a long dry period in arid and semi-arid ecosystems causes a large, abrupt change in water potential that can be both a severe physiological stress and a punctuated stimulus for the reawakening of soil microbial communities rendered inactive by low-water conditions. Using two California annual grassland soils collected following a typically dry Mediterranean summer, I simulated a wet-up comparable to the season's first rainfall. I extracted nucleic acids and monitored soil variables before and over a logarithmic time scale from 15 minutes through 72 hours after water addition. I looked at this experiment through three different lenses to address pressing questions in microbial ecology, soil nutrient cycling, and trace gas biogeochemistry.

To assess microbial resuscitation strategies to the wet-up event, I applied transformed RNA to a high-density (16S rRNA) microarray (PhyloChip). I identified three response strategies, rapid (within 1 hour of wet-up), intermediate (between 3 and 24 hours following wet-up), and delayed (24 to 72 hours post wet-up) and note that the taxa comprising these groups cluster phylogenetically and are relatively consistent between the two soils analyzed.

I then addressed the relationship between microbial functional gene transcription and activity. I followed the transcriptional response of functional genes in three groups of nitrifiers; ammonia-oxidizing bacteria, ammonia-oxidizing archaea, and nitrite-oxidizing bacteria of Nitrobacter spp. By comparing transcript abundances with soil ammonium and nitrate pools over the course of wet-up, I found strong correlations between the soil ammonium pool and functional gene transcripts of ammonia-oxidizing bacteria (amoA); induction of amoA continued until the rate of ammonia oxidation was greater than the resupply of ammonium, suggesting that transcription serves as a control point for ammonia oxidation in soil. I again measured similar time to response in both soils, this time with microorganisms grouped by metabolic function, with ammonia-oxidizing bacteria responding first. In addition, I again discovered that microorganisms traditionally thought to be slow-growing are capable of fast response, with transcriptional response detectable within 9 hours for the slowest group, ammonia-oxidizing archaea.

Finally, I related the lag in response of nitrite-oxidizers relative to ammonia-oxidizing bacteria to a significant nitrous oxide pulse and evaluated the implications of this decoupling to global nitrous oxide emissions. I reviewed the literature in light of this observation and discovered that increases in ammonia oxidation could explain the variability in and pulsed-behavior of nitrous oxide emissions reported from unsaturated soils.