Sensitivity of soil hydrogen uptake to natural and managed moisture dynamics in a semiarid urban ecosystem

Abstract

The North American Monsoon season (June-September) in the Sonoran Desert brings thunderstorms and heavy rainfall. These rains bring cooler temperature and account for roughly half of the annual precipitation making them important for biogeochemical processes. The intensity of the monsoon rains also increase flooding in urban areas and rely on green infrastructure (GI) stormwater management techniques such as water harvesting and urban rain gardens to capture runoff. The combination of increased water availability during the monsoon and water management provide a broad moisture regime for testing responses in microbial metabolism to natural and managed soil moisture pulses in drylands. Soil microbes rely on atmospheric hydrogen (H₂) as an important energy source in arid and semiarid landscapes with low soil moisture and carbon availability. Unlike mesic ecosystems, transient water availability in arid and semiarid ecosystems has been identified as a key limiting driver of microbe-mediated H₂ uptake. We measured soil H₂ uptake in rain gardens exposed to three commonly used water harvesting practices during the monsoon season in Tucson AZ, USA. In situ static chamber measurements were used to calculate H₂ uptake in each of the three water harvesting treatments passive (stormwater runoff), active (stored rooftop runoff), and greywater (used laundry water) compared to an unaltered control treatment to assess the effects of water management practices on soil microbial activity. In addition, soils were collected from each treatment and brought to the lab for an incubation experiment manipulating the soil moisture to three levels capturing the range observed from field samples. H₂ fluxes from all treatments ranged between -0.72 nmol m^-2 s^-1 and -3.98 nmol m^-2 s^-1 over the monsoon season. Soil H₂ uptake in the greywater treatment was on average 53% greater than the other treatments during pre-monsoon, suggesting that the increased frequency and availability of water in the greywater treatment resulted in higher H₂ uptake during the dry season. H₂ uptake was significantly correlated with soil moisture (r = -0.393, p = 0.001, df = 62) and temperature (r = 0.345, p = 0.005, df = 62). Our findings suggest that GI managed residential soils can maintain low levels of H₂ uptake during dry periods, unlike unmanaged systems. The more continuous H₂ uptake associated with GI may help reduce the impacts of drought on H₂ cycling in semiarid urban ecosystems.

Keywords: Aridlands; Biogeochemistry; Green infrastructure; Hydrogen fluxes; Microbial activity; Seasonal precipitation; Semiarid urban ecosystems; Soil hydrogen uptake; Water management.

Figure 1. Meteorological data measured hourly on site.

(A) Site level air temperature (°C) measured at 3 m aboveground. Light grey represents the hourly temperature. Dark grey line represents the smooth fit function from a generalized additive model estimation of air temperature. (B) Daily cumulative sum of precipitation (mm) at the site. Dashed red lines represent the static chamber hydrogen sampling dates.

Figure 2. The effects of green infrastructure management and time during the monsoon season on soil hydrogen fluxes (nmol H₂ m⁻² s⁻¹).

Boxplots displayed with point distributions for each batch within a treatment. The center line represents the median, and the lower and upper lines correspond to the first and third quartiles (25% and 75% quartiles). Whiskers correspond to the 95% confidence intervals. The two-way ANOVA presented F statistic, p-value and generalized eta squared (ges). The grey dashed line at 0 on the y-axis represents the transition between soil uptake (negative values) and emission into the atmosphere (positive values).

Figure 3. Microcosms assessing the effects of green infrastructure management and soil moisture on hydrogen fluxes (nmol H₂ m⁻² s⁻¹).

Boxplots displayed with point distributions for each batch within a treatment. The center line represents the median, and the lower and upper lines correspond to the first and third quartiles (25% and 75% quartiles). Whiskers correspond to the 95% confidence intervals. The one-way nonparametric Kruskal-Wallis ANOVA was performed independently for both treatment and moisture level on H₂. Corresponding p-values are recorded with treatment first and moisture level second. The grey dashed line at 0 on the y-axis represents the transition between soil uptake (negative values) and emission into the atmosphere (positive values).

Figure 4. Effect of selected environmental drivers during in situ measurements of soil hydrogen fluxes (nmol H₂ m⁻² s⁻¹).

(A) Positive correlation of hydrogen fluxes and soil temperature °C shows H₂ uptake into the soil is greater at lower temperature, and (B) negative correlation between hydrogen fluxes and soil moisture indicates that H₂ uptake increases with increased moisture. Boxplots display the median, 25th and 75th interquartile range, and whiskers show 1.5 times the interquartile for soil temperature (top left), soil moisture (top right), and soil hydrogen fluxes (right side) averaged during the in situ sampling period. Negative H₂ fluxes indicate uptake from the atmosphere into the soil.