Variability in Subalpine Wetland Carbon Flux
The C flux varied greatly over the study period as the system quickly shifted between a C source and sink in response to changing environmental conditions following Snow Melt . Considering all seasonal phases, the study period was a cumulative net sink of 63 gCm-2 and was comparable to snow-free period values reported in other mountain ecosystems, such as sub-alpine meadows (Kato et al. 2003), and wetlands (Zhao et al. 2010; Millar, Cooper, Dwire, Hubbard, and von Fischer, 2017) studies. Average growing season NEE in subalpine alpine wetlands varies from strong sinks at the Qinghai Tibet Plateau (46 to 212 gCm-2) (Kato et al. 2003; Kato et al. 2006; Zhao et al. 2010), to strong sources and sinks (-342 to 256 gCm-2) in the Rocky Mountains of Wyoming and Colorado (Knowles et al. 2015; Millar et al. 2017). Although alpine systems show varying results in ecosystem source/sink strength, winter C flux can cause alpine meadow and wetland ecosystems to be a strong C sources, due to the insulating properties of snow cover that keeps the ground surface unfrozen and soil respiration (Rs) active (Zhao et al. 2010; Knowles et al. 2015; Lange, Allaire, Castillo, and Dutille, 2016). In our study, snowpack kept soil temperature (2, 5, and 10 cm) above -5 °C during Snow Melt, meeting the threshold required for Rs to occur (Brooks, Schmidt, and Williams, 1997). This indicated that Reco was representative of Rs during the Snowmelt period, when vegetation was not productive. However, it was most likely that the warming period in Snowmelt promoted degassing from melted ice lenses in the snowpack (Lange et al. 2016), causing “burps” in Recoover time. Therefore, Rs likely continued over the winter months at Bonsai, possibly making it a small annual net source of CO2; however, an annual multi-year investigation will be required to quantify its average yearly net C contributions and confirm or refute this hypothesis.
The measurement period began with “pulses” of carbon emission duringSnowmelt because of degassing air pores in surface ice lenses that melted during rainfall and warm temperatures (Lange et al. 2016). Because of a lengthy melt process from shading (Hrach et al. 2021), theGreen Up period lasted longer at our study site compared to those in other alpine ecosystems. Green Up represented the increase in net seasonal CO2 flux, which led to peak production (maximum GPP and Reco) during the Peak Growing Season . The 2018 Green Up at Bonsai lasted from 24 June to 20 July, ending several weeks later than Green Up at similar ecosystems reported in literature (Kato et al. 2003; Zhao et al. 2010; Knowles et al. 2015), which began Green Up in June and ended by early July. Despite these subalpine wetlands being located at lower latitude (35-40N) compared to Bonsai (~50N) and experiencing fewer daylight hours due to less solar input in June and July, they greened up faster in comparison. Our results suggest that the microclimate resulting from horizon shade, extended the Green Upand shortened the Peak Growing Season periods at our site by approximately two weeks.
We found that in general, during the period of increasing shade, the C flux had a statistically significant relationship with horizon shade. During the period of Dynamic Shade , GPP and Recohad a statistically significant negative relationship with shade, while NEE had a statistically significant positive relationship with shade. Each hourly increase in shade per day during Dynamic Shade , decreased GPP by 15% and increased NEE by 18% (p<0.001, R2=0.65), indicating that shade negatively impacted C uptake at our subalpine wetland. Our finding found agreement with Cao et al. (2017), which concluded that NEE has a statistically significant negative linear relationship with Rg (p<0.05, R2=0.27). However, what we also found is that shade potentially increases or mitigates the overall reduction in available Rg at our site, by allowing plants to utilize more diffuse radiation during shaded periods (similar to cloudy conditions), thus increasing the overall photosynthetic rate of the ecosystem which would not be observed in direct light. However, further research on this topic is required at our study site.
Beyond shade and Rg, the literature has identified that other environmental variables like VMC (Kato et al. 2003), Ts(Kato et al. 2003; Zhao et al. 2010; Cao et al. 2017) , and Ta (Zhao et al. 2010; Cao et al. 2017) play a strong role in regulating the C flux. Past studies in mountain meadow (Knowles et al. 2015), wetland (Millar et al. 2017), and forest ecosystems (Monson et al. 2002) identified temperature and moisture as key controls of seasonal patterns in NEE. Our wetland was a net source in Snow Melt and early Green Up, but a sink at the end of Green Up and throughout Peak Growing Season. NEE increased alongside Ts through Green Up and Peak Growing Season , and quickly responded to an abnormal spike in Ta and Ts on September 7th in Late Growing Season. However, GPP at our site during Peak Growing Season was not related to any environmental variable, other than Rg.