DISCUSSION

Warming reduces dew formation and changes its seasonal patterns

Using three distinct measurement methods, our study showed that warming significantly reduces the dew amount (Fig. 1), which may have substantial impacts on plant growth, especially during the dry period in the alpine and dryland ecosystems (Stone 1957, Benasher et al. 2010, Rodney et al. 2013). Warming can reduce dew formation in two ways: by hindering dew condensation and shortening dew retention. Warming can hinder the dew condensation processes by decreasing the air humidity and increasing evaporation (Oliveira 2013, Li et al. 2018). Additionally, warming changes the air temperature, dew point temperature and dew point depression (Fig. 2b), which makes it more difficult for the air temperature to approach the dew point temperature (Beysens 1995, Jacobs et al. 2006). Warming can also accelerate the dew evaporation process (Xiao et al, 2013). Dew droplets lasted for a shorter period of time under warmer temperatures, which also led to a lower dew duration or amount (Xu and Tang 2015).
In this study, we found that warming not only reduces dew formation but also changes its seasonal variation (Fig. 2). Therefore, plants growing under water stress would have higher risks of not surviving under warming conditions (Rodney et al. 2013, Tomaszkiewicz et al. 2017). Overall, under the rapidly changing climate, changes in dew formation should be considered an important environmental factor and should not be neglected in arid and cold regions.

Dew formation varied among different functional groups under ambient and warming conditions

Functional groups create different microenvironments and have different water use strategies to influence the dew production and absorption by plants (Zhuang & Zhao 2017, Wang et al. 2019). Environmental conditions, such as temperature, relative humidity and wind speed, change due to various micromorphological features and distribution patterns among different functional groups (Agam & Berliner 2006), affecting dew formation and duration (Ninari & Berliner 2002). Our results showed that different functional groups had different degrees of dew formation, consistent with our expectations.
To date, few studies have investigated how biotic factors (e.g., plant traits and functional groups) affect dew formation. Here, we examined the effects of plant traits (i.e., plant height and aboveground biomass) on dew formation in different plant functional groups (sedges, forbs and grasses) and found that sedges and forbs with shorter heights are associated with less dew than grasses with taller heights under natural conditions (Fig. 3). Because under ambient conditions, the upper canopy air temperature is lower at night due to this area receiving less land-surface radiation, dew formation occurs earlier in higher leaves, such as those of grasses (Zhang et al. 2009, Wang et al. 2017). In addition, the dominant taller species (Stipa aliena , Elymus nutans , and Helictotrichon tibeticum ) usually have more aboveground biomass (Konrad et al. 2015, Ma et al. 2017) than shorter species, which can facilitate dew formation and retention (Pan et al. 2016). Additionally, the dew water stored within a dense canopy can be preserved for a longer period of time through the reduction in evaporation (Xiao et al, 2013).
Under warming conditions, the aboveground biomass and plant height increased, and the community composition changed with a higher prevalence of grass in the alpine ecosystems (Liu et al. 2018). Such changes should be beneficial for dew formation based on our findings under ambient conditions (i.e., results from the control plots, Fig. 4a). However, a substantial reduction in dew formation was observed under the warming treatments (Fig. 1 and Fig. 2). In addition, we found that warming resulted in a lower dew amount on taller plants, in contrast to the results under ambient conditions (Fig. 4). Warming changed the relationship between plant height and dew amount in both direct and indirect ways. Warming directly affected the air temperature profile and made dew formation more difficult (Wolkovich et al. 2012). In this case, the taller plants had less dew formation because artificial infrared heating made the temperature of the taller canopy higher than that of the lower canopy (Xiao et al, 2013). Warming indirectly caused the soil moisture to evaporate more quickly during the night (Tomaszkiewicz et al. 2015, Li et al. 2018). Therefore, the shorter plants experienced more dew collection than the higher plants during the night under warming conditions. Clearly, warming influenced the dew formation on plants and changed the ecosystem processes compared with those under natural conditions.

Infrared heater warming system reduces dew formation: an overlooked factor in climate change studies

There have been many studies about the response of ecosystem processes to climate change using various artificial warming methods in dry ecosystems (Kimball et al. 2018, Song et al. 2019, Korell et al. 2019), but the possible impacts from the differences between artificial and natural warming on the experimental results have often been overlooked. Our results showed that artificial warming (with an infrared heater warming system) affects dew formation, which likely affects ecosystem processes (Liu et al. 2016). However, it is worth noting that natural climate warming and the infrared heater warming system differ in terms of their heat-dissipating pathways (Korell et al. 2019). Artificial warming generates more heat radiation in the air and drier micro-environments than natural warming (Liu et al. 2018). This difference will affect a number of ecosystem processes and is often overlooked across simulated climate change experiments. Warming makes plants grow taller (Liu et al. 2018), but taller plants produced less dew under warming in our study (Fig. 3). This indicates that the dew formation was significantly reduced under the experimental warming conditions. In addition, the relationship between dew formation and plant height changed being positively correlated under the control treatment to being negatively correlated under the warming treatment (Fig. 4). For such cases, the conclusions of the impacts of warming obtained by artificial warming experiments may deviate from the actual impacts of warming on ecosystem processes. Under future climate warming, the changes in water condensation will also have an especially profound impact on the ecosystem patterns and processes in dryland ecosystems (Li et al. 2018, Wang et al. 2017). Therefore, we suggest that the impact of experimental warming on dew formation should be considered an important environmental factor affecting ecosystems processes during climate warming.