The toughest animals of the Earth versus global warming: Effects of long‐term experimental warming on tardigrade community structure of a temperate deciduous forest

Abstract Understanding how different taxa respond to global warming is essential for predicting future changes and elaborating strategies to buffer them. Tardigrades are well known for their ability to survive environmental stressors, such as drying and freezing, by undergoing cryptobiosis and rapidly recovering their metabolic function after stressors cease. Determining the extent to which animals that undergo cryptobiosis are affected by environmental warming will help to understand the real magnitude climate change will have on these organisms. Here, we report on the responses of tardigrades within a five‐year‐long, field‐based artificial warming experiment, which consisted of 12 open‐top chambers heated to simulate the projected effects of global warming (ranging from 0 to 5.5°C above ambient temperature) in a temperate deciduous forest of North Carolina (USA). To elucidate the effects of warming on the tardigrade community inhabiting the soil litter, three community diversity indices (abundance, species richness, and Shannon diversity) and the abundance of the three most abundant species (Diphascon pingue, Adropion scoticum, and Mesobiotus sp.) were determined. Their relationships with air temperature, soil moisture, and the interaction between air temperature and soil moisture were tested using Bayesian generalized linear mixed models. Despite observed negative effects of warming on other ground invertebrates in previous studies at this site, long‐term warming did not affect the abundance, richness, or diversity of tardigrades in this experiment. These results are in line with previous experimental studies, indicating that tardigrades may not be directly affected by ongoing global warming, possibly due to their thermotolerance and cryptobiotic abilities to avoid negative effects of stressful temperatures, and the buffering effect on temperature of the soil litter substrate.


| INTRODUC TI ON
Soil organisms such as nematodes, tardigrades, and rotifers-a microfauna that needs a film of water surrounding the body to be activeare generally abundant in most terrestrial biomes. They are relatively poorly studied, especially regarding their response to ongoing climatic changes. Most of the studies concerning the effect of experimental warming on this type of fauna are mainly focused on Antarctic nematode communities (e.g. Andriuzzi et al., 2018;Convey & Wynn-Williams, 2002;Newsham et al., 2020;Prather et al., 2019;Simmons et al., 2009), and contrasting results on their community have been found according to species and temperature increase (see Section 4). On one hand, these organisms might be particularly susceptible to changes in climate because their surface area relative to volume is great. On the other hand, some of these organisms have life stages that could buffer them from climatic changes. For example, tardigrades (water bears; Figure 1a) have the ability to enter ametabolic physiological states in all phases of their life cycle, allowing them to survive harsh environmental stressors such as desiccation (anhydrobiosis) and freezing (cryobiosis) (Guidetti et al., 2011). The ability of tardigrades to tolerate chemical and physical extremes in an anhydrobiotic state is so extraordinary (Rebecchi et al., 2007) that tardigrades have been called the "toughest animals on the Earth" (Copley, 1999).
Previous studies on the impacts of climate changes on tardigrade communities have found somewhat surprising results: no effect of increasing temperature on tardigrade communities has been detected (Briones et al., 1997;Sohlenius & Boström, 1999a). To better understand the effects of increasing air temperature on tardigrade communities of soil litter, we sampled a tardigrade community within an extensive set of experimental warming arrays in the eastern United States, performed in the Duke Forest in the Piedmont region of North Carolina (Pelini et al., 2011). Experimental warming has the potential to analyze the response of the entire tardigrade community compared to transplant experiments because the warming is performed in situ without the need to move the animals. Moreover, the use of open-top chambers minimizes the substrate disturbance and allows for longterm, consistent warming over a sustained period (Pelini et al., 2011).
Previous studies in the experimental warming array used in this study (Figure 1b) found that warming had differential effects on ants, other ground-living arthropods, and microbial communities, with some species benefiting from warming and others declining. In particular, ants have species-specific responses to temperature increase with some species benefiting from warming and others declining probably linked to physiological traits that differ between species (Diamond et al., 2013(Diamond et al., , 2016MacLean et al., 2017;Pelini et al., 2014;Penick et al., 2017), and the same is true for other ground-living arthropods (Fitzgerald et al., 2021), while the effect of warming on microbial community structure and function may become more pronounced as soil temperatures increase and carbon substrates are depleted through time at the Duke Forest site but were not affected at a more northern site where this experiment was replicated (Cregger et al., 2014).
The experimental design allowed the opportunity to test for the interactive effects between air warming and other environmental variables, such as soil moisture. We hypothesized two possible outcomes of this experiment: (a) we predicted that the tardigrade community (or part of it) would be influenced by both increasing temperature and soil moisture with regard to abundance (positive correlation between increased temperature and moisture on tardigrade abundances) and community composition (change in dominant species and/or diversity); (b) alternatively, we predicted that the tardigrade community (or part of it) will show no change in responses to the warming treatment, because they escape negative impacts of warming due to anhydrobiosis (avoiding period of substrate drying) or due to an extreme tolerance of high temperatures (see Giovannini et al., 2018;Li & Wang, 2005;Neves et al., 2020;Rebecchi, Boschini, et al., 2009). Those predictions come from the general trend observed for the effect of artificial warming on other terrestrial organisms belonging to the hydrobios, as the soil nematodes, that share the same habitat with tardigrades (Bakonyi & Nagy, 2000;Hiltpold et al., 2017;Yan et al., 2017).

| Experimental design and warming chambers
The experimental facilities used in this study are described in  Modena (Italy). The leaf litter was kept desiccated for 2 years until tardigrade extraction. Before animal extraction, desiccated leaf litter was kept at 45% relative humidity (RH) at 20°C for 24 hr, and then, 10 g of each sample (pseudoreplicate) for each chamber were rehydrated in tap water for 30 min. We then sieved the leaf litter and water using two sieves of different mesh sizes: 500 μm and 37 μm.

| Leaf litter sampling and tardigrade extraction
The leaf litter left in the 500 μm sieve was resubmerged in water and sieved a second time after 2 hr to be sure to extract all tardigrades and their eggs. The debris collected from the 37 μm sieve was resuspended in distilled water and tardigrades along with their eggs were collected manually with a glass pipette under a stereomicroscope.

| Species identification and data analysis
The tardigrades were fixed in Carnoy's fixative (¾ methanol, ¼ acetic acid) and mounted on microscopic slides in Faure's mounting medium. Animal observations were carried out with light microscopy (LM) under phase contrast (PhC) and differential interference contrast (DIC) up to the maximum magnification (100× oil objective) with a Leica DM RB microscope, at the Laboratory of Evolutionary Zoology (UNIMORE). Eggs were not included in the analyses due to their low number. Tardigrade classification was mainly done by following Ramazzotti and Maucci (1983) and Bingemer and Hohberg (2017).
Tardigrades are very sensitive to the moisture of the substrates being aquatic animals (although able to live in terrestrial substrates); soil moisture is one of the main environmental factors that could be effected by the air temperature increase. For these reasons, in addition to chamber air temperature, soil moisture was chosen as an additional variable. In this specific experimental setup, soil temperature F I G U R E 2 (a) Average air temperature by month over 5 years of the experimental chambers. (b) Average soil moisture by month over 5 years of the experimental chambers is correlated with air temperature (Cregger et al., 2014) but soil moisture is not correlated with the air temperature measures (Burt et al., 2014;Figure 2b). We chose to present the analysis done with air temperature as independent variable instead of soil temperature because within the experimental setup, air temperature was the directly manipulated variable and soil temperature was then causally dependent on it. We also performed the analysis with organic soil temperature instead of air temperature, and the results were very similar (analysis with soil temperature is available as Appendices S1 and S2). Average air temperature and soil moisture were calculated from the raw data file in Ellison and Dunn (2017); for statistical analysis, the 5-years average of each chamber was used.
Three community indices considering all tardigrade species  Cohen (1988). All of the analyses performed with their code are reported in Appendix S1.

| RE SULTS
Each of the 24 samples of leaf litter contained tardigrades. A total of 1,762 tardigrades, belonging to 13 taxa, were collected, for a mean of 14.6 (min 6, max 40) tardigrades/g of dry leaf litter (Table 1) Bayesian p-values and effect sizes (Table 2)  On the other hand, in another short-term transplant study, Briones et al. (1997) found that tardigrade population oscillations were positively correlated with temperature, though only in the colder sites.
Our findings come from a single tardigrade community from a single location; they may indicate that tardigrades are resilient to the increases in temperature. The comparisons of our results with those obtained with nematodes, with which share similar size, anhydrobiotic capabilities, and often existing in similar substrates, can be difficult due to the abovementioned reasons. Anyway, in an experiment conducted in a similar habitat (i.e., temperate forest), the effects of 2 years increased soil temperature recorded on nematode community were similar to those found in our experiment: no effect on total or trophic group abundances (Thakur et al., 2014).
We have three hypotheses, that are not necessarily mutually exclusive, that could explain the lack of response of tardigrades to extreme experimental warming. First, the tardigrades at the study site may have unusually high thermal optima and maxima, such that they thrive even as conditions warm. Even tardigrade species living in cold regions appear to have relatively high thermal limits. For example, the Antarctic species Acutuncus antarcticus (Richters, 1904) can withstand short heat shocks up to 33°C (Giovannini et al., 2018).
Similarly, the boreo-alpine tardigrade species Borealibius zetlandicus (Murray, 1907) showed a lethal temperature (LT50) of 33°C (Rebecchi, Boschini, et al., 2009) (Table 2). Leetham et al. (1982) and Hohberg (2006) also found a high intrasite variance in the number of tardigrades in soil habitats, and a similar level of patchiness has been shown for other substrates as well (Meyer, 2006;Sohlenius et al., 2004;Tilbert et al., 2019). In our experimental setup, soil moisture was not correlated with air temperature and was instead correlated with air relative humidity (Burt et al., 2014). The difference between soil moisture and air humidity is likely to create a moisture gradient in the leaf litter, generating microhabitats with different moisture levels. The potential indirect effect of warming (moisture change) could then be buffered by tardigrades moving to or living in different leaf litter areas/layers with a more suitable moisture, as tardigrades have been shown to be able to differentially colonize different soil and leaf litter layers according to the seasons (Briones et al., 1997;Guidetti et al., 1999, respectively).
Third, the ability of tardigrades to enter anhydrobiosis (Guidetti et al., 2011). The thermal tolerance of tardigrades in their desiccated state is known to be higher than when they are in their active state (Neves et al., 2020;. It is possible that tardigrades escaped the negative effects of high temperatures in our study, especially during summer; because they were in a desiccated state of anhydrobiosis, they could compensate the larger amount of time spent in anhydrobiosis in summer by taking advantage of a reduced frozen period in winter. During summer when temperatures peak in temperate environments, water availability is usually reduced in substrates (leaf litter, mosses, and lichens) inhabited by tardigrades, which dry out quickly. It is thus possible that, in our study site, tardigrades are most active in cooler times of year and spend the hottest time of the year in a state of anhydrobiosis, a state in which their thermal tolerance is increased (Neves et al., 2020). The co-occurrence of warmer temperatures with dry conditions could be an additional factor that increases tardigrade survival to warming in specific climatic areas.
The experimental data obtained in this work represent a small light to illuminate the effects of climate changes in the dark world of soil microfauna. It highlights that tardigrades in soil litter are part of those tolerant species able to survive the increase in air temperature due to global warming, and this may be related to their ability to enter anhydrobiosis or a higher tolerance of thermal extremes.

ACK N OWLED G M ENTS
We thank the anonymous referees and the editor for their comments that allowed us to improve the manuscript.

CO N FLI C T O F I NTE R E S T S
The authors have no competing interests to declare.