Snow Algae Blooming Are Benetable for Microinvertebrates Assemblages (Tardigrada and Rotifera) on the Seasonal Snow Patches in Japan

Although studies on snow algae and macroinvertebrates have been frequently conducted on snow patches, no attention have been paid to ubiquitous microinvertebrates which in other cold habitats reach high biomass and play various trophic roles. The aim of this study was to search microinvertebrates in seasonal snow patches in Mt. Gassan, in northern Japan, and identify factors determining their distribution associated with snow algal blooming of various coloration (red, green, and yellow). Microscopic observation revealed presence of two major groups of microinvertebrates Tardigrada and Rotifera. Tardigrades and rotifers were the most abundant and frequent in green snow formed by blooming of Chloromonas sp., but few in red or yellow snow. Body length of tardigrades increased through the melting season and animals laid eggs on colored snow. These results suggest tardigrades successfully grew and reproduced on snow patches. Taking into account the presence of tardigrades and rotifers mostly in green snow (only few found in red and yellow) with high densities, we may assume green snow patches constitute important and unique low-temperature ecosystems for microinvertebrates in a temperate mountainous forest. Area of snow algae blooming worldwide are unrecognized novel habitat for tardigrades and rotifers.


Introduction
Seasonal snow are one of the most common and at the same ephemeral cold environments in the world 1 . Owing to accumulation during winter and melting during the spring, seasonal snow in uence adjacent ecosystems such as terrestrial, freshwater and marine ecosystems as a source of water or nutrients 2 . In some cases snow signi cantly affects whole ecosystems i.e. reduction of seasonal snow might lead soil freezing and damage for forest ecosystems 3 , or melting snow play important role in supporting nutrient input to glacial ecosystem and biological production there 4 .
Although snow patches are important element supporting the functioning of other ecosystems, seasonal snow provide also speci c environments for microorganisms. Despite the harsh conditions of snow surface like a low temperature and high UV irradiation, there are many taxa inhabiting unique snow ecosystems including primary producers (snow algae and cyanobacteria), microbial heterotrophs like fungi 5 , and consumers represented mostly by insects 6 . Activity of microorganisms in seasonal snow could in uence their melting by reduction of snow albedo, change air composition and effects on climate by releasing e.g. hydrocarbons, halocarbons, aldehydes, mercury and 7 . Therefore, each step in understanding snow ecosystems, regrettably one of the fast disappearing, are highly needed to reveal their productivity and biodiversity.
During the melt season, snow algae blooming change the color of snow surface into red, green, yellow or orange 8 . One of the common groups of snow algae are Chloromonas and Sanguina species, reported in glacial and snow habitat worldwide including Japan; Sanguina sp. is common on open snow elds while Chloromonas sp. often appears of forest oor [9][10][11][12] . Snow algae blooming has been shown to in uence melt rates of snow due to reduction of albedo on the snow surface, at the same providing liquid water for further biological processes 13,14 , and act as mediator in nutrient and carbon cycles 7,15 . Potential consumers of algae on snow patches have been also reported 5,6,16 , but without detailed ecological data explaining links between their densities, biomass, sources and seasonality. Although invertebrates are common apex consumers in snow ecosystems, knowledge on their diversity is related mostly to macroinvertebrates in high mountains 17,18 with limited data on microinvertebrates particularly tardigrades and rotifers which are one of the most common faunal groups in polar and high mountain ecosystems 19 . Data on tardigrades and rotifers densities, biomass, and factors determining their distribution on snow are still unknown.
Tardigrada and Rotifera are small cosmopolitan microscopic (mostly < 1 mm) invertebrates inhabiting virtually most of terrestrial and aquatic ecosystems, from tropical rain forests to glaciers in polar regions 19,20 . Until now, ca.1400 and 2000 species of tardigrades and rotifers respectively have been described worldwide 21,22 . Their feeding mode and behaviors (herbivorous, omnivorous or carnivorous; grazers, lter-feeders, predators) and reproduction style (sexual, asexual and parthenogenetic) differ depending on the species 23 . Owing to their ability of cryptobiosis which enables them to survive in extreme conditions such as high UV radiation doses, low and high temperature or salinity 24,25 , they can inhabit various environments worldwide including the coldest ones like cryoconite holes on glaciers 26 . Although there are many studies devoted to ecology and diversity of tardigrades and rotifers in high mountain or polar bryophytes and lichen habitats 27,28 , their ecology in cryosphere is still underscore 29,30 .
Tardigrades and rotifers in snow habitats, although important, are particularly overlooked or just anecdotally recorded 5 .
Owing to in uence of monsoon, a large amount of snow accumulates during the winter in mountainous areas in Japan. This snow remains until late spring, sometimes occur through the whole year and play a role of a low temperature habitats for temperate psychrophiles (e.g. stone y 31 ). In Northern Japanese snow patches, there were no reports on these microinvertebrates. Such limited data on microinvertebrates on snow may hinder our understanding of trophic networks and functioning of low temperature habitats. In this study, we aimed to search microinvertebrates in the seasonal snow patches and identify the factors determining their distribution on snow in forest in Northern Japanese mountainous area (Fig. 1,   2). Microinvertebrates were collected in late-melt season over two years (2018-2019), and their population density, their body size, biomass, diet and habitat preferences are presented and discussed.

Microinvertebrates on snow surface
We found two groups of microinvertebrates on snow, Tardigrada and Rotifera ( Fig. 3 and see supplementary Fig. S1). Springtails (Collembola) and stone ies have been seen on the snow during sampling but they were absent in the collected samples. We did not detect other invertebrates (e.g. nematodes, planaria) but in samples we observed abundant ciliates. Tardigrades were represented by two morpho-species belongs to genus Hypsibius with smooth or reticular cuticle (see supplementary Fig. S2). Both taxa have typical buccal tube for herbivorous species. Rotifers have not been identi ed to species level; all specimens represented order Bdelloidea (Fig. 3b). Tardigrades found on the snow represented different size classes (Fig. 4), including juveniles. Exuvia of tardigrade with eggs have been also found, each contained 4-5 eggs (see supplementary Fig. S3a). We also found different sizes of rotifers (not measured due to their shrinking) along with exuvia contained eggs (see supplementary Fig. S3b).
In moss samples, tardigrades representing genus Hypsibius and Macrobiotus along with heterotardigrades have been found. However, based only on morphology we could not fairly say whether Hypsibius sp. in mosses and snow is the same. However, specimens of Macrobiotus and heterotardigrades have been not found in any kind of snow during two seasons.
Body length of tardigrades living in green snow were varied during sampling campaigns (Fig. 4). Chlorophyll a concentration largely varied across the snow surface. In particular there were signi cant difference in the concentration between colored snow and white snow (Fig. 6, 7 and Table 1). In samples in April and May 2018, chlorophyll a concentration were and in green snow (April and May respectively), in red snow (May), in yellow snow (May) and in white snow (May). Whereas there were no differences in the chlorophyll a concentration between green snow and other colored snow (Mann-Whitney test: green and red snow, U (4, 3) = 6, P > 0.05; green and yellow snow, U (4, 3) = 5, P > 0.05), there were signi cant differences between white snow and other colored snow (Mann-Whitney test: green and white snow, U (4, 4) = 0, P < 0.05; red and white snow, U (3, 4) = 0, P < 0.05; yellow snow and white snow, U (3, 4) = 0, P < 0.05). In samples from May 2019, chlorophyll a concentration were in green snow and in white snow, and there was signi cant difference between green snow and white snow (Mann-Whitney test, U (9, 21) = 0, P < 0.001).
Relationship between microinvertebrates and snow algae Tardigrades and rotifers were concentrated mainly in green snow and their population density were different between different snow colors. In April and May 2018, population density of tardigrades were (mean SD), in green snow (April and May respectively), in red snow (May), in yellow snow (May) and 0 in white snow (May) ( Fig. 6 and Table 1). In May 2019, population density of tardigrades and rotifers were and in green snow, and 0 in white snow respectively ( Fig. 7 and Table 1). Population densities of tardigrades, rotifers and concentration of chlorophyll a were signi cantly correlated in green snow (Pearson's correlation coe cient: tardigrades and rotifers, r = 0.91, P < 0.01; tardigrades and chlorophyll a, r = 0.87, P < 0.01; rotifers and chlorophyll a, r = 0.84, P < 0.01). Regarding to tardigrades, whole algae cells were observed in their intestine as well as had a green intestine color ( Fig. 3a and supplementary Fig. S1).

Discussion
In this study, we revealed that tardigrades and rotifers were concentrated on seasonal snow patches in beech forests in Japan, particularly in green snow. Our observation suggests that snow algae blooming is truly bene table for tardigrades and rotifers where they have stable and abundant food source. According morphology, we found two tardigrade taxa represented by genus Hypsibius and bdelloid rotifers on snow (Fig. 3). The high densities of microinvertebrates, their presence through more than one seasons and various size classes (Fig. 4, 6, 7) suggests that tardigrades and rotifers living on the seasonal snow patches and adapt to snow surface environment. Although snow algae blooming serves a good habitat for tardigrades, not all taxa found in mosses have been found on snow. Under laboratory conditions, tardigrades represented by tardigrade genus Macrobiotus which was found in mosses from trees at the study site are well known to feed on algae 32 , but was absent in snow. We expect that some physiological or dispersal constrains might limits their appearance on seasonal snow patches. Most probably it is a temperature regime, which seems shape tardigrade communities on glaciers 26 . Regrettably, we cannot say clearly where Hypsibius species on snow came from, whether mosses, lichens growing on ground or in tree canopies.
Most probably, microinvertebrates on snow origin from passive dispersal by wind. Hypothetically, if they would be passively transported to snow from surrounding mosses, more taxa should be found during sampling campaigns than only two Hypsibius. This observation suggests that even some tardigrades might appear on snow, they are not adapted enough to grow and reproduce in these ecosystems. In fact, previous work conducted in Akashibo snow patches in Japan, reported tardigrades and other microinvertebrates (e.g. Nematoda, Oligochaeta) in snow layer under the snow patches, which probably migrated from soil, however these microinvertebrates did not seem to be active in snow such as growing and reproduction 33 . Although the harsh conditions (low temperature and periodic freezing) occurred on seasonal snow patches in Mt. Gassan, due to presence of snow algae, these ephemeral ecosystems supports growth and reproduction of microinvertebrates. Tardigrades and rotifers found in green snow seems to not only be cold lovers, but taking into account high biomass and densities they must bene ts from being on snow. From the above, a simple, hypothetical estimation of the microinvertebrates life cycle is as follow (I): dispersal to snow surface by wind from moss growing on tree trunks or canopies (April), (II): grow and reproduce by eating green snow algae (May), (III): passive transport with melting snow to mosses and spend their life in there (no snowy seasons).
There were no differences of concentration of chlorophyll a between colored snow in 2018 (Fig. 6), however, it seems that green snow form the most favorable conditions for growth and reproduction of tardigrades and rotifers in the study area. In green snow, exuvia of tardigrades and rotifers were found (supplementary Fig. S3), body length of tardigrades in May 2018 were bigger than in April 2018 (Fig. 4) which suggest successful growth. At the same time concentration of chlorophyll a in the samples also increased in time, which means that along with available food sources tardigrades increased their size and numbers. Although seasonal variability may result in growth of tardigrades body size, such effect has not been widely accepted.
Microinvertebrates were found with the highest population density in green snow, none or low in red, yellow and white snow (Fig. 6, 7 and Table 1). Tardigrades in green snow are typical algae feeders 32,34 and indeed their intestine were full of algae cells ( Fig. 3a and supplementary Fig. S1). At this moment it is hardly to say whether tardigrades and rotifers utilize the same food source on snow, or they use different food; tardigrades eat algae and rotifers might eat some of the associated suspension bacteria or yeasts.
Although, the same two dominant algae genera on snow were found in Canada 5 (Chloromonas spp. or Sanguina spp.). Authors presented tardigrades, rotifers, mites and springtails in red snow, and showed that both tardigrades and rotifers have the same red intestine content. Regrettably, authors did not presented densities of animals, hence we cannot compare data. In Mt. Gassan, it is known that snow algae which compose green and red snow are mix of several species of genus Chloromonas, yellow snow are Ochromonas species 10,11,35,36 . Even though green and red snow may consist of same genus Chloromonas, still they are different species, the difference in their color is probably due to the difference in life cycle of snow algae. In green snow, algal cells were mostly motile vegetative cells, which have chloroplasts without any secondary pigment, and were actively swimming using agella. They can asexually reproduce in the melting snow surface and form visible green snow. In contrast, algal cells were mostly cysts with ribbed and thick cell walls, they contained orange or red colored secondary carotenoids, which are probably astaxanthin. Such secondary pigments are usually produced when snow algae are stimulated by light, for minimizing the amount of light available 36,37 . Tardigrades belong to Hypsibius also found in snow are considered to eat algae by sucking 34 , so this thick cell walls make it di cult for tardigrades to suck algae. Moreover, tardigrades in laboratory cultures feed on Chlorella vulgaris, Chlorella sp., 38,39 and green algae on snow have similar size ( 2-10 µm for C. vulgaris 40 . It was mentioned that size of algae is not important for their dietary preferences 34 , by far we cannot say whether size of snow algae affect their dietary preferences. For yellow snow algae, their cell walls and size do not seem to interfere with the tardigrade's dietary so we can assume that tardigrades get something bene cial from green snow algae pigments not from yellow snow algae. However, these relations required exact testing in the future.
Due to simplicity of snow ecosystems, low organic matter content, few primary producers and few consumers, snow ecosystems may facilitate recognition and understanding of ecological processes in the cryosphere. In cryoconite holes which are cylindrical water lled reservoirs contained cyanobacteria, dust, organic matter and mineral particles on glacial ice 41,42 , although some study showed correlation between tardigrades and rotifers 30,43,44 , the signi cant relation between tardigrades or rotifers and any organisms as a typical source of food in cryoconite have not been found 29,43 . There are two potential scenarios which might explain this unrecognized pattern in cryoconite holes. In comparison with areas of snow algae blooming, cryoconite holes host very diverse biota and rich organic matter, thus diet of invertebrates might be complex which may constrain understanding of food preferences 45 . Secondly, dynamic nature of cryoconite holes in Arctic and alpine regions like rapid ablation, inter-hole watersediment mixing and stochastic weather events disturb or sometimes destroy habitats on surface of ice like cryoconite holes 30,46 , hence in dynamic habitats ndings of any ecological relations are problematic.
In the seasonal snow patches, meltwater is considered the only way causing removal of biotic components including microinvertebrates. In this case, snow surface was mostly at and during eld works we could not see meltwater owing on the surface. Therefore animals might migrate along with the movement of snow algae, and maintain stable until they are supported by snow (as a liquid water) and algae (as a food). Nevertheless, population density of microinvertebrates in the seasonal snow patches was lower than that of in the cryoconite holes with note that snow patches and cryoconite holes have different conditions of water (solid or liquid) so the densities cannot be simply compared; in Greenland, average was approx. (looks as same as snow but reached up approx. for tardigrades, approx. for rotifers 29 , in European Alps, approx. (reached up approx. ) for tardigrades 26 , in Svalbard, reached up approx. for tardigrades, approx. for rotifers 47 , in Antarctica, reached up approx. for tardigrades, approx.
for rotifers 43 . In other cold environments, tardigrades were found in mosses in Antarctica with higher population density and biomass (reached up to approx. ) than that of in the seasonal snow patches 48 . Rotifers were found in weathered snow/ice in accumulation zone with lower population density (reached up to ) than that of in the seasonal snow patches 49 .
The need for holistic studies on snow and biosphere including microbial activity which change the chemical composition of snow is well acknowledged 7 . Many taxa inhabit the seasonal snow 5,6,16 , however in many cases it is hard to infer which faunal elements are true snow ecosystems element and which randomly feed on snow algae. Herein we showed that tardigrades and rotifers are abundant where snow algae blooming occurred with preferences towards green snow (Chloromonas sp.) over the two sampling seasons in Japanese mountain forests. Intestine contents which contained algae cells, various instars and size classes of animals suggests that microinvertebrates are virtual faunal element in snow blooming while snow patches serve as a new arena for studies on biodiversity and ecology of these ubiquitous microinvertebrates. Further studies on the consumption rate by snow algae consumers may bring a new data on how invertebrates suppress algae blooming and how these algae respond on abundant invertebrates. These studies also have possibility to contribute to understanding material cycle such as carbon and nitrogen in snow which related to microbial activity on the snow surface 50,51 . While it was also mentioned carbon and nitrogen cycle in snow cover environments and emphasized impact of presence of seasonal snow cover for such cycle in soil ecosystem under the snow 52 , microbial activity in and on snow were not considered. Understanding the relationship between snow and biosphere including microinvertebrates is important in anticipation future changes and biogeocycles in snow ecosystems.

Study site
The eld study was conducted at Yumihiradaira park (3830' N, 14000' E: altitude 770 m) in Mt. Gassan, Yamagata prefecture in Japan (Fig. 1). The peak of Mt. Gassan is an altitude of 1984 m above sea level (a.s.l.) and the main mountainous ridge extends from the north to the south along the west coast of the northern part of the Honshu Islands of Japan. This mountain range directly receives the strong winds during winter, which is originally from Siberia and blow over the Sea of Japan. This strong wind contain abundant moisture supplied from the Tsushima Warm Current in the Sea of Japan, and induce heavy snowfall in the mountains 53 (Fig. 1). Snow depth were approx. 320 cm, 163 cm, 90 cm respectively. Total of 52 samples have been collected. The dimension of the collected snow surface was 10×10×2 cm (length×width×depth; Fig. 2). During the study periods, there were various colors of snow surfaces, and each colored snow appeared patchy in the size of 10 cm to 30 cm in diameter. We classi ed the snow surface in 4 different colorations, which are visibly identi ed as white, green, red, and yellow snows as described before 35 . The different color of the snow is most likely to be shaped by abundance and species composition of snow algae. The 4 different colors of snow were collected in the season of 2018, white and green snow were collected in the season of 2019. In order to compare microinvertebrate taxa with those in the closest snow free habitats, moss samples attached to beech trees at the study site were collected in 2018 (April = 4 samples, May = 4 samples). All snow samples were collected by using sterile stainless-steel scoop, and all moss samples were collected by using small shovel. Immediately after sampling, material were kept frozen and preserved in 50 ml polypropylene bottle (AS ONE CORPORATION, Osaka, Japan) or Whirl-Pak ® bags (Nasco, Fort Atkinson, WI, USA). All samples were transported to the laboratory at Chiba University, Japan and stored in a freezer (-20℃) until further processing.

Microscopic observation of microinvertebrates
Snow and moss samples were melted gradually at refrigerator (5℃) in order to avoid potential thermal shock of psychrophilic invertebrates. After melting, liquid water were moved directly to a petri dish and examined using a MZ125 stereo microscope (Leica Microsystems, Wetzlar, Germany) with FLEXACAM C1 digital camera (Leica Microsystems, Wetzlar, Germany). Moss samples were put a petri dish with adding MiliQ water and left 2-3 days for waiting on animals' recovery. After 2-3 days, moss were removed after shaken using tweezers. Invertebrates have been counted and put into 6 ml glass tube (AS ONE CORPORATION, Osaka, Japan) with 70% ethanol for preservation. Some individuals were used for taking picture using a BX51 phase contrast microscope (OLYMPUS, Tokyo, Japan) with DP21 digital camera (OLYMPUS, Tokyo, Japan). Population density of tardigrades and rotifers in snow were calculated based on counted number of tardigrades or rotifers per volume of melted water.
Tardigrades were mounted on microscope slides in a small drop of Hoyer's medium then examined under the following light microscopes BX51 and Olympus BX53 with phase-contrast (PCM) and another with differential interference contrast (DIC). Pictures were taken using a DP21 digital camera, cellSens Entry 1.12 software (Olympus, Tokyo, Japan) or Quick PHOTO CAMERA 3.0 software (Promicra, Czech Republic). For species identi cation the keys and species descriptions were used 54,55 to tardigrades.
Rotifers presumably have been represented by more than one species, but have not been identi ed. For calculation of biomass of tardigrades, we measured their body length and width using pictures taken under Olympus BX53 and measured in Image J and Quick PHOTO Camera 3.0 software. We excluded from our analysis those animals not suitably orientated for morphometry (shrunken or bent). Body length was measured without length of legs and width of the body between legs and was measured.
Biomass (dry weight, W) of each specimens were calculated based on formula 56 : if body length (L) and width (D) were 4:1; , 5:1; . We used these formulae that takes the closer ratio.

Microscopy of snow algae and measurements of the chlorophyll a concentration
Melted snow was used for microscopic observation of snow algae and measurements of the chlorophyll a concentration which expressed concentration of algae 57 . Firstly, put 20-50 µm melted water on slide glass and observed snow algae by taking picture using a BX51 phase contrast microscope with DP21 digital camera. Length and width of snow algae were measured in Image J. For measurements of the chlorophyll a concentration, Welschmeyer method which quantify with uorescence was used; melted water (approx. 90-150 mL per sample) were ltered through a glass micro lter (Whatman ® glass micro ber lters gf/f 25 mm, Cytiva, Tokyo, Japan), and put them into 8 ml sterilized tube (60.542.024, SARSTED, Nümbrecht, Germany) with 6 ml of N,N-dimethylformamide (DMF) to extract chlorophyll. After storing tubes in refrigerator for 2-3 days to wait extracting pigment, uorescence intensity were measured by using a Trilogy Laboratory Fluorometer (TURNER, CA, USA). Chlorophyll a concentration were obtained from uorescence intensity based on molecular extinction coe cient 58 (88.74 L -1 g -1 cm -1 ), and a calibration curve created by measuring uorescence intensity and absorbance (664 nm) of chlorophyll a from spinach (C5753-1MG, Sigma-Aldrich Japan, Tokyo, Japan) with a uorometer and a spectrophotometer (UV-mini 1240, Shimadzu, Kyoto, Japan).

Statistical analysis
To test differences between invertebrate biomass and densities, and chlorophyll a in white and colored snow and seasonal changes of tardigrade body size, two statistical tests were used. The nonparametric Mann-Whitney U test for testing differences of population densities of tardigrades and rotifers or concentration of chlorophyll a between white and colored snow, and the Welch's t-test for testing differences in seasonally changes of body length of tardigrades. All statistical analysis were calculated using R software 59 .