Biological soil crusts (hereafter, biocrusts) are communities of microorganisms that regulate key ecosystem processes such as water distribution, soil erosion, and nutrient cycling in drylands worldwide. The nature of biocrust function can be influenced by multiple environmental factors, including climatic conditions (e.g., precipitation), interactions with plants, and anthropogenic disturbances. Animal regulation of biocrust function has received less research attention, focusing primarily on livestock trampling and to a much lesser extent on biocrust consumption by mesofauna. Deposition of animal waste products, carcasses, and other body secretions such as mucus may also affect biocrust function. Yet, this novel regulatory pathway, to our knowledge, has never been empirically tested. Our goal was to begin bridging this knowledge gap by exploring how snail mucus affects biocrust CO₂ efflux— using two distinct biocrust communities and three snail species. We found that snail mucus increased the CO₂ efflux of both cyanobacteria- dominated and lichen/moss- dominated biocrusts. However, the magnitude of snail mucus effects on biocrust CO₂ efflux varied between snail species— possibly due to species-level differences in snail diet. Our study highlights a novel interaction between animals and biocrusts and suggests that even small quantities of animal-derived nutrients can have important consequences for biocrust carbon dynamics.

Manuscript Highlights

1. Mucus increased the CO₂ efflux of cyanobacteria-dominated biocrusts by >20%.
2. Mucus enhanced the CO₂ efflux of moss/lichen-dominated biocrusts by > 86%.
3. Dietary differences likely underlie species-specific effects of mucus on biocrusts.

Study system

We focused our investigation on two LTER sites in the Negev desert that differ in climatic conditions and biocrust composition. The Avdat Research station (30°47’02” N, 34°46’09” E; hereafter, Avdat) is in the central Negev highland and receives ~93 mm of rainfall per year spread across 19-42 days (Israel Meteorological Survey, 2021; Station: 253052). Avdat is characterized by 1-2 mm thick cyanobacteria dominated biocrusts (also containing bacteria, algae, and fungi) that have a distinct flat, solid surface. This solid surface is due to cyanobacteria secreting polysaccharides that adhere soil particles (Mazor and others 1996). Sayeret Shaked Park (31°16’16” N, 34°39’03” E; hereafter, Shaked Park) is in the Northern Negev and receives 190 mm of rainfall per year spread across 25-52 days (Israel Meteorological Survey, 2021; Station: 251691). Shaked Park is characterized by moss/lichen dominated biocrust that also contain bacteria, cyanobacteria, algae, and fungi. The moss/lichen dominated biocrusts at Shaked Park reach thicknesses of 8-10 mm on south-facing slopes and 10-15 mm on north-facing slopes (Zaady and others 1996; Zaady and others 1998).

At Avdat and Shaked Park, snails are relatively abundant— reaching densities of 12 and 89 snails per m², respectively. Avdat has multiple common snail species, including Sphincterochila zonata (SZ), Xerocrassa simulata (XS), and Sphincterochila prophetarum (SP) that overlap in their distribution and use of biocrust surfaces (Bar, 1975; Degen and others 1992; Genot-Lahav, 1986). Meanwhile, Shaked Park is mainly dominated by XS, but also has a small population of SZ. The diets of all three snail species include biocrusts (Yom-Tov and Galun 1971; Shachak and Steinberger 1980; Shachak and Brand 1981). However, XS preferably consumes plant litter (Yom-Tov and Galun 1971).

The activity of snail species in the Negev is regulated by moisture, with all species only becoming active when the ground is damp (Yom-Tov, 1971; Hermony and others 1992). Consequently, snail movement along biocrusts in search of food, mates, and egg laying habitat is limited to short bouts following substantial precipitation events  (Shachak and Steinberger 1980). For example, S. zonata are active for only 8-27 days annually (Shachak and Steinberger 1980). These short bouts of snail activity likely have important consequences for biocrust function, as snails deposit considerable nutrient-rich mucus while moving along the biocrust surface— covering approximately 15% of the biocrust surface in mucus trails (Fig. 1).

General experimental approach

We used three complementary laboratory experiments to reveal how desert snails affect biocrust activity. To achieve a comprehensive answer, we used two different biocrust types, and three species of snails.  We also used two common biocrust cultivation methods to control for variation in biocrust performances that may reflect specific rearing conditions (Doherty and others 2015).  In experiment 1, we explored how a mix of mucus from three abundant snail species affects laboratory grown cyanobacteria-dominated biocrusts from Avdat. In experiment 2 we tested how X. simulata (XS) mucus from the Avdat and Shaked Park populations affect field collected moss/lichen dominated biocrust from Shaked Park. In experiment 3 we assessed how the mucus of each of the three common snail species at Avdat affect field collected cyanobacteria dominated biocrusts from Avdat. In all three experiments we measured the CO₂ efflux as a measure of biocrust activity.

Experiment 1

We collected biocrust (top 2 mm) and sediment (2-10 cm depth) from Avdat. After sieving the biocrust and sediment with a 2 mm metal sieve to remove rocks and plant litter, we added 63.7 ± 0.3g (mean ± SE) of sediment topped with 50.1 ± 0.2g (mean ± SE) of biocrust to 50, 145 mm (diameter) x 20mm (depth) plastic petri dishes. We grew these crusts in growing chambers at 16°C with 70% humidity and a 16:8 light: dark cycle for 111 days. For this first experiment, we chose to cultivate biocrusts using techniques designed for biocrust restoration (see Doherty and others 2015). This method may be less realistic but reduces variation within and between our biocrusts. Then, we randomly allocated 25 of our laboratory-grown crusts to mucus and no mucus (i.e., mucus- free) treatments. In the mucus treatment, we watered biocrusts with 1ml of our diluted snail mucus mixture (equal to 1 days’ worth of snail mucus) and an additional 2.5ml of DI every other day for 13 days. In the no mucus control, we repeated the same watering protocol but with 3.5ml of DI water. We chose this arbitrary protocol for logistical reasons. Yet, both the overall water addition and the distribution are well within the range of natural precipitation events in our study site, as the median number of rain events from 2008-2020 at Avdat in the wettest months (January and February ) is seven (Israel Meteorological Survey, 2021; Station: 253052).

To produce the mucus, we collected wild snails (Species: XS, SP, and SZ; n = 20 individuals/species) from Avdat and randomly placed groups of 10 individuals by species (n = 2 containers per species) in plastic containers (249 mm length x 190.5 mm width x 94.0 mm height). The containers were placed in a room, maintained at 14-17°C with a 11:13 light: dark cycle. We extracted mucus from housed snails twice weekly by placing single species groups of 10 snails on a 280 mm x 216 mm transparent plastic sheet (3M Write-on Overhead Projector Transparency Film) covered by a 140 mm x 115 mm x 50 mm height plastic lid. We wetted the snails daily with DI to ensure their activity. After 48 hours, we scrapped the mucus off the transparent plastic sheets and homogenized all collected snail mucus (across all species) in 60 ml of DI, generating a mixed mucus solution containing the equivalent of 60 days’ worth of snail mucus production (i.e., 60 snail^-1 days^-1, 1ml equals 1 snail^-1 day^-1).  Mixing all species’ mucus is representative of mucus deposition in nature, where all species co-exist on moist biocrusts (Yom-Tov 1971). Our mucus harvesting technique should capture mainly the water-soluble components of snail mucus. Diluted mucus was then frozen at -80°C until use. All snail species studied consume biocrusts. Thus, we chose to use pre-extracted mucus (rather than have snails directly deposit mucus trails on biocrusts) to isolate the effects of mucus deposition from biocrust consumption.

At the completion of the 13 days, we measured the effects of snail mucus on biocrust activity by measuring biocrust CO₂ efflux. We placed watered (3 ml of DI water) mucus and control laboratory grown crusts in airtight plastic chambers [155mm x 155mm x 61mm (length x width x height) Lock & Lock HPL 823; www.s-d.co.il]. We flushed the chambers with CO₂-free air at a rate of 2L minute^-1 for a total of 5 minutes. Laboratory grown crusts were incubated in the flushed, airtight chambers at 16°C with no light for a total of 24 hours. After the dark incubation, we used a LI-7000 CO₂/H₂O infrared gas analyzer (IRGA; LI-COR, Inc.) with a designated self- manufactured injection system to quantify the amount of CO₂ released by each biocrust.

Experiment 2

The goal of this experiment was to explore how XS mucus from two distinct populations affect field collected moss/lichen-dominated biocrust from Shaked Park. We used mucus from XS because this is the main snail species found at Shaked Park. To create the laboratory biocrusts, we collected 30 intact biocrusts from Shaked Park using 120 mm diameter x 20 mm deep plastic petri dishes (following Weber and others 2016). We housed the field collected biocrust in the lab for 25 days at 18°C, with 60% humidity and a 8:16 light: dark cycle. During this time, we watered our field collected biocrusts daily with 3.6-4.2 ml of DI water, using a hand-held spray bottle. Using field-collected biocrust for experiments 2 and 3 may provide more realistic, but more heterogeneous, results compared to those obtained from laboratory cultivated early successional biocrusts, complementing the approach used for experiment 1.

We randomly allocated field collected Shaked Park biocrusts (n =10 per treatment) to one of three treatments 1) XS-AV mucus, 2) XS-SS mucus, and 3) mucus-free control. In the XS-AV mucus treatment, we watered biocrusts with 1.5ml of the diluted mucus (equal to 1 days’ worth of snail mucus) extracted from XS snails from Avdat and 2.0 ml of DI water every other day for 16 days (until 01 April 2019). In the XS-SS treatment, we repeated the same procedures but with extracted XS mucus from Shaked Park. In the mucus-free control, we watered biocrusts with 3.5ml of DI water every other day for 16 days. Here, we chose the same arbitrary protocol as in experiment 1 but with eight watering days, which corresponds with the median number of rain events in January and February (the wettest months) from 2008-2020 in this region (Israel Meteorological Survey, 2021; Station: 251691).

To produce the mucus, we collected snails (Species: XS; n = 30 individuals per site) from both Avdat and Shaked Park. We randomly placed snails in 249 mm length x 190.5 mm width x 94.0 mm height plastic containers [in groups of 10 individuals by site (n = 3 containers per site)]. We placed all containers in a climate-controlled room, maintained at 15 ± 1.5 °C with a 11:13 light: dark cycle. Snails were fed biocrusts (Avdat and Shaked Park), Hammada scoparia litter, and Atractylis serratuloides litter ad libitum. We extracted snail mucus from housed snails once a week for two weeks, using the same protocol described for Experiment 1. However, at the end of each extraction session, we scrapped the mucus off the transparent plastic sheet and the small plastic container using 15 ml of DI water per box and homogenized all collected mucus (across boxes) in 50 ml falcon tubes. Falcon tubes containing diluted mucus were then frozen at -80°C until use. We did experience snail death during the snail extraction process (~10% mortality per week). When snails died, we replaced them with a new snail that was collected from the appropriate field site.

We quantified snail mucus production between XS populations (i.e., Avdat vs. Shaked Park) by creating six, 1.5ml pseudo-replicated samples of homogenized snail mucus. We then freeze-dried the mucus samples for 24 hours and weighed the remaining dried material.

To account for natural variation in field collected biocrust CO₂ efflux we wanted to quantify the pre-experimental CO₂ efflux of each biocrust but, due to incubator malfunctions, we were unable to measure biocrust CO₂ efflux prior to experimental manipulations. We quantified the CO₂ efflux using the same protocol as for Experiment 1.

Experiment 3

 The goal of this experiment was to reveal how interspecific variation in snail mucus affects field collected cyanobacteria dominated biocrusts from Avdat.  We collected 40 field collected biocrusts from Avdat and reared them in the lab using the exact same protocol as for Experiment 2. We randomly allocated 10 field- collected biocrusts to each of the four treatments: 1) XS mucus, 2) SP mucus, 3) SZ mucus, and 4) mucus-free control. In each mucus treatment, we watered field- collected Avdat biocrusts with 1ml of the corresponding diluted mucus (equal to 1 days’ worth of snail mucus) and an additional 2.5ml of DI water every other day for 16 days. In the mucus-free control, we watered Avdat biocrusts with 3.5ml of DI water every other day till the end of the experiment. We used identical watering protocol as for Experiment 2 to allow better comparisons of the XS mucus effect on biocrust CO2 efflux between biocrust types.  

To harvest mucus, we collected snails (XS, SP, and SZ; n = 60 individuals per species) and reared them using the same protocol as for Experiment 2. During the mucus excretion period, all snail species were able to feed ad libitum on biocrusts and Hammada scoparia litter collected from Avdat. We extracted snail mucus from housed snails once weekly for two weeks using a similar extraction protocol as in Experiment 2. At the end of each extraction, we scrapped the mucus off the transparent plastic sheet and the small plastic container and diluted the collected mucus from each box in 10ml of DI water and homogenized all mucus by snail species in 50ml falcon tubes. Falcon tubes containing diluted mucus were then frozen at -80°C until use. We did experience snail death between snail extractions (~10% mortality per week). When snails died, they were replaced with new snails also collected from Avdat. We quantified the production of snail mucus between the three species of snails (XS, SP, and SZ) by creating three, 1.5ml pseudo-replicated samples of homogenized snail mucus per species. We then freeze-dried the mucus samples for 24 hours and weighed the remaining dried material. We measured the CO₂ efflux of field collected biocrusts at the beginning and end of the two-week mucus addition period using the protocol described in Experiment 1.

Data analysis

In Experiment 1, we compared the final CO₂ efflux (ug C day^-1 g^-1) of laboratory grown biocrusts between mucus treatments using a two-sample T-test. In Experiment 2, we compared the dry mass of snail mucus between snails from Avdat and Shaked Park using a two-sample T-test. We compared the final CO₂ efflux of biocrusts between treatments using a Generalized Linear Model (GLM) with a Tukey’s HSD test. We used a GLM because they accommodate variance in heterogeneity, non-normal distributions, and uneven sample sizes (Venables and Dichmont 2004; Bolker 2008). Additionally, we used goodness-of-fit statistics to determine the best distribution for each model. Prior to data analysis, we removed two outliers that were 1.5-times the interquartile range above the third quartile [No mucus (n = 8); XS-AV (n = 10); XS-SS (n = 10)]. In Experiment 3, we compared the snail mucus dry mass between our three snail species using a one-factor ANOVA with snail species (i.e., XS, SP, and SZ) as a fixed factor.  We compared the change in CO₂ efflux of Avdat biocrusts between treatments using a GLM and a Tukey’s HSD test. Additionally, we used goodness-of-fit statistics to determine the best distribution for each model. Prior to data analysis, we removed five outliers that were 1.5-times the interquartile range above the third quartile or below the first quartile [No mucus (n = 7); XS mucus (n = 9); SP mucus (n = 9); SZ mucus (n = 10)]. We ran all our statistical analysis in jamovi version 1.217 using the “jmv” and “gamlj’ modules (Gallucci 2019; R Core Team 2019; The jamovi project 2020). 

We calculated the effect sizes of mucus impacts on biocrusts to compare across snail and biocrust type using OpenMee software (Wallace and others 2017). Specifically, we calculated the Hedges d by comparing the means of the mucus treatment(s) to the mean of the mucus-free control in each of our experiments. We interpreted our effect sizes using the benchmarks set by Cohen (1988), who suggested that an effect size of 0.2 is small, 0.5 is moderate, and anything greater than 0.8 is large. While Cohen’s benchmarks are general, we have no other standard at which to compare our effect sizes too—as mucus effects on biocrusts are a novel interaction in the literature. Thus, we have chosen to use Cohen’s benchmarks to simply consider the relative magnitude of each interaction in our study.

Please see metadata within each file for details.