Final Report Summary - BIOSOC (CONTRIBUTION OF BIOLOGICAL SOIL CRUSTS TO THE CARBON BALANCE IN DRYLANDS) Monitoring terrestrial ecosystems in an attempt to quantify and accurately predict the C-balance from local to regional scale is one of the highest European Commission research priorities (see ESFRI and ICOS EC FP 7 Projects) as these predictions are essential for identifying and quantifying CO2 sinks and sources. Europe is already a leader in carbon cycle research in temperate forests and agricultural systems (Central Europe), but not yet in arid and semiarid regions. In such areas, biological soil crusts (BSCs) are one of the predominant land covers, covering up to 70% of the surface in some areas, and possibly they are the main resource for soil organic carbon (SOC). Nevertheless, the role of BSCs as a carbon source or sink has hardly been studied, and this is one of the priorities of the present BIOSOC proposal. The main objective of BIOSOC is to study of the role of BSCs at different development state in the C-balance in arid and semiarid zones. The specific tasks for achieving this goal are: (i) Monitor in situ CO2 fluxes in BSCs at different successional stages in a representative catchment and relate these fluxes to key environmental factors controlling the CO2 exchange rates; (ii) Map the distribution of different BSC types (at different successional stages) in the catchment based on their spectral signals; (iii) Assess the BSC C-balance in a small catchment by combining the CO2 flux function from the different crust types with their spatial coverage. The CO2 fixation was measured by using permanent plots installed on the most representative BSCs (lichen crusts: Diploschistes diacapsis and Squamarina lentigera and cyanobacteria) and on bare soil or physical crust with incipient colonization by cyanobacteria at El Cautivo (Tabernas desert, Spain) and D. diacapsis, cyanobacteria, mosses, incipient cyanobacteria and annual plants at Las Amoladeras (Cabo de Gata Natural Park, Spain). Four PVC soil-borne collars were installed for each biocrust type and site and the net CO2 fluxes (NP) was periodically measured in these collars using an IRGA LI-6400 associated to a portable camera designed by my research group at Almería (EEZA-CSIC, Spain). Simultaneously, soil respiration fluxes (DR) were measured by using an IRGA PP-systems EGM-4 and gross photosynthesis (GP) was calculated from NP and DR fluxes. Several full – day measurement field campaigns were carried out in different environmental conditions to evaluate the effect of environmental variables. In addition, because BSCs are poikilohydric organisms, which are in an almost latent state when dry conditions, CO2 fluxes during different drying stages were also measurement in situ some days after rainfall when they were active. Microclimatic parameters and key environmental variables (air temperature and relative humidity just above the biocrusts, soil moisture, photosynthetically active radiation (PAR) and rainfall) were also continuously monitored during the gas exchange measurements. Our results showed that biocrusts behave as important CO2 sinks during certain periods in drylands, as it has been raised in most of the previous literature, due to their physiological characteristics that allow them to be active and photosynthesize at very low humidity, offering a photosynthetic advantage compared to higher plants. The water availability was a decisive variable activating both photosynthesis and respiration. Our study demonstrated that biocrusts were able to fix CO2 during periods of soil water availability as well as mild temperatures; while a net CO2 release was generally found for annual plants. Below a soil moisture threshold, DR and GP showed low values, although both fluxes increased after rainfall, especially soil respiration, which was triggered by the first rainfall after a long dry period. During the drying curves, we verified also that CO2 fixation occurs only just after rainfall events but only one, two or three days after precipitation, as the soil begins to dry out and soil moisture decreases, then the biocrusts photosynthetic rate was lower than the soil respiration. The significant differences in the biocrust CO2 fluxes throughout the day were due to the interplay of environmental variables such as temperature, air humidity and PAR. The combination of these environmental variables along with the physiological characteristics of biocrusts and soil types determines their ability to photosynthesize and fix atmospheric CO2 when the soil moisture was available. Nevertheless, it is very rare that the optimal range of different environmental variables for biocrust photosynthesis occurs simultaneously under field conditions in drylands, as these regions are characterized by their adverse environmental conditions. The rarity of optimal conditions could significantly limit the potential CO2 sink nature of biocrusts to sporadic days during the year, making soils covered by biocrusts act either as net CO2 sources (following rainfall after long drought periods) or neutral during most part of the year. Notwithstanding the negative net CO2 balance during a large part of the year, biocrusts occupy a large surface and they have higher OC content than bare soils. This suggests that the biocrusts are able to activate using water from other sources than rainfall such as dew or fog, thus ensuring a high relative humidity coupled to a low soil moisture that enable photosynthesis that guarantees their essential functions. In these cases, despite the biocrusts retains organic carbon guaranteeing their nutritional requirements, the net CO2 balance was negative (CO2 emission to the atmosphere) indicating that the respiration processes of the whole soil and crusts exceeded the photosynthesis of the crust. This underlines the crucial role of microbial communities growing below biocrusts, which have an essential role in the carbon cycle and net CO2 emission in crusted ecosystems. During wet periods, when the biocrust was active, late successional biocrusts (i.e. lichens and mosses) had significantly higher photosynthetic activity than early successional biocrusts (developed and incipient cyanobacteria crusts). Nevertheless, these late successional biocrusts showed also higher respiration rates than early successional ones. After all, both the biocrusts and the whole soil profile under biocrusts with their microbial communities associated should be considered for a C balance. Therefore, although late successional biocrusts, such as lichens and mosses, showed higher gross photosynthetic rates than cyanobacteria, initially pointing to a better CO2 sink potential than the cyanobacteria, there were no significant differences between the net CO2 fluxes of these successional stages. These results highlight the importance of considering all contributions to C fluxes (both the biocrusts and the microorganisms of the whole soil profile) in real field conditions for modelling C-balance of biologically crusted soils from drylands. Our results highlighted the important role of microbial communities in soils below BSCs modulating CO2 emission fluxes. For this reason, we investigated possible associations between specific bacteria genus below BSCs, enzymatic activities involved in the C, N and P cycles and physico-chemical soil variables (TOC, TN, P, pH, carbonates....). Thus, we studied the microbial communities structure in soils under four different biocrusts types from Tabernas Desert (Spain): (i) bare soil with an incipient cyanobacteria colonization (IC, representing the earliest successional stage), (ii) cyanobacteria (C), (iii) lichens D. diacapsis and S. lentigera (L-D+S) and (iv) lichen lepraria issidiata (L-L), this last representing the latest successional state. In addition, for each biocrust type, we sampled soils in four locations representing a gradient in biocrust covere. Soil bacterial communities were studied extracting DNA from the soils, then the quantification of bacteria was carried out by qPCR and the composition of microbial communities was analysed by pyrosequencing [sequencing of amplicon libraries of 16S V4-V5 rRNA gene by Illumina MiSeq platform (Reagent Kit v3 -2x300 cycles)] and bioinformatics analysis [QIIME and MOTHUR]. Statistical Analysis (one-way ANOVA, Principal Component Analysis, Generalized Linear Models) were used to search associations between the bacteria genus in soils underlying different biocrust types, with different percentage of coverage and physico-chemical soil variables and to find general trends of the samples. Our results showed that in all biocrust types the number of bacteria per gram of soil decreased as the crust coverage decreased, confirming the biocrust role increasing the soil microbial communities. Surprisingly, the number of bacteria per gram of soil decreased as the biocrust succession state progressively increased, following the sequence: IC > C > L-D + S > L-L. Nevertheless, several diversity indices for 16S rDNA sequences in soils below biocrust types (SOBS, SHAO, INVSIMPSON, SHANNON, J´), indicated that biocrusts with higher successional state increased the bacterial diversity in their underlying soils, which decreased as the biocrust succession successively decreased, following the sequence: L-L > L-D+S > C > IC. The Phylogenetic analysis showed 21 phylum and 197 genus in biocrust soils. All crusts had the same phylum, but the relative abundance varied significantly between biocrust type (i.e. lichens had higher percentage of Acidobacteria and Bacterioidetes and cyanobacteria had higher percentages of Proteobacteria). Statistical analysis showed contrasting microbial composition between soil below each biocrust type. Soils under lichens (L-L and L-D+S) have greater number of bacteria genus associated than cyanobacteria (C and IC). Some bacteria genus comparatively more abundant in soils under L-L were: Gemmatimonas, Gemmata, Acidobacteria Gp6, WPS-1_genera_incertae_sedis; meanwhile in soils under L-D+S were Adhaeribacter, Devosia, Truepera and under cyanobacteria and cyanobacteria incipient were Armatimonas/Armatimonadetes_gp1 and Rubrobacter. In soils under L-L there was a weak effect of pH possibly due to organic acids excreted by this lichen and in soils under L-D+S the OC and N content were associated with the bacterial communities of their underlying soils. In contrast, carbonates and electrical conductivity were associated with the bacterial communities in soils under incipient cyanobacteria.
