Data for: Past, present, and future spatial distributions of deep-sea coral and sponge microbiomes revealed by predictive models

This dataset includes output rasters of the spatio-temporal model presented in Busch et al. 2024 (Data folder 1) giving cumulative microbial richnesses in deep-sea sponges and corals for the present, and two future scenarios. We also provide point data on the environmental variables used in our predictions (Data file 2): Depth, Slope, Bottom Stress (“BtmStress”), Bottom Current Speed (“BtmCur”), Mixed Layer Depth (“MLD”), Bottom Salinity (“BtmSal”), Sea Surface Salinity (“SSS”), Sea Surface Temperature (“SST”), and Bottom Temperature (“BtmTmp”) averaged for each of the seven time periods (extracted from the Simple Ocean Data Assimilation, SODA, for the past periods 1871-1900, 1901-1930, 1931-1960, 1961-1989; extracted from the BNAM Ocean Model for the present period 1990-2015 and for the future periods 2046-2065 and 2066-2085 under a RCP8.5 scenario). This dataset also contains raw occurrence data compiled from different indicated sources (Data file 3), and presence/absence data (true absences as well as pseudo absences) gridded to be used as model input (Data file 4) of the analysed host species Weberella bursa, Stryphnus fortis, Lophelia pertusa, Desmophyllum dianthus, and Vazella pourtalesii. Another data file (Data file 5) contains an overview of the ENA (European Nucleotide Archive) accession numbers, the original literature sources, and some basic ecological metadata of the used microbial data (16S amplicon data) drawn from the host species. After assigning the microbial abundance status (HMA – high microbial abundance, and LMA – low microbial abundance) to ten key sponge species in the Flemish Cap area area (Asconema foliatum, Geodia barretti, Geodia macandrewii, Geodia parva-phlegraei, Mycale lingua, Stelletta normani, Stryphnus fortis, Stylocordyla borealis, Tentorium semisuberites, and Weberella bursa) we correlated our spatial predictions of status occurrence with predictions of overall ecosystem function, i.e. here nutrient cycling and habitat provision (derived from Murillo et al. 2020, Diversity and Distributions). For our study the original rasters (created by Murillo et al. 2020) were resampled to a 0.088 cell size using Bilinear interpolation as the resampling technique in ArcGIS Pro. Data 6 contains the resampled predictions of overall ecosystem function, as well as our predicted occurrences of the HMA and LMA status, and the respective summed biomasses. In our article (above) we show a biomass network, integrating the generated information on HMA and LMA sponge biomasses with biomass measurements of other sessile filter feeding invertebrates, which occur in high abundances at the Flemish Cap (data taken from Murillo et al. 2020, Diversity and Distributions). Data file 7 contains the underlying biomass data used for the network, covering 116 different species, which are classified according to size (small, medium, medium large, large), functional (passive “PFF”, and active “AFF” filter feeders) and taxonomic (8 phyla) groups.

本数据集囊括了Busch等人在2024年提出的时空模型（数据文件夹1）的输出栅格，其中包含了深海海绵和珊瑚当前及两种未来情景下的累积微生物多样性。此外，我们还提供了用于预测的环境变量点数据（数据文件2）：深度、坡度、底部应力（“BtmStress”）、底部流速（“BtmCur”）、混合层深度（“MLD”）、底部盐度（“BtmSal”）、海表盐度（“SSS”）、海表温度（“SST”）、以及底部温度（“BtmTmp”），这些数据均为七个时间段内的平均值（过去时期1871-1900年、1901-1930年、1931-1960年、1961-1989年数据来源于简单海洋数据同化系统，SODA；当前时期1990-2015年及未来时期2046-2065年和2066-2085年数据来源于BNAM海洋模型，均基于RCP8.5情景）。本数据集还包含了从不同来源汇编的原始发生数据（数据文件3），以及用于模型输入的网格化存在/不存在数据（包括真实不存在以及伪不存在）（数据文件4），这些数据是针对分析的宿主物种Weberella bursa、Stryphnus fortis、Lophelia pertusa、Desmophyllum dianthus和Vazella pourtalesii。另一个数据文件（数据文件5）包含了ENA（欧洲核苷酸档案）的存取号、原始文献来源以及所使用微生物数据（16S扩增子数据）的基本生态元数据概览。通过对弗拉芒角地区十种关键海绵物种（Asconema foliatum、Geodia barretti、Geodia macandrewii、Geodia parva-phlegraei、Mycale lingua、Stelletta normani、Stryphnus fortis、Stylocordyla borealis、Tentorium semisuberites和Weberella bursa）的微生物丰度状态（HMA – 高微生物丰度，LMA – 低微生物丰度）进行赋值后，我们将空间预测的发生状态与整体生态系统功能预测（即本例中的养分循环和栖息地提供，参见Murillo等人2020年的《Diversity and Distributions》）进行了相关性分析。针对本研究，原始栅格数据（由Murillo等人2020年创建）在ArcGIS Pro中使用双线性插值作为重采样技术，重采样至0.088单元格大小。数据6包含了重采样后的整体生态系统功能预测，以及我们预测的HMA和LMA状态的发生，以及相应的总生物量。在上述文章中，我们展示了一个生物量网络，该网络整合了关于HMA和LMA海绵生物量的生成信息，并与在弗拉芒角高丰度出现的其他固着性滤食性无脊椎动物的生物量测量值相结合（数据来源于Murillo等人2020年的《Diversity and Distributions》）。数据文件7包含了用于网络的底层生物量数据，涵盖了116种不同的物种，这些物种根据体型（小、中、中大型、大型）、功能（被动“PFF”和主动“AFF”滤食者）和分类（8个门类）进行分类。