REE patterns of Sinemurian (Lower Jurassic) automicrite-sponge mounds (Djebel Bou Dahar, High Atlas, Morocco)

The Sinemurian-Pliensbachian Djebel Bou Dahar (DBD) carbonate platform (High Atlas, Morocco) evolved from a low-relief depositional system into a high-relief platform. This study focuses on geochemical data (stable isotopes, trace elements) from decametre-scale siliceous sponge mounds (upper Sinemurian) developed in a mid- to outer ramp setting. Trace element analyses were carried out on thin sections using laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS). Stable Oxygen and Carbon isotopes of mound automicrites have mean values (\u3b418O -2.780 \u2030, SD 0.55; \u3b413C 2.34 \u2030, SD 0.28) comparable to those from early marine, radial fibrous (RF) cement (\u3b418O -1.6 \u2030, SD 0.26; \u3b413C 2.29, SD 0.08), brachiopods (\u3b418O -2.70 \u2030, SD 0.29; \u3b413C 2.23, SD 0.72) and published Early Jurassic marine values. Blocky sparite (BS) cement of burial origin is characterized by mean \u3b418O -5.54 \u2030 (SD 0.33) and \u3b413C 1.77 \u2030 (SD 0.85). Precipitation of automicrites occurred in equilibrium with sea water without major enzymatic fractionation and they were not subjected to significant meteoric diagenesis. The RF cement has shale-normalized (sn) REE+Y patterns comparable to modern open-ocean seawater with: 1) light REE (LREE) depletion (average Ndsn/Ybsn = 0.19, SD 0.05; Dysn/Ybsn = 0.42, SD 0.07); 2) superchondritic Y/Ho ratio (mean Y/Ho = 81.29, SD 15.12); 3) positive La and Gd anomalies, and 4) negative Ce anomaly consistent with oxygenated waters. The BS cement has: 1) bell-shaped, MREE-enriched REE+Y patterns (mean Ndsn/Ybsn = 1.77, SD 0.38; Dysn/Ybsn = 2.03, SD 0.29); 2) low Y/Ho ratios (36-64); 3) mostly small negative Ce anomalies; 4) variable, mostly positive La anomalies; and 5) positive Eu anomalies (mean Eu/Eu* = 1.25, SD 0.33). Allomicrite (i.e., particulate fine carbonate sediment) has relatively high REE concentrations (\u3a3REE = 19-26 ppm) with sea water-like, but LREE enriched, flat REE+Ysn patterns (mean Ndsn/Ybsn = 0.79, SD 0.016; Y/Ho = 19-26; mean Ce/Ce* = 0.87, SD 0.08). Their reduced LREE depletion and high \u3a3REE cannot be explained by shale contamination and may reflect incorporation of particulate matter that scavenged LREE from the overlying water column. Automicrites have variable REE+Y signatures. Three samples have patterns representative of sea water (mean Ndsn/Ybsn = 0.29, SD 0.011; Y/Ho = 54-78; mean Ce/Ce* = 0.64, SD 0.04), however most samples have reduced LREE depletion and more variable anomalies (mean Ndsn/Ybsn = 0.76, SD 0.21; Y/Ho = 34-62; variable positive or negative Ce anomalies: mean Ce/Ce* = 1.03, SD 0.31; mean \u3a3REE = 9.4 ppm; mean Al2O3% = 0.36%, mean Zr = 4.0 ppm; mean Th = 0.34 ppm). REE distributions of automicrites are interpreted as being variably contaminated by LREE-enriched allomicrite and early diagenetic processes. Hypothetical mixing lines between early marine cement and shale (MUQ) show that shale contamination cannot explain the elevated LREEs. The positive Ce anomalies in some automicrites may reflect precipitation in dysoxic surface or early diagenetic microenvironments. Time-equivalent mounds in Morocco were interpreted as being deposited near the OMZ. Regardless, REE distributions of most automicrites were enriched in LREEs, leaving few samples with pristine seawater signatures. Stable isotopes suggest that the diagenetic alteration occurred in a burial, rather than meteoric, setting. The better stability of REE patterns in RF cements probably reflects their lower porosity relative to the microporous automicrites and lack of particulate matter. The RF cement REE patterns confirm that pre-Cenozoic (Early Jurassic) REE patterns were comparable to modern ones. Our data also show that the robust marine REE patterns of automicrites may be significantly contaminated in relatively clean carbonate environments, in this case by LREE-enriched allomicrite. In situ LA-ICP-MS analyses are superior to bulk rock analyses in isolating different components for evaluation of hypothetical mixing lines