Carrizo Arroyo, Central New Mexico - A New Late Paleozoic Taphotype of Arthropod Fossillagerstätte
Joerg W Schneider1, Spencer G Lucas2, Steffen Trümper1, Christiane Stanulla3, and Karl Krainer4
1Technical University Bergakademie Freiberg, Russia
2New Mexico Museum of Natural History, USA
3GeoWiD GmbH, Germany
Submission: December 12, 2017; Published: April 26, 2018
*Correspondence author: Joerg W Schneider, Technical University Bergakademie Freiberg, Cotta-Str. 2, D-09596 Freiberg, Germany; Kazan Federal University, Kremlyovskaya str. 18, 420008 Kazan, Russia, Email: Joerg.Schneider@geo.tu-freiberg.de
How to cite this article: Joerg W S, Spencer G L, Steffen T, Christiane S, Karl K. Carrizo Arroyo, Central New Mexico - A New Late Paleozoic Taphotype of Arthropod Fossillagerstätte. Oceanogr Fish Open Access J. 2018; 7(2): 555710.DOI:10.19080/OFOAJ.2018.07.555710
At Carrizo Arroyo, southwest of Albuquerque, New Mexico, an approximately 100m-thick section of the latest Pennsylvanian (latest Gzhelian) to Early Permian (early Asselian) Red Tanks Member of the Bursum Formation is exposed. This sedimentary succession is interpreted as a coastal plain on a very shallow shelf affected by repeated transgressions and regressions. Besides the marine marls and limestones, the most common lithotypes in the nonmarine fossiliferous intervals are greenish-gray and gray, variably sandy fine clastics. Lithology and facies architectures together document a low energy floodplain environment crossed by very shallow but wide flood channels. In the floodplain deposits, three basic taphotypes were observed:
a. The common plant bed type,
b. The rare conchostracan bed type, and
c. The insect bed type, which is not as rare as previously assumed.
Plant beds are commonly formed by single layers of dm-long branches and leaves, as well as cm-sized plant fragments. Consequently, it is assumed that the Carrizo Arroyo plant beds were deposited by waning flood in shallow and wide floodplain channels. Conchostracan and insect beds have several features in common. Bedding planes with enrichments of conchostracans, freshwater pelecypods, insects, and, in places, eurypterids, contain tiny plant detritus of mm- to cm-size only. They form a sub-mm to mm-thick layer only, and have a restricted lateral extent of several meters to decameters. Altogether, this points to autochthonous assemblages of aquatic arthropods and molluscs preserved in short-lived freshwater puddles and ponds on the floodplain. The common but generally isolated insect wings were most likely transported by winds and trapped at the water surface of those freshwater accumulations on the floodplain. Obviously, fossiliferous deposits at Carrizo Arroyo contain an assemblage of autochthonous and allochthonous elements of the insect fauna, covering environments from the hinterland down to the seacoast. This makes the Carrizo Arroyo Fossillagerstätte exceptional.
In contrast to the highly diverse earliest Pennsylvanian (Bashkirian) marine insect fossil-Lagerstätten of Ningxia, China, and Hagen-Vorhalle, Germany, Middle Pennsylvanian to Early Permian insect sites worldwide are dominated by fossils of cockroachoids . Most of these sites are situated in the roof shales and interbeds of coal seams, such as Mazon Creek, Writhlington, Commentry, Wettin, Kuznetsk, etc. [5-8]. They mainly reflect the environments of wet coal forests of the Late Pennsylvanian and earliest Permian . The earliest Permian Carrizo Arroyo site in New Mexico (Figure 1) , investigated during the last decade, represents a new taphotype of arthropod Fossillagerstätten, which will be discussed here. The general depositional environment of this site is interpreted by Lucas & Krainer  as a coastal plain on a very shallow shelf during repeated transgressions and regressions, as evidenced by the identification of six depositional sequences (DS) in the Bursum Formation section. DiMichele et al.  regarded the paleoflora as representing a seasonally dry biome of the Pennsylvanian- Permian transition. Here, we evaluate the taphonomy of nonmarine Fossillagerstätte in the Bursum For-mation at Carrizo Arroyo.
Since the 1970s, the fossil content of the Red Tanks Member of the Bursum Formation at Carrizo Arroyo has become an in-creasing focus of publications (e.g., [12,13]). Kues & Kietzke  discriminated and numbered 29 units in the Carrizo Arroyo section, whereas Krainer and Lucas (2004) recognized 55 units arranged in six DS (Figure 2). A really exact, bed-by-bed sampling only became possible after Krainer & Lucas  measured the approximately 100-m-thick profile of the Red Tanks Member at Carrizo Arroyo in great detail (Figure 2). Based on this work from 2005 to 2009, one of us (JWS) documented and sampled the lithology and fossil content of five narrow stratigraphic intervals, each between 3.5 to 5m in thickness with cm-scale resolution:
a. The coal bed profile, NMMNH (New Mexico Museum of Natural History) locality 3428, units 17 to 21 of Krainer & Lucas  in the lower half of DS 2 (completed by R. Werneburg in 2009), (Figure 3);
b. NMMNH locality 7724 in the lower third of unit 27 in DS 3;
c. NMMNH locality 3437 in the upper half of unit 27 in DS 3, (Figure 4);
d. NMMNH locality 3433 in unit 37 of DS 4 (Figure 5); and
e. NMMNH locality 5123, units 53 to 54 at the top of DS 6, directly below the Abo base as well as the basal part of unit 55 at the Abo base.
The several hundred plant and animal fossils that are correlated to specific beds in these stratigraphic intervals are stored in the collection of the NMMNH. In the following text, the term "unit" followed by a number refers to the measured section documented by Krainer and Lucas  (Figure 2), and the term "bed" followed by a number refers to the profile documentations of JWS (Figures 3-5).
Conchostracans form a paraphyletic group of Branchiopo-da, now divided into the monophyletic Laevicaudata, Spini-caudata and Cladoceromorpha . Here, we deal with Spinicaudata and Laevicaudata, but for convenience we use the term Conchostraca.
Located on the eastern edge of the Colorado Plateau, Carrizo Arroyo (Figure 1) is 50km southwest of Albuquerque, New Mexico, USA (~34° 45'N, 107° 07'30"W). Here, an approximately 105m-thick section of upper Paleozoic clastic and carbonate rocks yields extensive fossil assemblages of marine and nonma-rine origin [13,16-18]. At the base of the section, marine limestones of the upper part of the Atrasado Formation of unquestioned Late Pennsylvanian (Virgilian) age are exposed. Most of the section at Carrizo Arroyo belongs to the Red Tanks Member (~100m thick locally) of the Bursum Formation [14,18-21]. At Carrizo Arroyo, the Red Tanks Member is mostly gray and very rarely reddish shale, mudstone and siltstone of nonmarine origin, intercalated with several beds of limestone and shale of marine origin (Figure 2) . The Red Tanks Member is overlain by nonmarine wet red beds of the Abo For-mation, consisting of alluvial plain to floodplain fine clastics with intercalated fluvial channel deposits [22,23].
The stratigraphic architecture of the Red Tanks Member at Carrizo Arroyo has been interpreted to indicate the presence of six DS  Figure 2). The base of each DS is drawn at the base of beds of conglomerates or sandstones sharply incised into underlying mudrock, and each sequence then fines upward into mudrockdominated floodplain deposits. Marine limestone beds cap each sequence. Derived from these limestone intervals, six marine flooding events are dis-criminated (Figure 2). The nonmarine, floodplain deposits are mostly composed of mudstone/siltstone beds, some of which contain abundant calcrete nodules and other evidence of im-mature pedogenesis. A thin cordaitalean-leaf coal bed, unit 19 at NMMNH locality 3428 (Figure 3), in the middle of DS 2, is underlain by fossiliferous, thin-bedded to laminated marly silt-stone to claystone (plants, lingulid brachiopods, "Spirorbis" (?microconchids), ostracods, isolated fish remains) and over-lain by marly mudstone containing brackish marine bivalves (myalinids), lingulids, "Spirorbis" incrustations on dm-thick tree trunks, cordaitalean leaves up to 40cm long and plant debris. Carbonate conglomerates at the bases of DS 3 and 4 probably represent upper shoreface deposits, and, thin layers in DS 3 and DS 4 are small side- channel fills of several m thick conglomeratic and sandy major channels.
The biostratigraphic age of the Red Tanks Member at Car-rizo Arroyo is determined by conodonts [21,24] and insects [21,25]. The presence of Streptognathodus virgilicus in the uppermost part of the Atrasado Formation con-strains its age to the middle to upper part of the Virgilian and to a comparable position in the Gzhelian. The only biostrati-graphically-significant conodont assemblage in the Red Tanks Member comes from horizon marine D, units 30-32, at the top of DS 3, and the assemblage is probably equivalent in age to the Midcontinent Streptognathodus nevaensis Zone, which is early to middle Asselian in age. In DS 3, nonmarine horizon A, at roughly 43m in the upper half of unit 27, NMMNH locality 3437, fragments of the cockroachoids Sysciophlebia ilfeldensis and Spiloblattina weissigensis were discovered. Based on the insect correlation of Lucas et al.  and the precised isotopic age calibration of the insect zonation of Schneider & Werneburg [26,27], in Schneider et al. , the Syscio-phlebia ilfeldensis-Spiloblattina weissigensis- insect zone strad-dles the Gzhelian/Asselian boundary. This correlation strongly supports assigning an early Asselian age to nonmarine horizon A in the upper half of unit 27, as is inferred from conodonts.
At Carrizo Arroyo, the Red Tanks Member yields fossils from many beds; insects as well as other nonmarine arthropods are scattered through the whole section. In the course of our prospecting for fossils, Lagerstätten of insects, conchostracans, eurypterids and other fossils were found at three levels (Figure 2):
I. NMMNH locality 7724, situated in the basal part of DS 3, lower third of unit 27 at 32.5 to 37.5m;
II. NMMNH locality 3437 near the middle of DS 3 in the upper half of unit 27 at 40 to 45.5m (Figure 4);
III. NMMNH locality 3433 in the first third of DS 4, upper half of unit 37 at 66.5 to 68.5m (Figure 5).
The most common lithotypes in these fossiliferous intervals are greenish-gray and gray, variably sandy mudstones and silt-stones, silty sandstones and marls (Figure 6). Bedding is devel-oped at the larger scale as dm- to several dm-thick horizontal beds (Figure 7); rare, m-thick, trough-cross-bedded channel fills are intercalated. At the smaller scale, the widely horizontal bedsets consist of several cm- to dm-thick and some meter- to decimeter-wide shallow, stacked, and mostly internally hori-zontally bedded (sub-mm- to cm-scale) lenticular bodies (Figure 8). Bedding planes often expose layers of larger plant detritus; dm-size fern leaves and cordaite leaves are common (Figure 9); the latter can form coaly layers up to 1 or 2cm thick. Trunks are rare, and diameters range from cm to dm. Varying pedo-genic overprint caused destruction of bedding, resulting in a completely structureless and massive appearance.
In DS 3, several thin conglomerate beds (~1dm to 1m thick), made up of limestone clasts with thin intercalated sand-stone layers, are intercalated in the greenish-gray mudstones and siltstones (Figure 10). At NMMNH locality 3437, one of these beds could be traced from the documentation trench laterally to a several m high cliff (base not exposed). This cliff consists of stacked limestone conglomerate channels at the bottom and of pebbly sandstone at the top. It is interpreted as a fluvial main channel that pinched out laterally in the trench. The conglomerate is grain supported; the subrounded clasts are poorly sorted granules and pebbles in a sandy matrix (Figure 11). Most of the clasts are carbonates, mainly pedogenic nodules, reworked from immature calcic soils and calcretes. Red rims of some peb-bles point to intermittent subaerial exposure and oxidation.
Variably fragmented plant remains are scattered throughout the section. Accumulations of cm- to dm-size plant fragments, forming plant beds, occur in cm- to dm-thick, planar-bedded, fine sandy siltstones and silty fine sandstones of mainly len-ticular shape. Most common are cordaitalean leaves, walchian conifer twigs and compound leaves (fronds) of pteridosperms, mainly with sphenopteridian foliage, as well as callipterid and odontopterid leaves and fronds, all preserved as carbonaceous compressions (Figure 9). The horizons are often dominated by single plant taxa. Plant beds have been observed at different levels, with changing dominant floral elements.
The thus-far known lowermost plant bed, NMMNH locality 3428 (Figure 3), unit 18, immediately below the coal bed (unit 19) in DS 1, contains, based on Kues & Kietzke  and our own observations, mainly larger walchian twigs and rarely cordaitalean leaves. Plants occur in silty lay-ers of a 30cm-thick, fine-bedded ostracod marlstone (smooth shelled Darwinula-type); the uppermost 1cm contains, in ad-dition to masses of ostracods, lingulids up to 1cm long. The 10cm-thick dirty coal bed is mainly made up of cordaitalean leaves. Sediments above belong to the brackish A horizon of Lucas et al. .
The middle of DS 3 in the upper half of unit 27 between 40m to 45.5m, NMMNH locality 3437 (Figure 4), contains several plant layers and insect beds. This is the original plant locality of Kues & Kietzke  in their unit 10 . Several leaf layers are absolutely dominated by cordaitaleans (Figure 9), and large fronds of other plants occur at one level in the higher part of the profile only (Figure 4, section II, boundary bed 7/8). The relatively high diversity of the flora reflects the so-called "rare elements" of DiMichele et al. , which are mainly preserved as leaf fragments. The stratigraphically highest plant bed is situated in NMMNH locality 3431, only dms below the insect bed NMMNH locality 3433, in the first third of DS 4, upper half of unit 37 at 66.5m to 68.5m (Figure 5). Walchians are absolutely dominant; cor-daitalean leaves are rare. Besides walchians, one bed contains common, dm-long fronds with neuropterid pinnules resembling Odontopteris .
Accumulations of conchostracan valves on distinct bedding planes, forming mass occurrences in places, may have different causes. Monospecific accumulations could originate from the dieoff of one population at the end of the ontogenetic cycle. Such accumulations are easily recognizable and characterized by adult individuals only, which show the typical crowding of growth lines at the borders of the valves. If a population dies because of any lethal environmental factor (e.g., sudden oxygen deficit, sudden elevated salinity, drying up), accumulations of valves are formed by one nonadult growing stage with a nearly identical number of growth lines on the valves.
Other than the coal bed profile, NMMNH locality 3428, the horizon brackish A, in DS 2, where conchostracans are seem-ingly absent, all ofthe stratigraphically higher fossiliferous layers contains scattered conchostracans. Most likely, they are allochthonous. But, at some levels of the section, there appear accumulations that are regarded as autochthonous populations. One of them occurs in bed 6 at NMMNH locality 3433, in the first third of DS 4, upper half of unit 37. There, single bed-ding planes of claystones are covered with tiny plant detritus and contain masses of Lioestheria valves (Figure 13G) together with single valves of laevicaudate conchostracans (Figure 12, & Figure 13C). Additional, smooth-shelled ostracods of the Darwinula type are very common, and insect wings (Figure 13K) are not rare at some levels (see below - insect bed type). One of those bedding planes exhibited a large number of Carbonicola- like freshwater pelecypods, which are concentrated in a small area of dm diameter.
Additional mass occurrences were found in different levels of bed 8, section II, of NMMNH locality 3437. There, one bed-ding plane of the siltstone is covered with tiny plant detritus and Pseudestheria valves, and among them single insect wings and some fragments of eurypterids (Figure 14). Unusually large conchostracans, up to 30mm long, identi-fied as Palaeolimnadiopsis (Figure 13I), occur as single individu-als in all detailed measured sections in DS 3 and DS
The sedimentology and taphonomy of several of the insect beds in the above-mentioned localities can be exemplified by bed 8, section II, at NMMNH locality 3437 (Figure 4) in the orig-inal "insect zone" of Kues & Kietzke  (Figure 2) of this locality. One bedding plane in the middle of unit 8, section II, that pinches out laterally over a strike of about 3 m, exhibited patchily arranged, very fine plant detritus in places together with common conchostracans and isolated and often fragmented insect wings (Figures 14, 13J, L). In order of decreasing abun-dance follow relatively small Carbonicola- like bivalves and eurypterids (Figure 13E). Single, complete eurypterids occur as early juveniles of mm-size (Figure 13A) up to semi-adults of dm length (comp. ); the same size spec-trum is covered by common eurypterid body fragments (Figure 14). Not rare are complete specimens of the wingless aquatic monuran insect Dasyleptus (Figure 13F); very rare are xiphosu-rids (Figure 13B) and diplopods (Figure 13H).
The lowermost insect bed yet discovered is situated in the basal part of DS 3, in the lower third of unit 27 at 32.5m to 37.5m, NMMNH locality 7724, in bed 7 of JWS documenta-tion. Bedding planes in the lower part of the 18cm-thick bed are covered by very fine plant detritus and bear common insect wings and eurypterid remains, but rarely only conchostracans. Very similar arthropod beds are bed 7a, section II, at NMMNH locality 3437, and bed 6, at NMMNH locality 3433, described above as one of the conchostracan bed types (Figure 15).
The sedimentary successions of the Red Tanks Member exposed at Carrizo Arroyo are interpreted by Krainer & Lucas  as a coastal plain on a very shallow shelf affected by repeated transgressions and regressions of decreasing intensity to the top. Besides the marine marls and limestones, the most common lithotypes in the nonmarine fossiliferous intervals are greenish- gray and gray, variably sandy mudstones and silt-stones, silty sandstones and marls. Bedding is developed at a larger scale as dm- to several dm-thick horizontal beds; rarely, intercalated are m-thick, trough-cross bedded silty and sandy channel fills. At the smaller scale, the widely horizontal bedsets consist of several cm- to dm-thick and some meter- to decime-ter-wide, shallow, stacked, and mostly internally horizontally bedded (sub-mm- to cm-scale) lenticular bodies. Lithology and facies architectures together document a low energy floodplain environment crossed by very shallow but wide flood channels. Plant beds are commonly formed by single layers of dm-long branches and leaves as well as cm- sized plant fragments. Consequently, it is assumed that the Carrizo Arroyo plant beds were deposited by waning floods in shallow and wide flood plain channels after rainstorms, during which winds had been able to snap off larger twigs of plants.
Conchostracan and insect beds have in common the fol-lowing features. Bedding planes with enrichments of con-chostracans, insects, and, in places, eurypterids, contain tiny plant detritus of mm- to cm-size only. They form a sub-mm to mm-thick layer only and have a restricted lateral extent of sev-eral meters to decimeters. Accumulations of conchostracans represent one growth stage only (i.e., the die-off of one popula-tion). Some of these conchostracan and/or insects beds contain freshwater pelecypods of the Carbonicola-type, scattered on the bedding plane or, in places, forming patches. Their small sizes of up to 1cm indicate, compared to modern unionids, juveniles of not more than one year in age. Altogether this con-figuration points to autochthonous assemblages preserved in short-lived freshwater puddles and ponds on the floodplain.
The question to be answered is, why is there an enrichment of isolated insect wings in these beds? It is preliminarily as-sumed that these isolated wings were blown by the wind across the floodplain and were finally trapped on the water surface of floodplain puddles and ponds. The unusual, very high diversity of those insect beds at Carrizo Arroyo with 30 insect genera of 17 families and 13 orders  could be ex-plained only by the following scenario: Fossiliferous deposits, such as at Carrizo Arroyo, contain an assemblage of autochtho-nous and allochthonous elements of the insect fauna, covering environments from the hinterland down to the seacoast. This makes the Carrizo Arroyo Fossillagerstätte exceptional [33,34].
We are grateful to Larry Rinehart, Joshua Smith, Justin Spielmann, Matt Celesky, Ralf Werneburg, Sebastian Voigt, and DanS.Chaney for collaborative and delightful field work. Hans Kerp and William DiMichele are thanked for discussions of the flora and determinations of some plant remains. Olaf Elicki gave support in the interpretation of marine bioclasts, and Vladimir Silantiev in the interpretation of nonmarine pe-lecypods. Michael Magnus and his team at the Technical Uni-versity Bergakademie Freiberg, Germany, produced the thin sections. The reviewers Bill DiMichele and Larry Rinehart are acknowledged for helpful comments. JWS thanks the German Research Foundation for financial support in the framework of grants DFG Schn 408/12 and DFG Schn 408/21.