DISCUSSION

Consideration of the arvicoline rodent fauna from SCC in a broad biochronologic context raises interesting issues regarding the age of the reddish-pink silt stratum. With the exception of Microtus paroperarius and M. meadensis, the arvicoline rodent taxa identified from SCC are known to occur as recently as the early Holocene (four-triangle morphotype of Lemmiscus curtatus), or as components of the extant biota (Lemmiscus curtatus, Microtus sp., Mictomys sp., and Phenacomys sp.). However, records of Phenacomys and Mictomys at SCC and in central portions of the Great Basin represent disjunct geographic occurrences given the known modern distribution of these taxa in areas to the north, east, and west of the Great Basin (Grayson 1981; Mead et al. 1982, 1992).

One of the two Microtus meadensis specimens came from a layer of rodent dung and brown silt that was situated stratigraphically above the reddish-pink silt. There are no radioisotopic dates associated with this layer, but the color of this specimen, and specimens of Microtus sp. from the same level, suggest that they were mixed from the reddish-pink silt zone. Extensive rodent burrows were noted during the excavation (Bryan 1979; Miller 1979) and may explain the presence of a red-stained tooth above the reddish-pink silt.

Recent biochronologic summaries report the known age distributions of Microtus paroperarius as ~840 ka to 252 ± 30 Ka and that of Microtus meadensis as 820 ka to 252 ± 30 Ka (Bell et al. 2004a, 2004b). Given updated last occurrences from Cathedral Cave (146.02 ± 2.584 - 153.7 ± 6.4 Ka; Jass 2007; Jass and Bell in press), the presence of M. paroperarius and M. meadensis at SCC imply an age between 820 ka and 146.02 ± 2.584 - 153.7 ± 6.4 Ka for the reddish-pink silt, based on maximum and minimum ages of all arvicolines from the site. A strictly biochronologic age estimate for the reddish-pink silt suggests a minimum difference of approximately 117 Ka between that stratum and the overlying archaeological deposits. A disconformity of 117 Ka does not seem particularly daunting or improbable in the context of an open-air deposit. However, within the confined context of a cave deposit, even a relatively open rock shelter like SCC, such a disconformity in vertical sequence would be rare.

Conversely, there is no explicit reason to rule out an alternate interpretation whereby SCC contains chronologic range extensions for both Microtus paroperarius and Microtus meadensis to as recently as 28,650 ± 760 ¹⁴C yr BP. Although I was not able to reconstruct confidently the vertical relationships of the specimens from the reddish-pink silt because provenience data varied in scope and detail, I consider all taxa from the reddish-pink silt to be derived from a single stratigraphic level that has an associated radiocarbon age of 28,650 ± 760 ¹⁴C yr BP. The stratigraphic information provided by Bryan (1979) indicates that the reddish-pink silt represents a discrete stratum, and the unique color of bone from the reddish-pink silt supports this interpretation. As mentioned above, the radiocarbon date for the reddish-pink silt level was based on bone collagen. In the absence of additional radioisotopic testing using newer collagen extraction techniques or testing on more reliable materials (e.g., charcoal, twigs; see Meltzer and Mead 1985 for discussion), an outright dismissal of the radiocarbon age would be premature.

Another possibility would be that both biochronologic and radioisotopic (¹⁴C yr BP) age assignments are incorrect and another age represents the "true" age (or ages) of the red silt zone. In many respects, the chronologic quandary regarding arvicoline rodent biochronology versus radioisotopic data from SCC is similar to that encountered at Cathedral Cave, a paleontological site situated directly across the canyon from SCC, where there were significant differences between biochronologic age estimates and initial radioisotopic ages for the site (see Mead et al. 1992; Bell 1995; Jass 2005; Jass and Bell in press). Fine-scale (5 cm levels) excavation, uranium-series analysis, and paleomagnetic analysis in relation to the occurrence of arvicolines throughout the excavated sedimentary levels has contributed to resolving chronological discrepancies at Cathedral Cave (Jass 2007; Jass and Bell in press). If comparable sedimentary levels still exist within SCC, additional radioisotopic dating (e.g., AMS on bone collagen and charcoal) and further fieldwork focused on micro-sampling the reddish-pink silt zone might help clarify the distribution of Microtus meadensis and Microtus paroperarius within the sedimentary successsion of SCC.

A fourth possibility is that the specimens represent population variants of Microtus and are chronologically uninformative. Variation in the dentition of extant and fossil Microtus is common and widely recognized within species (e.g., Paulson 1961; van der Meulen 1978; Guilday 1982; Weddle and Choate 1983; Martin 1987, 1993; Harris 1988; Barnosky 1990, 1993; Pfaff 1990; Bell and Repenning 1999; Gordon 1999). Because morphologies that resemble Microtus meadensis and Microtus paroperarius are known from extant populations, the possibility that the SCC specimens represent population variants of extant taxa must be acknowledged. I am hesitant to accept such an interpretation because of the presence of similar morphotypes found in association with other extinct taxa (e.g., Mictomys cf. M. meltoni or M. kansasensis) at Cathedral Cave, situated across the canyon from SCC, and observations of arvicoline m1s in the back dirt pile of SCC.

Both Microtus meadensis and Microtus paroperarius are known in higher percentages (2.4% and 6.6%, respectively) relative to the total number of arvicolines at nearby Cathedral Cave (Jass 2007; Jass and Bell in press). Independent age estimates indicate a maximum age between 146.02 ± 2.584 - 153.7 ± 6.4 Ka for the Cathedral Cave fauna (Jass 2007; Jass and Bell in press). The data from Cathedral Cave indicate the presence of both taxa in Ssmith Creek Canyon during the Pleistocene. Although the abundance of these taxa in SCC is lower than in Cathedral Cave, it seems reasonable to infer that Microtus meadensis and M. paroperarius may be represented by the morphologies described above.

The recognition of abundant arvicoline m1s in the backdirt pile of SCC (Jass personal observation; Mead et al. 1982, 1992; Bell and Mead 1998) suggests that many excavated arvicoline specimens were missed in the original collection process. As pointed out by Miller (1979), much of the osteological material was recovered using 0.25-inch (6.35 mm) mesh screens, and this likely contributed to the loss of many arvicoline specimens. Therefore, the sample examined for this study is likely biased toward larger specimens. There is no explicit reason to suspect that this contributed to bias in the recovery of one morphotype versus another, but it does increase my reluctance to classify the low percentages of Microtus meadensis and Microtus paroperarius morphotypes as morphological variants of extant species of Microtus. At the least, the use of this taxonomy calls attention to the specimens and the variation in the m1 of Microtus from SCC. Future research at SCC and/or further evaluation of dental variation in Microtus may alter this taxonomic interpretation. Re-screening of the backdirt piles for arvicoline specimens would also produce additional materials that could be used to further evaluate the rarity of m1s of Microtus meadensis and Microtus paroperarius.

A final possibility to explain the conflicting age data is that the reddish-pink silt layer represents a broadly time-averaged unit where older fossil forms (e.g. Microtus paroperarius) occur with more recent specimens. Multiple processes (e.g., mixing, slow accumulation rates) can produce broadly time-averaged deposits, and scales of time-averaging in cave deposits can vary dramatically from hundreds of years (Hadly 1999; Hadly and Maurer 2001) to thousands of years (Barnosky et al. 2004). Because time-averaging can vary over such a broad scale in cave deposits, the evaluation of independent data sets with respect to questions concerning site chronologies is essential.

Specific resolution of the non-congruence of biochronologic and radiocarbon data at SCC is not possible at this time, but the recognition that there is disagreement in these data sets has broad implications for other studies that would seek to incorporate paleontological data from the site, and the reddish-pink silt stratum in particular. The inconsistency in the chronological data sets from SCC illustrates the complexity in attaining chronologic resolution throughout fossil sequences preserved in caves.

Caves represent complex depositional settings (Sutcliffe 1970; Gillieson 1996) and SCC is no exception. Multiple processes (anthropogenic, biologic, chemical, and geologic) contributed to the deposition of sediments at the site, and served to alter and modify the deposit (Bryan 1979). In a sense, sedimentary deposits like the ones preserved in SCC represent a microcosm of the larger-scale sedimentary processes that shape larger, open-air deposits. Some geologic features observed on broad scales in outcrop (e.g., unconformities) may occur within sedimentary deposits within caves. However, because of spatial restrictions in caves, the geologic processes that impact cave deposits seem magnified relative to similar processes in larger, open-air, sedimentary systems. When geologic processes, such as bioturbation, occur in the restricted geographic space of a cave deposit like SCC, the complexity of the deposit potentially increases more rapidly than a similar process occurring over a larger area (e.g., extensive, open-air deposits).

Similarly, the impact of conducting an excavation within a cave deposit can have more significant long-term impacts on future work than is seen in many other field settings. For this study, my capability for accurately interpreting the chronologic implications of the arvicoline fauna was directly related to previous work on the site, including collection methods, the reporting of site stratigraphy, and the retention of provenience information associated with individual specimens. Returning to SCC to conduct additional fieldwork may not result in comparable data to those considered here. Once a portion of a cave deposit is removed, a unique piece of geologic history is taken out of the record, with no guarantee that sediments left behind for future generations to examine represent comparable or even time-equivalent data sets. Previous research has clearly shown that caves may contain multiple unique deposits of differing ages across distinct portions of a cave (e.g., Lundelius 1985). Therefore, careful attention to detail in the reporting of cave excavations and recording of associated data (e.g., provenience data) is important to ensure that new analytical techniques or approaches can be accurately considered within the context of previously developed stratigraphies or chronologies.

My re-examination of the arvicoline fauna from SCC does suggest a potentially unique situation with respect to North American cave deposits, in that the site may preserve fossiliferous sediments of disparate age in vertical succession. This sequence may represent something real about the nature of fossil preservation in cave deposits, or may reflect a relatively short (geologically speaking) existence of such deposits in western North America. Methodological limitations for attaining accurate radioisotopic age control on sites that pre-date radiocarbon (i.e., > 60,000 yr B.P.) may be contributing to a skewed understanding of the age of fossils preserved in North American caves. However, I hypothesize that it may be (in part) an artifact of the abundance of cave deposits that contain terminal Pleistocene-Recent sediments and the heavy research emphasis from both paleontologists and archaeologists on questions relating to deposits of this age. Perhaps the paucity of sites containing pre-terminal Pleistocene sediments is partially an artifact of research bias or interest. Whatever the reason, the presence of Microtus paroperarius and Microtus meadensis at SCC should serve as a reminder that caves are complex depositional systems, and that researchers must at least be aware of the potential for significant time-averaging and/or disconformities in the vertical sequences excavated from them.