The Denisovans are a hominin lineage, first discovered in the Siberian cave of Denisova in 2010 (Reich et al., 2010). Confirmed Denisovan human remains are scarce and mostly morphologically non-indicative. Examples include a phalanx, a few teeth and a presumed mandible (Slon et al., 2017). Most Denisovan remains to date still originate from Denisova cave, where they are dated to 200–60 kya (Douka et al., 2019). While the remains are small and partial, the climate conditions in the cave enabled excellent preservation of genetic material, resulting in several low-coverage and one high-coverage Denisovan genomes (Reich et al., 2010, Meyer et al., 2012). 

​

The extracted DNA sequences revealed a distinct population, closely related genetically to Neanderthals. It is estimated that these two sister taxa diverged between 400–440 kya (Prufer et al., 2017). Unprecedentedly, this lineage was defined on the basis of genetic identity alone, with barely any known morphology. The main indicative phenotype are the extremely large molars, with the crown area being an outlier among the familiar variations of the Homo genus (Reich et al., 2010).

​

Genomic studies show that current human populations have both Neanderthals and Denisovans DNA, around 2% and 0.5% respectively (Mallick et al., 2016). The populations with the highest rates of Denisovan introgression live in eastern Asia and Oceania (up to 5% introgression in Oceanic populations). This led researchers to believe that the Denisovan population inhabited eastern Asia at the same time as the Neanderthal population inhabited Europe and western Eurasia. 

​

The lack of data regarding Denisovan morphology has proven problematic for taxonomic classification. The usual morphological practice cannot be used to confidently categorize Denisovans, and we must rely on molecular analyses. Unfortunately, DNA and protein preservation is not sufficient for such analyses in most specimens. In fact, only one human remain outside of Denisova cave has been classified as Denisovan – a partial mandible from Xiahe, China (Chen et al., 2019). Its classification was based on dental proteomic analysis, which determined the Xiahe’s Collagen to be phylogenetically closest to the confirmed Denisovan individual. 

​

The disappointing number of Denisovan specimens stands in clear contrast to the effective population size of Denisovans. Genome heterozygosity indicates very similar population size for Neanderthals and Denisovans (Prufer et al., 2014), yet Neanderthal remains are two magnitudes more abundant in the fossil record. What could cause such a discrepancy?

​

A possible reason is a historical archaeological bias towards western Eurasia and the African continent, places that were not likely to be inhabited by Denisovans. In addition, archeological finds from east Asia were often studied within Asian academic circles, sometimes not even published in English. This limited the access to east Asian specimens and prevented most researchers from directly comparing them to the rest of the fossil record. However, recent decades have seen an impressive surge of excavations and archeological fieldwork in east Asia. New and old specimens were shared with the global academic community, enhancing the understanding of the east Asian fossil record. Despite the expansion of available data, almost no additional Denisovan findings were reported, hinting at another underlying reason for the paucity in remains. 

​

The most likely reason we are left with for this discrepancy therefore, is the inability to identify the genetically defined Denisovans within the morphologically classified fossil record. This means that the current fossil record is likely to contain more Denisovan specimens, hiding in plain sight. Assuming this is the case, where should we look for them? Based on the genetic data available, the most likely candidates are Middle Pleistocene (dating after the Neanderthal-Denisovan divergence) east Asian specimens, preferably with loose taxonomic affiliation. Fortunately, there are several such cases, including the crania of Dali (Wu et al., 2013), Harbin (Ni et al., 2021), Hualongdong (Wu et al., 2019), Jinniushan (Lu 2003), Xuchang (Li et al., 2017) and Xujiayao (Wu et al., 2022). The Harbin specimen was also noted to have especially large molars, with a crown size close to that of the known Denisovan specimens (Ni et al., 2021). Another possible candidate is the mandible of Penghu (Chang et al., 2015), which was found out of context near the shores of Taiwan. 

​

Most of these specimens are regarded as archaic Homo sapiens, in accordance with the highly controversial multiregional theory. This model is held persistently by east Asian academics despite overwhelming genetic (e.g., Cann et al., 1987, Su et al., 1999) and morphological (e.g., Manica et al., 2007) evidence. I believe that the adherence to this model is also a subtle “under the surface” reason for the possibility of Denisovan identity not gaining traction. Declaring such specimens as Denisovans would contradict the multiregional model, making it problematic for many academics. 

​

Of course, the possibility of these specimens being Denisovan is based merely on circumstantial evidence. To test the identity of these specimens, we must look further than spatiotemporal overlap with the hypothesized Denisovan population. Unfortunately, without molecular analyses this is not yet possible. 

​

This intriguing possibility of Denisovan identity will remain a hypothesis until more solid evidence surfaces. The obvious key is a Denisovan fossil with both genetic material and morphological preservation. As the east Asian fossil record keeps expanding, such a specimen is likely to be found. The environmental conditions for adequate preservation are found in parts of east Asia, such as the Tibetan plateau. The protein preservation state of the Xiahe dentition is a prime example of this potential.

​

With the supervision of Prof. Liran Carmel (HUJI) and Dr. David Gokhman (Weizmann Institute), and in collaboration with Dr. Gadi Herzlinger (HUJI) and others, we are pursuing a slightly more indirect approach to detect this elusive hominin. Based on the epigenetic anatomical reconstruction of Denisova 3 (Gokhman et al., 2019), we are currently scanning the available fossil record for crania that match the Denisovan profile. With exciting initial results, I believe our study could make the so called “muddle in the middle” slightly less muddy.