The authors would have been well advised to stick to one classification system, not to mix two (Nixon 1993, Flora of China) that are not congruent (and using additional concepts, adding to confusion).

I understand, it may have been politically difficult given that the authors are Chinese, but choosing one systematic concept and sticking to it is crucial for such a study. The main conclusion, subtropical (montane?) forests draws mainly from the co-occurrence of all four sections found today in East Asia. The more temperate, and deciduous ones (sect. Cerris of subgenus Cerris, sect. Quercus of subgenus Quercus), the climatically most tolerant evergreen section Ilex, and the warm-loving section Cyclobalanopsis (addressed both as infrageneric group and subgenus) of subgenus Cerris, contributing more than half of the investigated (found?) pollen.

Already the first molecular data has challenged the putative reciprocal monophyly of 'subgenus Cyclobalanopsis' (evergreen, characteristic pollen type) and 'subgenus Quercus' (all other oaks, evergreen to deciduous, diverse pollen, some lineages with same principal pollen type – red and white oaks, sects Quercus and Lobatae include evergreen and deciduous species in the New World, modern-day) used by Nixon and some earlier classifications.

Noting that you also follow our 2009 concept ("Pollen terminology and group classification of Quercus pollen were based on references [17, 27].") of pollen-molecular defined infrageneric lineages for characterising your pollen types and have cited (hence, are aware of) the updated classification for oaks (Denk et al. 2017), both of which re-defined the subgenera including the former 'subgenus Cyclobalanopsis' as one of three sections in subgenus Cerris (at the subgeneric level, Cerris has – unfortunately – priority over Cyclobalanopsis), the mix up with the outdated traditional concept makes it difficult to cherish the notable results of the paper.

The new concept (2009/2017-papers) highlighted further that one cannot distinguish evergreen from deciduous oaks by pollen per se. [The paper also includes a bit of strange English phrasing, e.g. "this sculpture was also occurred in some evergreen pollen types", "has been detailed investigated"; not a native speaker myself, so maybe my problem only, one, I occasionally have when reading botanical articles published by PLoS ONE. With respect that PLoS charges 500$ more than e.g. PeerJ as APC for a PLoS ONE article, it may be an idea to include a proof-reading service.]

“Deciduous” vs “evergreen” pollen only applies to Eurasia (the Old World) because section Quercus (sensu Denk et al. 2017 and earlier classifications other than Nixon's) has (today) no (fully) evergreen species in this geographic region. The other three Eurasian sections, all from the same evolutionary lineage, are either consistently evergreen (sections Cyclobalanopsis, Ilex) or deciduous (the two-three species of section Cerris found in East Asia). A related remark: Molecular data also has clearly established that Cerris and Ilex are sister lineages, thus, deciduous-ness seems to be a derived feature within the (exclusively) Old World subgenus Cerris as also found in the the New World (Hipp et al. 2017). The unique (semi or sub)evergreeness of Q. suber and Q. crenata may be due to introgression from section Ilex, plesiomorph or a reversal in adaption to summer-dry climates (not found in East Asia).
[Side note, we have many fossils that may belong to section Lobatae in western Eurasia into the Ice Ages, and they may have included both evergreen and deciduous species with very similar pollen morphs. But this is no problem for East Asia].

Some needed clarifications and further comments (to make future papers more consistent and readable)

"while abundant evergreen pollen types of subgenura Quercus and Cyclobalanopsis" – should have just read "of subgenus Cerris, sections Cyclobalanopsis and Ilex" (evergreen species of Nixon's/FoC "subgenus Quercus" in East Asia are exclusively species of subgenus Cerris section Ilex)
Note, not whether evergreen or not, but the high proportion of section Cyclobalanopsis are an indication for (humid) subtropical conditions (southern Cwa, Cfa climates according the Köppen-Geiger system). This was the whole idea of our 2009 study, giving researches a tool to identify the actual oak lineages (reflecting molecular clades) instead of the somewhat diffuse differentiation between evergreen and deciduous oaks of uncertain biological affinity.

"some arboreal pollen (e.g., Quercus sensu lato)" - why "sensu lato"? Quercus is a well-defined, also pollen-wise, genus (even though the FoC long treated Cyclobalanopsis as its own genus).

"Quercus pollen (mainly of the subgenus Cyclobalanopsis) is a good indicator" - "Pollen of section Cyclobalanopsis is ..."

"identification of Quercus subgenus Lepidobalanus pollen" - "Quercus section Quercus".
The Japanese paper probably still followed the classification of Schwarz (1936), many Old World researchers rejected Nixon's (1993) because he wrongly included Ilex and Cerris oaks in the "white oaks", his section Quercus. The supplement to our 2017 paper, openly accessibly on figshare (https://doi.org/10.6084/m...), includes a table for synonymy of the updated (2017, fitting the molecular data) and earlier classifications.

"China is home for more than 110 species ... regions of China [26] (Fig 1)." - It would have been more sensible to give the numbers and info (evergreen, deciduous) for the sections, fitting the found and diagnosed pollen types.

"Pollen of subtropical and tropical taxa, primarily composed of Quercus pollen, is a dominant arboreal component of the pollen record of the MD05-2906 core" – you mean subtropical 'to' tropical taxa? Very few oaks penetrate the tropics (A-climates following Köppen-Geiger), but many (Chinese) tropical taxa extend into the subtropics (matching fairly Cfa and Cwa climates, in China southern Cwa only). In this context, see also Grímsson et al. (2016; open access), providing 'Köppen signatures' and Schroeder interpretations for various Fagaceae including infrageneric oak lineages (now sections, except for Group Quercus, which we divided into three sections, Quercus, Virentes and Ponticae, to better reflect the situation seen in oligogene/phylogenomic trees).

"Infrageneric groups of Quercus pollen, as well as their type, evergreen or deciduous, were tried to determined based on the comparisons with size and sculpture of modern oak pollen." - the determination whether evergreen or deciduous is (and should be) in this case a direct consequence of determining the infrageneric group (i.e. sections formalised in Denk et al. 2017).

During Results and Discussion “infrageneric group XXX” should have been generally replaced by “section XXX” with respect to the cited Denk et al. (2017), which formalised (including a substantial update) the used informal infrageneric groups of 2009. (Side-note: Being informal, the labels should not be italicised for “infrageneric group XXX”, says the Botanical Code).

"one pollen is with very scabrate verrucate sculpture" - this makes no sense, pollen-nomenclature wise. Scabrate means the sculptural elements are too small to be classified. Verrucate means well visible (>1 µm) 'warts'. Probably should be "scabrate to micro-verrucate" or “faintly verrucate”

"This sculpture corresponds to the Quercus infrageneric group Quercus [17]." - "... is typical for Quercus section Quercus in East Asia"
(Note, in the Americas, the related, exclusively evergreen section Virentes has the same basic ornamentation; as has the deciduous sister section Ponticae, a rare western North America-western Eurasia disjunct with two threatened species; verrucate pollen are also found in section Lobatae, the ‘red oaks’ and fourth section in subgenus Quercus sensu Denk et al. not Nixon etc.)

"Current studies have shown that ornamentation of Quercus pollen can be used for accurate identification up to the level of infrageneric groups or sections but not species [16–17, 19–20, 29]. Therefore, we focused on distinguishing the subgenera Quercus and Cyclobalanopsis...". This is illogical, why focussing on (obsolete) subgenera (highest intrageneric taxonomic level), when determination can be accurately done to sectional level (the next lower taxonomic level)? Sectional level is important because the (partly distant-related) groups within the traditional subgenus Quercus (sensu Nixon etc.) have different climate preferences in East Asia.

"The fact that these species are all evergreen suggests that the group Ilex is indicative of the evergreen subgenus Quercus (Fig 3) [15]." - This sentence makes no sense. "Thus, pollen of (subgenus Cerris) section Ilex reflects/indicates evergreen oaks."

"A few species are semi-evergreen or evergreen, such as Q. crenata and Q. suber" – This implies there are more, but there aren’t: "Two Mediterranean species (Q. crenata, Q. suber) are semi-evergreen to evergreen"

"was attributed to the Quercus infrageneric group Quercus, derived from the subgenus Quercus [17]" - "assigned to Quercus section Quercus." (the second half-sentence makes no sense)

"This group includes pollen from mostly deciduous species such as Q. aliena, Q. dentata, and Q. fabrei" - I may be wrong, but I think all East Asian species of subgenus Quercus sensu Denk et al. non Nixon/FoC (including section Quercus) are deciduous.

"In our samples, most of pollen with verrucate and micro-verrucate sculpture is relatively small in size (~25 μm) and likely not of a typical deciduous type. However, the diagnostic character of pollen sculpture strongly suggested that both pollen types belonged to deciduous." – This shows that the traditional distinction of evergreen vs. deciduous oaks should be discontinued (as you did, by using SEM-determination for higher taxonomic resolution, and assigning the pollen to the 2009 infrageneric groups, now sections).

"The rodlike vertical structure is typical for the infrageneric group Cyclobalanopsis [20]. Many species in the subgenus Cyclobalanopsis have this type, such as ..." - "... is typical for section Cyclobalanopsis, including ... "

"... is attributed to the section Protobalanus, most of which is mainly distributed in the New World [17]" - "... is assigned to section Protobalanus, with today c. 5 species that are exclusively found in western North America". Pollen of section Protobalanus is, however, distinct from what you show here as "rodlike masked".

"In China, grouping the Quercus fossil pollen into evergreen and deciduous types is significant for interpreting the Quaternary dynamics of temperate and subtropical climate" and later on "Subgenus Cyclobalanopsis and subgenus Quercus pollen with rodlike and rodlike masked sculptures were evergreen (66%) (Fig 7C); therefore, such pollen types indicate the presence of subtropical forests." - what would be significant is what you did, identifying four lineages, two evergreen (sects Cyclobalanopsis and Ilex of subgenus Cerris) and two deciduous (in East Asia) lineages (sect. Cerris of subgenus Cerris, sect. Quercus of subgenus Quercus), because
— the first evergreen component is a subtropical, marginally tropical as you also point out, lineage (found in Köppen's A, Cfa, Cwa climates),
— the second evergreen component is a subtropical-temperate one (Cfa/Cwa to ~Dwb, e.g. Q. semecarpifolia), and
— the last two deciduous components fairly temperate to cold-temperate (extending into Cfa and especially the northern Cwa, mostly Cfb to Dwc).
At mid- to high elevations in the Himalaya foothills, you will mainly find evergreen oaks, too, despite the (more) temperate climate. Because section Ilex has here a wider distribution (altitudinal and in range) than section Cerris (in contrast to Northeast Asia, where Ilex is much rarer and much more restricted). But no/very few Cyclobalanopsis. Only the other way the evergreen/deciduous argument works: You wouldn't expect a pollen of the (in East Asia deciduous) sections Cerris and Quercus in (near)tropical setting. All four, with a dominance of Cyclobalanopsis, are indeed typical for subtropical settings (as e.g. found up to c. 1500 m in the Cfa zone of central China or the southern Cwa zone; cf. Feng et al. 2009, see also the altitudinal ranges recorded for the Shennongja Forest District, Zhu & Song 1999, which lists 13 Quercus spp. covering all four sections occurring in East Asia, with Cyclobalanopsis only found in the subtropical altitudinal zone up to 1800 m, but others extending into temperate montane climates up to 2500 m)

In this context, Fig. 7, summarising your main result, should have been better organised: B puts c. 25 "subgenus Cyclobalanopsis" (= section Cyclobalanopsis; all evergreen) against c. 15 "subgenus Quercus" (one evergreen, two deciduous lineages); C has c. 15 deciduous visually replacing the subgenus Cyclobalanopsis bar vs. c. 25 evergreen (nearly all Cyclobalanopsis, including 2 Ilex, hence, the little difference)
Fig. 7 could have just shown the types viz sections, and as B (or just by bracketing) the summation of the deciduous vs. evergreen sections.

A last remark, the numbers strike me quite low for absolute pollen count numbers (however, it says in results "two grains", "six grains", etc.) “Accounting for 59% of total pollen” and “66%” evergreen oaks with a total sample size of c. 40 is not very meaningful, statistically speaking.

References
Schwarz O. 1936. Entwurf zu einem natürlichen System der Cupuliferen und der Gattung Quercus L. Notizblatt des Botanischen Garten und Museum, Berlin-Dahlem Bd. 13 Nr. 116:1–22.
Nixon KC. 1993. Infrageneric classification of Quercus (Fagaceae) and typification of sectional names. Annales scientifiques forestières 50:25s–34s.
Zhu Z, Song Z. 1999. Scientific Survey of Shennongjia Nature Reserve. [in Chinese script]: [in Chinese script].
Fang J, Wang Z, Tang Z. 2009. Atlas of Woody Plants in China. Volumes 1 to 3 and index. Beijing: Higher Education Press.
Grímsson F, Grimm GW, Meller B, Bouchal JM, Zetter R. 2016. Combined LM and SEM study of the Middle Miocene (Sarmatian) palynoflora from the Lavanttal Basin, Austria: Part IV. Magnoliophyta 2 – Fagales to Rosales. Grana 55:101–163. http://dx.doi.org/10.1080...
Denk T, Grimm GW, Manos PS, Deng M, Hipp AL. 2017. An updated infrageneric classification of the oaks: review of previous taxonomic schemes and synthesis of evolutionary patterns. In: Gil-Pelegrín E, Peguero-Pina JJ, and Sancho-Knapik D, eds. Oaks Physiological Ecology. Heidelberg, New York: Springer, p. 13–38. Free Pre-Print at https://doi.org/10.1101/1...
Hipp AL, Manos PS, González-Rodríguez A, Hahn M, Kaproth M, McVay JD, Valencia Avalos S, Cavender-Bares J. 2017. Sympatric parallel diversification of major oak clades in the Americas and the origins of Mexican species diversity. New Phytologist DOI:10.1111/nph.14773.