The nutty world of hazel names – a critical taxonomic checklist of the genus Corylus (Betulaceae)

Hazelnuts (Corylus L.) are the source of one of the globally most important nut crops. Despite their economic and cultural importance, taxonomic knowledge is poor, even the number of species is equivocal. Weak morphological differentiation, the inconsistent taxonomic treatment of horticultural selections and cultivars, and uncritical regional treatments generated a multitude of names. The situation is further complicated by an ancient history of use (at least 10 400 years), trade (at least 4000 years) and domestication (at least 2000 years). Here, we present an annotated checklist of the taxa in the genus Corylus based on an extensive literature review, electronic database research, and visits to some European herbaria. Full citations are given for all names, typifications are provided for the majority of them. Cultivars are listed if described under the rules of the ICN. We designate lectotypes and neotypes for 28 names, and discuss the identity of enigmatic C. maxima Mill., a taxon not known from the wild.


Introduction
Species of the genus Corylus L. (Betulaceae Gray) are distributed in the Northern Hemisphere and grow as shrubs or trees.The genus is readily recognized by its characteristic fruits, with a large, oil-rich seed enclosed in a brown pericarp, which in turn is enveloped by a large, tubular or foliose involucre.The detailed morphology of the involucre, the shape and size of the nut, and the leaf indument are the most important morphological traits in species delimitation.Human use of hazelnuts has an ancient history across its distribution range, and dates back at least to 8400 years BC in western and central Europe (Holst 2010;Zohary et al. 2012;Antolín & Jacomet 2015), North America (Abrams & Nowacki 2008;Fine et al. 2013) and China (Keightley 1983;Liu & Chen 2012).There is evidence that hazelnuts have been traded at least regionally as early as 4000 years ago (Fairbairn et al. 2014) and cultivation and domestication probably developed independently in the western and the eastern Mediterranean at least 2000 years ago (Pliny the Elder 77-79; Boccacci & Botta 2009).Today, hazelnuts are one of the most important nut crops world-wide with more than 850 000 tons produced in 2013 (FAOstat 2013)

R e s e a r c h a r t i c l e
most of the commercial production in Turkey.In spite of the economic importance, the taxonomy of Corylus is poorly understood and even the wild ancestors of the cultivars are still enigmatic.One of the most mysterious species in this context is C. maxima Mill.It has been variously reported as a wild species occurring in Lombardy (hence the name "lamberts nut") (Gesner in Cordus 1561: 249r) and in the Balkans (Bobrov 1936a;Săvulescu 1952;Tutin & Walters 1993), yet this name is neither typified nor are unequivocal wild populations known.This is especially surprising because C. maxima is the species name employed for a wide range of commercial cultivars and is commonly regarded as (one of the) progenitor(s) of the cultivated hazelnut in Europe (Kosenko 2002).Also, in the context of wild crop relatives, for example, in attempts to use resistant strains for breeding against the various diseases, such as eastern filbert blight or hazel dieback "moria", a clear understanding of the systematics of Corylus would be highly desirable.
The delimitation of Corylus is undisputed due to the easily recognizable nut.Corylus is sister to the other genera of the Coryloideae: Carpinus L., Ostrya Scop., and Ostryopsis Decne.(Forest et al. 2005;Li 2008;Grimm & Renner 2013).The infrageneric classification varies, however, depending on whether tree-like growth is regarded as an important character (Schneider 1916) or if involucre morphology is considered more important (Spach 1841;Candolle 1864;Bobrov 1936a;Li & Cheng 1979).Undisputed, however, is the acceptance of the group of hazels with spiny bracts (C.ferox Wall., C. ferox var.tibetica (Batalin) Franch.) as C. sect.Acanthochlamys Spach (Spach 1841).Whitcher & Wen (2001) place the remainder into a single second section, C. sect.Corylus, based on ITS data with three subsections: tree hazels (Colurnae C.K.Schneid.),hazels with tubular bracts (C.subsect.Siphonochlamys (Bobr.)P.C. Li), and Corylus (shrubs with leafy bracts).These relationships are also supported by crossing experiments: Erdoğan (1999) and Erdoğan & Mehlenbacher (2000) found incompatibility or only unilateral compatibility especially between species of different subsections, while C. americana This study aims at clarifying the published names in the genus Corylus as another step towards a taxonomic re-evaluation of Betulaceae subfam.Coryloideae (Holstein & Weigend 2016a, 2017).We aim at a complete listing of the published species and infraspecific names under the provisions of the ICN (McNeill et al. 2012) for Corylus, providing typifications where required and where possible.Since many cultivars have been described as biological taxa, we also list cultivars.However, this paper does not intend to list all cultivars or to evaluate which cultivars are to be accepted.It intends to provide their correct names according to the rules of the ICN or ICNCP (Brickell et al. 2009) respectively if the cultivars are to be accepted.

Material and methods
We adopted the list of names from the World Checklist of Selected Plant Families (WCSP 2016) and supplemented it with overlooked names by an extensive internet search and names given on herbarium specimens.The acceptance status was compared with those on the World Checklist and Bibliography of Fagales (Govaerts & Frodin 1998) and the relevant floristic treatments: Flora of China (Li & Skvortsov 1999), Flora Sichuanica (Editorial Board of the Flora Sichuanica 2012), Flora Europaea (Tutin & Walters 1993), Flora of Japan (Ohwi 1965), Flora of North America (Furlow 1997), the Flora of the USSR (Bobrov 1936a), and the last genus-wide revision (Bobrov 1936b).For further names, we surveyed The Plant List (http://www.theplantlist.org),IPNI (http://www.ipni.org),Tropicos (http://www.tropicos.org),and IFPNI (http://fossilplants.info/).Synonyms are listed here according to WCSP (2016), or if not listed there, then according to the species the taxon is currently assigned to.We also list cultivar names if published under the provisions of the ICN (McNeill et al. 2012).The names of Dochnahl (1860) applicable to Corylus are ignored, as the work was added to the opera utique oppressa (cf.Greuter & Holstein 2016;Holstein & Greuter 2016;Wilson 2017).
European Journal of Taxonomy 409: 1-75 (2018) C. subg.Phyllochlamys sect.Monophyllon Bobrov, Sovietskaia Botanica 1936 (1): 32 (Bobrov 1936b).-Type: not designated (the section included C. colchica Albov, C. maxima Mill. and C. pontica K.Koch).C. subg.Siphonochlamys Bobrov, Sovietskaia Botanica 1936 (1): 32 (Bobrov 1936b) Ward (2007Ward ( : 1099) ) and here corrected: A00247927!; isoneo-: NY01185413!).-Note: Walter's Herbarium was in the hands of John Fraser's family and later sold to BM (Stafleu & Cowan 1976-1988), but apparently lost.Ward therefore designated the neotype at the Gray Herbarium (GH), but the specimen is actually part of the collection of Arnold Arboretum (A).As both collections are housed in the same building, we correct Ward's minor mistake.There is no type locality mentioned in the protologue.The name is recognized as a species in the Flora of North America (Furlow 1997) and World Checklist and Bibliography of Fagales (Govaerts & Frodin 1998 Petzold was in contact to Schlechtendal in HAL (Stafleu & Cowan 1976-1988), possibly original material may be found there.(Jarvis et al. 1993: 38): Fuchs (1542: 397, 398) [icon]).-Note: The lectotypification by Browicz (1972: 97) in Rechinger's Flora Iranica (LINN 1132.1) is superseded because: 1) the chosen specimen featured the note "byzantinus", which Sell interpreted to be in conflict with the protologue, and 2) it may be post 1753 material.However, it would have been a better match for interpretation, because it actually shows the one character Linnaeus uses to separate the species from C. colurna, this being the ovate, obtuse stipules.The lectotype chosen by Sell, on the other hand does not show stipules, but instead a short cupula, a character that is accepted as typical for the wild C. avellana.As the type of the genus, C. avellana L. is to be accepted as long as the genus itself is accepted.
It apparently differs from C. avellana var.avellana by double-serrate bracts that surpass the nut and are apically prominently veined.This morphology is also widely found elsewhere and particularly common in cultivars.Therefore Kosenko (2002) hypothesized that the domesticated hazel might be a hybrid involving or directly selected from C. avellana var.pontica.The eastern origin of hazel cultivation seems to find support in the ancient Roman name "nux pontica", supposedly first introduced from Pontus (Pliny the Elder 77-79).Koch neither mentions the Romans, nor any cultivars.Koch mentions that the nuts resemble those of C. colurna and refers to a wild collection.However, given that hazels freely intercross within the subsections (Erdoğan 1999;Erdoğan & Mehlenbacher 2000) and that hazels have been bred for centuries, the persistence of a "pure" genotype "var.pontica" in the wild (if it ever existed) is more than doubtful.C. pontica var.glandulifera K. Koch (Li & Skvortsov 1999), in the Flora Sichuanica (Editorial Board of the Flora Sichuanica 2012), and in the World Checklist and Bibliography of Fagales (Govaerts & Frodin 1998).The material for the description of this "species" was only observed in cultivation and according to the author, it already had disappeared 10 years before publication.(Bobrov 1936a) and in the World Checklist and Bibliography of Fagales (Govaerts & Frodin 1998).According to Bobrov (1936a), the type is in G, whereas a duplicate is found in LE.However, it is not mentioned, which of the two syntype specimens is to be regarded as the lectotype (Bobrov 1936a: 267), therefore, we designate one.The existence of a LE duplicate could not be confirmed (LE, pers.com.).

Corylus colurna L.
Species Plantarum: 999 ( (neo-, designated here: DWC!).-Note: Humphry Marshall's original collection is located in DWC with some specimens also in BM (Stafleu & Cowan 1976-1988).However, there is apparently no extant specimen that can be regarded as original material (S.E.Bartholomew-Began, pers.com.).The specimen chosen here as neotype is from Marshall's Garden, but collected about 50 years later.
As it contains a fruit in its striking horn-like involucre, it allows unambiguous identification.
European Journal of Taxonomy 409: 1-75 (2018) 7a.Corylus cornuta Marshall subsp.cornuta Note: it has been recognized as having two subspecies by Flora of North America (Furlow 1997) and World Checklist and Bibliography of Fagales (Govaerts & Frodin 1998).-Note: the first author did not find any original material in the relevant herbaria.In order to explore which material might be taken into account, we briefly discuss the origin of the name here.As there is no explicit indication of an author, the name is to be attributed to Aiton (Art 46.8. ICN, McNeill et al. 2012).However, the original author of C. rostrata is apparently Daniel Carlsson Solander.Although the third volume was mainly edited by Dryander (Stafleu & Cowan 1976-1988), Dryander largely used Solander's manuscripts (Britten 1912) (Gronovius 1762).Material from Gronovius is usually found in the Clayton Herbarium (Stafleu & Cowan 1976-1988), but there appears to be no original material of this taxon.Also, the number cited for Clayton's herbarium in Gronovius' Flora Virginica ed. 2 is erroneous, because Clayton 747 is a Juglans sp., which is already cited on the previous page.There is a specimen in BM (BM001191197) by Peter Collinson (1694-1768) from Mill Hill, where Collinson lived from 1749 on (Living & Braithwaite 2013).Collinson obtained material from Gronovius for cultivation (Reveal & Pringle 1993), and Collinson was also in contact with Solander.Thus, the specimen might be derived from Gronovius' collection.In addition, as a student of Linnaeus, Solander had access to his herbarium, so the LINN specimen (LINN 1132.2) might also be taken into consideration.However, neither Collinson's specimen nor the one from the Linnean herbarium can unambiguously linked to Solander, so they are not original material, making a neotypification necessary.Due to be possible link and the completeness of the specimen, we choose the specimen from Collinson's garden as neotype.C. cornuta f. inermis Fernald, Rhodora 38: 76 (Fernald 1936) were likely collected at different localities.Burkill, however, only cited one of these gatherings, being "gorges of Lanho near Lankong".The collections from "les bois à Yang in Chan près Lankong" from the same date (K!, P06810941!) are not syntypes, but may be regarded as original material.However, they differ in being somewhat tomentose and lacking glandular trichomes on the petioles and the lower leaf lamina, linking to C. yunnanensis.The specimen chosen as a lectotype here shows the double-serrate bract margin as a name-giving character for this taxon, as well as the young male catkins that typically develop in C. heterophylla as the fruits mature.C. kweichowensis Hu, Bulletin of the Fan Memorial Institute of Biology n.s., 1: 149 (Hu 1948) (Bobrov 1936a), Flora Europaea and World Checklist and Bibliography of Fagales (Govaerts & Frodin 1998).The absence of both a physical specimen and the direct citation of an illustration (Art. 7.7, McNeill et al. 2012) makes the interpretation of this name difficult.The rather cryptic description of the stipules being intermediate between C. avellana and C. colurna makes it rather hard to understand the exact intention of the author.Miller wrote that "red and white Filberts, both which are so well known, as need no description" (and was definitely wrong in that point).Solely his citation of Bauhin's descriptive name "Corylus sativa fructu oblongo," gives some indication of what was meant, but the most important character, the involucre of the nut, is not described by him.Miller's idea of C. maxima is clearly connected to the filbert.The word is said to be either derived from "full beard" (Loudon 1838; Goeschke 1887) or from an old English name (Loudon 1838).There are two Saints of the name Philibert: Philibert of Jumièges, whose feast day is HOLSTEIN N. et al., Taxonomic checklist of Corylus on 20 th August, and Philibert of Toledo, whose feast day is on 22 nd August (The Benedictine Monks of St. Augustine's Abbey, Ramsgate 1921), both at the time when filberts are supposed to ripen (Bunyard 1920).Filberts are generally regarded as having a long involucre, much exceeding the ripe nut in length.However, there are two morphs: 1) the involucre is often deeply split and the lobes are striate to reticulate and often bend outwards, and 2) a smooth tubular involucre enclosing the nut and producing a short beak with a laciniate apex, similar to what is found in C. chinensis.Brookshaw (1812, pl. 73) illustrates both kinds, the former as "Barcelona filbert" (also with an oblong nut) and the latter as "white and scarlet filbert."The citations under Bauhin's descriptive name all mention that Miller's "filbert" is cultivated and that the fruit is oblong and is either white or red.White and red filberts were described separately as C. avellana var.alba Aiton and avellana var.Bauhin (Art. 7.7, McNeill et al. 2012).As there are no corresponding specimens in either Kew or BM, nor a figure in Bauhin's Pinax Theatri Botanici, there is no original material.According to Bauhin's phrase name, this taxon has oblong reddish nuts, corresponding to the red filbert of Miller (1768).Rea (1665) already described the "wellknown" filberts and indicates that the "red filbert" has a red testa, but he did not mention red leaves, although he discussed leaf characters for other names.Also Dodoens (1757: 515) mentioned a red pellicule.Therefore, the mention of the red color is thus not referring to the foliage, but the testa.The application and interpretation of Borkhausen's C. rubra is confusing in this context, because

C. rostrata
European Journal of Taxonomy 409: 1-75 (2018) the same descriptive epithet was published before, but it was eventually used to describe something different.In all cases, the authors deal with filberts (in the sense of C. maxima Mill.) as they refer to the same Pre-Linnean polynomina by Bauhin.However, on the one hand, the epithet in C. avellana var.rubra Aiton (1789: 363) describes the red testa.On the other hand, Borkhausen understood the epithet to describe a plant with red or brownish leaves, while Aiton did not mention the leaves at all.A red-leaved filbert was mentioned by Du Roi (1771: 178) for the first time, but without giving it a valid name.When Du Roi dealt with Miller's Corylus maxima, he distinguished white filberts and red filberts as did Miller and later Aiton; he even cited the same polynomina by Bauhin.But in contrast to Miller and Aiton (and any other author to our knowledge), already Du Roi understood his "red filberts" to have red or brownish leaves.Since there is no information on the leaf color of Miller's and Aiton's "red filbert", it is doubtful whether Borkhausen actually meant his name to be something new.Even Goeschke (1887) mentions that the "Rote Lambertsnuss"/red filbert has slightly brownish leaves, while he accepted a purple-leaved filbert as something different.Corylus rubra is therefore interpreted best a new combination, because Borkhausen meant the same "red filbert" that also Miller and Aiton cited by using Bauhin's polynomen, although Aiton and Borkhausen referred the epithet to different characters.Since Borkhausen does not discuss green-leaved filberts or cited any filberts under another name, there is no apparent intent to describe a new species but rather to use a better-fitting name.Therefore, Art.41. 4 (McNeill et al. 2012) applies.However, Borkhausen's new combination is illegitimate, as he also cites a "Corylus maxima Münchh." in synonymy.The present authors are not aware of any publication of this name, but rather interpret this name as an "updated" citation.Münchhausen does not use Miller's binomial in the third volume of his Der Hausvater (Münchhausen 1768: 827), but refers to Miller's polynomen (Miller 1759) that Miller took up later.Münchhausen also did not publish that name when he dealt with hazelnuts again in the fifth volume of the Der Hausvater (Münchhausen 1770: 142).He omitted the discussion of the filberts at all by referring to his publication in the third Hausvater volume.After Miller adopted the binomials in the eigth edition of his The Gardeners Dictionary and published C. maxima (Miller 1768) The fourth label bears the number 2, locality as in the protologue, and the Latin and the Japanese name.It is unclear which label with a number corresponds to which specimen or where the "no.2" comes from, therefore the lectotype citation here is given rather crudely.The "f" in the transcriptions in the protologue is easily explained as misreading of the "h" of the original label in the Kurrent handwriting, which is similar to the "f".The indument of the tubular involucre varies among the different type specimens of the synonyms.) and led to missing four species in a major regional floristic treatment in Carpinus (Holstein & Weigend 2017), the consequences in Corylus are minor.But then again, the omission of Corylus potaninii, described from Sichuan, in the Flora of China (Li & Skvortsov 1999) and the Flora Sichuanica (Editorial Board of the Flora Sichuanica 2012) may be a result of nescience.
The vast majority of heterotypic synonyms is found in C. avellana var.avellana (85) followed by C. colurna (14), both from Europe, while recognition of the major taxa in North America and East Asia is less ambiguous.Many of the taxon definitions go back to characters of the involucre and the shape of the nut and combinations of the two.Thus, Henriksson described 61 legitimate subspecific taxa (plus several illegitimate and invalid names) of C. avellana to record all differences found in Scandinavian hazels.While his taxa are hardly of taxonomic value, he 1) followed the early 20 th century fashion of splitting, and 2) nicely documented the variability of fruit and involucre characters in what today is treated as a single species without hesitation.
One pervasive problem, especially in the horticultural varieties, is their large-scale negligence in botanical collections and especially herbaria.Few of the varieties have any herbarium vouchers associated, making their unequivocal identification highly problematic especially if, as in much of the older literature, neither detailed descriptions nor illustrations are provided.This is a problem that not only affects Corylus systematics.There is generally an under-documentation of wide-spread and cultivated plants (A.Rockinger (M), F. Luebert (BONN), pers.com.), with C. maxima, a species widely recognized in floras, but only known from cultivation, being a particularly striking example.The observation that Hedera helix L. (Linnaeus 1753: 202), a wide-spread cultivated and feral climber, had not been documented for more than 50 years in the vast herbarium of the Botanische Staatssammlung München (F.Schuhwerk, pers.com.) is a point in case.
While the existence of four major groups in Corylus (Whitcher & Wen 2001;Forest et al. 2005 Bobrov (1936b), and there has been a major regional update for China (Li & Skvortsov 1999).Much molecular work has been done, but so far with a focus on cultivars (e.g., Boccacci et al. 2006;Gökirmak et al. 2009;Bassil et al. 2013).The present study attempts to provide the basis for further studies on Corylus phylogeny and taxonomy by providing a detailed list of taxa recognized at one time or another.There are two overall layers to the taxonomic problems, the natural diversity in the wild, which sometimes leads to the recognition of more or fewer species and subspecies across the range of the genus, and the multiplicity of horticultural forms and selections, which have often been treated as HOLSTEIN N. et al., Taxonomic checklist of Corylus varieties and subspecies and even species under the ICN in the past.Stable and unequivocal naming of cultivars based on Latin names is going to be much facilitated by the comprehensive list of names here provided.Resolving the remaining taxonomic issues for the wild taxa outside western Eurasia might be relatively straightforward, based on critical herbarium and field studies plus modern, molecular tools.
The clear definition of the cultivated forms, virtually all from western Eurasia, will also find a new basis now, since the clear definition of the wild taxa will help to re-investigate the origins of cultivated forms.There is some overlap between the two, where the identity and origin of cultivated forms remains obscure.The majority of leaf and color variants can be easily assigned to C. avellana in the strict sense, but the large-fruited forms bred for the production of hazelnuts and "filberts" remain highly problematic.
Fruits of Corylus have here been extensively used by humans for at least 10 400 years (Holst 2010;Zohary et al. 2012;Antolín & Jacomet 2015), and there is evidence that they have been traded regionally for at least 4000 years (Fairbairn et al. 2014).They had been domesticated and were widely traded during the Roman empire 2,000 years ago (Pliny the Elder 77-79; Boccacci & Botta 2009).As active trade and breeding have such a long tradition in Europe and Minor Asia, different morphs and mutations were most likely taken up independently in different regions.Boccacci & Botta (2009) argue for an independent domestication of C. avellana in the western and the eastern Mediterranean, which would imply that the present-day large fruited forms do not have a single, common origin.The fruit-hazels essentially differ in two sets of characters: nut shape (spherical versus elliptical in outline) and involucres (shorter than or equal to the nut, flaring and about twice as long, or much longer and more or less beaked, i.e., contracted above the nut).These characters are more or less freely combined in different varieties.Corylus, and especially the large-fruited forms and the wild C. avellana, have been shown to be fully interfertile except when the parents share self-incompatibility alleles (Erdoğan 1999;Erdoğan & Mehlenbacher 2000), which must have contributed to some of the diversity observed.However, the large fruited forms do not seem to become naturalized anywhere to any extent, as all national and regional floras clearly emphasize (Bobrov 1936a;Săvulescu 1952), nor is there any clear evidence that wild and domesticated Corylus do hybridize noticeably where they co-occur.
Corylus maxima is a particularly curious case, because the name is widely accepted, but there is neither type material, nor are any wild collections.When Miller described the species, he mentioned that it was so widely known, that a detailed description would be superfluous.Poiteau & Turpin (Poiteau 1846) mention and illustrate it (under the illegitimate name C. sativa), yet, herbarium specimens are unknown.We only found the red-leaved C. maxima 'Purpurea' in herbarium collections, although green-leaved plants are still in cultivation (as 'Aveline Blanche Longue', 'White Filbert', and 'Aveline rouge' or 'Red Filbert').Many "Corylus maxima" collections, however, are misidentified and might be treated under the name C. avellana var.grandis in so far as they have large, but round nuts.The application of formal Latin names for cultivated plants causes a lot of taxonomic uncertainty, i.e., because they tend not to be associated with the documentation usually provided for new species, such as detailed diagnoses and herbarium vouchers.Dochnahl's singular attempt of naming cultivars separate from biological taxa (Greuter & Holstein 2016) was not a productive effort to resolve this issue.By tracking the sparse evidence from old literature due to absence of type material, we were able to reconstruct the concept behind Miller's C. maxima.
This work is licensed under a Creative Commons Attribution 3.0 License.
(Govaerts & Frodin 1998)morin believed that the seeds were from China and had been originally collected by P.G.Farges in 1904.The lectotype was chosen because the upper right leaf is the one shown in the illustration accompanying the protologue.The other specimen bears the unpublished name "Corylus papyrifera", but does not differ from the lectotype.Possibly, that name was a working name.C. chinensis var.macrocarpaHu,Bulletin of the Fan Memorial Institute of Biology, Botany 8: 32(Hu 1937).-Type:China,Yunnan,Wai-SiHsien, Yeh Chih wa-rung, 3400 m a.s.l., Aug. 1935, C.W. Corylus chinensis var.macrocarpa is listed as a synonym of C. chinensis in the Flora of China(Li & Skvortsov 1999), Flora Sichuanica (Editorial Board of the Flora Sichuanica 2012), and World Checklist and Bibliography of Fagales(Govaerts & Frodin 1998).
(Korallova & Panova 1971)nowledge, there is no critical revision or phylogeographic analysis existing that has tested the significance of the involucre indument.andexcludenda,butincludingillegitimateandinvalidnamesasfar as unambiguously assignable to a valid name were not listed before.The nescience (and rediscovery) of older names, such as C. avellana var.microphyllaLej., C. avellana f. peltata Buser, and C. avellana f. monstrosa Henriksson rendered later names illegitimate.Although not of significant consequence in Corylus, old names might pose a threat to the stability of currently used names.Corylus mongolica Burchardt might have challenged the well-established C. heterophylla for priority if Burchardt had accepted his species fully.Furthermore, some epithets were used on species rank and infraspecific rank despite being based on different types, but not illegitimate according toArt.53.4 (McNeill et al. 2012).Corylus avellana var.heterophyllaLoudon (1838)and C. heterophylla Fisch.exTrautv.(Trautvetter1844)andC. avellana var.tenuisLoudon (1838) and Korallova's fossil hazel C. tenuis V.V.Korallova(Korallova & Panova 1971)are quite similar; and the present authors did not find validly published combinations of the older names that would render the later name illegitimate.Although overlooked names caused the creation of a new epithet for an existing taxon that simply would have needed a status change in Ostryopsis(Holstein &  Weigend 2016a Two basic forms can be found in different described varieties of C. sieboldiana.The involucre is glabrous or very finely hairy (C.sieboldiana var.mitis and C. hallaisanensis), while it consists of dense bristly trichomes in C. sieboldiana var.brevirostris, C. sieboldiana var.mandshurica, and C. sieboldiana var.sieboldiana.The type of C. hallaisanensis differs from C. sieboldiana var.mitis in having apically coarsely irregular serrated leaves, like C. sieboldiana var.mandshurica.
; Li 2008; Grimm & Renner 2013) is barely in debate, morphological identification within these groups appears to be more challenging.On the other hand, Bassil et al. (2013) found evidence for genetic separation of only some morphologically hardly distinguishable taxa, for example C. cornuta subsp.californica (vs C. cornuta subsp.cornuta) and C. jacquemontii (vs C. colurna).The latest global taxonomic treatment of Corylus is by