Buckley Et Al 2018 Ignotornis Canadensis

Cretaceous Research 84 (2018) 209e222

Contents lists available at ScienceDirect

Cretaceous Research

journal homepage: www.elsevier.com/locate/CretRes

First report of Ignotornidae (Aves) from the Lower Cretaceous Gates Formation (Albian) of western Canada, with description of a new ichnospecies of Ignotornis, Ignotornis canadensis ichnosp. nov.

* Lisa G. Buckley a, , Richard T. McCrea a, Lida Xing b a Peace Region Palaeontology Research Centre, Tumbler Ridge, British Columbia V0C 2W0, Canada b School of the Earth Sciences and Resources, China University of Geosciences, Beijing 100083, China article info abstract

Article history: Ignotornidae is an avian vertebrate ichnofamily known from the Lower to middle Cretaceous (Aptian e Received 1 August 2017 Cenomanian) of North America (Colorado) and South Korea. The ichnogenus Ignotornis was first described Received in revised form from North America; however, in the eight decades since no Ignotornidae have been described from North 21 November 2017 America, and hitherto none have been reported from Canada. Here we describe the first ichnospecies of Accepted in revised form 23 November 2017 Ignotornis, Ignotornis canadensis ichnosp. nov. (Gates Formation, Lower Cretaceous, Albian), the first new Available online 24 November 2017 ichnospecies of Ignotornis to be described from North America since the original description of Ignotornis mcconnelli, and the first report of Ignotornidae from Canada. Ignotornis canadensis ichnosp. nov., while Keywords: Ignotornis canadensis morphologically similar to Ignotornis mcconnelli, is larger and has a relatively longer digit I impression than Early Cretaceous Ignotornis mcconnelli, and has digit I impressions with a similar orientation to those of the ichnogenus Albian Goseongornipes. Discriminant analyses show that Ignotornis canadensis ichnosp. nov. is most similar to the Gates Formation ichnogenera Goseongornipes and Hwangsanipes due to the divarication of digits IeII and IIeIII. The Ignotornidae apparent sinusoidal digit III impression in Ignotornis canadensis and other Ignotornidae is due to water Avian ichnology saturated sediment not completely preserving digital pad impressions, rather than a morphologic feature of digit III. While the majority of ichnospecies of Ignotornidae are from South Korea, and while Asia currently contains the largest diversity of avian ichnotaxa from the Cretaceous Period, the apparent lower diversity of Cretaceous avian ichnotaxa from North America may be preservationally controlled, and will be resolved with a greater focus on Early Cretaceous ichnofauna in North America. © 2017 Published by Elsevier Ltd.

1. Introduction (Kim et al., 2006), were assigned. Ignotornidae was established to distinguish these ichnotaxa from other tetradactyl Mesozoic bird Ignotornis mcconnelli was the ﬁrst bird track described from the tracks, as Koreanaornipodidae have an inconsistently preserved, Mesozoic (Mehl, 1931; Lockley et al., 1992) from the upper part of short digit I impression (Lockley et al., 2006; Xing et al., 2016)To the regressive-transgressive cycle of the Kassler Sandstone and Van date, only one ichnospecies of Ignotornidae, Ignotornis mcconnelli, Bibber Shale of the Dakota Sandstone (Lockley et al., 1992). The has been described from North America (Table 1). unique morphology of Ignotornis mcconnelli (wide divarication, asymmetric semipalmate webbing traces attached more mesially 1.1. History of avian track descriptions from western Canada on digit impressions IIIeIV than on digit impressions IIeIV, prom- inent posteriomedially-directed digit I) prompted establishing a The ﬁrst avian track type described from Canada was Aqua- new ichnofamily, Ignotornidae, in to which Goseongornipes mark- tilavipes swiboldae from the Gaylard Member of the Gething For- jonsei (Lockley et al., 2006), Hwangsanipes choughi (Yang et al., mation of British Columbia (Currie, 1981). The Gething Formation 1995), Ignotornis gajinensis (Kim et al., 2012), and Ignotornis yangi (Lower Cretaceous: Aptian) is stratigraphically below the Gates Formation. Aquatilavipes swiboldae is a small tridactyl avian trace: to date, no hallux impressions have been described from any ich- * Corresponding author. nospecies of Aquatilavipes (Buckley et al., 2016). An ichnospecies of E-mail address: [email protected] (L.G. Buckley). Aquatilavipes was described from the Grande Cache Member of the https://doi.org/10.1016/j.cretres.2017.11.021 0195-6671/© 2017 Published by Elsevier Ltd. 210 L.G. Buckley et al. / Cretaceous Research 84 (2018) 209e222

Table 1 Geographic distribution of Ignotornidae from the Cretaceous. While Ignotornis mcconnelli (Mehl, 1931)wasﬁrst described from North America, the majority of Ignotornidae are known from Lower Cretaceous deposits of South Korea, with the exception of Hwangsanipes choughi (Upper Cretaceous.) The Uhangri Formation is described by Yang et al. (1995) as being Upper Cretaceous, and the deposits from which Hwangsanipes choughi was recovered reﬁned by Kim et al. (2006) as being Cenomanian or younger. Although AptianeCenomanian deposits are not uncommon in North America, Ignotornidae (and Cretaceous bird tracks in general) remain underreported.

Ichnotaxon Geographic Location Formation, Age

Hwangsanipes choughi Hwangsan Basin, South Korea Uhangri Formation, Upper Cretaceous (Cenomanian) Goseongornipes markjonesi Kosong County, South Korea Jindong Formation, Lower Cretaceous (Albian) Ignotornis mcconnelli Golden, Colorado, USA Dakota Group, Lower Cretaceous (AptianeCenomanian) Ignotornis canadensis ichnosp. nov. Peace Region, northwest Alberta and Gates Formation (Grande Cache Member), Lower Cretaceous northeast British Columbia, Canada (Albian) Ignotornis gajinensis Gajin, South Korea Haman Formation, Lower Cretaceous (Aptianemiddle Albian) Ignotornis yangi Changseon and Sinsu Islands, South Korea Haman Formation, Lower Cretaceous (Aptianemiddle Albian)

Institutional abbreviationsdPRPRC, Peace Region Palaeontology Research Centre, Tumbler Ridge, British Columbia, Canada; TMP, Royal Tyrrell Museum of Palaeontology, Drumheller, Alberta, Canada; UCM, University of Colorado at Boulder Museum, Denver, Colorado, USA.

Gates Formation, Aquatilavipes curriei (McCrea and Sarjeant, 2001). preserves several small tridactyl avian tracks. During further However, as Aquatilavipes curriei is a trace from a larger trackmaker investigation, a second track type was identified that, while heavily than that of Aquatilavipes swiboldae, and because the only similarity eroded, has digit I impressions (Fig. 4). between the two track types is the lack of hallux impressions, McCrea et al. (2014) reassigned A. curriei to a new ichnofamily 2. Materials and methods (Limiavipedidae) and ichnogenus (Limiavipes). To date, there are only two avian ichnotaxa known from the Cretaceous-age deposits 2.1. Specimen collection and documentation of Canada: Aquatilavipes swiboldae and Limiavipes curriei. One slab with two prints of Limiavipes curriei was found in the Gaylard PRPRC 2014.05.001 was discovered prior to 2000 (Helm, 2000, Member of the Gething Formation (McCrea et al., 2014). A specimen 2001) near the summit of Roman Mountain, northeastern British of Aquatilavipes swiboldae was also identified from the Gladstone Columbia. It was not until 2014 that the slab was recovered by Formation (¼ Gething Formation) along the highway outside of helicopter airlift. The slab is deposited at the PRPRC. TMP Grande Cache, Alberta (McCrea et al., 2014 and references therein). 2016.036.0003 was discovered in the fall of 2016 at the base of the There are more Cretaceous avian ichnotaxa known from west- W2 track site the footwall of a former open pit coal mine (Fig. 1). ern North America, but the number is still low in comparison to Specimen data (Table 2) was collected as outlined in McCrea et al. other Cretaceous-age localities worldwide, especially in Asia (2014). Photogrammetric images of TMP 2016.036.003 (Fig. 2) (Lockley and Harris, 2010; Xing et al., 2016). In addition to Ignotornis were obtained using multiple images (61 photographs taken at an mcconnelli (Cretaceous: AlbianeCenomanian, Mehl, 1931; Lockley average of 0.472 m altitude from the subject) with a Canon EOS 70D et al., 1992), Gruipeda vegrandiunus from the Cantwell Formation camera (focal length 18 mm, resolution 5472 3648, pixel size of Denali National Park (Upper Cretaceous: Campanian- 0.00417 0.00417 mm). The photographs were processed by Agi- Maastrichtian, Fiorillo et al., 2011) and Sarjeantopus semipalmatus soft Photoscan Professional (v.1.0.4) with the model having an error from the Lance Formation of Wyoming, as well as unnamed avian of less than 0.152 pix. These models were converted to color tracks (Upper Cretaceous: Maastrichtian, Lockley et al., 2004)have topographic profile images using CloudCompare (v.2.5.3). been described from North America. There are reports of Kore- anaornis ichnosp. from the Dakota Group of Utah (Cretaceous: 2.2. Multivariate statistical analyses Albian-Cenomanian, Anfinson et al., 2009). Aquatilavipes ichnosp. has been reported from the Poison Strip Member of the Cedar Multivariate statistical analyses were performed using Paleon- Mountain Formation of Utah (Early Cretaceous: Aptian; Lockley tological Statistics (PAST, Hammer et al., 2001). Discriminant ana- et al., 2015), and the Lakota Formation (Lower Cretaceous, Lockley lyses and multivariate analyses of variance (MANOVA) were et al., 2001a). The first description of the ichnogenus Magnoavipes performed on linear and angular data of Cretaceous tetradactyl (Magnoavipes lowei, Lee, 1997) described it as the track of a large avian ichnotaxa Hwangsanipes choughi (Yang et al., 1995), Goseon- avian trackmaker. However, Lockley et al. (2001b) describe Mag- gornipes markjonesi (Lockley et al., 2006), an unassigned ichno- noavipes as an ichnogenus of non-avian theropod in their descrip- species of Goseongornipes (Lockley et al., 2006), Ignotornis gajiensis tion of Magnoavipes caneeri. Fiorillo et al. (2011) describe (Kim et al., 2012), Ignotornis mcconnelli (Lockley et al., 1992) Igno- Magnoavipes denaliensis as the track of a large avian trackmaker. tornis yangi (Lockley et al., 2006), and Uhangrichnus chuni (Yang However, Matsukawa et al. (2014) demonstrate that the ichnogenus et al., 1995). Dongyangornipes sinensis (Azuma et al., 2013)is Magnoavipes is most likely from an ornithomimid trackmaker. Also, considered by Buckley et al. (2016) as a subjective junior synonym based on comparisons of ranges of total divarication and footprint of Uhangrichnus chuni, based on the emendation of Uhangrichnus length to pace length ratios of bird and theropod trackways, the chuni by Lockley et al. (2012a, b) that included a description of a ichnogenus Magnoavipes is most likely that of a non-avian short posteriormedially-directed hallux impression (Lockley et al., theropod (McCrea et al., 2014; Xing et al., 2015). 2012a), and described the slight differences in webbing impres- McCrea (2000) and McCrea et al. (2014) outlined several novel sions and digit divarication as due to preservational variation avian track types from the Lower Cretaceous (AptianeAlbian) of (Buckley et al., 2016). Uhangrichnus chuni, although possessing western Canada whose descriptions are in progress. One specimen palmate webbing impressions as compared to Ignotornidae (sem- is a track-bearing block from the former Grande Cache Coal Mine ipalmate), was included in the analyses based on similarities in near the community of Grande Cache, Alberta (Fig.1). The specimen overall size and digit morphology. Although no hallux impressions (TMP 2016.036.003, Figs. 2e3) was collected in September 2016. were reported for Gyeongsangornipes lockleyi (Kim et al., 2013), it Two avian ichnotaxa are present on the track slab: a small tridactyl was also included in the analyses based on similarities in asym- track type and a slightly larger tetradactyl track type. A second metric semipalmate webbing and digit morphology. Shandongor- specimen was collected in August 2014 (PRPRC 2014.05.001) that nipes muxiai (Li et al., 2005) was not included in the analyses as it is L.G. Buckley et al. / Cretaceous Research 84 (2018) 209e222 211

Fig. 1. Locality map of specimens of Ignotornis canadensis TMP 2016.036.003 from the Grande Cache Coal Pit near Grande Cache, Alberta, and PRPRC 2014.05.001 from Roman Mountain, northeastern British Columbia. a zygodactyl ichnotaxon, whereas the other Cretaceous avian ich- collected from the W2 Site of the formerGrande Cache Coal oper- notaxon are anisodactyl. Ichnospecies of Koreanaornis (Kore- ation (McCrea, 2000; McCrea et al., 2014). anaornis hamanensis, Koreanaornis lii) were also included due to the The stratigraphy of the Gates Formation is well-documented. presence (although somewhat inconsistent) of hallux impressions. The strata of the Gates Formation were deposited in the Western Koreanaonris dodsoni (Xing et al., 2011) was not included due to the Canada Foreland Basin during the second of three clastic deposi- consistent lack of a hallux impression. However, ichnospecies of tional cycles in the Lower Cretaceous (Plint and Hart, 1988; Leckie Koreanaornis are morphologically distinct from those of Igno- and Smith, 1992). The Gates Formation is part of the Fort St. John tornidae due to its relatively smaller size, consistent lack of Group, which conformably overlies the Lower Cretaceous Bullhead webbing impressions, and relatively short hallux impressions (Kim, Group (Langenberg et al., 1987)(Fig. 5). The Gates Formation is 1969; Lockley et al., 2006; Xing et al., 2016). Jindongornipes kimi was divided into three members (lower Torrens Member, middle included in preliminary statistical analyses despite its larger size Grande Cache Member, upper Mountain Parks Member), with the and morphological differences from other Cretaceous tetradactyl vertebrate track-bearing Grande Cache Member as a source of tracks. For the analyses discussed herein, Jindongornipes kimi was economic coal seams in western Alberta (Langenberg et al., 1987; not included due to its much larger footprint size, and the small McCrea, 2000). sample size data (n ¼ 2). Gruipeda vegrandiunus (Fiorrillo et al., In the Grande Cache coal mines of western Alberta, the verte- 2011) and Sarjeantopus semipalmatus (Lockley et al., 2004) were brate track-bearing layer was exposed after the removal of the also excluded from the analyses due to their small sample sizes. Number 4 Coal Seam in the Grande Cache Member (McCrea and Size as a confounding factor was removed from the data prior to Currie, 1998; McCrea, 2000). The paleoenvironment is interpreted analyses, as per Farlow et al. (2013): linear data were ﬁrst log- to be that of a coastal plain or deltaic complex (Langenberg et al., transformed, and then the mean was subtracted from each of the 1987; Leckie and Smith, 1992; McCrea, 2000). The Grande Cache log-transformed linear and angular variables. Member of the Gates Formation has been assigned to the lower Albian based on the ostracod Cytheridea bonaccordensis (Langenberg et al., 1987). Paleobotanical studies on the Gates For- 3. Geological setting mation (Wan, 1996) is interpreted as warm with abundant rainfall but with possible cold seasons (McCrea, 2000). Both TMP 2016.036.0003 (Figs. 2e3) and PRPRC 2014.05.001 (Fig. 4) originate from the Gates Formation of British Columbia and Alberta, respectively (Fig. 1). The stratigraphic context of the lo- 4. Systematic ichnology cality for PRPRC 2014.05.001 is uncertain, as the block was recovered ex situ from the summit of Roman Mountain (10U 0631373 AVES 6082782, WGS 84) within the coal lease tenure of the formerPeace River Coal Trend Mine (McCrea et al., 2014). TMP 2016.036.0003 Ignotornidae (Lockley et al., 2006) was recovered by RTM under the aegis of the Royal Tyrrell Museum Ignotornis Mehl, 1931; emend. Lockley et al., 1992 in 2016 as an ex situ track slab from the Grande Cache Member of the Gates Formation (Lower Cretaceous: Albian). The specimen was Ignotornis canadensis ichnosp. nov. 212 L.G. Buckley et al. / Cretaceous Research 84 (2018) 209e222

Fig. 2. Three-dimensional photogrammetric images of TMP 2016.036.0003 showing true color rendering (top) and false color depth image (bottom). Photogrammetric images, presented here were obtained using multiple images (61 photographs taken at an average of 0.472 m altitude from the subject) with a Canon EOS 70D camera (Focal Length 18 mm, resolution 5472 3648, pixel size 0.00417183 0.00417183 mm). The photographs were processed by Agisoft Photoscan Professional (v.1.0.4) with the model having an error of less than 0.151929 pix. These models were converted to color topographic proﬁle images using CloudCompare (v.2.5.3).

Type specimensdTMP 2016.035.0003, natural cast track slab con- DiagnosisdAnisodactyl tetradactyl avian tracks of large size (foot- taining six tetradactyl tracks in a consecutive trackway, deposited print length 39.0 mme57.2 mm; footprint length with hallux at Royal Tyrrell Museum of Palaeontology, Drumheller, Alberta. 63.8 mme71.8 mm), hallux impression posteriomedially directed Referred specimen, PRPRC 2014.005.001, natural mold track slab and larger than 25% of digit III impression length. Digit II and IV containing ﬁve tetradactyl tracks in a consecutive trackway, impressions weakly curve towards digit III impression. Claws short deposited at Peace Region Palaeontology Research Centre, Tumbler and acuminate. Asymmetric semipalmate webbing, impressions Ridge, British Columbia. with larger webbing impressions between digits IIIeIV than IIeIII. Etymologyd“Ignotornis,” likely a derivative of “ichno” and “ornis” Horizon and localitydHolotype specimen is an ex situ block from from Mehl (1931), bird trace; “canadensis,” from Canada. the 12 Mine South, W2 site at base of footwall, about 21 km L.G. Buckley et al. / Cretaceous Research 84 (2018) 209e222 213

Fig. 3. TMP 2016.036.0003, natural cast track slab of type specimen of Ignotornis canadensis ichnosp. nov., from the Gates Formation of western Canada. A, Photograph of trackway showing tracks 1e5, with two partial tracks that may be part of the trackway. B, line drawing of TMP 2016.036.0003, showing the type trackway of Ignotornis canadensis ichnosp. nov. (black), and several small tridactyl bird tracks (grey) that are likely Aquatilavipes ichnosp. northwest of Grande Cache, Alberta; Grande Cache Member of the preserved, webbing impressions between digit IIIeIV impressions Gates Formation, lower Albian (Lower Cretaceous) just below the extend more distal than webbing impressions between digit IIeIII No. 4 coal seam (Langenberg et al., 1987). impressions. No webbing impressions are detected between digits DescriptiondUnless otherwise stated, the descriptions are based on IeII. Individual tracks are weakly rotated towards the midline of the tracks from TMP 2016.035.0003, as those are best preserved trackway (footprint rotation (FR), also known as angle of divarica- (Figs. 2e3, 6e7; Table 2). Tracks are between 44.7 mm and 57.4 mm tion from midline: 7 to 12). Track placement in the trackway is in footprint length (FL). Footprint lengths including the hallux relatively straight (pace angulation (PA): 160e180). Pace length impressions (FLwH) are 65.0 mme65.6 mm. Digit impressions are (PL) is short compared to footprint length (FL) (FL/PL: 0.30e0.48), narrow with inconsistently preserved digital pads. Digit III im- as is typical with bird tracks and trackways (Xing et al., 2015). pressions are the longest, with digit impressions II and IV subequal Digit III impressions appear sinusoidal in shape for track natural in length. The distal ends of digit II and IV impressions curve casts that preserve the entire depth of the track. However, this si- slightly towards the digit III impression (Fig. 5). Digit I impressions, nusoidal digit III impression is not observed in tracks that have where preserved, are greater than 25% the length of the total track been eroded to the level of the track surface: digits are straight with length. Digit I impressions are 31.8%e41.7% the lengths of digit III partially preserved digital pads (Figs. 3, 6). impressions. Average digit I impression is 27.9% of total footprint length with hallux (FLwH). Digit I impressions are posteriomedially 5. Comparative ichnology oriented, with digit IeII divarication between 55 and 71. Total divarication (IIeIV) is between 109 and 129, with the proximal Ignotornis mcconnellidThe emended description of Ignotornis edge of the track (where the digit impressions converge) appearing mcconnelli (Lockley et al., 1992, 2009) reveals that its morphology is as an obtuse angle with a rounded margin, rather than pointed. similar to that of Ignotornis canadensis ichnosp. nov. Divarication Webbing impressions are inconsistently preserved. Where IeII of Ignotornis canadensis ichnosp. nov. (average digit IeII 214 L.G. Buckley et al. / Cretaceous Research 84 (2018) 209e222

Fig. 4. PRPRC 2014.05.001, natural mold track slab with faint impressions of a tetradactyl avian trackmaker referred to Ignotornis canadensis ichnosp. nov. A, photograph of trackway, and B, line drawing of trackway. While heavily eroded, the hallux impression is still visible in Track 2. To date only tridactyl avian tracks have been described from western Canada.

Table 2 Linear and angle measurements of Ignotornis canadensis, ichnosp. nov. FR, footprint rotation; FL, footprint length; FLwH, footprint length including hallux impression; FW, footprint width; PA, pace angulation; PL, pace length; SL, stride length.

Track Footprint Digit lengths (mm) Digit divarication () PL (mm) SL (mm) PA ()FR() measurements (mm)

FL FLwH FW I II III IV IeII IIeIII IIIeIV

PRPRC 2004.05.001 .001 ? ? ? ? ? ? ? ? 64 60 ? ? ? 6.5 .002 48.5 63.9 68.8 20.24 27.1 36.4 22.9 108 65 55 115 ? ? 12 .003 50.23 71.8 67.8 21.6 32.1 39.8 40.3 105 51 50 82 185 170 24 .004 39 ? ? ? ? 31.8 ? ? 42 ? 153 230 150 18 .005 ? ? ? ? ? 32.6 ? ? ? ? 75 230 184 ? TMP 2016.035.003 1 57.42 ? ? ? 28.1 47.7 ? ? ? 42 ? ? ? 2 2 50.96 ? 57.4 ? 32.3 51.0 37.1 ? 55 54 107 ? ? 12 3 49.5 ? 61.1 ? 34.0 49.5 33.6 ? 70 55 148 251 165 7 4 48.56 65 68.8 17.6 25.6 42.7 28.5 97 59 70 158 310 180 7 5 47.3 65.6 66.7 15.1 40.4 47.3 33.1 82 55 70 160 258 163 3 6 44.72 ? ? ? 28.1 31.9 ? ? 71 57 103 260 160 7 L.G. Buckley et al. / Cretaceous Research 84 (2018) 209e222 215

Fig. 5. Geologic context of Gates Formation in western Canada, from McCrea (2000). Modiﬁed from Plint and Hart (1988). divarication ¼ 98) is close to that of Ignotornis mcconnelli (average canadensis ichnosp. nov. is 118.7 (IIeIII ¼ 57.6; IIIeIV ¼ 56.9). The digit IeII divarication ¼ 97.5;63e133). However, Ignotornis divarication between digit impressions IIeIII is relatively larger in canadensis ichnosp. nov. is larger (average FL ¼ 48.5 mm) than Ignotornis gajinensis than in Ignotornis canadensis ichnosp. nov. those described for the type of Ignotornis mcconnelli (average While the absolute values of digit divarication have potential for FL ¼ 40.9 mm) by approximately 7.6 mm on average, and 11.4 mm large variation (Buckley et al., 2015), relative differences in digit larger in track length including the hallux impression. The hallux divarications within tracks may be useful. In Ignotornis gajinensis, impressions of I. canadensis ichnosp. nov. are relatively longer the divarications between digit IIeIII are consistently larger than (average 18.6 mm, N ¼ 4, 45.3% length of digit III impression) than divarication between digits IIIeIV (Table 3). In Ignotornis canadensis those of Ignotornis mcconnelli (average 14.2 mm, 34.7% length of ichnosp. nov., divarication between digits IIeIII and IIIeIV are digit III impression, N ¼ 58; Table 3). almost equal in size. The relatively larger divarication of digits IIeIII, Ignotonis yangidIgnotornis yangi is smaller (average when compared to divarication of digits IIIeIV, differentiates FL ¼ 32.6 mm) than Ignotornis canadensis ichnosp. nov. The hallux Ignotornis gajinensis from both Ignotornis canadensis ichnosp. nop. impression of Ignotornis yangi (36.1% of digit III impression) is and Ignotornis mcconnelli. Ignotornis gajinensis also differs from smaller in relative length than that of Ignotornis canadensis ichnosp. other described ichnospecies of the ichnogenus Ignotornis in the nov. Digit IeII divarication of Ignotornis yangi (average digit IeII presence of feeding traces, similar to that seen in extant spoonbills divarication ¼ 83.9) is smaller than that of Ignotornis canadensis (Ciconiformes, Kim et al., 2012). (average digit IeII divarication ¼ 98); however, there is a large Goseongornipes markjonesidGoseongornipes markjonesi is amount of variation in digit IeII divarication (50e120). In different from described ichnospecies of Ignotornis in that the described specimens of Ignotornis yangi (Kim et al., 2006) there are hallux impression is consistently shorter in Goseongornipes than obvious webbing impressions between digits IIeIII. Ignotornis reported for Ignotornis (Kim et al., 2006; Lockley et al., 2006). canadensis ichnosp. nov. does have a visible webbing impression Average total divarication for Goseongornipes markjonesi is 124.5 between digits IIeIII (Fig. 6), but it is not as robust as in Ignotornis (IIeIII ¼ 62.9; IIIeIV ¼ 61.5; total divarication IIeIV ¼ 101e152). yangi. Ignotornis canadensis ichnosp. nov. also has a large range of total Ignotornis gajinensisdIgnotornis gajinensis is smaller in footprint divarication (101e129). The webbing impressions (when pre- length (average FL ¼ 58.1 mm) than Ignotornis canadensis ichnosp. served) are more symmetric between digit impressions IIeIII and nov. (average FL ¼ 66.6 mm); however, there is limited information IIIeIV in Goseongornipes markjonesi than in Ignotornis canadensis available on the proportions and the divarication of the hallux ichnosp. nov. impression for Ignotornis gajinensis. Average total divarication Hwangsanipes choughidYang et al. (1995) described Hwangsa- (IIeIV) for Ignotornis gajinensis is 137.5 (IIeIII ¼ 73.8; nipes choughi from a trackway consisting of four consecutive tracks IIIeIV ¼ 63.7), whereas the average total divarication of Ignotornis on a slab with many other associated Uhangrichnus chuni tracks. 216 L.G. Buckley et al. / Cretaceous Research 84 (2018) 209e222

Fig. 6. Close view of tracks 2e4 of TMP 2016.036.003, type trackway of Ignotornis canadensis ichnosp. nov. Aquatilavipes ichnosp. is also present on the track slab (grey.) Comparison of straight (tracks 2, 4) and sinusoidal (track 3). Incompletely preserved digit III impressions appear to be sinusoidal (B), whereas digit III impressions that have been eroded or broken to the surface of the slab (the surface that would have been in contact with the upper most level of the track-bearing surface) appear straighter, but with partial digital pads (A, C).

While larger in size than Ignotornis canadensis ichnosp. nov., impressions are similar to those of the ichnogenera Ignotornis and Hwangsanipes choughi has a relatively longer hallux impression, Hwangsanipes. Gyeongsangornipes lockleyi is relatively much and larger digit divarications IIeIII and IIIeIV. Another morpho- smaller in footprint length (32.1 mm, Deokmyeongri sample; logical difference cited between the ichnogenera Hwangsanipes, 30.8 mm Donghae-myeon sample; Kim et al., 2013) than tracks of Goseongornipes, and Ignotornis, is the size of the webbing impres- Ignotornis, Hwangsanipes,andIgnotornis canadensis ichnosp. nov. sion between digit impressions III and IV. In Hwangsanipes choughi, However, Gyeongsangornipes lockleyi is similar in size to Goseon- the digit IIIeIV webbing impression extends more distal on the gornipes markjonesi, but is described as distinct from Goseongor- digit impressions than in both ichnogenera Goseongornipes and nipes markjonesi due to Goseongornipes markjonesi possessing a Ignotornis. Based on observations of modern bird tracks, the pres- clear hallux impression, web impressions only between digits III ence and size of webbing impressions can vary depending on and IV, and a larger total divarication than Gyeongsangornipes substrate consistency. However, the authors have not observed lockleyi (Kim et al., 2013). While smaller than Ignotornis gajiensis extramorphologic features that would be misinterpreted as and lacking a hallux impression, Gyeongsangornipes lockleyi is morphologic webbing, as described in Falkingham et al. (2009),in similar to Ignotornis gajiensis in that it has a proportionally larger modern semipalmate and palmate bird tracks. digit IIeIII divarication than digit IIIeIV divarication. One of the Jindongornipes kimidJindongornipes kimi (Lockley et al., 2006)is features consistent in all Ignotornidae is the slight asymmetry relatively longer (including the hallux impression) than other ani- between the divarication of digits IIeIII and IIIeIV, with the sodactyl, tetradactyl bird tracks from the Cretaceous Period divarication angle being slightly larger between the impressions (average FLwH ¼ 80 mm). Also, when the hallux impression is of digits II and III (Table 3). preserved, it is posteriorly directed (180 away from digit III impression), rather than posteromedially directed as in other tet- 6. Multivariate analyses radactyl bird tracks described from the Cretaceous Period (Lockley et al., 2006). The main similarity between Ignotornis canadensis Tracks of Ignotornis canadensis ichnosp. nov. group in morpho- ichnosp. nov. and Jindongornipes kimi is the relatively long hallux space with Goseongornipes markjonesi based on the relative pro- impression; however, the absolute size difference of Jindongornipes portions of the divarication of digit impressions IeII, IIeIII, and kimi as compared to Ignotornis canadensis ichnosp. nov., and the IIIeIV (Table 4). Ignotornis canadensis is not signiﬁcantly different different orientation of the hallux impression, suggests that these from Goseongornipes markjonesi (psame ¼ 0.364), Goseongornipes are different track types produced by different trackmakers. ichnosp. (psame ¼ 0.551), Hwangsanipes choughi (psame ¼ 0.674), and Gyeongsangornipes lockleyidAlthough not a tetradactyl avian Ignotornis gajinensis (psame ¼ 0.053) (Fig. 7, Table 5.) However, ichnotaxon, Gyeongsangornipes lockleyi is included in the Ignotornis canadensis ichnosp. nov. is signiﬁcantly different from comparative ichnology because the asymmetric webbing Ignotornis mcconnelli and Ignotornis yangi (Table 4.) The separation L.G. Buckley et al. / Cretaceous Research 84 (2018) 209e222 217

Fig. 7. Discriminant analysis plot of Ignotornis canadensis ichnosp. nov. with tetradactyl avian ichnotaxa from the Cretaceous Period. Ignotornis canadensis shares a morphospace with Goseongornipes ichnosp. and Hwangsanipes choughi due to the relative proportions of digit divarications IeII and IIeIII. Gyeongsangornipes lockleyi was included due to the similarity of the webbing impressions to that of Ignotornidae, but it forms a separate group due to its lack of a digit I impression.

Table 3 Comparison of measured data averages for described Ignotornidae. Hwangsanipes choughi, although larger in footprint length with hallux (FLwH) than Ignotornis canadensis ichnosp. nov., the webbing impression between digit impressions IIIeIV extends more distal in Hwangsanipes choughi, and a larger digit divarication IIeIII and IIIeIV than Ignotornis canadensis. DIV, digit divarication; FL, footprint length; FLwH, footprint length with hallux; FW, footprint width.

Ignotornis Ignotornis Ignotornis Hwangsanipes Goseongornipes Gyeonsanipes lockleyi Ignotornis canadensis mcconnelli gajiensis yangi choughi markjonesi (2 samples) ichnosp. nov.

Average FL (mm) 40.9 58.1 32.6 48.7 33.76 32.1/30.8 48.5 Average FLwH (mm) 55.2 58.1 51.1 75e80 39.4 ? 66.6 Average FW (mm) 47.6 53.9 44.7 62.9 46.4 41.4/40 65.1 Average Digit I Length 14.2 (25.7%) ? (NA) 11.9 (23.3%) ? (NA) 11.4 (28.9%) N/A (NA) 18.6 (27.9%) (% of Digit III length) DIV IeII 97.4 ? 83.9 ? ? ? 98 DIV IIeIII 63.5 73.8 60.6 56 62.9 71.1/68.9 58.5 DIV IIIeIV 60.8 63.7 63 57 61.5 58.4/55.8 56.6 of these four ichnospecies from Ignotornis mcconnelli and Ignotornis tetradactyl Cretaceous bird tracks. Analyzes with digit I data yangi was reported by Buckley et al. (2016) based on divarication of removed (in other words, treating all tetradactyl tracks as though digit impressions IeII and IIeIII, with Ignotornis mcconnelli and their hallux impressions were not preserved) revealed that Ignotornis yangi having relatively larger digit IeII and IIeIII divari- Gyeongsangornipes lockleyi shares a morphospace with Ignotornis cation values (Table 5.) The inclusion of Ignotornis canadensis ich- mcconnelli and Ignotornis yangi based on the relative proportions of nosp. nov. shows that this ichnotaxon has relative digit proportions divarication between digit impressions IIeIII and IIIeIV (Fig. 8). and digit divarication values consistent with Ignotornidae, and is Gyeongsangornipes lockleyi is significantly different from both 41 similar in said proportions to Ignotornis gajinensis. Ignotornis mcconnelli (psame ¼ 1.13 10 ) and Ignotornis yangi 06 Comparison of Gyeongsangornipes lockleyi to Ignotornidae (psame ¼ 1.50 10 )(Fig. 8). dInitial discriminant analysis shows that Gyeongsangornipes lock- Goseongornipes markjonesi, Goseongornipes ichnosp., and leyi, coded as having hallux data as “missing,” groups separately Hwangsanipes choughi are not significantly different from Ignotornis and is significantly different from all other tetradactyl Cretaceous canadensis ichnosp. nov. when analyzed after the removal of digit I avian ichnotaxa. However, vector loadings indicate that the variable data (digit I length, divarication IeII.) However, Ignotornis gajinensis with the largest amount of variation along Axis 1 is divarication becomes significantly different from Ignotornis canadensis ichnosp. between digit impressions IIeIII (2.82) and IeII (2.72). The vari- nov. (psame ¼ 0.005). It is not surprising that there are different ables along Axis 2 with the strongest loadings are the divarication results obtained by removing of digit I data, as this data contributes values between digits IeII (9.27) and digits IIIeIV (1.44). It is along the most variation when included in the previous analyses Axis 2 that Gyeongsangornipes lockleyi separates from the rest of the (Table 4). 218

Table 4 Multivariate analysis of variance (MANOVA) results for linear and angular data (Materials and Methods) comparing Ignotornis canadensis to functionally tetradactyl Cretaceous avian ichnospecies of Ignotornidae (Goseongornipes markjonesi, Goseongornipes ichnosp., Hwangsanipes choughi, Ignotornis gajinensis, Ignotornis mcconnelli, Ignotornis yangi), Koreanaornipodidae (Koreanaornis dodsoni, Koreanaornis hamanensis, Koreanaornis lii), Uhangrichnus chuni (including Dongyangornipes sinensis), with Gyeongsangornipes lockleyi included because of its morphologic similarity to Ignotornidae despite the absence of hallux impressions. Ignotornis canadensis is not signiﬁcantly different from Goseongornipes, Hwangsanipes, and Ignotornis gajinensis. Interestingly, the tridactyl Gyeongsangornipes is also not signiﬁcantly different from Hwangsanipes or Goseongornipes ichnosp., and is morphologically similar to Hwangsanipes choughi. Gates, Ignotornis canadensis ichnosp. nov.; GoMa, Goseongornipes markjonesi; Go isp, Goseongornipes ichnosp.; GyLo, Gyeongsangornipes lockleyi; HwCh, Hwangsanipes choughi; IgGa, Ignotornis gajinensis; IgMc, Ignotornis mcconnelli, IgYa, Ignotornis yangi; KoDo, Koreanaonris dodsoni; KoHa, Koreanaornis hamanensis; KoLi, Koreanaornis lii; UhCh, Uhangrichnus chuni.

IgMc IgYa IgGa HwCh Gos isp. GoMa KoDo Ko Ha UhCh GyLo KoLi Gates 95 11 04 23 10 65 22 45 42 60 27 8.86 10 2.58 10 1.75 10 1.57 10 5.34 10 4.47 10 2.64 10 4.86 10 4.36 10 3.31 10 IgMc 7.21 10 IgYa 7.21 10 95 3.52 10 23 8.10 10 16 4.37 10 03 2.31 10 22 5.80 10 18 3.08 10 44 2.51 10 90 9.67 10 21 3.18 10 11 4.55 10 19 IgGa 8.86 10 11 3.52 10 23 Fail Fail Fail 2.30 10 12 0.936 5.97 10 05 8.11 10 04 1.02 10 10 0.053 HwCh 2.5804 10 8.10 10 16 Fail Fail Fail 7.85 10 08 0.478 0.543 0.041 6.37 10 07 0.674 Gos isp. 1.7523 10 4.37 10 03 Fail Fail Fail 1.00 3.08 10 06 1.74 10 20 0.118 0.837 0.551 GoMa 1.57 10 10 2.31 10 22 Fail Fail Fail 3.91 10 11 9.89 10 05 9.68 10 15 0.002 1.81 10 09 0.364 KoDo 5.3465 10 5.80 10 18 2.30 10 12 7.85 10 08 1.00 3.91 10 11 9.74 10 26 8.19 10 59 9.90 10 12 0.018 1.47 10 10 KoHa 4.47 10 22 3.08 10 44 0.936 0.478 3.08 10 06 9.89 10 05 9.74 10 26 8.44 10 08 1.23 10 13 2.11 10 23 2.51 10 10 UhCh 2.64 10 45 2.51 10 90 5.97 10 05 0.543 1.74 10 20 9.68 10 15 8.19 10 59 8.44 10 08 3.87 10 39 4.21 10 56 1.72 10 31 209 (2018) 84 Research Cretaceous / al. et Buckley L.G. GyLo 4.86 10 42 9.67 10 21 8.11 10 04 0.041 0.118 2.30 10 03 9.90 10 12 1.23 10 13 3.87 10 39 8.78 10 11 8.07 10 05 KoLi 4.36 10 60 3.18 10 11 1.02 10 10 6.37 10 07 0.837 1.81 10 09 0.018 2.11 10 23 4.21 10 56 8.78 10 11 5.03 10 08 Gates 3.31 10 27 4.55 10 19 0.053 0.673 0.551 0.364 1.47 10 10 2.51 10 10 1.72 10 31 8.07 10 05 5.03 10 08

Table 5 Discriminant analysis loadings for linear and angular data (Materials and Methods) comparing Gates Formation Ignotornidae to functionally tetradactyl Cretaceous avian ichnospecies of Ignotornidae (Goseongornipes markjonesi, Goseongornipes ichnosp., Hwangsanipes choughi, Ignotornis gajinensis, Ignotornis mcconnelli, Ignotornis yangi), Koreanaornipodidae (Koreanaornis dodsoni, Koreanaornis hamanensis, Koreanaornis lii), Uhangrichnus chuni (including Dongyangornipes sinensis), with Gyeongsangornipes lockleyi included because of its morphologic similarity to Ignotornidae despite the absence of hallux impressions. Digit divarication has the greatest amount of variation in the dataset, and are the vectors along which the ichnotaxa group. DL, digit length; DIV, digit divarication; FL, footprint length; FR, footprint rotation; FW, footprint width; PA, pace angulation; PL, pace length; SL; stride length.

Axis 1 Axis 2 Axis 3 Axis 4 Axis 5 Axis 6 Axis 7 Axis 8 Axis 9 Axis 10 Axis 11 e

FL 0.028 0.009 0.033 0.009 0.001 0.019 0.004 0.004 0.004 0.006 0.009 222 FW 0.018 0.006 0.015 0.024 0.002 0.019 0.004 0.025 0.012 0.010 0.006 DLI 0.004 0.003 0.008 0.005 0.013 0.017 0.016 0.043 0.008 0.024 0.027 DLII 0.001 0.003 0.003 0.012 0.010 0.001 0.018 0.017 0.011 0.001 0.020 DLIII 0.014 0.014 0.038 0.010 0.006 0.017 0.010 0.007 0.009 0.006 0.010 eII 2.41 9.21 3.60 1.37 0.660 0.199 0.316 2.72 1.61 1.43 0.379 DIVI DIVIIeIII 2.84 0.470 1.04 4.15 4.84 0.005 3.46 0.007 2.21 1.39 1.97 DIVIIIeIV 2.34 1.28 0.437 2.77 0.865 1.19 0.364 2.20 4.91 4.72 0.049 FR 0.18 0.129 0.482 1.41 0.328 0.183 1.81 0.953 0.559 2.03 1.14 PA 0.195 0.847 2.46 2.63 0.118 2.63 4.05 5.01 4.90 7.48 4.53 PL 0.004 0.004 0.004 0.046 0.032 0.020 0.002 0.007 0.003 0.010 0.010 SL 0.003 0.004 0.006 0.042 0.030 0.023 0.003 0.009 0.011 0.006 0.021 L.G. Buckley et al. / Cretaceous Research 84 (2018) 209e222 219

Fig. 8. Discriminant analysis of Ignotornis canadensis ichnosp. nov., with tetradactyl avian ichnotaxa from the Cretaceous Period. Removal of the digit I data (digit I impression length, digit divarication IeII) shows that, while the ichnospecies of Ignotornidae retain their relative positions in morphospace (including the relationship between Ignotornis canadensis ichnosp. nov. and Goseongornipes ichnosp.), Gyeongsangornipes lockleyi groups with Ignotornis mcconnelli and Ignotornis yangi. Ignotornis canadensis ichnosp. nov. still not signiﬁcantly different from Goseongornipes markjonesi, Goseongornipes isp., and Hwangsanipes choughi, when analyzed after the removal of digit I impression data (digit I length, divarication IeII.) However, Ignotornis gajinensis becomes signiﬁcantly different from Ignotornis canadensis (psame ¼ 0.005). It is not surprising that there are different results obtained by removal of digit I data, as this data contributes the most variation when included in the previous analyses (Table 4).

7. Discussion Ignotornis canadensis ichnosp. nov. in its proportionally longer hallux impressions and larger webbing impressions. Webbing im- 7.1. Statistical results pressions between digits IIeIII and IIIeIV are equal in size in ichnospecies of Goseongornipes, whereas webbing impressions The multivariate statistical results (MANOVA) show that Igno- between digits IIeIII are smaller than those between digits IIIeIV in tornis canadensis ichnosp. nov. is not significantly different than Ignotornis canadensis ichnosp. nov. Goseongornipes ichnosp., Hwangsanipes choughi, and Ignotornis gajinensis. This is an opportunity to offer a caveat for multivariate 7.2. Occurrences of Ignotornidae and Early Cretaceous avian statistical analyses. While a useful tool for testing hypotheses of ichnodiversity distinctiveness, they can only use the data that are available. Linear and angular data that capture some distinctive morphologic fea- The large diversity of avian footprints from the Lower Creta- tures, such as webbing impression shape and digit curvature, have ceous of China and Korea is suggested as evidence of provincialism yet to be collected for all of the taxa discussed herein. As such, there and endemism in the Lower Cretaceous track record in Asia are morphologic data, traditionally used to differentiate among (Lockley et al., 2012b). However, it is possible that the current avian ichnotaxa, that are missing from these analyses. Other tools, apparent low diversity of bird tracks in Lower Cretaceous deposits such as landmark analyses, may have potential in future analyses of of North America is artificial (Xing et al., 2016), at least in part due avian ichnotaxa. to preservational factors. Reporting and describing avian ichnotaxa While Ignotornis canadensis ichnosp. nov. is not significantly from Lower Cretaceous deposits, in what was once the western part different than Ignotornis gajinensis, or ichnospecies of Goseongor- of Laurasia, is in the process of “playing catch-up” (Xing et al., 2016), nipes, and Hwangsanipes choughi, there are consistent morphologic as the known diversity of avian footprints from the Lower Creta- differences between Ignotornis canadensis ichnosp. nov. and these ceous of North America is growing. McCrea et al. (2015) described ichnotaxa that justify their distinction. Ignotornis gajinensis is Paxavipes babcockensis from the Boulder Creek Formation morphologically distinct from Ignotornis canadensis ichnosp. nov. (AptianeAlbian), and Limiavipes curriei (McCrea and Sarjeant, by the consistently larger divarication of digit impressions IIeIII 2001; McCrea et al., 2014) was described from the Gates Forma- than digit impressions IIIeIV. Hwangsanipes choughi is distinct from tion (Albian), both from western Canada. As more research is 220 L.G. Buckley et al. / Cretaceous Research 84 (2018) 209e222 conducted on the avian ichnology of the Lower Cretaceous strata of 7.3. Interpretation of sinusoidal digit III impressions North America, it is possible the diversity of North American avian ichnotaxa may approach that known from the Lower Cretaceous of With only the information provided by tracks of the Ignotorni- China and Korea. At present, assuming the validity of named ich- dae, there would be justification in assuming that the sinusoidal notaxa from the two regions, there are some similarities in the shape of digit III impressions is due to the morphology of the digit. Lower Cretaceous ichnogenus diversity: i.e., the ichnogenera However, comparisons with neoichology specimens reveal that the Aquatilavipes, Koreanaornis, and now Ignotornis, have been found in sinusoidal appearance of digit III is likely due to preservational both regions. However, Limiavipes curriei and Paxavipes babcock- variability rather than digit morphology. Curved digit impressions ensis are known only from North America, while the ichnogenera are not unknown in the ichnogenus Ignotornis: Ignotornis yangi is Goseongornipes, Gyeongsangornipes, Jindingornipes, Shandongornips, described as having digit III impressions that are straight or slightly and Wupus are known only from Asia, with Limiavipedidae (Wupus curved (Kim et al., 2006). Ignotornis mcconnelli also shows slightly agilis, Limiavipes curriei; McCrea et al., 2014; Xing et al., 2015) sinusoidal digits (Fig. 9). Casts of tracks of Actitis macularius present in both North America and Asia. (Spotted Sandpiper) show that, when individuals walk on a fine- Ignotornis mcconnelli is present in Lower Cretaceous deposits of grained surface with high water content, the sediment does not North America in Colorado (AlbianeCenomanian). The Gates For- have the structural integrity to preserve the shape of the digital mation is early Albian in age and within the age range of Ignotornis pads, which may differentially collapse. This makes the digit III mcconnelli. However, Ignotornis canadensis ichnosp. nov. is larger impression appear as a line that can present as both straight and than Ignotornis mcconnelli, indicating that the trackmaker for curved within the same trackway (Fig. 9). Ignotornis canadensis ichnosp. nov. was distinct from the trackmaker of Ignotornis mcconnelli. The trackmaker of Ignotornis cana- 8. Conclusions densis ichnosp. nov. also possessed a pes that had digit proportions and digit divarication similar to that of Goseongornipes ichnosp. and While still small, the number of avian ichnotaxa known from the Ignotornis gajinensis. Cretaceous deposits of North America is growing. Here, we describe

Fig. 9. Comparison of sinusoidal digit III impressions of Ignotornis canadensis ichnosp. nov. (Fig. 6) to extinct and extant avian traces. A, UCM 203.5, replica of Ignotornis mcconnelli, showing a sinuosity to digit III impressions; B, PRPRC NI2010.004, plaster cast of tracks of the modern Actitis macularius (Spotted Sandpiper), showing the transition from a straight digit III impression to a digit III impression that is laterally curved. The only difference between the tracks is the substrate consistency: the tracks show that the curved digit III traces were made on sediment that was saturated. Scale in centimeters. L.G. Buckley et al. / Cretaceous Research 84 (2018) 209e222 221 a novel avian ichnotaxa, Ignotornis canadensis ichnosp. nov., which Hammer, Ø., Harper, D.A.T., Ryan, P.D., 2001. PAST: paleontological statistics soft- represents the second report of Ignotornidae from North America ware package for education and data analysis. Palaeontologia Electronica 4, 1e9. and the first report of Ignotornidae from Canada. Ignotornis cana- Helm, C., 2000. Beyond rock and coal: the history of the Tumbler Ride area. Jasper densis ichnosp. nov. is larger than other reported Ignotornidae, yet Printers, Edmonton. possesses digit divarication that is similar to that of ichnogenera Helm, C., 2001. Tumbler Ridge: enjoying its history, trails and wilderness. MCA Publishing, Tumbler Ridge. Goseongornipes and Hwangsanipes. The apparent sinusoidal digit III Kim, B.K., 1969. A study of several sole marks in the Haman Formation. Journal of impressions, while common in Ignotornidae, is due to preserva- the Geological Society of Korea 5, 243e258. tional variation rather than the morphology of digit III. Compari- Kim, J.Y., Kim, S.H., Kim, K.S., Lockley, M.G., 2006. The oldest record of webbed bird and pterosaur tracks from South Korea (Cretaceous Haman Formation, Chang- sons with neoichnology track casts show that incomplete seon and Sinsu Islands): more evidence of high avian diversity in East Asia. preservation of digital pads in water-saturated sediment can result Cretaceous Research 27, 56e69. in digit III impressions appearing curved or sinusoidal. A high level Kim, J.Y., Lockley, M.G., Seo, S.J., Kim, K.S., Kim, S.H., Baek, K.S., 2012. A paradise of Mesozoic birds: the world's richest and most diverse Cretaceous bird track of endemism is hypothesized for the ichnofauna of the Lower assemblage from the Early Cretaceous Haman Formation of the Gajin Tracksite, Cretaceous of Asia (Lockley et al., 2012b). The diversity of bird Jinju, Korea. Ichnos 19 (1e2), 28e42. tracks in Asia, particularly that of South Korea, is likely high based Kim, J.Y., Kim, M.K., Oh, M.S., Lee, C.Z., 2013. A new semi-palmate bird track, on optimal preservation conditions (lacustrine basin and lake Gyeongsangornipes lockleyi ichnogen. et ichnosp. nov., and Koreanaornis from the Early Cretaceous Jindong Formation of Goseong County, Southern Coast of margin paleohabitats). In comparison, the Dakota Formation, while Korea. Ichnos 20 (2), 72e80. also optimal for bird track preservation, has fewer bird track sites Langenberg, C.W., Kalkreuth, W., Wrightson, C.B., 1987. Deformed Lower Cretaceous than in South Korea (Lockley et al., 2012b). However, there are coal-bearing strata of the Grande Cache area, Alberta. Geological Survey Department, Alberta Research Council Bulletin 56, pp. 1e56. several Lower Cretaceous bird track sites that await description Leckie, D.A., Smith, D.G., 1992. Regional setting, evolution, and depositional cycles of (McCrea et al., 2014), which may add to the paleodiversity of the Western Canada Foreland Basin. In: Macqueen, R., Leckie, D.A. (Eds.), e Cretaceous bird tracks from North America. With increased focus American Association of Petroleum Geologists Memoir 55, pp. 9 46. Lee, Y.-N., 1997. Bird and dinosaur footprints in the Woodbine Formation (Cen- on avian traces from Lower Cretaceous deposits, we will have a omanian), Texas. Cretaceous Research 18, 849e864. better understanding of the paleogeographic distribution of Lower Li, R., Lockley, M.G., Liu, M., 2005. A new ichnotaxon of fossil bird track from the Cretaceous avian ichnotaxa and the global distribution of birds in Early Cretaceous Tianjjialou Formation (BarremianeAlbian), Shandong Prov- ince, China. Chinese Science Bulletin 50 (11), 1149e1154. the Early Cretaceous. Lockley, M.G., Harris, J.D., 2010. On the trail of early birds: a review of the fossil footprint record of avian morphological and behavioral evolution. In: Ulrich, P.K., Willett, J.H. (Eds.), Trends in ornithology research. Nova Science Publisher, Hauppage, New York, pp. 1e63. Acknowledgements Lockley, M.G., Yang, S.Y., Matsukawa, M., Fleming, F., Lim, S.K., 1992. The track record of Mesozoic birds: evidence and implications. Philosophical Transactions of the Specimens were collected from the ancestral lands of the Treaty Royal Society of London B 336, 113e134. Lockley, M.G., Janke, P., Theisen, L., 2001a. First reports of bird and ornithopod 8 First Nations in British Columbia and Kelly Lake First Nations in tracks from the Lakota Formation (Early Cretaceous), Black Hills, South Dakota. British Columbia, and Kelly Lake, Tsuu T’ina, and Aseniwuche In: Carpenter, K., Tanke, D. (Eds.), Mesozoic vertebrate life: new research Winewak First Nations in Alberta. The authors wish to thank D. inspired by the paleontology of Philip J. Currie. Indiana University Press, Bloomington, Indiana, pp. 443e452. Spivak and B. Strilinsky (TMP) for access to TMP 2016.035.0003; Lockley, M.G., Wright, J.L., Matsukawa, M., 2001b. A new look at Magnoavipes and Ridge Roters for helicopter support for collection of PRPRC so-called “big bird” tracks from Dinosaur Ridge (Cretaceous, Colorado). The 2014.05.001; M. Lockley for access to cast material of Ignotornidae. Mountain Geologist 38, 137e146. This manuscript was greatly improved by the contributions of M. Lockley, M.G., Nadon, G., Currie, P.J., 2004. A diverse dinosaur-bird footprint assemblage from the Lance Formation, Upper Cretaceous, eastern Wyoming: Lockley, an anonymous reviewer, and the editorial staff. implications for ichnotaxonomy. Ichnos 11 (3e4), 229e249. Lockley, M.G., Houck, K., Yang, S.-Y., Matsukawa, M., Lim, S.-K., 2006. Dinosaur dominated footprint assemblages from the Cretaceous Jindong Formation, Hallyo Haesang National Park area, Goseong County, South Korea: evidence and References implications. Cretaceous Research 27, 70e101. Lockley, M.G., Chin, K., Houck, K., Matsukawa, M., Kukihara, R., 2009. New in- Anfinson, O., Lockley, M.G., Kim, S.H., Kim, K.S., Kim, J.-Y., 2009. First report of the terpretations of Ignotornis, the first-reported Mesozoic avian footprints: im- small bird track Koreanornis from the Cretaceous of North America: implica- plications for the paleoecology and behavior of an enigmatic Cretaceous bird. tions for avian ichnotaxonomy and paleoecology. Cretaceous Research 30, Cretaceous Research 30, 1041e1061. 885e894. Lockley, M.G., Li, J.-D., Kim, J.Y., Kim, K.S., Huh, M., Hwang, K.-G., 2012a. Tracking Azuma, Y., Lu,} J., Jin, X., Noda, Y., Shibata, M., Chen, R., Zheng, W., 2013. A bird Korea's early birds: a review of Cretaceous avian ichnology and its implication footprint assemblage of the early Late Cretaceous age, Dongyang City, Zhejiang for evolution and behavior. Ichnos 19 (1e2), 17e27. Province, China. Cretaceous Research 40, 3e9. Lockley, M.G., Lucas, S.G., Matsukawa, M., Harris, J.D., 2012b. Cretaceous tetrapod Buckley, L.G., McCrea, R.T., Lockley, M.G., 2015. Birding by foot: a critical look at the footprint biostratigraphy, biochronology, and ichnofaces. Journal of Stratigraphy synapomorphy- and phenetic-based approaches to trackmaker identification of 36 (2), 503e551. enigmatic tridactyl Mesozoic traces. Ichnos 22 (3e4), 92e207. Lockley, M.G., Buckley, L.G., Foster, J.R., Kirkland, J.I., Deblieux, D.D., 2015. First Buckley, L.G., McCrea, R.T., Lockley, M.G., 2016. Analyzing and resolving avian ich- report of bird tracks (Aquatilavipes) from the Cedar Mountain Formation (Lower notaxonomy using multivariate statistical analyses: approaches and results. In: Cretaceous), eastern Utah. Palaeogeography, Palaeoclimatology, Palaeoecology Falkingham, P.L., Marty, D., Richter, A. (Eds.), Dinosaur tracks: the next steps. 420, 150e162. Indiana University Press, Bloomington, Indiana, pp. 259e308. Matsukawa, M., Lockley, M.G., Hayashi, K., Korai, K., Peiji, C., Haichun, Z., 2014. First Currie, P.J., 1981. Bird footprints from the Gething Formation (Aptian, Lower report of the ichnogenus Magnoavipes from China: new discovery from the Cretaceous) of northeastern British Columbia, Canada. Journal of Vertebrate Lower Cretaceous inter-mountain basin of Shangzhou, Shaanxi Province, Cen- Paleontology 1, 257e264. tral China. Cretaceous Research 47, 131e139. Falkingham, P.L., Margetts, L., Smith, I.M., Manning, P.L., 2009. Reinterpretation of McCrea, R.T., 2000. Vertebrate palaeoichnology of the Lower Cretaceous (lower palmate and semi-palmate (webbed) fossil tracks; insights from finite element Albian) Gates Formation of Alberta (Unpubl. MSc thesis). University of Sas- modelling. Palaeogeography, Palaeoclimatology, Palaeoecology 271 (1e2), katchewan, Saskatoon, Saskatchewan, 184 pp. 69e76. McCrea, R.T., Currie, P.J., 1998. A preliminary report on dinosaur tracksites in the Farlow, J.O., Holtz Jr., T.R., Worthy, T.H., Chapman, R.E., 2013. Feet of the fierce (and lower Cretaceous (Albian) Gates Formation near Grande Cache, Alberta. New not so fierce): pedal proportions in large theropods, other non-avian dinosaurs, Mexico Museum of Natural History and Science Bulletin 14, pp. 155e162. and large ground birds. In: Parrish, J.M., Molar, R.E., Currie, P.J., Koppelhus, E.B. McCrea, R.T., Sarjeant, W.A.S., 2001. New ichnotaxa of bird and mammal footprints (Eds.), Tyrannosaurid paleobiology. Indiana University Press, Bloomington, from the Lower Cretaceous (Albian) Gates Formation of Alberta. In: Tanke, D.H., pp. 88e132. Carpenter, K., Skrepnick, M.W. (Eds.), Mesozoic vertebrate life: new research Fiorillo, A.R., Hasiotis, S.T., Kobayashi, Y., Breithaupt, B.H., McCarthy, P.J., 2011. Bird inspired by the paleontology of Philip J. Currie. University of Indiana Press, tracks from the Upper Cretaceous Cantwell Formation of Denali National Park, Bloomington and Indianapolis, pp. 453e478. Alaska, USA: a new perspective on ancient northern polar diversity. Journal of McCrea, R.T., Buckley, L.G., Plint, A.G., Currie, P.J., Haggart, J.W., Helm, C.W., Systematic Palaeontology 9 (1), 33e49. Pemberton, S.G., 2014. A review of vertebrate trackebearing formations from 222 L.G. Buckley et al. / Cretaceous Research 84 (2018) 209e222

the Mesozoic and earliest Cenozoic of western Canada with a description of a Xing, L., Harris, J.D., Jia, C.-K., Luo, Z.-J., Wang, S.-N., An, J.-F., 2011. Early Cretaceous new theropod ichnospecies and reassignment of an avian ichnogenus. In: bird-dominated and dinosaur footprint assemblages from the northwestern Lockley, M.G., Lucas, S.G. (Eds.), Fossil footprints of western North America. New margin of the Junggar Basin, Xinjiang, China. Palaeoworld 20, 308e321. Mexico Museum of Natural History and Sciences Bulletin 62, pp. 5e93. Xing, L.D., Buckley, L.G., McCrea, R.T., Lockley, M.G., Zhang, J., Pinuela,~ L., Klein, H., McCrea, R.T., Buckley, L.G., Plint, A.G., Lockley, M.G., Matthews, N.A., Noble, T.A., Wang, F., 2015. Reanalysis of Wupus agilis (Early Cretaceous) of Chongqing, Xing, L., Krawetz, J.R., 2015. Vertebrate ichnites from the Boulder Creek For- China as a large avian trace: differentiating between large bird and small non- mation (Lower Cretaceous: middle to ?upper Albian) of northeastern British avian theropod tracks. PLoS One 10 (5), e0124039. https://doi.org/10.1371/ Columbia, with a description of a new ichnotaxon, Paxavipes babcockensis, journal.pone.0124039. ichnogen. et isp. nov. Cretaceous Research 55, 1e18. Xing, L.D., Buckley, L.G., Lockley, M.G., Zhang, J., Marty, D., Wang, Y., Li, J., Mehl, M.G., 1931. New bird record from the Dakota Sandstone, of Colorado. McCrea, R.T., Peng, C., 2016. A new bird track, Koreanaornis lii ichnosp. nov., Geological Society of America Bulletin 42, 331. from the Lower Cretaceous Hekou Group in the Lanzhou-Minhe Basin, Gansu, Plint, A.G., Hart, B.S., 1988. Field guide to the Upper Cretaceous Dunvegan (Cen- northwest China, and implications for Early Cretaceous avian diversity. Creta- omanian) and Cardium (Turonian) formations in the Dawson Creek-Fort St. ceous Research 66, 141e154. John area, British Columbia. Canadian Society of Petroleum Geologis's ﬁeld Yang, S.-Y., Lockley, M.G., Greben, R., Erickson, B.R., Lim, S.-K., 1995. Flamingo and guide to sequences, stratigraphy, and sedimentology. Surface and Subsurface duck-like traces from the Late Cretaceous and early Tertiary: evidence and Technical Meeting, September 14e16, Calgary, Alberta, 51pp. implications. Ichnos 4 (1), 21e34. Wan, Z., 1996. The Lower Cretaceous ﬂora of the Gates Formation from western Canada (Unpubl. PhD thesis). University of Saskatchewan, 304 pp.