Systematics of 12 Jurassic, Cretaceous, and Paleogene squat lobster taxa (Galatheoidea)

Abstract Squat lobsters within Galatheoidea are very diverse with over 1,300 extant species that live in all marine ecosystems, but their fossil record, starting in the Middle Jurassic, consists of only ~200 species. Consequently, much remains to be learned about their biodiversity, phylogeny, and paleobiogeography. We describe five new species (Galatheites sforum n. sp., Kimmeridgian, Late Jurassic, Germany; Vasconilia zapotitlanensis n. sp., Barremian, Early Cretaceous, Mexico; Eomunidopsis texcalaensis n. sp., Barremian, Early Cretaceous, Mexico; Protomunida bennickei n. sp., Danian, Paleocene, Denmark; and Protomunida eurekantha n. sp., Danian, Paleocene, Denmark) and one new genus (Tethysgalathea n. gen., Ypresian, Eocene, Italy). We further reassign Munida cretacea (Albian, Early Cretaceous, Texas, USA) to Galathea?, return Eomunidopsis? cobbani (Campanian, Late Cretaceous, Colorado, USA) to its original genus, report on the second occurrence of Vetoplautus latimarginus Robins et al., 2013 (Tithonian, Late Jurassic, Czech Republic), and reinstate Palaeomunidopsis moutieri (middle Bathonian, Middle Jurassic, France) as the oldest galatheoid known to date. The five new species, all found in limestones containing corals, increase galatheoid diversity in the fossil record by 2.5%. Finally, the common yet hitherto unrecognized Protomunida eurekantha was discovered by making casts of external molds, revealing distinct spines on the posterior margin that are more difficult to see in internal molds and specimens with cuticle due to breakage. Collecting and studying the external molds for galatheoids and other fossil decapods could yield additional cryptic species. UUID: http://zoobank.org/68a25381-65c6-4773-a4f2-71d9b0a25304

The sparse fossil record of Galatheoidea suggests many more species remain to be discovered. New taxa and records will help to advance the study of their phylogenetic relationships, paleobiogeography, biodiversity, and environmental preferences now and in the future. To this end, we here describe and reassess 12 Mesozoic to early Cenozoic taxa.

Materials and methods
The studied material comprises 78 specimens from various Jurassic, Cretaceous, and Paleogene localities in Europe and North America. Most specimens (64) originate from the Paleocene (Danian) of Denmark. For specifics per taxon, referral is made to the materials sections for each taxon. The specimens were prepared by using air scribes (Compressed Air Hammer HW 70 and PaleoTools) to remove the matrix around the specimens as needed. For some of the Danish material, a casting technique was used to obtain morphological details that could not be seen easily on internal molds. Low-viscose silicone rubber (Silastic RTV 9161) was put into external molds in a vacuum environment for about a minute, after which the casts were pulled from the external mold. We have indicated when casts were used.
-Carapace excluding rostrum about 10-20% longer than maximum width; widening posteriorly, maximum width in posterior third of carapace. Rostrum slightly downturned, moderately wide at base, 20-25% of maximum carapace length including rostrum, with axial keel; narrowing anteriorly and ending in three spines, middle spine longest. Cervical groove strong, U-shaped; curves transversely near lateral margin. Ornamentation consists of strong tubercles on anterior carapace and transversely elongated ovate tubercles to ridges on posterior carapace on internal mold.
Description.-Carapace excluding rostrum about 10-20% longer than maximum width; weakly convex longitudinally, moderately convex transversely; widening posteriorly, maximum width in posterior third of carapace; lateral margins not rimmed. Rostrum slightly downturned, moderately wide at base, 20-25% of maximum length including rostrum, with axial keel; narrowing anteriorly and ending in three spines, middle spine longest; margins flared up somewhat; tubercles at base. Upper orbital margin directed forward, concave. Epigastric regions distinctly bordered anteriorly by inverse V-shaped rim with axial indent. Hepatic region small. Anterior part of gastric region with weak groove perpendicular to lateral margins and row of tubercles posterior to it. Mesogastric region well defined at posterior border, moderately so on lateral sides and at process. Cervical groove strong, U-shaped but curves transversely near lateral margin. Groove branching off cervical groove near posterolateral edges of mesogastric region laterally oriented, strong. Anterior branchial region triangular in dorsal view; posterior branchial region not subdivided. Cardiac region triangular, poorly defined anteriorly and posteriorly, with weak obliquely oriented lateral groove extending to groove branching off cervical groove. Intestinal region not defined by grooves, bears less ornamentation. Lateral margins ornamented with tubercles; spines do not appear to be present. Posterior margin rimmed, concave. Ornamentation consists of strong tubercles on anterior carapace and ovate tubercles to ridges on posterior carapace on internal mold. Ventral surface, abdomen, appendages, cuticle, and muscle scars unknown.
Etymology.-sforum is referring to the decapod paleontology team from Kent State University, Carrie Schweitzer (s) and Rodney Feldmann (f), for their enormous contributions to the taxonomy and systematics of fossil Decapoda in >150 joint scientific papers. Remarks.
-Intraspecific variation appears limited, but only three specimens are known. We cannot determine the degree of ontogenetic variation with only a single, incomplete smaller specimen available. Compared with the type species, G. zitteli, the new species exhibits coarser ornamentation in general. The elongated ornamentation on the cardiac region, the middle part of the mesogastric region, and the lateral parts of the branchial region in dorsal view is more interrupted in G. sforum n. sp. Relative to G. aiola, Galatheites sforum has a more U-shaped cervical groove, the transverse groove in the gastric region is less obvious, and it bears coarser ornamentation. The new species is close to G. britmelanarum, but the ornamentation of the new species is coarser, and more elongated ridges are present at the base of the mesogastric region and in the posterior carapace. The ornamentation of the new species is coarser and has diverging lateral margins toward the posterior carapace, whereas it is straight in G. diasema. Also compared with G. obtecta, the ornamentation of the new species is coarser. In addition, the lateral margins of the rostrum are straight, but they are slightly concave in G. obtecta. The ornamentation is weaker in G.? royoi. The use of ornamentation to differentiate this new species from congenerics is warranted because ornamentation shows limited intraspecific variation in general and does not differ substantially depending on the presence or absence of cuticle in galatheoids (e.g., Klompmaker et al., 2012;Robins et al., 2016 Description. -Carapace excluding rostrum about as wide as long; moderately convex longitudinally, strongly convex transversely; widening posteriorly, maximum width in posterior third of carapace; lateral margins not rimmed. Rostrum downturned, moderately wide at base, approximately 25% of maximum length including rostrum; narrowing anteriorly and ending in three spines, middle spine longest; rostral surface appears flat. Upper orbital margin directed forward and concave, with possible spine at outer orbital margin. Epigastric regions distinctly bordered anteriorly by inverse V-shaped rim. Protogastric, hepatic, and mesogastric regions not delimited, except for tip of mesogastric process. Cervical groove shallow, widely V-shaped but rounded at axis, appears to disappear at lateral third of dorsal carapace; no groove branching off cervical groove. Branchial, cardiac, and intestinal regions not clearly defined. Posterior margin nearly straight, may be rimmed. Ornamentation consists of small tubercles on anterior carapace and long ridges on posterior carapace that bend forward on flanks, on both cuticular layers visible. Muscle scar impressions may be discernable at base of mesogastric region (posterior gastric muscles) and adjacent to approximate position of cardiac region. Ventral surface, abdomen, and appendages unknown.
Etymology.-Named after the Zapotitlán Formation, in which specimens were found.
Remarks.-Although Paragalathea ruizi was assigned as the type species of Vasconilia by Robins et al. (2016), Vega et al. (2019) assigned two specimens to P. ruizi. The specimens from Mexico are separated geographically from the specimens from Spain by ∼9,000 km today. In addition, a long temporal gap exists between the late Albian specimens from Spain and the late Barremian specimens from Mexico (ca. 25 million years). Although long stratigraphic ranges have been reported for fossil decapods (see Klompmaker et al., 2012, p. 792-793, for references), such ranges are rare. Above all, the specimens differ morphologically from Vasconilia ruizi in that the carapace of the new species widens much more toward the posterior carapace, the rostrum is more downturned, and the flanks appear to be longer, giving the carapace a more convex appearance transversally. The latter characters are similar to Mesogalathea Houša, 1963(see Robins et al., 2016), but the anterior carapace with the anterior portion of the epigastric regions and the tip of the mesogastric region clearly defined, in addition to the narrower rostrum, suggest this species does not fit Mesogalathea. Compared with V.? miyakoensis, Vasconilia zapotitlanensis n. sp. does not possess a groove branching off the cervical groove and is wider posteriorly. Vasconilia straeleni bears moredistinct tubercles on the anterior carapace and has a higher length-width ratio (∼1.15 [see Vía Boada, 1982;Klompmaker et al., 2012;Robins et al., 2016] versus ∼1.00 for IGM-11372). Vasconilia xystosa does not exhibit transverse ridges on the posterior carapace as in the new species, and the cervical groove appears less V-shaped axially. The use of ornamentation to differentiate species can be problematic for brachyurans as the cuticular layers and internal mold may show different ornamentation (summarized by , but this possible issue does not appear to exist for galatheoids (summarized by Robins et al., 2016).
The two specimens are nearly identical, but the paratype does not show an equally expressed cervical groove as seen in the holotype, although it is present (Vega et al., 2019, fig. 4.1, right of axis). Whether this is a matter of preservation or intraspecific variation cannot be determined at this stage.
Muscle scars are not often reported for squat lobsters, but some have been recorded at the base of the mesogastric region . Such posterior gastric muscle scars are also present in the new species, but another set lateral to the approximate position of the cardiac region (branchial muscles, see Figure 2.2) has not been recorded in the literature from fossil galatheoids thus far. ornamentation with spines mostly on anterior carapace and transverse ridges mostly on posterior carapace, depression present posterior to cardiac region.
Etymology.-Tethysgalathea is a merger of Tethys and Galathea. The specimens of the genus known so far inhabited the western part of the Tethys during the early Eocene. Gender is feminine.
Remarks.-The type species was originally placed in Eomunidopsis, but members of this genus do not bear strong spines on the anterior part of the dorsal carapace. Moreover, the axial point of the cervical groove is located more posteriorly (see Table 1). The taxon also superficially resembles Acanthogalathea, but Acanthogalathea bears two distinct spines at the base of the rostrum, which Tethysgalathea n. gen. lacks. A further difference separating this from Acanthogalathea is a higher length-width ratio (see Table 1). Of the other fossil galatheid genera, Bolcagalathea, Lessinigalathea, and Palaeomunida bear prominent spines on the lateral margins of the rostrum, which are absent in the new genus. Moreover, Palaeomunida bears a keel on the rostral axis from the base to tip, whereas such a keel is not present in the new genus. Unlike Tethysgalathea, Luisogalathea does not exhibit distinct spines on the anterior carapace.
All galatheid genera with modern representatives can also be distinguished from the new genus. Alainius bears no spines on the anterior carapace, has indistinct grooves, and bears two small accessory spines adjacent to the central rostral spine. Allogalathea exhibits a longer rostrum with small spines along the rostral margins and more rounded lateral carapace margins. Allomunida possesses an indistinct cervical groove, the cardiac region is poorly delineated, and the carapace is widening posteriorly.
Coralliogalathea has wide-based rostrum with distinct lateral spines and a poorly delineated cardiac region. Fennerogalathea exhibits a narrower, spine-like rostrum and a less-defined cardiac region. Galathea and Janetogalathea both have a rostrum with distinct lateral spines and a less-spinose anterior carapace. Lauriea has poorly defined regions and a rostrum with distinct lateral spines. Macrothea exhibits a longer carapace that is widening posteriorly, and the rostrum bears spines on its lateral margins. Nanogalathea possesses more-rounded lateral margins of the carapace and bears fewer spines on the anterior carapace. Phylladiorhynchus has a less-defined cardiac region and a lateral margin with at least two prominent spines. Finally, Triodonthea possesses a rostrum with distinct lateral spines and bears no transverse ridges on its carapace.
We also considered placing Acanthogalathea paucispinosa Beschin et al., 2016, in the new genus because this species is proportionally longer than other species within Acanthogalathea, but this taxon bears two distinct spines at the base of the rostrum not figured previously (Fig. 4)   Emended diagnosis.-Carapace without rostrum somewhat longer than wide. Rostrum long, triangular, ending in one or, usually, three spines, with median keel. Cervical groove strong, U-shaped. Transverse groove branching off cervical groove present. Tip of mesogastric process and epigastric regions well defined. Ornamentation consists of transverse ridges that are variously interrupted, without large spines dorsally. Lateral margins of carapace usually lined with row of tubercles/spines.
Remarks.-Some disagreement over the family placement of Eomunidopsis has arisen within the recent literature. Ahyong et al. (2010) and Beschin et al. (2016) placed this genus in Munidopsidae, whereas Klompmaker et al. (2012) favored placement in Galatheidae. Robins et al. (2013) pointed out that this genus does not belong to Munidopsidae. Ahyong et al. (2010) placed Eomunidopsis within the Munidopsidae, as well as pointing out similarities between Calteagalathea, Paragalathea, and Munidopsis. Although Calteagalathea is a munidopsid, the species of Paragalathea and Eomunidopsis they examined have mostly been reassigned from Eomunidopsis and Paragalathea to other genera in new families by Robins et al. (2016). The type species of Eomunidopsis does not possess a circumgastric groove and has transverse ornamentation, which are far more common in Galatheidae than in Munidopsidae. The paper by Beschin et al. (2016), who put Eomunidopsis in Munidopsidae, was  published at a similar time to Robins et al. (2016), and thus did not consider the new generic placements in their work. The galatheid Eomunidopsis limonitica was placed in this genus by Schweitzer et al. (2010), Klompmaker et al. (2012), and Vega et al. (2019), but Franţescu (2014) moved it to a different genus and family (Munida Leach, 1820, and Munididae, resp.). This change was based primarily on similarities in the sternum of Eomunidopsis limonitica with Munida, but a comparison with venters of Galatheidae was not provided. Munididae are diagnosed by a trifid frontal margin with a central rostral spine and one or two pairs of supraorbital spines close to the axis (Ahyong et al., 2010;Robins et al., 2012). Eomunidopsis limonitica bears a single triangular rostrum. All other characters match Eomunidopsis and Galatheidae, which is why we maintain placement within that genus and family.
Vía Boada (1981) assigned Eomunidopsis orobensis as the type species upon the erection of the genus but incorrectly (International Commission for Zoological Nomenclature, 1999, article 70.2) listed E. navarrensis as the type species in Vía Boada (1982), in which he provided a diagnosis of the genus. The suggested change may relate to not having a specimen of E. orobensis with a complete rostrum available, nor did Ruiz de Gaona (1943) and Bataller (1950). Vía Boada (1982, p. 13) provided a diagnosis of the genus, indicating that rostra would have a tridentate tip. However, Klompmaker et al. (2012, fig. 5E) showed a specimen of E. orobensis that seems to have only a single spine at the tip. As multiple taxa were ascribed to and removed from Eomunidopsis since 1982, we provide a new genus diagnosis. Bishop, 1985 Figure 5 Types. Remarks.-The holotype and one of the paratype carapaces have been figured in dorsal view. We here also provide new figures of the holotype from different angles. The species has been placed in three genera (see synonymy list), so the taxon merits further discussion. Karasawa and Hayakawa (2000) transferred the species to Luisogalathea on the basis of several characters, including the lack of a median rostral ridge. However, the figured type material has only a small part of the base of the rostrum preserved, and Bishop (1985, p. 602) expressed doubt about the presence or absence of a ridge on the rostrum ("unridged (?)"). The impression of the lower side of the rostrum with the medial groove of the holotype (  Klompmaker et al., 2012, fig. 5D, 5E). Franţescu (2014) placed this species in Munida without further explanation. Subsequently, Nyborg and Garassino (2015) and Feldmann et al. (2015) noted that this species lacks the supraorbital (= supraocular) spines. The latter authors also suggested a reevaluation and expressed that placement within Galatheidae might be possible on the basis of the wide rostrum and ornamentation. Munididae are characterized by supraorbital spines, so placement in Munida is not tenable. We questionably place the species in its original genus within Galatheidae because the preserved characters match the original (Vía Boada, 1982, p. 13) and emended diagnoses of Eomunidopsis herein. New material with a complete rostrum is needed to further evaluate this species.
Remarks.-Vega et al. (2019) presented two specimens they ascribed to Luisogalathea cretacea (Stenzel, 1945). Although this species is known from a not-too-distant region (Texas, USA), there is a substantial temporal gap of ca. 25 million years between the specimens, warranting a reassessment. Specimen IGM-11362 (Vega et al., 2019, fig. 3.1-3.4) does not represent L. cretacea because of the ridge on the axis of the rostrum, whereas L. cretacea has a groove here as described and figured by Stenzel (1945, p. 430-431, pl. 43.3) and Schweitzer and Feldmann (2000, fig. 2) and further implied by Franţescu (2014, fig. 3). This specimen does not fit the diagnosis of Luisogalathea (see Karasawa and Hayakawa, 2000) because the rostrum bears an axial ridge and shows lateral spines near the tip. Instead, this specimen has many similarities to specimens of the late Albian species Eomunidopsis navarrensis (Van Straelen, 1940) found in Spain (see Klompmaker et al., 2012, fig. 4). The substantial temporal gap alone already may argue against conspecificity, which is confirmed by a morphological difference. The   Klompmaker et al., 2012Klompmaker et al., , 2014, but the posterior margin of this region is more rounded in IGM-11362, and the anterior margin of the cardiac region of IGM-11362 appears less well defined. Eomunidopsis aldoirarensis bears more-frequently interrupted ridges than the new species, has fewer spines in the posterior part of the lateral margin, and the tip of the mesogastric process is wider. Compared with E.? cobbani, the new species has less-curved posterolateral margins. Karasawa et al. (2008) fig. 3.5-3.7. Eomunidopsis limonitica (Stenzel, 1945; Vega et al., p. 3, fig. 3.11-3.21. Remarks.-The second of the two specimens of Luisogalathea cretacea (IMG-11363) presented by Vega et al. (2019) appears to be not conspecific to L. cretacea either, once again because of the presence of a ridge on the axis of the rostrum. Moreover, the mesogastric region is clearly defined on all sides, which is not the case for L. cretacea (Stenzel, 1945, pl 43.3;Schweitzer and Feldmann, 2000, fig. 2; Franţescu, 2014, fig. 3). Mexican specimens identified as Eomunidopsis limonitica by Vega et al. (2019) are ca. 25 million years older than the specimens from the United States (Texas) (Stenzel, 1945;Franţescu, 2014). Although the Mexican specimens are close to this species morphologically, there are some differences compared with the two figured specimens from the United States. One specimen (Vega et al., 2019, fig. 3.16, 3.17) is much more tubercular on the anterior carapace. Others show more ridges with granules on that region (Vega et al., 2019, fig. 3.13, 3.18, 3.19), which may or may not be influenced by preservation, the method of photography, and/or intraspecific variation. One difference that appears to be consistent is the short groove running from the anterolateral corners of the cardiac region to the groove branching off the cervical groove, which is present in the Mexican specimens but apparently absent in the E. limonitica from the United States (Franţescu, 2014, fig. 4). Such a groove is present in the type species of Eomunidopsis, E. orobensis, from the late Albian of Spain (Vía Boada, 1982, pl. 2.8, 2.9; Klompmaker et al., 2012, fig. 5). Our comparisons between the Mexican and US specimens are complicated by the fact that the Mexican specimens are smaller (2-4 mm wide) than the two US specimens (5-6 mm wide), implying that ontogenetic change is more difficult to consider. We conclude that the specimens are highly unlikely to represent L. cretacea or E. limonitica and refer them to open nomenclature within Eomunidopsis. More complete, well-preserved specimens are needed to determine the range of intraspecific variation to determine whether there is one new species with variable ornamentation (compare Vega et al., 2019, fig. 3.5-3.17)  Holotype.-BEG0002118.000.
Remarks.-Galathea cretacea from the Early Cretaceous (Albian) of Texas, USA, has been placed in a variety of genera since its description (see preceding synonymy list). Feldmann et al. (2015) called for a reevaluation of the placement of the species in Munida. The reassignment to a different family (Munididae) by Franţescu (2014) was based primarily on similarities in the sternum of Galathea cretacea with Munida, but a differentiation to venters of Galatheidae was not provided. Moreover, Munididae are diagnosed by a trifid frontal margin with a central rostral spine and one or two pairs of supraorbital spines close to the axis (Ahyong et al., 2010;Robins et al., 2012), but G. cretacea bears a single triangular rostrum without adjacent spines. Thus, placement of G. cretacea in Munida is not tenable. Luisogalathea was defined as possessing no lateral spines on the rostrum, but Schweitzer and Feldmann (2000) noted small spines on this margin, more consistent with Galathea. The medial groove on the rostrum is also in agreement with the type species of Galathea, extant G. strigosa. Eomunidopsis bears a rostrum with a median keel. Additional specimens of G. cretacea with complete rostra are required to further assess its generic placement. For now, we questionably assign it Galathea. The species has been described first from the upper Albian Pawpaw Formation in Texas, USA (Stenzel, 1945;Franţescu, 2014). A specimen ascribed to G. cretacea from the Santonian of Texas was figured by McCall et al. (2012, fig. 4-U1) and is refigured herein (Fig. 8). This specimen is ca. 15 million years younger than the holotype, which is why we restudied it. Unfortunately, the rostrum is not preserved, and the ventral side is not as well preserved as those in Franţescu (2014). The dorsal side is  Journal of Paleontology 96(5):1087-1110 comparable to the holotype, but the rostrum is a key feature. It is difficult to ascertain whether the rostrum is truly narrow or whether it was wider, but breakage caused it to appear narrower. We did not observe obvious bases of supraocular spines as in Munididae, which suggests placement in Galatheidae is most likely. Specimens with complete rostra are necessary to determine the species.
Family Munididae Ahyong et al., 2010Genus Protomunida Beurlen, 1930 Type species.-Protomunida munidoides (Segerberg, 1900 Varol (1998), which is equivalent to lower Zone NP3 (Martini, 1971  Remarks.-Despite this species being commonly cited in the literature (see synonymy lists), no specimens from the previous literature other than the type specimen can be confidently attributed to this species. The posterior margin of the lectotype is incompletely preserved, but it is preserved to the extent that we can confirm the absence of a tubercle/spine or a base thereof. This statement is supported by visual inspection combined with a quantitative comparison with the position of spines on this margin for Protomunida eurekantha n. sp. (Fig. 9.6; Table 2  spines marking the anterior border of the epigastric regions appears to vary from two to five and the number of spines on the epibranchial regions varies from two to three. However, taphonomy may have affected the preservation of these spines. Because this species is diagnosed and described anew here, a comparison with congenerics is necessary. Differences between P. munidoides and the new species are described in the remarks sections of those taxa. The central part of the rostrum of P.? pentaspinosa appears to be wider at its base, and the spines on the lateral margins appear to extend more posteriorly. Protomunida primaeva does possess spines just posterior to the transverse groove branching off the cervical groove, unlike P. munidoides. In addition, the base of the rostrum has a shorter keel in P. primaeva. Protomunida spitzbergica bears transverse ridges that are more interrupted than in P. munidoides. Protomunida primaeva (Segerberg, 1900) Figure 10 v 1900 Munida primaeva Segerberg, p. 353, pl. 7.6. ? 1901 Galathea munidoides;Woodward, p. 489, pl. 12.8. ? 1969 Protomunida munidoides;Glaessner, p. R482, fig. 290.1. v 1994 Munida primaeva;Collins and Jakobsen, pl. 10.2. v 1997 Munida primaeva;Jakobsen and Collins, pl. 2.8. ? 2010 Munida primaeva;Damholt et al., fig. on  Diagnosis.-Carapace excluding rostrum ∼15-25% longer than wide; maximum width about mid-length. Rostrum with narrow, sharp, triangular central rostral spine; fairly flat but with faint keel on axis only at base rostrum. Lateral margins of carapace with spines on anterior two-thirds. Posterior margin slightly concave, rimmed, without spines. Spines present just posterior to transverse groove branching off cervical groove.
Description.-Carapace excluding rostrum ∼15-25% longer than wide, up to ∼6 mm wide; weakly convex longitudinally, moderately convex transversely, maximum width about mid-length. Rostrum with narrow, sharp, triangular central rostral spine; fairly flat but with faint keel on axis only at base rostrum; slightly downturned at distal half; one spine at base rostrum on either side of central rostral spine. Upper orbital margin concave. Forward spine at anterolateral angle. Epigastric regions distinctly bordered, triangular, with row of spines near anterior border. Hepatic regions small, weakly delimited. Protogastric and mesogastric regions not or poorly delimited, except for sharp tip of mesogastric process that may show two small tubercles. Cervical groove broadly U-shaped and distinct, somewhat sinuous more laterally, intersects with lateral margin posterior to hepatic region. Transverse grooves branching off cervical groove distinct, separating anterior and posterior branchial regions. Cardiac region appears straight at anterior border and straight to rounded posterior to it. Intestinal region not clearly defined. Lateral margins with spines on anterior two-thirds. Posterior margin slightly concave, rimmed, without spines. Ornamentation consists of fairly continuous, transverse ridges on most of carapace; tubercles/spines on epibranchial regions and near anterior margin; ornamentation appears similar on internal molds and casts. Spines present just posterior to transverse groove branching off cervical groove. Muscle scar impressions generally not discernable, except for posterior gastric muscles in some specimens (MMH 249, OESM 11199). Posterior gastric pits within cervical groove expressed as slits. Ventral surface, abdomen, and appendages unknown. Remarks.-No specimens from the previous literature other than the type specimen can be attributed to this species with certainty. No obvious ontogenetic variations are observed within P. primaeva. In terms of intraspecific variation, the number of spines marking the anterior border of the epigastric regions appears to vary from four to five, and the number of obvious spines on the epibranchial regions varies from two to three. Taphonomic processes may have influenced the preservation of these spines.
A comparison with the two new species of Protomunida is provided in the following and a comparison with P. munidoides Table 2. Relative position of the spines on the posterior margin for Protomunida eurekantha n. sp. For the lectotype of Protomunida munidoides (Segerberg, 1900)   Journal of Paleontology 96(5):1087-1110 was given in the preceding. Compared with P.? pentaspinosa, the base of the central rostral spine appears narrower in P. primaeva, and the latter species bears spines just posterior to groove branching off cervical groove, which appear absent in P.? pentaspinosa. Protomunida spitzbergica bears transverse ridges that are much more interrupted than in P. primaeva.
Diagnosis.-Carapace excluding rostrum ∼5-15% longer than wide; maximum width in posterior half of carapace. Rostrum triangular and sharp, slightly downturned, with strong median ridge along entire axis. Lateral margins of carapace with spines on anterior half. Posterior margin slightly concave, rimmed, without spines. Spines absent just posterior to transverse groove branching off cervical groove.
Description.-Carapace excluding rostrum ∼5-15% longer than wide, up to ∼4 mm wide; weakly convex longitudinally, moderately convex transversely; somewhat widening posteriorly, maximum width in posterior half of carapace. Rostrum with narrow, sharp, triangular central rostral spine and strong median ridge along entire axis; slightly downturned at distal half; one spine at base rostrum on either side of central rostral spine. Upper orbital margin concave, rimmed. Forward spine at anterolateral angle. Epigastric regions distinctly bordered, triangular, with row of spines near anterior border. Hepatic regions small, weakly delimited. Protogastric and mesogastric regions not or poorly delimited, except for sharp tip of mesogastric process, which may show two small tubercles. Cervical groove broadly U-shaped and distinct, somewhat sinuous more laterally, intersects with lateral margin posterior to hepatic region. Transverse grooves branching off cervical groove distinct, separating anterior and posterior branchial regions. Cardiac region appears straight at anterior border and straight to rounded posterior to it. Intestinal region not clearly defined. Lateral margins with spines on anterior half. Posterior margin slightly concave, rimmed, without spines. Ornamentation consists of fairly continuous, transverse ridges on posterior carapace and on central gastric region; tubercles/spines on epibranchial regions and near anterior margin; ornamentation appears similar on internal molds and casts. Spines absent just posterior to transverse groove branching off cervical groove. Muscle scar impressions generally not discernable, except posterior gastric muscles in holotype. Posterior gastric pits within cervical groove expressed as slits. Ventral surface, abdomen, and appendages unknown.
Etymology.-Named after Peter Bennicke, sole collector of the specimens.
Measurements ( (Jakobsen and Collins, 1997); L. affinis (Jakobsen and Collins, 1997); and Raniliformis baltica (Segerberg, 1900). All these species except P. eurekantha are known from the middle Danian Faxe Formation of Faxe, Denmark (nannofossil zone NNTp2G-3, ca. 64.5 Ma, see Hvid et al., 2021). One of us (E.S.) investigated the nannofossils assemblage preserved in the carbonates of the glacial erratics to determine the age. Stratigraphically important nannofossil taxa (for nannofossil taxonomy, see Young et al., 2021) include Biantholithus sparsus Bramlette and Martini, 1964;Biscutum harrisonii Varol, 1989; Coccolithus pelagicus (Wallich, 1877); Cyclagelosphaera alta Perch-Nielsen, 1979; Neocrepidolithus cruciatus (Perch-Nielsen, 1979); Neocrepidolithus fossus (Romein, 1977); and Zeugrhabdotus sigmoides (Bramlette and Sullivan, 1961) (see Appendix 1 for all taxa), which assigns the sample to nannofossil subzone NNTp2C (uppermost lower Danian or ca. 65.5 Ma, see Hvid et al., 2021). This age is ca. 1 Myr older than the Faxe decapods, extending the stratigraphic range of many of these decapod taxa. A principal difference between the new species and all other species of Protomunida is the strong keel present along the entire axis of the rostrum, a feature that is not seen to that extent in congenerics. Some other differences include the following. Protomunida? pentaspinosa bears spines located more posteriorly on the lateral margins. Protomunida spitzbergica has transverse ridges that are more interrupted. Unlike P. bennickei n. sp., P. primaeva and P. eurekantha show spines just posterior to where the posterior cervical groove branches off the cervical groove. The carapace of P. primaeva appears proportionally longer. Protomunida eurekantha also exhibits spines on the posterior margin, absent in P. bennickei. The spines on the lateral margin appear to extend more posteriorly in P. munidoides, and the carapace of P. bennickei appears proportionally wider than that of P. eurekantha. No obvious ontogenetic and intraspecific variations are observed within P. bennickei.
Diagnosis.-Carapace excluding rostrum ∼20-25% longer than wide; maximum width about mid-length. Rostrum with narrow, sharp, triangular central rostral spine; fairly flat but with faint keel on axis only at base rostrum. Lateral margins of carapace with spines on anterior two-thirds. Posterior margin slightly concave, rimmed, with spines. Spines present just posterior to transverse groove branching off cervical groove.
Description.-Carapace excluding rostrum ∼20-25% longer than wide, up to ∼5 mm wide; weakly convex longitudinally, moderately convex transversely, maximum width about mid-length. Rostrum with very narrow, sharp, triangular central rostral spine; fairly flat but with faint keel on axis only at base rostrum; slightly downturned at distal half; one spine at base rostrum on either side of central rostral spine. Upper orbital margin concave. Forward spine at anterolateral angle. Epigastric regions distinctly bordered, triangular, with row of spines near anterior border. Hepatic regions small, weakly delimited. Protogastric and mesogastric regions not or poorly delimited, except for sharp tip of mesogastric process that may show a small tubercle. Cervical groove broadly U-shaped and distinct, somewhat sinuous more laterally, intersects with lateral margin posterior to hepatic region. Transverse grooves branching off cervical groove distinct, separating anterior and posterior branchial regions. Cardiac region appears straight at anterior border and straight to rounded posterior to it. Intestinal region not clearly defined. Lateral margins with spines on anterior two-thirds. Posterior margin slightly concave, rimmed, with at least two spines near position of lateral edges of cardiac region. Ornamentation consists of fairly continuous, transverse ridges on most of carapace; tubercles/spines on epibranchial regions and near anterior margin; ornamentation appears similar on internal molds and casts. Spines present just posterior to transverse groove branching off cervical groove. Muscle scar impressions generally not discernable, except for posterior gastric muscles in OESM 11172. Posterior gastric pits within cervical groove expressed as slits. Ventral surface, abdomen, and appendages unknown.
Etymology.-The species name (eurekantha) is a merger of the Greek Eureka or ϵὕρηκα (heúrēka) meaning "I have found (it)" and the Greek άκανθα or acantha meaning spine. This name refers to the spines on the posterior margin of the carapace, which are diagnostic for this species and were first noted by one of us (S.L.J.) after casting external molds.
Several specimens from the previous literature can be attributed to this new species with certainty because they bear spines on the posterior margin. Such spines were not previously recognized in galatheoids from Faxe, but spines at or near the posterior margin can be of prime taxonomic importance in distinguishing species in extant galatheoids (Macpherson, 2004, p. 249;Osawa and Takeda, 2007, p. 140;Macpherson et al., 2020, p. 24). The presence or absence of spines on the posterior margin is relatively constant in the same species in extant galatheoids (never zero to two or zero to four) (E. Macpherson, personal communication, 2021), and sexual dimorphism in spines on the dorsal carapace has not been observed (E. Macpherson, personal communication, 2020). When broken, the bases of these spines in fossil specimens can be observed after careful examination with a hand lens or microscope.
The main difference between all other species of Protomunida and P. eurekantha n. sp. is the presence of spines on the posterior margin, which are lacking in other species. Some additional differences per species are discussed briefly. A comparison with P. bennickei n. sp. was provided in the preceding. Protomunida munidoides and P.? pentaspinosa lack the spines just posterior to the groove branching off the cervical groove, which are present in P. eurekantha. Moreover, P. munidoides bears a longer keel on the axis of the rostrum, and the base of Journal of Paleontology 96(5):1087-1110 the central rostral spine appears wider in P.? pentaspinosa. Differences other than the presence/absence of spines on the posterior margin between P. eurekantha and P. primaeva are minimal. Protomunida spitzbergica bears more interrupted transverse ridges.
Description.-"The carapace is hardly wider behind than in front, its lateral margins are straight and almost parallel. The rostrum is very wide, appreciably triangular and appears to have been quite long, tending to be styliform; it is a little depressed compared to the dorsal surface, at its base it has a small keel and its edges are marginalized. The antero-lateral angles of the carapace are rounded, they are attached to the rostrum by a smooth surface, limited on the internal side by a small groove. This surface corresponds to a small bead. The cervical groove is very deep, strongly concave towards the front, this depth is held in the antennal groove which follows it, this one is hardly marked. The gastric region is very prominent. The latero-posterior angles of the carapace are broadly rounded, projecting a little over the posterior marginal edge. Scaly tubercles, more or less arranged in rows, cover almost all parts of the cephalothorax. These tubercles have a tendency to become spiniform towards the lateral edges" (translated from Van Straelen, 1925, p. 308).
Remarks.-Palaeomunidopsis moutieri from the Middle Jurassic (middle Bathonian) of France was considered the oldest fossil representative of Galatheoidea, but Fraaije et al. (2019a) declared it a nomen dubium in the same paper in which they described a newly proposed oldest galatheoid, Gastrosacus raboeufi, from the late Bathonian of France. We disagree with this nomen dubium as P. moutieri was published with a detailed description, a line drawing, and a photo. While the illustrations may not be the high quality preferred today, they fulfill the rules set within the ICZN (International Commission on Zoological Nomenclature, 1999, article 12). Although the holotype was presumed to be destroyed in World War II, it is clearly described by Van Straelen (1925). The line drawing is inaccurate because it does not match the photo closely (compare the course of the circumgastric groove for example), but the photo shows hallmark characters of a munidopsid: the strong circumgastric groove and the base of a wide rostrum accompanied by a medial keel. These characters were also described. Other characters described by Van Straelen (1925) such as the tubercular ornamentation, the well-defined gastric region, and lateral margins with spines are also mentioned in the diagnosis of Munidopsidae (Robins et al., 2013, p. 175). Van Straelen (1925) inferred the rostrum would be long, a character that is common in Jurassic Munidopsidae (Robins et al., 2013). A similar circumgastric groove would also be present in Gastrodoridae, Parapylochelidae, and Pylochelidae according to Fraaije et al. (2019a). However, the groove in the gastric region in Gastrodoridae is generally more V-shaped, more sinuous, and less circular than in P. moutieri, and gastrodorids do not have a wide rostral base (Van Bakel et al., 2008;Klompmaker et al., 2011;Fraaije et al., 2013;Krzemińska et al., 2020). Members of Parapylochelidae do not possess a circular groove in the gastric region and bear a short rostrum without a keel (Fraaije et al., 2012a(Fraaije et al., , 2019bFraaije, 2014). Pylochelidae exhibit a gastric groove that is much more sinuous and V-shaped, and the rostrum appears short (Van Bakel et al., 2008;Fraaije et al., 2012bFraaije et al., , 2019bFraaije, 2014). In conclusion, we reinstate P. moutieri as a munidopsid and oldest galatheoid to date. Gastrosacus raboeufi is only slightly younger than P. moutieri and is from the same region in France, warranting a further comparison between these two munidopsids. According to the photos available, some differences may be present. Van Straelen (1925) mentioned a wide rostrum, and it may indeed be wider than in G. raboeufi. Moreover, the tubercles of P. moutieri are said to have been arranged in rows, which is not obvious in G. raboeufi. Munidopsid specimens from the stratigraphic horizon in which P. moutieri was found and additional specimens of G. raboeufi would be most welcome for further comparisons and to evaluate the generic status of Palaeomunidopsis because the only specimen has many similarities to Gastrosacus von Meyer, 1851. For now, we keep the taxa separate within their respective genera.  (Vašíček and Skupien, 2016;Vašíček et al., 2018).
In addition, rostrum with broad base, triangular, with moderately strong keel along entire axis, lateral margins of rostrum containing spines, rostral surface with granules.
Remarks.-The new specimen represents the second occurrence and the third specimen of this species in addition to the holotype and paratype from the Tithonian of Ernstbrunn, Austria. The Czech specimen is slightly smaller than the holotype and shows additional features not seen in the type specimens, such as a complete rostrum.

Discussion
As of July 2021, 199 galatheoid species have been recorded from the fossil record, mostly originating from Europe. We add five species, thereby expanding their diversity by 2.5%. The Early Cretaceous (Barremian) Mexican galatheoid fauna was initially thought to consist of three known species (Vega et al., 2019), but the fauna is more endemic and consists of at least two new species as described herein. Consequently, the North American fossil record of galatheoids now consists of 17 species. As for most fossil galatheoids, most taxa discussed herein were found among reef-associated limestones. Although coral reefs are suggested to have played a key role in the high diversity and radiation of galatheoids in those environments during the Late Jurassic (e.g., Robins and Klompmaker, 2019), the Journal of Paleontology 96(5):1087-1110 relative preservation potential of galatheoids in reef environments versus other habitats has been insufficiently studied. However, squat lobsters are not exquisitely preserved in these environments generally because appendages and ventral surface are nearly always disarticulated from dorsal appendages. Moreover, many extant galatheoids are found in the Indo-West-Pacific Coral Triangle region (Schnabel et al., 2011) and various galatheoids live in symbiosis with corals (Baeza, 2011). Therefore, although a preservational signal cannot be fully excluded, a biological explanation for their high diversity in reef-associated settings is plausible. Another aspect of preservation relates to anatomical details on the carapace. Whereas the fossil record of galatheoids relies heavily on internal molds and carapace specimens with cuticle preserved, external molds are often not collected. A substantial part of extant galatheoids show spines on multiple carapace regions, which are used to diagnose taxa (Baba et al., 2009;Macpherson and Robainas-Barcia, 2015;Rodríguez-Flores et al., 2018;Schnabel and Ahyong, 2019). The discovery of Protomunida eurekantha n. sp. from the Danian of Denmark, principally by making casts of external molds, led to the discovery of distinct spines on the posterior margin by one of us (S.L.J.) in many specimens (Fig. 12). Small spines in the already small galatheoid carapaces often break, making them harder to observe on internal molds and specimens with cuticle. Whereas the Faxe decapod fauna has been very well studied systematically (von Fischer-Benzon, 1866;Segerberg, 1900;Jakobsen and Collins, 1997;Jakobsen et al., 2020), this relative common galatheoid species remained undescribed till now. Therefore, more frequent collecting and studying of external molds for galatheoids and other decapods could aid in the discovery of more cryptic species.

Acknowledgments
We thank E. MacPherson (Centro de Estudios Avanzados de Blanes, Spain) for sharing insights on morphological details of modern galatheoids; A. Busulini (Società Veneziana di Scienze Naturali, Museo di Storia Naturale) helped to get permission to reuse images in Beschin et al. (2016) and kindly provided Figure 4; M. Nose (Bayerische Staatssammlung für Paläontologie und Geologie, Germany) provided access to specimens and museum numbers; A. Nützel (Bayerische Staatssammlung für Paläontologie und Geologie, Germany) provided access to photography equipment for the same specimens in 2014. We are grateful to the two anonymous reviewers and the editors for providing comments that improved this paper. The specimen of Vetoplautus latimarginus was collected in 2013 by the authors during a field trip as part of the 5th Symposium on Mesozoic and Cenozoic Decapod Crustaceans, so we thank the organizing committee; R. Portell (Florida Museum of Natural History) arranged a loan of the specimen, photographed using equipment of the University of California Museum of Paleontology. L. Appleton (University of Texas at Austin) arranged the loan of the holotype of Galathea? cretacea and a specimen previously assigned to this species. We thank M. Florence (Smithsonian National Museum of Natural History) for access to the holotype of Eomunidopsis cobbani after the 9th International Crustacean Congress in 2018 and J. Milàn (Geomuseum Faxe) for providing museum numbers. S.B. Andersen collected part of the Protomunida specimens from Faxe and P. Bennicke collected all Protomunida specimens from Hesnaes Beach. A.R. Bashforth (Natural History Museum of Denmark) provided information on the lectotype specimen museum numbers. C. Olsen (GEUS) is thanked for preparing the nannofossil sample from the Hesnaes Beach chalk.