Morphological structure of head and comparison between Agraulos longicephalus Hicks, 1872 and A. ceticephalus Barrande, 1846, from the of Spain and Czech Republic.

Daniel Eduardo Rojas Ariza

Proyecto de Grado para optar por el título de Geocientífico Asesor: Jorge Esteve Serrano

Universidad de los Andes Facultad de Ciencias Departamento de Geociencias Bogotá, Colombia – diciembre 2020 Contents

1. Abstract 2. Introduction 2.1. Background 2.2. Ptychopariid morphologic features 3. Objective 3.1. General 3.2. Specifics 4. Study Material and Methodology 4.1. Fossils 4.2. Methodology 5. Results: Visualization of morphological variance 5.1. Regression Analysis 5.1.1. Individual regressions 5.1.1.1. A. longicephalus Hicks, 1872 5.1.1.2. A. ceticephalus Barrande, 1846 5.2. Principal Component Analysis (PCA) 5.3. Discriminant Function Analysis (DFA) 5.4. Procrustes ANOVA 6. Discussion 7. Conclusions 8. Acknowledgments 9. References 1. Abstract

Order has always been a problematic group concerning the taxonomic classification of , and even though the temporality of this group is well known, the relation between this order with others, and, the difficulty for classifying ptychopariids have let to diverse problems in the phylogenetic classification of this, and other primitive orders of trilobites. Agraulos is one of the genera belonging to this order and is a common group of the middle Cambrian, it is a well-known inside ptychopariids, therefore, morphological knowledge about this taxon could be helpful to analyze morphological patterns among other from the genus and even other genera inside the family Agraulidae, and even more, about the superfamily Ellipsocephaloidea. This paper provides a morphological statistical comparison based on landmark geometric morphometric techniques performed on specimens of two species of this genus, Agraulos longicephalus Hicks, 1872, and Agraulos ceticephalus Barrande, 1846, from the Drumian stage of Spain and the Czech Republic, respectively. The findings of this study suggest that, even though these species are morphologically similar, there are significant differences as to facial suture and glabella shape and size. In general, allometry in the samples seems to be an irrelevant factor related to morphological variation, but, individually, it is a significant source of variation in morphological patterns of A. ceticephalus. The results obtained from this study may contribute to future similar studies interested in analyzing the morphology of other species, genders, or even families inside order Ptychopariida.

2. Introduction

2.1. Background

Morphological variation represents a fundamental study phenomenon in phylogenetic classification, as well as in the task of following the routes of evolution through inter and intraspecific variations. Particular evolutionary trends can be understood to a greater extent by supporting hypotheses and conclusions with knowledge regarding degrees and patterns of variation within the studied taxa (Webster, 2011). In the case of the fossil record, morphological variation applied in paleontological analyses may be caused by different additional factors different from the biological ones, unlike the common case, when the study object is an actual non-extinct organism. These factors are related directly to taphonomy, as Webster & Hughes (1999) explain, which can generate shape distorts, as a consequence of compaction, increase in the shape variation, and possible biological misinterpretations. A similar situation happens with geographic variation (morphologic changes range of taxa spread in different locations), because of the lack of enough samples in some cases, a common problem in paleontology, which causes the frequent difficulty of distinguishing between geographical intraspecific variation from interspecific disparity (Webster, 2011).

On the other hand, one of the most impactful biological phenomena in morphological variation is allometry, which refers to the changes of shape associated with the size of the organism (Klingenberg, 2009, 2016). Allometry may have great unpredictable effects on the behavior of morphological analysis results, therefore, it is always useful to address the magnitude of the impact that this characteristic has on the morphological variation.

Different authors agree with the fact that the Order Ptychopariida has been significantly problematic concerning the taxonomic classification of the families within it, its paraphyletic character among the general phylogeny, and the general implications that this group has regarding the Cambrian and post-Cambrian trilobite classification (Cotton, 2001; Fortey, 2001; Webster, 2011; Rasetti, 1951 in Cotton, 2001). Nevertheless, Webster (2011) states that analyzing ptychopariid trilobites can provide a substantial knowledge base for understanding the first steps in the evolution of this diverse group of , and, that is the reason why this study is a small contribution to the palaeobiological knowledge about this group.

Order Ptychopariida appeared in the late lower Cambrian, as a descendant group of Redlichiids (Fortey, 1990), and, currently is thought that this order comprises 4 suborders, from which only 3 are partially known: Ptychopariina, Olenina, Harpina and a last uncertain suborder (Whittington et al., 1997). The common morphological features that characterize this order are the following: i) A tapering glabella; ii) preglabellar field; iii) presence of natant hypostome; iv) opisthoparian sutures; v) rim-like cephalic borders extending into genal spines; vi) commonly more than 12 thoracic spines; vii) a micropygidium and viii) a subdued surface sculpture, (Fortey, 2001; Fortey, 2009). Within this group, the genus Agraulos Hawle and Corda, 1847, is mostly known from middle Cambrian, commonly it’s preservation is relatively poor, and most specimens have been collected from Spain, England, Poland, Czechoslovakia, France, North America among other countries (Samson et al., 1990).

Agraulos is an effaced Cambrian ptychopariid very common in the Cambrian strata of West Gondwana as well as in Laurentia and Baltica (Fletcher, 2017; Geyer & Landing, 2001; Gozalo et al., 2011; Liñán et al., 2008). However, the taxonomy of this group is highly complicated given the relatively simple cranidial shape and body pattern. Taxonomic studies of early ptychoparioid trilobites would be improved thanks to a compressive understanding of their morphological variation. These Cambrian trilobites have been assigned to the family Ptychopariidae Matthew, 1887, according to Fortey in Whittington et al. (1997) and were considered the plesiomorphic ancestral stock from which more derived trilobite clades arose (Whittington et al., 1997; Fortey, 2001). However, its general body pattern and relatively minor morphological variation make this group a challenge in the trilobite systematic. Further, there is no doubt that analyzing morphological variation in organisms of this, and similar groups, is always a complex task, as taphonomic factors, as mentioned above, may play a fundamental role in the generation of biases regarding the interpretation of morphological traits, therefore, is necessary to understand very well how taphonomy can distort fossils dimensions and shape, as well as the information- related limitations that the investigator has when describing morphological comparisons (Fletcher 2017). Researchers have done detailed studies regarding ptychoparioid cladistics (Fortey, 1990, 2001; Lochman et al., 1947; Rasetti, 1963; Webster, 2011; Whittington et al., 1997), and, even though there is still an immense amount of progress to be made, to describe and understand, with great precision, the morphological variation of this group, advances had allowed works like the present to be developed. Thus, the present paper proposes to address the assessment and comparison of the cranidial structure of two very well-known species from the Drumian Stage of Spain and the Czech Republic, Agraulos longicephalus Hicks, 1872, and Agraulos ceticephalus Barrande, 1846, respectively. This, by using landmark-based geometric morphometric procedures and statistical analyses, to compare the degree and patterns of interspecific morphological variation, in samples preserved in siliciclastic rocks.

2.2. Ptychopariid morphological features

Figure 1: General morphological division of trilobites. Taken and modified from Gon III (2009). From a general morphological approach of the class Trilobita, the exoskeleton of this organisms is divided into three elements: cephalon, thorax, and pygidium, additionally, the lateral portions of these parts are called gena (cephalon), and the pleural region, composed of thorax and pygidium (Whittington, 1997). About this study, the morphological analyses are focused on the cephalon, as it is the only part that is preserved for both taxa in the studied record, and, is the most important feature for trilobite species classification (Gon III, 2009). For this reason, Figures 2, 3, and 4 show the four cephalic elements included in the seven typical morphologic features of ptychopariids, mentioned above. Figure 2 shows the outlines of the right cephalic borders (green curves) and the common glabellar tapering shape, Figure 3, the natant hypostome, and Figure 4 show the opisthoparian suture.

Figure 2: Hypostome position from the lateral view of Agraulos ceticephalus Barrande, 1846, cephalon (Whittington, 1997).

Figure 3: Typical morphological traits of families of the suborder Ptchopariina, order Ptychopariida. After Gon III (2009).

Figure 4: The three main facial sutures in trilobites, with special attention to the opisthoparian one. After Gon III (2009). 3. Objectives 3.1. General

i) Assessment and comparison of the cranidial structure between Agraulos longicephalus Hicks, 1872 and Agraulos ceticephalus Barrande, 1846 3.2. Specifics

i) Digitize the reference points (landmarks and semi-landmarks), already defined in the photo collection to differentiate the body parts, to be studied, in the samples ii) Construct a data set with the morphological characteristics of interest and their respective measurements, apply a Procrustes ANOVA adjustment to the data, followed by a Principal Component Analysis (PCA), and discriminant function analysis. iii) Evaluate and identify those characteristics with the greatest influence on the general morphological variability of the organisms. iv) Compare the morphological features of A. longicephalus with species of the genus described originally in the Czech Republic, A. ceticephalus. 4. Study Material and Methodology 4.1. Methodology

As the first process for this study, the 93 cephalons were digitized by marking their relevant morphological parts (specifically for this project), with points in a coordinate system called landmarks, which some authors (Bookstein, 1991; Palaniswamy et al., 2006), agree to define them as discrete and precise points which maintain correspondence among all the specimens. The digitizing was done using the software tpsDig2 (Rohlf, 2017), and the morphometric and statistical analyses were performed using the software MorphoJ (Klingenberg, 2011). A total of 7 landmarks were placed in each fossil, only on the right side of each cephalon, because for this study, it is already known that these organisms are symmetric, and also, it is considered that only one side of the cephalon is enough to evaluate, with detail, the morphological variation, as well as, it is helpful to avoid taphonomic effects that are not part of the objectives of this work. Additionally, as said before, the population of this study is part of a bigger collection, and, another reason why some specimens were not taken into account is that they are incomplete or have one fragmented side, therefore, it was determined that applying missing landmarks, for example, to complete those missing structures was not reliable, as the additional information that they would provide, was not significantly relevant for the analysis of morphological variation in this case.

Figure 4 summarizes the landmark-based geometric morphometrics techniques applied in this study and the order of the analyses that were performed.

Figure 4: Morphological analysis flow diagram

4.2 Fossils

Figure 5: Specimens of cephalons of Agraulos ceticephalus Barrande, 1846, (left) and Agraulos longicephalus Hicks, 1872 (right), with digitized landmarks.

93 cephalons were studied in this work, from which 81 correspond to the species A. longicephalus, collected by Prof. Esteve in the Iberian Chain (NE Spain) and 12 to A. ceticephalus, housed at the National Museum of Natural History, Prague (Czech Republic) photographed by Dr. Lukáš Laibl (see Fig. 5). These samples were selected as the best well-preserved ones from a bigger collection of fossils and represent the most suitable material for this morphological study. 5. Results 5.1. Regression Analysis

As the first set of results, Regression Analysis allowed quantifying the influence of the allometry in the morphological variance of both taxa in the dataset. This, to evaluate the influence of the changes in the relative size of cranidial parts for the total size of the specimens and correct the effects of size on the shape, related to the general morphological variance that has been described in the previous sections. The results show that around 5.5% of the variance in the dataset was explained by allometry, which means that this percentage of the variance has its cause in the size changes of individual parts of the cephalon, concerning the total size of it. Therefore, according to Benitez (2020), it is not a substantial influence of allometry, as a remarkable effect is considered to be greater than 10% of the variance. This means that, in this particular case, is not truly necessary to perform a correction of size, as the difference between the datasets (the original and the corrected), will not be significantly relevant, apart from removing the anomalous points in the Principal Components graphs. This is evident in Figure 6, where it can be seen that the differences regarding shape (vertical axis), between both groups, are reasonably small.

Figure 6: Regression score vs. Logarithm of the centroid size. The logarithm of the centroid size is used instead of its original value due to the difference in sizes between specimens. Nevertheless, it is also useful to observe the way this regression analysis results behave when the dataset is divided into two, one dataset for each group. Analyzing both taxa separately will provide even more clear information about the effects of allometry in the morphological variance for each species. This will help to evaluate the results given before, and also, will allow recognizing some possible limitations regarding the interpretations that can be done about all the information that has been obtained. Additionally, these individual regressions are convenient for observing the shape variation differences across the specimens for both groups.

5.1.1. Individual Regressions 5.1.1.1. Agraulos longicephalus Hicks, 1872 For this species, the shape variation behaves similar to what is expected, being the one with more samples available, it shows that bigger cephalons (higher values of Log Centroid Size) tend to have a slightly wider border (value of regression score between 0 and 0.10), concerning transversal axis, however, this difference is considerably minimum, as the slope of the data shown in Figure 7 is significantly small.

Figure 7: Regression Analysis results for Agraulos longicephalus Hicks, 1872, samples

The results presented that around 8.2% of the variance of this individual dataset was explained by the allometry of these specimens, which, as same as before, indicates that there is not a significant influence of allometry in the morphological variance of A. longicephalus.

5.1.1.2. Agraulos ceticephalus Barrande, 1846

In contrast with the other group, from the regression analysis done for samples of A. ceticephalus, it was found that allometry caused approximately 33% of the morphological variance analyzed in this study. This implies that, individually, to observe the real variance, it would be necessary to remove the effect of allometry from the dataset, however, as this study focused on the joint analysis of data sets of both taxa, and the number of specimens of this species is much less than those of A.longicephalus, it is not related with the purposes of this work.

Nonetheless, by observing both, the patterns of variation without and the morphospace occupied by this group without allometry influence in a principal component scatterplot (PC1 vs. PC2), it is verified that, individually, the changes are remarkable, as the morphospace is reduced mainly in the PC1 axis, where the samples are located between -0.09 and 0.09 (Fig. 9). But, even with this, the difference in the PC1 vs. PC2 graph for the whole dataset, with the effect of allometry being removed for both groups, is not distinguishable (see Figure 10).

In addition to the Principal Component graphs, the shape diagrams in Figures 11 and 12, show the morphological patterns in A. longicephalus samples for both Principal Component Analyses done, with, and without allometry. Each diagram corresponds to a limit value of PC1 (the principal component which represents the biggest fraction of variance), positive or negative, in the ranges of scores shown in Figures 8 and 9 (horizontal axis). The general trend suggests that the influence of allometry in the morphological patterns of variation is higher for the positive extreme values of PC1, as the differences in the size and shape of the glabella, as well as the facial suture and part of the cephalic margin, are more significant than the ones related with the negative limit values of this principal component. These differences are much more evident if one observes the preglabellar area and preocular area sizes contrast between both pairs of diagrams.

Figure 8: PC1 vs. PC2 Agraulos ceticephalus Barrande, 1846, with allometry effect.

Figure 9: PC1 vs. PC2 Agraulos ceticephalus Barrande, 1846, without allometry effect

Figure 10: PC1 vs. PC2 complete dataset without allometry effect

A. B.

Figure 11: Agraulos ceticephalus Barrande, 1846, shape changes diagrams for negative scale factor limit. A. With allometry influence (scale factor = -0.1) B. Without allometry influence. (scale factor = -0.09) A. B.

Figure 12: Agraulos ceticephalus Barrande, 1846, shape changes diagrams for positive scale factor limit. A. With allometry influence (scale factor = 0.15) B. Without allometry influence (scale factor = 0.09).

5.2. Principal Component Analysis

As a first approach, principal components allow observing the relationships between the morphological variables and their maximum variance tendencies (forgive the repetition), which, are represented by each principal component. Thus, the principal component graphs, together with the shape changes diagrams, will provide a clear first morphological variation pattern (Figs. 13-16). In this case, the first 3 principal components were considered for describing the total morphological variation (composed of 48 total principal components), as they, added, representing 70.048% of the total variation of the dataset (see Table 1).

Principal Component Eigenvalues % Variance Cumulative % 1 0.00458336 41.355 41.355 2 0.00218732 19.736 61.091 3 0.00099276 8.958 70.048 Table 1: General results of Principal Component Analysis PC1 comprise 41.355% of the variance, the range of scores oscillates between -0.1 and 0.18 approximately, without counting the anomaly present at 0.36 (see Figure 14), and, in Figure 13.A., it is evident that greater values of PC1 imply a wider and larger glabella [1] (regarding sagittal and transverse axis respectively), as well as a narrower superior part of the opisthoparian suture [2] (facial suture with the abbreviation “s” in Figure 2), in the exsagittal direction. The last, also results in narrower preocular and preglabellar areas with greater values of PC1, not only because of facial suture shortening but also because the glabellar growth rate is higher near to the top side of this structure than at the bottom part, adjacent to the palpebral and postocular areas. On the other hand, a negative increase in the PC1 values is associated with wider facial sutures (in sagittal direction) and smaller glabellas, which results in bigger preglabellar, preocular, palpebral, and postocular areas of the cephalon.

[2] [2] [3] [1] [1]

[4]

Figure 13: Morphological variation direction for positive increases of principal components 1 (A) and 2 (B).

Figure 14 and Figure 13.B., show that Agraulos longicephalus Hicks, 1872, tend to have smaller glabellas and broader preocular facial sutures than Agraulos ceticephalus Barrande, 1846, even though both groups share a significant morphospace (scores between -0.4 and 0.1 in PC1), and, some anomalous A. longicephalus specimens are located out of this range of values.

Figure 14: Principal Component 1 vs. Principal Component 2.

Principal Component 2 represents 19.736% of the variance and shows a much delimited shared morphospace than the previous case. Opposite to what was previously described for negative values of PC1, Fig. 13 exhibit that the glabellar size decreases with the increase of the PC2 values, which means that PC2 is inversely proportional to glabellar length in sagittal axis direction [4] and also to glabellar width (transverse direction) [1]. As for the relation between the PC2 variance proportion with the preocular facial suture part, something interesting happens near the boundary between the preocular area and the preglabellar field, since a pivot-like point is located there, which divides the contour into two parts, by their morphological variation behavior. Along the preocular area, the facial suture widens [3] with greater values of PC2, however, as the contour approaches the preglabellar area, where the facial suture joins with the cephalic margin, this border shrinks [2] with the increase of PC2 values, which generates a proportionality relation between preocular area size and the PC2 values, but an inverse relation between preglabellar area size and these principal component values.

PC3 accounts for 8.958% of the total variance, and even though this principal component represents a considerably low percentage of the variation, its relationship with PC2 scores is presented in Figure 16.

[1] *

[2] [3]

[4]

Figure 15: Morphological variations directions for principal component 3.

Figure 15 shows that the positive scores increase of PC3 values implies a growth in the preglabellar area, due to the shortening of the glabella [2] and the elongation of the anterior border [1] in the direction of the sagittal axis. However, the width of the glabella [4] also increases with the PC3 values, while part of the facial suture [3], adjacent to the limit between the anterior area and the palpebral area, and also the eye position, tapers diagonally. It is worth noting that the opposite of the above happens for the negative increase in PC3 values, which, in general, would mean that negative values of PC3 are associated with a longer, but less wide glabella, smaller glabellar area, and wider palpebral, and probably, posterior areas. By considering this, it is possible to notice that, in Figure 16, A. longicephalus, samples tend to have a bigger glabellar area (anterior area), and narrower palpebral area than A. ceticephalus, which means that the palpebral suture, and with this, the eye, is closer to the glabella.

Figure 16: Principal Component 2 vs. Principal Component 3.

5.3. Discriminant Function Analysis (DFA)

Together with the information that Principal Component Analysis has provided, the Discriminant Function Analysis (DFA) will allow finding those specific traits that differentiate morphologically these two species. Concisely, the Discriminant Function Analysis projects multivariate data onto a one-dimensional space, which maximizes the separation between two taxa (Fisher 1936; Anderson, 1984; Davis, 1986, as cited in Hammer & Harper, 2006). Hammer & Harper (2006) explain that it identifies the linear combination of measurements of morphological traits that give the maximal separation of both groups. The success of this separation can be evaluated by observing the position of each specimen in the discriminant axis (horizontal), calculating the percentage of samples correctly assigned to one of the taxa (considering that the degree of the correct assignment must be better than 90%).

Figure 17: Morphological variation for Agraulos longicephalus Hicks, 1872, (dark blue), and Agraulos ceticephalus Barrande, 1846 (cyan blue), following discrimination function (linear combination of morphological variables measured associated with a maximum separation of both species). S.F. refers to the scale factor Then, considering the previous description of the analysis and comparing the morphological variation patterns shown in Figure 17 with the initial discriminant scores graph (Fig. 18), is very evident that both groups are greatly well separated, with A. longicephalus specimens corresponding to positive values for the discriminant function, while the ones belonging to A. ceticephalus species are located in the negative values of the function

The morphological traits, associated with the most significant differences, focus on three regions of the cephalons: the preglabellar area, the preocular area, and the palpebral area. Agraulos ceticephalus Barrande, 1846, species tend to have smaller preglabellar areas but wider preocular and palpebral areas, which means that their facial sutures show to be larger in the transverse axis. On the other hand, A. longicephalus shows a pattern of longer cephalons in the sagittal axis direction, and narrower in the transverse axis direction. Also, their glabella is shorter than the one of A. ceticephalus (as it is evident in Figure 17), and because the cephalic margin extends, their anterior and preglabellar regions are bigger. The above means that, according to the Discriminant Function Analysis, the samples of both species differentiate mainly by the length and width proportions of the facial suture.

Figure 18: Frequency values of specimens for discriminant scores Even though the morphological separation seems to be well done by this algorithm, it is necessary to evaluate the reliability of the discrimination, and this can be done using the leave- one-out cross-validation. This algorithm, according to (Gingerich et al., 2014), removes one specimen from the dataset and trains the remaining data using each of the specimens, this is repeated a number of times equal to the total specimens (n), such that each of the specimens is classified once by the algorithm as a validation sample. Thus, the results of this validation are shown in Table 2 and Figure 19, and it is conspicuous that, based on the dataset, it is easier to classify an unknown specimen as A. longicephalus, than A. ceticephalus, since the specimens of the first are much more, as it can be seen on the blue highlighted areas of Table 2.

Table 2: Discriminant Function Analysis results

Figure 19: Frequency values of specimens for cross-validation scores

The table presents that the specimens who were correctly classified (yellow highlighted numbers) are the significant majority, and the number of specimens wrongly classified (red highlighted) for both species differs only by one. The above, under a parametric p-value significantly small (<0.0001), which indicates that samples from both groups are well differentiated. 5.4. Procrustes ANOVA

The morphological variation of both species is quantified employing the Analysis of Variance (ANOVA), regarding size (centroid size in the landmark configuration), and shape, associated with Procrustes coordinates. (Klingenberg & McIntyre, 1998; Savriama, 2018; Savriama & Klingenberg, 2011) See tables 3 and 4.

Centroid size Effect SS MS df F P (param.) Individual 17312468.05 17312468.05 1 89.41 <.0001 Residual 17619437.02 193620.1871 91 - - Table 3: ANOVA for measurement error for size

Shape, Procrustes ANOVA: Effect SS MS df F P (param.) Pillai tr. P (param.) Individual 0.12821715 0.002671191 48 13.09 <.0001 0.86 <.0001 Residual 0.89141814 0.000204079 4368 - - - - Table 4: Procrustes ANOVA for measurement error for shape Overall, it is found that the parametric p-value related to the main effect of individuals (variation among them); centroid size and shape, is smaller than 0.0001, which implies that the morphological difference between A. longicephalus and A. ceticephalus is statistically significant.

6. Discussion

Overall, the results obtained from Principal Component Analysis and Discrimination Function Analysis evidenced that, based on the cranidial morphology, the morphological variation patterns for both groups were greatly different, and, as a consequence, imply a significant intraspecific variation between these two groups. The patterns of variation observed mainly in Figures 13,14, and 17, agree that the principal cranial structures that distinguish A. longicephalus from A. ceticephalus are the anterior branch and the glabella, as the first species tend to have a much narrower anterior branch (or superior part of facial suture), as well as a more defined tapering glabella than A. ceticephalus, which glabella, apart from being is less transversely tapered, is longer (in the sagittal axis direction). The fact that the anterior branch of the opisthoparian suture of A. longicephalus bends to a much greater degree than the one of the other group (see Figure 17), makes this morphological trait a key factor (and possibly the most important one), in the contrast of both taxa.

Furthermore, the above also indicates that each group morphology has certain concordances with some, but not all, of the common morphological features of ptychopariids. This finding is considered to be very striking because A. ceticephalus seems to have the less complex opisthoparian suture, and, at the same time, the specimens from this species present a significantly less evident intraspecific variation regarding this structure, in comparison with A. longicephalus, which facial suture morphology varies remarkably among samples.

On the other hand, as to the allometry, although it was shown that the overall influence of shape changes, with size, on the morphological variation of the specimens were not enough significant to be considered, it was found that the number of fossils of A. ceticephalus could be an important factor in the great fraction of variance explained by allometry for this individual dataset. Even so, the samples of this group have a greater diversity of sizes, compared with the samples of A. longicephalus, which makes think that study a bigger amount of these specimens could be helpful to address this effect with more precision.

7. Conclusions

The morphological comparison between Agraulos longicephalus Hicks, 1872, and Agraulos ceticephalus Barrande, 1846, demonstrated that these two trilobites have almost opposite morphological variation patterns, and, the two most influential structures in this interspecific disparity are the facial suture (and its junction with the cephalic margin), and the glabella, which are greatly different despite the relatively simple shape of the cephalon of this primitive organisms. The results obtained about the samples from Spain of A. longicephalus denoted a major concordance than A. ceticephalus with the typical cranidial features of ptychopariids, therefore these findings could be useful to improve the knowledge about the Ellipsocephaloidea group and the implications that this superfamily has in the relationship between and order Ptychopariida. Patterns of variation in other species inside the genus Agraulos (or even other genders), should be analyzed, considering this work, for the sake of looking for similarities in other groups, to address possible concordances among taxa. 8. Acknowledgements

To Jorge Esteve for the help and support in this work, and also for motivating my interest and passion for paleontology.

9. References

Barrande, J. 1846.Notice préliminaire sur le Système Silurien et les trilobites de Bohême. vi 1–97. Leipzig.

Benítez, Hugo [EME lab]. (2020, Julio 11). Curso MG: Clase Práctica 03 - Alometría [Archivo de video]. www.youtube.com/watch?v=iyWF3EExJWE.

Bookstein, F. (1992). Morphometric Tools for Landmark Data: Geometry and Biology. Cambridge: Cambridge University Press. doi:10.1017/CBO9780511573064

Cotton, T. J. (2001). The phylogeny and systematics of blind Cambrian ptychoparioid trilobites. Paleontology 44(1). pp. 167-207

Fletcher, T. P. (2017). Agraulos ceticephalus and other Cambrian trilobites in the subfamily Agraulinae from Bohemia, Newfoundland and Wales. Papers in Palaeontology, 3(2). https://doi.org/10.1002/spp2.1071

Fortey, R. A. (1990). Ontogeny, Hypostome Attachment and Trilobite Classification. Palaeontology, 33(3), 529–576

Fortey, R. A. (2001). Trilobite Systematics: The Last 75 Years. Journal of Paleontology 75(6), 1141–1151.

Geyer, G., & Landing, E. (2001). Middle Cambrian of Avalonian Massachusetts: Stratigraphy and correlation of the Braintree trilobites. In Journal of Paleontology. https://doi.org/10.1017/S0022336000031942

Gingerich, J. A. M., Wärmländer, S., Sholts, S., & Stanford, D. J. (2014). Fluted point manufacture in eastern North America: An assessment of form and technology using traditional metrics and 3D digital morphometrics Fluted point manufacture in eastern North America: an assessment of form and technology using traditional metrics and 3D digital morphometrics. World Archaeology, 46(1), 101–122. https://doi.org/10.1080/00438243.2014.892437

Gon III, S. M. (2009). A pictorial guide to the orders of Trilobites. Adaptation of the award-winning website: http://www.trilobites.info.

Gozalo, R., Martorell, J. B. C., Esteve, J., & Liñán, E. (2011). Correlation between the base of Drumian stage and the base of middle Caesaraugustan stage in the Iberian Chains (NE Spain). In Bulletin of Geosciences. https://doi.org/10.3140/bull.geosci.125

Hammer, Ø., & Harper, D. A. T. (2006). Paleontological Data Analysis. Blackwell Publishing. https://doi.org/10.1002/9780470750711 Hawle, I. and Corda, A. J. C. (1847). Prodrom einer Monographie der böhmischen Trilobiten. Abhandlungen Kongligischen Böhemischen Gesellschaft der Wiossenschaften, V. Folge 5(5):1-176

Hicks, H. (1872). On some undescribed fossils from the Menevian Group. Quarterly Journal of the Geological Society of London 28, 173–185.

Klingenberg, C. P. (2011). MorphoJ: an integrated software package for geometric morphometrics. Molecular Ecology Resources 11: 353-357. doi: 10.1111/j.1755-0998.2010.02924.x

Klingenberg, C. P. (2009). Morphometric integration and modularity in configurations of landmarks: Tools for evaluating a priori hypotheses. Evolution and Development, 11(4), 405–421. https://doi.org/10.1111/j.1525-142X.2009.00347.x

Klingenberg, C. P. (2016). Size, shape, and form: concepts of allometry in geometric morphometrics. Development Genes and Evolution, 226(3), 113–137. https://doi.org/10.1007/s00427-016-0539-2

Klingenberg, C. P., & McIntyre, G. S. (1998). Geometric Morphometrics of developmental instability: Analyzing patterns of fluctuating asymmetry with Procrustes methods. 52(5), 1363–1375.

Liñán, E., Gozalo, R., Dies Álvarez, M. E., Gámez Vintaned, J. A., Mayoral, E., Chirivella, J. B., Esteve, J., Yurievich, Z. A., Andrés, J. A., & Zamora, S. (2008). New lower ovetian trilobites (low lower Cambrian) from Sierra Morena (Spain) | Nuevos trilobites del Ovetiense inferior (Cámbrico Inferior bajo) de Sierra Morena (España). Fourth International Trilobite Conference.

Lochman, C. (1947). Analysis and Revision of Eleven Lower Cambrian Trilobite Genera. Journal of Paleontology 21(1), 59–71. http://www.jstor.org/stable/1299294.

Palaniswamy, S., Thacker, N. & Klingenberg, C. (2007). Automatic Identification of Morphometric Landmarks in Digital Images. BMVC 2007 - Proceedings of the British Machine Vision Conference 2007. 10.5244/C.21.77. Ptychopariidae Matthew, 1887 in Paleobiology Database (2018). The Paleobiology Database. Checklist dataset https://doi.org/10.15468/zzoyxi accessed via GBIF.org 2020-06-23.

Rasetti, F. (1963). Middle Cambrian ptychoparioid trilobites from the conglomerates of Quebec. Journal of Paleontology, 37(3), 575–594. http://www.jstor.org/stable/10.2307/1301374

Rohlf, F. J. (2017). tpsDig, digitize landmarks and outlines, version 2.31. Department of Ecology and Evolution, State University of New York at Stony Brook.

Samson, S., Palmer A.R., Robinson R.A., Secor D.T. (2014). Biogeographical significance of Cambrian trilobites from the Carolina slate belt. GSA Bulletin; 102 (11): 1459–1470. doi: https://doi.org/10.1130/0016-7606(1990)102<1459:BSOCTF>2.3.CO;2

Savriama, Y. (2018). A Step-by-Step Guide for Geometric Morphometrics of Floral Symmetry. 9(October), 1–23. https://doi.org/10.3389/fpls.2018.01433

Savriama, Y., & Klingenberg, C. P. (2011). Beyond bilateral symmetry: Geometric morphometric methods for any type of symmetry. BMC evolutionary biology. 11. 280. 10.1186/1471-2148-11-280.

Webster, M. (2011). The structure of cranidial shape variation in three early ptychoparioid trilobite species from the Dyerian-Delamaran (Traditional “Lower-Middle” Cambrian) boundary interval of Nevada, U.S.A. 85(2), 179–225. Webster, M., & Hughes, N. C. (1999). Compaction-related deformation in Cambrian olenelloid trilobites and its implications for fossil morphometry. Journal of Paleontology, 73(2), 355–371. https://doi.org/10.1017/S0022336000027827

Whittington, H. B., Chatterton B.D.E., Speyer, S.E, Fortey, R.A., Owens, R.M, Chang, W.T., Dean W.T., Jell, P.A., Laurie, J.R., Palmer, A.R., Repina, L.N, Rushton, A.W.A., Shergold, J.H., Clarkson, E.N.K., Wilmot, N.V., & Kelly, S.R.A. (1997). Treatise on Invertebrate Paleontology: Part O, Trilobita. In Treatise on Invertebrate Paleontology (Vol 1.)