Did Human Dispersal into Europe Cause the Continent-Wide Extinction of the Pig Sus strozzii at 1.8 Ma?—Review of a Debate

Abstract

For many years, the temporal distribution of pigs in the Pleistocene of Europe drew little attention. This changed when, what became known as, the “suid gap” hypo-thesis was published. Subsequent publications added elements to this hypothesis, while others questioned the hypothesis and even the existence of a “suid gap”. In its most complete form the hypothesis consists of a chain of arguments: (1) pigs are r-selected (a life history trait), (2) therefore fossils of their deciduous teeth are particularly abundant, (3) because being r-selected, pigs are abundant, (4) sites without pig fossils are sure indicators of their absence, (5) at 1.8 Ma, humans dispersed into Europe driving Sus strozzii to extinction in all the continent by competitive displacement, but not in the Middle East, (6) around 1.2 Ma pigs appeared again in Europe, either Sus strozzii, coming from the Middle East, or another species. The proposed link between human and pig ecology increased the interest of this hypothesis. Recently parts of this hypothesis were questioned and a polemic arose. It is the aim of this paper to review the literature and arguments used in favour and against this “suid gap” hypothesis. The hypothesis is rejected, but the life history traits of pigs may prove to be of interest for comparison with humans.

1. Introduction

Since long, it was assumed that there were two pig species in the Pleistocene of the mainland of Europe [1]: Sus strozzii and later the living species Sus scrofa. Though the appearance of the latter was usually situated around the beginning of the Middle Pleistocene, the last appearance of Sus strozzii was less known. Recently, it became a matter of debate whether there was a period in Europe without pigs or not.

In 2015, Martínez Navarro and colleagues [2] proposed the hypothesis that: (1) pigs are r-selected and that therefore they are abundant in present and fossil faunas, and that therefore, if their fossils are not found, they were certainly absent; (2) around 1.8 Ma, humans dispersed into Europe, causing the extinction of Sus strozzii in all the continent, resulting in a period without Suidae; (3) around 1.2 Ma, Sus scrofa or a similar form dispersed into Europe and that this is an important biostratigraphic marker. Later, the term “suid gap” was coined and it was suggested that it might be merely a lack of documentation. Medin and colleagues [3] added to this that humans, bears and pigs competed for underground resources and nuts (acorns, nuts, chestnuts, hazelnuts, almonds and pistachios) and that climatic change may have caused fragmentation of the suid populations. In the meantime, other elements were added to the “suid gap” hypothesis [4,5].

In 2024, Iannucci [6], in a paper published in Quaternary, discussed these hypotheses and concluded that: (1) there is no direct relationship between reproductive strategy and abundance in the fossil record, (2) Sus is present in Peyrolles and other sites with ages between 1.8 and 1.2 Ma, (3) the unequal distribution of sites and lack of dating are an impediment. A polemic arose [7,8].

The “suid gap” hypothesis is attracting much attention, and it certainly is a merit of Martínez Navarro and colleagues [2] to start the debate on the temporal distribution of the suid species in the Pleistocene of Europe. However, some of the arguments against this hypothesis [6] have to be taken more seriously than its proponents are willing to admit [7]. Therefore, it is the aim of this paper to review this debate and in particular the validity of the arguments. The “suid gap” hypothesis as a whole is rejected. Suids were rare and the “suid gap” corresponds to a time when the European fossil record was poor. Since doubts have been raised about the reported pig fossils of this time, it is important that these are published in a way that the identifications can be checked. The “suid gap” hypothesis drew attention to one life history trait in the Suidae. The evolution of life history traits in pigs, merits more attention, since there are various parallels to human evolution.

2. Materials and Methods

Though most of the paper is based on the literature and published data, some specimens of Sus strozzii from the Early Pleistocene of the Cromer Forest-bed Formation at East Runton are described. The specimens are kept in the Natural History Museum in London (NHM). The metrics of these specimens are compared to those of their homologues of Sus strozzii, Sus sp. and Sus scrofa from other localities. The raw data, localities and provenance of the data are indicated in the Supplementary Information. The acronyms of the collections are explained in the Abbreviations section.

The measurements (reported in mm) follow the standard protocol according to Van der Made [9]. The position of upper teeth is indicated with superscript (e.g., M³, P⁴) and that of lower teeth with subscript (M₃, P₄). If both are meant, or if upper/lower is irrelevant, neither super-, nor subscript is used (M3, P4).

3. The History of the “Suid Gap”: Relevant Literature That Has Not Been Cited

In 1926, Pilgrim [10] accompanied his “Fossil Suidae of India” with a phylogenetic tree of the known species of pigs at the time. The tree has a temporal scale, which is very coarse, showing a temporal gap between Sus strozzii and the first Sus scrofa, corresponding to what he considered to be the “Lower Pleistocene”. In fact, no pig species was indicated anywhere for this time.

Though in 1954, Azzaroli [11] dedicated a monograph to Sus strozzii, he was never explicit about whether this species overlapped in age with Sus scrofa, or whether there was a time gap between them. He indicated a possible presence of Sus strozzii at the end of the late Villafranchian or in the Farneta unit, but did not indicate on which fossils or site this was based [12,13]. However he stated that Sus scrofa arrived during the Galerian and that Sus strozzii went extinct at the end of the Villafranchian or in the early Galerian [14].

In 1968, Kurtén [15] was probably the first one to give detailed range charts of the Pleistocene fauna of Europe. Again, the stratigraphy differs from that of today. In his chart, he indicated a gap between the two species, corresponding to the Günz and Waalian, situated at the beginning of the Middle Pleistocene. In his text, he stated that pigs were probably present in Europe throughout the Pleistocene.

In 1982, Guérin introduced a biozonation for the Late Miocene till the end of the Pleistocene [16]. Later these zones became called the MNQ zones. He indicated that the first Sus scrofa appeared in zone 20, with an estimated age of 1.0–0.6 Ma, and that Sus strozzii was present in MNQ19 (zone of Peyrolles, with an estimated age of 1.4 to 1.0 Ma), but not whether it persisted into MNQ20. Faure & Guérin [17] listed 10 sites with Sus strozzii attributed to MNQ16 to 20. Among these sites are Mugello (tentatively placed in MNQ17, 2.5–1.9 Ma) and Tegelen (MNQ20). However, the large mammal fauna from Tegelen is older than the small mammal fauna from there and much older than what they believed [18]. Azzaroli et al. [14] placed Mugello in the latest Villafranchian, close to Selvella and shortly before the end-Villafranchian event. In various publications it was indicated that both species were present in MNQ20 (1.0–0.75 Ma) and that the first Sus scrofa priscus is from Chalôn-Saint-Cosme, Ceyssaguet, Mosbach, Vallonnet and Untermassfeld and Gombasek [19,20,21,22]. Others identified the pigs from nearly all of these sites as Sus strozzii or a related form [5,6,23,24]. Sus scrofa was indicated to be present in Mosbach 1 (Jaramillo) and 2 (Middle Pleistocene) [25]. I studied the pigs from Mosbach in the NMMi and found only one remain indicted to be from Mosbach 1 (the lower level). It is an I₃ of a pig, but does not allow a secure species identification.

When I prepared a paper on the Suinae from Europe, I did not have enough data on the Pleistocene and asked the help of Salvador Moyà Solà [18]. He was of the opinion that there was a period without pigs in Europe. This is reminiscent of the “cricetid vacuum”, a period without Cricetidae in Europe [26]. This cricetid vacuum was much debated at the time and was given biostratigraphic importance. Salvador’s argument for a Sus vacuum was the absence of this genus in Cueva Victoria and various sites in the Orce area, including Venta Micena. These sites were then placed in biozone MmQ2 of a biozonation that was published shortly before [27]. When the paper was published in 1989, a “Sus vacuum” was not mentioned in the text, but showed up in the range chart [18]. One thing is confirming the presence of a species in a timespan, because of the fossils you know, but it is a step further to infer the absence of a species because you do not know fossils of that time and species. When I published range charts for the Suidae from Europe, these inherited the “Sus vacuum” [28,29]. My range charts for the Suidae from Spain have also other gaps, because the Spanish record is less complete than the European one [30,31]. Both Martínez Navarro and Moyà Solà worked at that time at the Institut de Paleontologia Miquel Crusafont in Sabadell and some of my range charts were published in the journal of that institute [29,30].

In 2015, Martínez Navarro et al. [2] published a paper, which postulates a suid gap exactly where it appeared in the range charts I published. They reasoned: (1) pigs are r-selected; (2) therefore they are abundant in present and fossil faunas, hence not finding them is a sure indication of their absence; (3) at 1.8 Ma, Sus strozzii was driven to extinction all over Europe because of competitive displacement by early humans; (4) the appearance of pigs at 1.2 Ma in Europe is an important biostratigraphic marker indicative of the beginning of the Epi-Villafranchian. Cherin et al. [4] coined the term “suid gap” and suggested that it might be merely a lack of documentation. Medin et al. (2019) [3] made the addition to the “suid gap” hypothesis and specified that humans and pigs competed for nuts (acorns, nuts, chestnuts, hazelnuts, almonds and pistachios) and underground resources. They also proposed that climatic change fragmented suid populations, which then became vulnerable to extinction. Cherin et al. [5] appeared to be more willing to accept the “suid gap”, but were of the opinion that it was not Sus scrofa that appeared in Europe after the “suid gap”, but Sus strozzii, which dispersed from the Middle East, where it was present in Ubeidiya.

Other publications contradicted the existence of a “suid gap”. In 2017, I published a range chart for the species of Sus of the Pleistocene of Europe, showing several sites filling the “suid gap”, but did not discuss the “suid gap” hypothesis [23]. Kostopuolos and colleagues mentioned the existence of a pig from Krimni with an age right in the “suid gap” was published and suggested that, if the gap existed, it was more local [32].

In 2024, Iannucci [6] described a pig fossil from Peyrolles, a site dated right in the middle of the “suid gap”, and discussed records from: Mugello, Selvella, Ubeidiya, Pirro Nord, Ceyssaguet, Krimni-1 and Krimni-3. He also discussed other elements of the “suid gap” hypothesis, found that Sus is much rarer in the fossil record than would be expected on the basis of its large litter size and drew attention to the problems of dating localities from the time of the “suid gap”. Immediately, Martínez Navarro et al. [7] responded and discarded for one or another reason all the sites with suid fossils discussed by Iannucci [6]. They re-iterated that fossils of Sus are common in the fossil record and went so far as to state: “the remains of juvenile pigs, especially teeth, use to be relatively abundant in the fossil collections”. Further, they stated about their 2015 paper [2]: “Our study was also complemented with an extensive revision of the literature and chronological estimates available for all late Early Pleistocene sites of Western Eurasia”. Of all the papers previous to 2015 cited above, they only cited Guérin & Faure [22] and did not cite “equivalent” papers, which would make the citation of these papers redundant. In addition, they ignored all the sites with potential suid records discussed by Iannucci [6], who responded and reaffirmed his opinion [8].

4. Life History Traits in the Suoidea

The “suid gap” hypothesis holds that pigs are r-selected and this is probably the first time that a life history trait was inferred for a fossil species of pig [2]. To show that pigs are r-selected, a series of graphs were shown comparing body mass with the life history features: litter size, teat number, gestation length, neonate mass, interbirth interval, weaning age and sexual maturity age. All these data are on living species and were extracted from the PanTHERIA database.

It is relevant to note that litter size is not uniform across the Suidae, and not even across the genus Sus [33]. Babirousa tends to have litter size of 1–2, occasionally 3. Sus cebifrons tends to have 2–4, occasionally 5. Sus scrofa tends to have 4–9, occasionally 13. Sus strozzii was probably between those extremes. Any discussion on this subject should take this information into account.

Life history traits are much debated with respect to human evolution. While apes have infancy, juvenility, and adulthood, modern humans have two more, childhood and adolescence, inserted between the other three phases [34]. The sequence in which teeth appear holds information on life history. The delay in the change from deciduous to permanent premolars is related to the appearance of the childhood. In apes, lactation lasts till the juvenile phase, but in humans, childhood is the phase of dependence, but without lactation [35]. Lactation interferes with procreation and therefore the appearance of childhood has been interpreted to increase descendance [36]. These changes occurred in the context of a change in diet and the evolution of larger brains.

A delay in changing from deciduous to permanent teeth also happened in the Suoidea: in Tayassuidae and early Suidae, like the Listriodontinae, the deciduous molars are still in use after the appearance of the M3, while in later Suidae, like Sus and Phacochoerus, the premolars are in use before the M3 appear [9,37]. Many or all ruminants change their teeth like in primitive pigs, so it is plausible that this feature evolved within the Suoidea. The observation of this tooth eruption pattern in Artiodactyla is very coarse, while much more precision is obtained by the study of perikimata in human fossils. This change in the Suoidea happened in the context of the appearance of larger litter sizes, a change in digestive system from foregut fermenter to hindgut fermenter, affecting the protein requirements [38], and an increase in brain size, with Sus scrofa being the most evolved species [39,40].

There are parallels between human and suid evolution in the appearance of life history traits, a change in nutrition and increase in brain size. Whether the relationship between the evolution of these features is similar in humans and pigs, or different, remains to be seen. In any case [39,40], the study of the evolution of these traits in pigs might be helpful to understand of our own evolution.

5. r-Selection and Pig Abundance in the Fossil Record

According to the “suid gap” hypothesis, pigs are r-selected and therefore abundant in the present and in fossil faunas [2]. The graphs on reproductive traits [2] (Figures 3 and 4) do show that pigs are capable of reproducing faster than many other ungulates, but do not provide proof of high population densities and abundance of fossils. Species that are r-selected have also higher mortality rates. The balance of reproduction and mortality rates decides whether there is population growth, or not, and whether high population densities are reached and maintained. R-selection favours fast population growth, if the environment becomes more favourable or in the case of dispersal into new areas. It also might lead to fluctuations as in the Lotka-Volterra model.

Iannucci [6] gave various reasons why r-selection does not necessarily translate into abundance in the fossil record. In the reply to his paper, it was stated that: “…, the remains of juvenile pigs, especially teeth, use to be relatively abundant in the fossil collections.” [7] As a direct proof for the abundance of deciduous teeth in the fossil record, a paper was cited on just one deciduous suid molar from Hungary [41]. In addition, the indirect argument was used, that both pigs and rodents are r-selected and common and therefore used in biostratigraphy. Harris & White [42] were cited in support of this statement. However, Suidae are used in African biostratigraphy, because of the rapid evolution of their M3, not because they are r-selected and not because they are well represented by deciduous teeth. In fact, Harris and White [42] gave tables of many measurements of the M3 and none of deciduous teeth, because they did not use them in biostratigraphy.

Species that are r-selected have different mortality profiles from species that are K-selected: many juveniles die with their deciduous dentition in function and with their third molars not formed, partially formed or not erupted. It is possible to test whether teeth of juveniles are abundant in the fossil record.

Table 1. The frequency of the different tooth types in selected samples. Each individual has 12 D2-4, 12 P2-4, 8 M1-2 and 4 M3. The % expected is based on the number of teeth of a type in an indivi-dual and the best represented tooth type (without taking right and left into account).

			Number Specimens				% of Expected
Species	Locality	Source	D2-4	P2-4	M1-2	M3	D2-4	P2-4	M1-2	M3
Choermorus	Sansan	[43]	6	56	65	31	6	57	100	95
Propalaeochoerus sp.	Tomerdingen	[44]	36	75	130	40	18	38	100	62
Listriodontinae	>350 occurrences	[9]	258	1192	1033	733	12	54	70	100
Bunolistriodon meidamon	Pasalar	[9]	30	82	71	44	23	62	81	100
Listriodon splendens	Pasalar	[9]	72	186	126	87	28	71	72	100
Hyotherium meisneri	Cetina de Aragón	[45]	5	30	32	18	9	56	89	100
Hyotherium major	Ulm-Westtangente	[46]	21	57	72	30	19	53	100	100
Hyotherium soemmeringi	Styria	[47]	46	164	149	69	21	73	100	93
Conohyus simorrensis	Styria	[47]	5	73	55	33	5	74	83	100
All Suidae	Styria	[47]	15	128	134	66	4	32	100	99
Hyotherium soemmeringi	Sandelzhausen	[48]	46	164	149	69	21	73	100	93
Hyotherium soemmeringi	South German Molasse	[49]	1	16	31	15	2	34	100	97
Parachleuastochoerus steinheimensis	Deinotheriensande	[49]	12	45	38	41	10	37	46	100
Propotamochoerus palaeochoerus	Deinotheriensande	[49]	5	65	132	156	1	14	42	100
All Suidae	Deinotheriensande	[49]	17	110	170	197	3	19	43	100
Parachleuastochoerus	Rudabánya	[50]	1	16	17	9	4	59	94	100
Propotamochoerus palaeochoerus	Rudabánya	[50]	43	55	72	28	40	51	100	78
Propotamochoerus wui	Lufeng	[51]	14	69	87	43	11	53	100	99
Hippopotamodon	Dorn Dürkheim	[52]	29	121	70	88	11	46	40	100
Hippopotamodon	Spain	[53]	15	57	54	40	13	48	68	100
Sus strozzii	many	*	33	160	140	81	14	66	86	100
Sus sp.	Vallonnet	*	2	50	38	9	4	88	100	47
Sus sp.	Vallparadís	[5]	2	15	12	8	8	63	75	100
Sus scrofa	Mosbach	*	6	23	38	23	9	33	83	100
Sus scrofa	Taubach	*	3	23	52	35	3	22	74	100
Sus scrofa	Pinilla del Valle—Camino	*	3	32	23	15	7	71	77	100
mean (no doubles)							11	52	85	93

* Source: my notes.

, shows the number of deciduous premolars (12 in one individual), premolars (12), first and second molars (8) and third molars (4) of different species and samples, taken mostly from the literature. The data in this table are from hundreds of localities, dating from about 23 to less than 0.1 Ma and from an area extending from western Europe to China and southern Africa. In some cases, many fossils from many sites are taken together (e.g., Listriodontinae), which were collected in different ways. In other cases the data are from old collections (Styria, Deinotheriensande, Mosbach) and in still other cases from modern systematic excavations (Pasalar, Dorn Dürkheim, Pinilla del Valle, etc.). With minor differences all samples show the general pattern that third molars are better represented than deciduous molars; on average eight times better. Adult pigs are best represented in the fossil record not juveniles. This is not restricted to pigs, but is also the case in Cervidae, Bovidae and many other groups.

For the present discussion, it is not necessary to explain why M3 are overrepresented and deciduous teeth under-represented, but some possible explanations come to mind. Deciduous molars are small and have thin enamel. Part of the time that they are functional, they have large pulp cavities and thin dentine, their roots are not fully formed, and as a result they are fragile. At a later stage, their roots are resorbed, when the permanent premolars push up. By contrast the M3 have the thickest enamel of all teeth and with ageing, the pulp cavities retract and dentine increases in thickness. Another reason could be that deciduous teeth are shed isolated, while the death of the individual supposes already a concentration of bones and teeth. Individuals may die at particular moments or places and become part of bigger concentrations that later, when fossilized, are detected and collected.

6. Are Pigs Abundant in the Pleistocene European Fossil Record?

According to the “suid gap” hypothesis, pigs were abundant before the “suid gap” and “Later, the genus Sus is recorded everywhere in Europe as a ubiquitous member of the Epi-Villafranchian/Early Galerian and posterior faunas” [2,7].

Iannucci tested this statement [6]. He used a published dataset and compared sites from the periods before and after the “suid gap” (2–1.8 and 1.1–1 Ma). He found that Sus was present in 57.1% and 37.5% of the sites, respectively, while the more common species were present in 85.7–100% and 37.5–75% of the sites. This contradicts part of the “suid gap” hypothesis [2]. I used a different dataset [54] and looked at the period after 1 Ma. Sus scrofa (with the largest litter sizes among pigs) and Cervus elaphus appear in his zones MNQ20 to 26. A priori, there is no reason to suppose that Cervus elaphus is more abundant in sites with Rhinocerotidae. Figure 1 gives the percentage of sites per biozone that have Sus scrofa and Cervus elaphus. The number of sites per biozone vary between 4 and 257. About this more in the next section, but the higher the number of sites, the more reliable are the percentages. Especially the last three biozones have many sites and there Cervus elaphus is found in more than twice as many sites than Sus scrofa.

Another way to look at abundance is to compare the number of individuals (MNI) of Sus and the most abundant cervid in a series of localities. Alternatively, the element that is usually the most abundant, the M₃, could be used, and this is much easier.

Table 2. Comparison of the number of M₃ of Sus and the most abundant species of Cervidae in selected localities. Data from my notes. For the collection acronyms see the Abbreviations section.

	M3 Sus	M3 Cervidae	Cervidae/Sus	Species	Collection
Pinilla del Valle	8	15	1.88	Dama dama	UCM
Taubach	22	5	0.36	Capreolus	IQW
Ehringsdorf	2	34	17.00	Cervus elaphus	IQW
Bilzingsleben	3	126	42.00	Cervus elaphus	FBFSUJ
Petralona	4	56	14.00	Dama	AUT
Mosbach	18	>>31 *	>>1.722	Cervus elaphus	NMMi
Mauer	4	55	13.75	Cervus elaphus	SMNK
Voigtstedt	1	12	12.00	Cervus elaphus	IQW
Süssenborn	2	16	8.00	Cervus elaphus	IQW
Untermassfeld	1	48	48.00	Eucladoceros giulii	IQW
Vallonnet	9	69	7.67	Dama vallonnetensis	MPRM
Ubeidiya	2	33	16.50	Dama	HUJ
Upper Valdarno	25	46	1.84	Dama nestii	IGF
Olivola	4	17	4.25	Dama nestii	IGF
Tegelen	2	16	8.00	Dama rhenana	NBC, MNB, NMMa
mean			13.94

* I studied only a small part of the Cervus elaphus fossils from Mosbach.

shows these numbers for various localities, ranging in age from about 2 Ma till less than 100 ka, covering the “suid gap”. The data in Table 2 are from my notes and the main criterion for inclusion in Table 2 was that I studied both Sus and Cervidae from these sites, but sites without one of both were omitted as well as sites where one of the two was represented only by elements other than the M₃. In the cases that not all M₃ of a sample could be studied, it is more likely that all Sus M₃ were studied than that all cervid M₃ were studied. For instance, this is the case with the Upper Valdarno collections, of which I studied Sus strozzii in the AVPM, but no Cervidae at all. The ratio values in Table 2 are thus in some cases infra estimations. On average, the most common cervid in a site is represented by fourteen times more M₃ than Sus. Sus is not abundant and even if it is present in a site, it tends to be a rare species compared to cervids. Such comparisons could also be made with Equus or Bovini and the result would be similar.

The number of species in a sample (n) increases with sample size till approaching the number of species in the biocenosis (N), an asymptote [55]. This relationship depends on taphonomy and N. The larger samples tend to have more of the rare species. For instance, Sus is represented by only one tooth from Atapuerca unit TD6 (out of 104 ungulate fossils [56]), one tooth from TD8 (out of ≈600 fossils [23]) and one from TD10 (out of many thousands of fossils), while there is just one tarsal in TD4-5, and Sus is not represented in Atapuerca units TD7, TG10-11, TE10-14 and TE18-19. The case of unit TE7 appears to be different with the remains of several skeletons. In the case of small samples, the rare species appear only in few of the samples. Of about 70 sites in the terraces of the Manzanares and Jarama rivers, Sus is reported from only three of these sites [57]. In at least two of them this is by a single specimen.

All evidence points in the same direction: Sus was rare. Being a rare species, it may first turn up after many years of excavation. If, after many years of excavation and after thousands of excavated fossils, a pig molar was found in TD10, this could also happen in Venta Micena, Barranco León or Fuente Nueva 3. In fact, this happened in Dmanisi [6,7]. The assumption that pigs were abundant and recorded “everywhere” is essential to the “suid gap” hypothesis and its use in biostratigraphy [2,7]. Iannucci [6] tested the assumption that pigs were abundant and found it wrong. Here, it is tested in different ways and with different data sets and again in each case it is found to be wrong.

7. Pigs Became Rare in Europe Long Before the Pleistocene

Martínez Navarro and colleagues contrasted the European “suid gap” with Africa and Asia where pigs were well represented at that time [7]. Apparently, this comparison is made to support the argument that if pigs were present in Europe, they were abundant and if not found, they were absent. However, pigs in Europe were rare and were so for about eight million years.

Suoidea originated in the Eocene of Asia and appeared in Europe at the beginning of the Oligocene, where during the Middle and early Late Miocene their species richness peaked with up to seven or eight coeval species and up to five species in a single locality (Figure 2) [29,30,58]. At about 9.7 Ma it dropped to one or two coeval species and rarely more than one per site in continental Europe. Also, the proportion of pigs of the whole faunas of the localities may have decreased and the proportion of localities without pigs may have increased. This decrease in suoid species richness forms part of an event called the Vallesian Crisis [59]. For the Suoidea, this event is also noted in Anatolia [38]. Around the same time, Hominoidea and other primates went extinct in western and central Europe, there was also a major turn over in other large and small mammals and a change in the vegetation, probably related to an increase in seasonality [60,61].

At the time of the drop in species richness in Europe, no similar drop occurred at other latitudes [62,63,64]. Figure 2 shows that no such decline occurred in Africa and China, but it suggests that in the Indian Subcontinent may have been a decline, though not so abrupt as in Europe. In addition, some suoid subfamilies that went extinct in Europe continued to live at lower latitudes [48,63,65]). This is a common pattern that occurs in many groups of mammals and was known already in 1779 [66] and continued to attract attention [67]. It has recently been documented for hipparions [68]. The most likely explanation for this pattern in the pigs is, that suid species richness declined in Europe and other areas at the same latitudes when the climate became more seasonal. When the climate became cooler and the glacial cycles more pronounced, suid diversity dropped further. This happened at different times at different latitudes. This is a more likely explanation for pigs being rare (or even absent on higher latitudes) than competition with humans.

The observation that suoid species richness in the Early Pleistocene of Europe is much lower than in Asia or Africa is not new. This was documented over 30 years ago. Many species of Suidae were described from the Early Pleistocene of Liucheng [69]. My own interpretation of the material is that there were probably three species of Sus and one of Hippopotamodon. Other sites have three or all four of these species (e.g., Longupo [70]). Several occur together at least as late as in the latest Early Pleistocene of Yunxian [71]. During the Middle Pleistocene, they were replaced by Sus scrofa, or at least so in the north of China. This means that, before and after the “suid gap”, species richness in China was higher than in Europe. Rather than supporting a European “suid gap”, this diversity of Chinese species of Sus, coexisting with early humans, suggests that humans did not displace species of Sus. Like in Europe, the subsequent decline in suid diversity in China is probably due to climate change and not to human competition.

Figure 2. The number of contemporaneous species of Suoidea (pigs and peccaries) in the mainland of western Eurasia [58], Africa [64], China (data mostly from Liu [72]), and the Indian Subcontinent (various sources, including my notes). The latter two are based on less elaborate range charts and are less precise, but are included for the overall picture. Light green colour in the figure for Europe indicates possible presence of species in this time.

Figure 2. The number of contemporaneous species of Suoidea (pigs and peccaries) in the mainland of western Eurasia [58], Africa [64], China (data mostly from Liu [72]), and the Indian Subcontinent (various sources, including my notes). The latter two are based on less elaborate range charts and are less precise, but are included for the overall picture. Light green colour in the figure for Europe indicates possible presence of species in this time.

8. The Time of the “Suid Gap” Is Under-Represented in the Fossil Record

In the previous section, it was mentioned that the fossil record is not evenly distributed. This occurs at different scales and in different times. There are some gaps or periods with a poor record that have attracted much attention. A much-debated case is the African record of fossil mammals, which is poor for the late Middle and Late Miocene, including for Hominoidea [73]. By coincidence, this is the time that the Eurasian record of fossil Hominoidea is rich. David Begun [74] proposed the hypothesis that Hominoidea spread out of Africa to Europe, evolved there, spread back into Africa and gave there rise to the gorillas, chimpanzees, and us. Begun’s cladogram of the Hominoidea has the middle section with all Eurasian species of this age. He has been criticised because the gap in the African fossil record may be the reason for the gap in the African record of Hominoidea [74]: if African Hominoidea from this time become known, the cladogram may change and the Eurasian Hominoidea may turn out to be a side branch.

Turning back to the “suid gap” hypothesis, the Pleistocene of Europe has notable gaps. The north of Europe virtually does not have a Pleistocene fossil record, presumably because of the glaciations. At mid-latitudes, there are many Middle and Late Pleistocene sites in the UK, Germany and Poland, but very few Early Pleistocene ones [75,76,77].

Though there appears to be a continuous record in Italy [78], the part between about 1.8 and 1.2 Ma is in reality much poorer. The range chart of this paper shows the faunal units (FA) and the main Italian sites. The “suid gap” is covered by the Farneta FU and the Pirro FU. The site of Farneta, which gives its name to the faunal unit, has just 8 species [79]. Another locality included of this faunal unit is Mugello. In fact this is a basin with ten known sites, most of them with short faunal lists, either of different ages or of an age transitional between the Tasso and Farneta units [80]. The two other sites of this unit are Pietrafitta and Selvella. The former is rich and the latter has 8 taxa, but only 3 of them were identified without reservation (aff., “ex. gr.”) at the species level [81]. The other faunal unit is defined on the fossil locality Pirro Nord. This is the only one of these sites with independent age control [82], but that contradicts its position in the “suid gap”. The only other site of this faunal unit in the charts for Italy [78] is Capena. This site has just two species [79]. There is no country in Europe with a much better and well dated fossil record of the time of the “suid gap” than Italy, so the fossil record of this time is not particularly well known.

The sites studied by Guérin [54] were used in the previous section to test the abundance of Sus. They show very well the fluctuations in the numbers of sites in time or per MNQ unit. There is no reason to believe that the sites were selected according to age, resulting in sites dating to the “suid gap” being under represented. Figure 3 shows the numbers of sites studied per MNQ unit, as well as the number of sites of each MNQ unit per 100 ky duration of this unit. Two scales are used, a plain one, and a logarithmic one, because of the great differences in numbers of sites. It is obvious that the time corresponding to the “suid gap” has the poorest fossil record. In the context of the discussion on human dispersal into Europe, attention was drawn to a period with a poor fossil record in Europe [83], which happens to be nearly identical to the “suid gap”.

If we combine the information that pigs are rare and that sites dating to the “suid gap” are rare, it is obvious, that sites with suids dating to the “suid gap” are very rare. This does not mean that pigs did not live in Europe, it only means that it is difficult to find their fossils. It is not enough to mention a few sites without pigs.

While many palaeontologists may be familiar with the uneven distribution of the fossil record, not everyone is or needs to be. The paper on the “suid gap” is directed to a wider public that is not familiar with the fluctuations in the fossil vertebrate record. For instance, biologists and archaeologists need not know about fluctuations in the density of the fossil record anterior to the times they study. It would have been useful to provide this information to them.

9. Is the Link with Human Ecology Real?

The “suid gap” hypothesis explains this gap as the result of pigs being ecologically displaced by humans, who are also bunodont and have an opportunistic trophic behaviour [2]. Later, it was indicated that humans and pigs may have competed during winter for nuts (e.g., acorns, nuts, chestnuts, hazelnuts, almonds and pistachios) and belowground resources, with the result that: “Europe was empty of pigs during more than 600 Ka, until the arrival of Sus scrofa around 1.2 Ma” [2]. This means that, all over Europe, Sus strozzii would be driven to extinction because humans did not leave them enough nuts or underground resources in winter.

We do not know the diet of Sus strozzii, but given the documented diets of other species of the genus [33], it would have been similar to that of Sus scrofa. Studies of the stomach contents of Sus scrofa shows that a great variety of plants and animals are eaten, as well as algae, fungi and even soil [84,85]; 153 genera of plants and 117 genera of animals were reported from its stomach contents [86]. Mid-latitude Europe has soils with a thick upper horizon rich in organic material and invertebrates, including earth worms. In favorable conditions, as many as 1200 individuals or 120 g of the most abundant species may be found in one square metre [87]. That is over 100 metrical tonnes per square kilometre. It is difficult to imagine that humans exploited all over Europe all these resources to such an extent, that Sus strozzii was left with nothing. For this, a dense human population would be required, also all over Europe. Let us do a mental exercise to see whether this is a likely scenario.

Melis et al. [88] studied population densities of Sus scrofa in western Eurasia and, found that densities varied from 10 to 0.01 individuals/km², depending mostly on January temperature and minimally on the presence or absence of wolves. The colder locations in this study were situated close to 60° N and in W Siberia, while those in France and Spain were among the warmest. The geographic distribution of Sus strozzii is far less known and its population density is not known. If it were similar to that of Sus scrofa there would have been up to 10 individuals/km² in Spain. Of course, this is in environments with optimal conditions. Differences with the Pleistocene are that there are fewer other ungulate species, but more humans. Spain, with a surface of about 500,000 km², would have much less than 5 million pigs, but in any case, they may have been many. A dense human population would be needed not to leave sufficient nuts or earthworms for pigs to survive. Even if only one human would displace one pig, this would mean a numerous Early Pleistocene human population for Spain. For the whole of Europe this would be much more. Of course, this is a very coarse comparison, but it is only meant to provoke thinking about the implications of the “suid gap” hypothesis.

It is estimated that early Homo sapiens lived at 10 times higher population densities than the neanderthals [89]. West European population sizes between 650 and 360 ka were estimated as 13,000 to 25,000 [90]. Human population densities for the Early Pleistocene of Europe are not known, but humans were present “in many areas but at low densities” [91]. The more primitive humans with more primitive technology at 1.8 or 1.2 Ma must have had still smaller populations and lower population densities. Pigs reproduce fast [2,7], which implies that their populations recover fast and that they could re-colonize fast from areas with less dense human populations. Sus scrofa lives today as far north as the south of Finland and so may have done Sus strozzii in the Early Pleistocene. Driving Sus strozzii to extinction by food competition would require unrealistic human population sizes and densities. Based on the literature, the west European suid population size could have been millions and the human population thousands.

If we take another approach and use the number of sites with pig and human fossils as a proxy for their respective population sizes, we can make a direct comparison. Published and universally accepted human fossils are from Barranco León, Atapuerca TE7, TE9, and TD6 and Mauer. Pig fossils of the same time range are from: Ceysaguet, Atapuerca TE7 and TE9, Vallonnet, Untermassfeld, Vallparadís, Atapuerca TD6 and TD8, Pakefield, West Runton, Voigtstedt, Süssenborn, Isernia, Mauer and still other sites from the same period [5]. Based on this comparison, it is very likely that human population density was much lower than that of the pigs and that humans were numerically unable to outcompete pigs. It is more likely that both were relatively rare compared to other mammals and that their rareness limited competition between the two.

If primitive humans using Oldowan technology displaced Sus strozzii from all of Europe, why was it able to coexist in Ubeidiya with humans with the more developed Acheulian technology? If the Acheulian has advantages over the Oldowan, this is probably reflected in greater human population densities and a more intense use of the resources in Ubeidiya than in Europe at 1.8 Ma. That Ubeidiya is outside Europe is no explanation; the explanation has to be on the basis of ecology. And why were pigs able to colonize Europe repeatedly after 1.2 Ma and, in words of Martínez Navarro and colleagues [2], be “everywhere”?

It is not only that after 1.2 Ma, pigs colonized Europe again one time. First Sus strozzii (if it were absent during the “suid gap”), next Sus sp. (in my opinion different from Sus strozzii) and finally Sus scrofa. These species pig did not live together with only one species of Homo with one type of technology. They lived together with progressively more evolved species: Homo aff. erectus (Atapuerca TE7 and TE9 [92]), Homo antecessor (Atapuerca TD6 [56,93]) H. heidelbergensis (Mauer, [76]), H. neanderthalensis and finally with H. sapiens. They lived together with humans with each time more advanced technology: Oldowan, Acheulian, Mousterian, Aurignacian, Solutrean, Magdalenian, … Each new technology meant a more efficient use of the environment and probably higher human population densities. Pigs survived hunting with spears, bows and arrows and with guns. The Late Pleistocene and Holocene extinctions have been much debated [94]. After the initial overkill hypothesis, a much more nuanced picture has appeared, in which other factors than hunting play a role. Many large mammal species went extinct in Europe, but Sus scrofa survived. So why would Sus strozzii be driven to extinction by competition by primitive humans with primitive technology and very likely living in lower population densities? The most likely answer is that pigs did not go extinct between 1.8 and 1.2 Ma, but that we have difficulties to find their fossils and date them well.

10. If the “Suid Gap” Existed, When Would It Have Started?

As Iannucci [6] already noted, dating sites from the “suid gap” is difficult. To date the onset of the “suid gap”, it is not only necessary to date the last site with a pig, but also the first sites without pigs.

Sus strozzii is present in Il Tasso and this site is situated at or just above the top of the Olduvai (1.775 Ma) [95]. Its extinction is thus after 1.775 Ma.

The oldest sites, discussed as a proof of the absence of pigs, that have been dated are Barranco León and Venta Micena, dated to 1.43 ± 0.38 [96] and to 1.37 ± 0.24 Ma [97], respectively. It should be noted, that some serious problems with the ages of these sites have been pointed out [98,99]. The large error margins could place the sites after the “suid gap”. In the case of Venta Micena, there are two previous amino acid racemisation dates, of 983 ± 58 ka and 1095 ± 55 ka [100,101]. The latter AAR age is compatible with the age of 1.37 ± 0.24 Ma, resulting in an age of about 1.14 Ma. This places Venta Micena after the “suid gap”. Accepting the age of Barranco León, the beginning of the “suid gap” may be as late as 1.43 Ma.

Whether there would be a “suid gap” starting at 1.8 or 1.4 Ma would be of little importance, if it were not, that a 1.8 Ma date would support human presence all over Europe and a much older age for Barranco León. The 1.43 ± 0.38 Ma date for Barranco León is also the oldest date for a human fossil from Europe and the absence of pigs does not make it older.

11. What Is the Importance of the “Suid Gap”?

As we will see, the importance has much to do with ages of events and sites. When criticising Iannucci [6] for not citing the 1.2 Ma age for Peyrolles [102], Martinez Navarro and colleagues [7] apply a double standard, since they published many papers on Venta Micena and systematically omitted citing the amino acid racemisation ages of 983 ± 58 ka and 1095 ± 55 ka, obtained for that site [100,101]). They stated that Venta Micena has an age of around 1.6–1.5 Ma [2]. However, Martínez Navarro and Palmqvist are co-authors of Duval et al. [97], who dated the site to 1.37 ± 0.24 Ma and 1.6–1.5 Ma is in the lower end of the one sigma error margin. One might expect that they use the numerical date they published, or else explain why they assume that date to be too young. Venta Micena, near Orce, was mentioned as particularly relevant since no pigs were found in an excavated surface of 370 m² [2].

Inflating the ages of the sites in the Orce region has a long story, of which only glimpse can be shown here. Early human presence has been claimed in various sites there. When publishing the magnetostratigraphy of Fuente Nueva 3 and Barranco León, the geochronologists prudently concluded that those sites are situated between the Olduvai and Jaramillo subchrons [103]. However, the text on biochronology of the large mammals (presumably by Martínez Navarro) stated that Fuente Nueva 3 has the same fauna as Venta Micena [104,105], which in turn is similar to that of Dmanisi [106]. The cited list of Dmanisi [106] includes 20 species of large mammals, identified till at least the genus level, as well as small mammals and other vertebrates. The cited list of Venta Micena [105] includes 18 species of large mammals identified till at least the genus level. The species that coincide (although some placed in different genera) are: Canis etruscus (about 2.1–1.2? Ma), Ursus etruscus (about 2.5–1.1 Ma), Mammuthus meridionalis (about 2.5–0.7 Ma), Equus altidens/E. aff. altidens (about 1.2?–0.5 Ma), and Stephanorhinus etruscus (about 3.5–0.7 Ma). If only five out of 18 or 20 species coincide, the faunas are not very similar. Moreover, the species that coincide tend to have long temporal distributions and cannot be used to correlate Venta Micena to Dmanisi. At present, several of the identifications have changed, but this does not affect the argument used in 2000. Twenty-four years later, and despite numerical dating, Venta Micena is still said to have an age (1.6–1.5 Ma) close to Dmanisi, but this is not supported by biochronology, nor by dating.

The “suid gap” hypothesis states that humans drove pigs to extinction all over Europe by 1.8 Ma. If this hypothesis is accepted just as it was presented [2], this implies that humans occupied all of Europe by 1.8 Ma. Making a short-cut, this means that the lack of suid localities dates human dispersal. The “suid gap” hypothesis seems to be meant as an argument where numerical dating of sites with human presence falls short.

12. Are There Pigs from the “Suid Gap”?—Whom to Believe?

Proposing the hypothesis of a “suid gap” implies having done an effort to ascertain that the gap really exists. This supposes knowing or searching the literature and preferentially also having experience with the study of fossil Suidae. However, several potential suid records were not mentioned [2]. After being confronted by Iannucci’s [6] paper, attention was paid to the literature that cited pigs dating to the “suid gap”. All sites mentioned by Iannucci were rebuffed for one or another reason [7]. The situation would have been different if they would have discussed these records in 2015. Now their arguments are less convincing.

It is peculiar that Martínez Navarro and colleagues [2] cited Pirro Nord as an example of a site dating between 1.8–1.2 Ma without pigs, while various authors cited a pig from there. Freudenthal [107] gave a faunal list, including Sus, from “cave fillings”, but without specifying which one. De Giuli et al. [108] stated that they previously [109] studied micromammals from a fissure filling called Pirro Nord 1 and that this was probably the same fissure as from which Freudenthal collected. They studied more material from different fissures in the University of Bari, collected some fossils by themselves, and identified Sus from a fissure called Pirro Nord 5. This suggests, that De Giuli and colleagues did not know the material collected by Freudenthal, but that they did see one or more fossils belonging to Sus. So, several authors independently identified Sus from possibly two different fissures: Pirro Nord 1 and Pirro Nord 5. Martinez Navarro et al. [7] stated that they revised the old collection from Pirro Nord in 2000 and did not find the pig, but why did they not mention this in 2015 [2]? In any case, both Freudenthal [107] and De Giuli, Masini and Torre [108] did see a fossil, which they identified as a pig. Whom should we believe: the persons who collected and studied the fossils and who stated that there is a pig, or Martínez Navarro et al. [7] who did not find it? Iannucci [8] reported that recently boxes with Freudenthal’s fossils turned up in the University of Bari. The material is under study and we may have an answer soon.

Recently, one of the fissure fillings of Pirro Nord (P13) has been dated to around 0.8 Ma [82]. It is not the intention to discuss this surprisingly young age for P13. There are many fissure fillings at Pirro Nord and the ages need not be identical. The pig fossils are said to come from P1 and P5. However, the faunas described from the fissure fillings [108] seem homogenous and fit well an age of over 1.2 Ma. In fact, for nearly 40 years no one doubted such an age on the basis of biochronology [78,82,110].

De Giuli [81] figured a bone from Selvella as a suid humerus. It is a diaphysis and the figures do not permit a sure identification. Martínez Navarro et al. [7] stated that it is the radius of a different taxon, but did not provide figures that allow to verify this new identification. In view of the conflicting opinions, the diagnostic features of the specimen should have been clearly described and figured, so that this is not a matter of credence.

Faure & Guérin [17] discussed the sites from which they identified Sus strozzii, and Mugello is one of them. They assigned the site to MNQ 17, but later it was placed the site in the Farneta Unit around 1.5–1.4 Ma [78]. The suid teeth from there have been identified as two M² and one M₃ [7] or as M¹, M² and M₃ [8]. When I studied these specimens in 1988, I identified them M¹, M² and M₃. There are not so many papers on Sus strozzii and it might have been expected that the paper by Faure & Guérin [17] would be cited and that the pig from Mugello and its age would have been discussed. However, this paper was not cited at all [2].

The suid record from Ubeidiya was dismissed, because it is not in Europe. It is true, that the site is not in Europe, but, like Dmanisi, it is in western Eurasia and both have faunas of European affinities [83] (Figure 2). It was not explained why humans drove Sus strozzii to extinction all over Europe, while the two coexisted in Ubeidiya. Saying that Ubeidiya is not in Europe [7], is not an explanation.

According to Martínez Navarro and colleagues [7], the suid fossil from Peyrolles was “without any reference to the stratigraphic provenance of the fossil specimens apart from the name of the town, Peyrolles”, but Iannucci [8], saw now reason to doubt its provenance. In this discussion, I observe an inconsistency in Martínez Navarro’s attitude towards the provenance of fossils from old collections. He is co-author of Lacombat et al. [111], who published fossils of Canis sp., cf. Eucladoceros sp. and Equus supposedly from Vialette, which would then be much older than the normally accepted first appearance of these taxa in Europe. These fossils were found during a large-scale revision of the fossils of the collections of the Musée Crozatier in Le Puy-en-Velay, which resulted in a book [112]. In the introduction, Guérin [113] described the penurious situation of the collections during a long time and particularly mentioned problems with labels and the provenance of specimens. In other chapters of this book, fossils were described that were found with the collections of Vialette and were attributed to Listriodon splendens and cf. “Microstonyx” major [114,115]. These species have their last appearances at about 9.7 and 6 Ma. If these species really formed part of the Vialette fauna, they would have survived up to 6 My longer than normally accepted. However, it is more likely that these fossils either were resedimented or mislaid and the latter may also have been the case with the fossils of Canis sp., cf. Eucladoceros sp. and Equus. There are many fossils from Vialette in other museums [54,116] and Canis sp., cf. Eucladoceros sp. and Equus have not been reported from these collections [111]. The presence of these taxa in Vialette is suspect, because in all three cases they are out of the known range for these taxa. The pig fossil from Peyrolles is not suspect, since its age is within the known temporal distribution of Sus strozzii [5,23]; it is only suspect to Martínez Navarro et al. [7], because it does not fit their “suid gap” hypothesis. Two sets of rules are applied here.

Iannucci was criticised [7] for not mentioning the ages of 1.20 and 1.54 Ma for Peyrolles obtained long ago [102,117]. The date of 1.20 Ma would remove Peyrolles from the centre of the “suid gap” to its upper limit, saving thus the hypothesis, in case the provenance of the metapodial would be accepted to be from the classic fossil site of Peyrolles.

Iannucci [6] mentioned pig fossils reported by Kostopoulos and colleagues [32] from Krimni-1 and Krimni-3 with estimated ages of about 1.5 Ma. The latter authors proposed that the “suid gap” was shorter or absent in SE Europe. Martínez Navarro et al. [7] did not mention the pigs from Krimni, but maintained that a “suid gap” exists.

For the “suid gap” to exist the following conditions have to be met: (1) Freudenthal [107] and De Giuli, Masini and Torre [108] were mistaken in their identifications of Sus from different fissures at Pirro Nord; (2) De Giuli [81] was mistaken in the identification of Sus from Selvella; (3) Kostopoulos and colleagues [32] were mistaken in the identification of Sus from Krimni 1 and Krimni 3; (4) the suid metapodial comes from a different place as the other fossils from Peyrolles and has a different age; (5) the site of Pulicciano in the Mugello basin is older than the “suid gap”; (6) for some reason, Sus strozzii could live together with humans in Ubeidiya, while they could not in Europe. These are many conditions.

13. Sus strozzii from East Runton

There are three fossils of Sus from East Runton in the Natural History Museum. They were acquired by the museum after 1885, when Lydekker’s [118] catalogue was published and are not included in it. Stuart [75] stated that Sus strozzii has not been found in Britain, except in the Red Crag Nodule Bed. The specimens from East Runton are described here.

Specimen NHM M6623 is a left M³ (Figure 4b). It was collected by Savin (“Savin coll. 1332”) and has the indication on its provenance: “Sandy clay in lower Forest Bed 40 yds from cliff of Gangway East Runton Norfolk”. There is another label stating: “M6623/Sus strozzii/Det. A. Azzaroli 30.3.1987”. The third lobe of the specimen is very simple with one main cusp (pentacone), placed clearly on the lingual side. Lingually and labially anteriorly directed ridges descent from this cusp. The crests of these ridges have little cusplets. This is a primitive structure, found in older species. In Sus scrofa, the third lobe tends to be more elongate, there tends to be a cusp intermediate between the pentacone and the second lobe and there tend to be many smaller cusps that are separate from other cusps nearly till the base of the crown. The measurements are as follows: length (DAP) = 43.2, width of the first lobe (DTa) = 26.3, width of the second lobe (DTp) = 23.3, width of the third lobe (DTpp) = 13.9, height at the paracone (Ha) = 18.1, and height at the metacone (Hp) = 17.8 mm. Metrically, the tooth is situated well within the Sus strozzii cluster and is larger than all other Sus known from Europe. It is also larger than the M³ of any other Suoidea from Europe, save for Hippopotamodon, but it differs in having a higher crown height.

Specimen NHM M6627 is a left third metacarpal of a juvenile, lacking the distal epiphysis (Figure 4a). It is accompanied by the following information: “Sandy Gravel Forest Bed 30 yds from cliff N of gangway East Runton”, “Purchd 1897” and “Savin coll. 534”. The anteroposterior diameter of its proximal end (DAPp) is 18.9 mm and the width (DTd) is 22.1 mm.

Specimen NHM M3799 is a left M_1/2 (Figure 4c). Because of this size it is probably an M₂. It is accompanied by the following information “Forest Bed, East Runton, Norfolk” and “Pred E.R. Dodds, Esq., 1888”. The length (DAP) = 28.4, the width of the anterior lobe (DTa) = 17.3, of the posterior lobe (DTp) = 19.5, enamel thickness at the metaconid (Ta) = 1.6, and enamel thickness at the entoconid (Tp) = 1.5 mm. Metrically, this tooth clusters well with Sus strozzii and is larger than the M₂ of other European species of Sus. Again, the size is similar to the Hippopotamodon M₂.

Lister [119] studied the antlers of deer from the Forest Bed. Savin, who collected many of those specimens, had recorded the distance from the cliff. Lister discussed this and found that in each locality there is indeed a general relationship, the further from the cliff, the older, but he also noted that the signal is “blurred”. The specimens of Sus strozzii were found far from the cliff, particularly the M³. The metacarpal was found at a distance, where eight antlers were collected, most of them being Eucladoceros tetraceros, but some specimens, like one Cervus elaphus, were of Middle Pleistocene origin (or a least much younger than E. tetraceros). Lister [119] discussed East Runton in particular and offered different explanations of the apparently mixed nature of the East Runton assemblage and concluded that the most likely one is that the Pastonian fossiliferous deposits span a longer time than normally recognized. The main two sites with Eucladoceros tetraceros are Peyrolles and East Runton, suggesting a broad similarity in age and this may also apply to the pig from both sites.

Under normal circumstances, supposing that the Sus strozzii from East Runton is close in age to Peyrolles, would not raise eyebrows. However, given the present discussion, the situation is different.

14. The Sites That Should Prove the Absence of Pigs

Martínez Navarro and colleagues [2] indicated in their Figure 2, the approximate age of the localities they took into consideration. The sites situated in the “suid gap” are (from old to young): Pietrafitta, Av. Marcel, Farneta, Libakos, Venta Micena, Apollonia-1, Trlica, Pirro Nord, Barranco León, Capena, Incarcal I, and Sainzelles. However, they did not discuss all these sites in their text and centred mostly on Venta Micena and Barranco León.

Some of the sites are poor and not very indicative: Capena has just two species and the Farneta fauna does not have many species either [79]. I was not able to find any information on the site “Av. Marcel”.

Several sites have numerical dates that do not place them with certainty in the “suid gap”. Though based on biostratigraphy, Pirro Nord is older than sites such as Atapuerca TE7 and TE9, dated to 1.13 ± 0.18 and 1.22 ± 0.16 Ma respectively [92], the recent dates obtained for Pirro Nord suggest the much younger age of 0.8 Ma [82], well outside the “suid gap”. The age of Venta Micena has been discussed above (Section 10) and considering the dates of 1.095 ± 0.055 Ma [101] and 1.37 ± 0.24 Ma [97], the site might be younger than the “suid gap”. Barranco León (1.43 ± 0.38) and the lithostratigraphically higher Fuente Nueva 3 (1.19 ± 0.21 Ma) have ages with error margins that allow to place them in, but also after, he “suid gap”. Whereas “absolute” dating has long error margins, on the basis of biostratigraphy, Venta Micena and Barranco León are older than sites that were cited as being after the “suid gap”, such as Atapuerca TE7 and TE9.

Based on biochronology, other sites could be younger than the “suid gap”. Incarcal is now considered to date to 0.9–0.8 Ma [120]. There are different interpretations of the age of Trlica [121], and it is well possible that the site is after the “suid gap”. Apollonia-1 was originally assigned an age in the range 0.9–0.6 Ma [122] and, though some assign it an older age [2], an age around 0.9 Ma fits well [23].

The rich sites without pigs and with likely ages in the range 1.8–1.2 Ma are Venta Micena, Barranco León, Pietrafitta, Pirro Nord, Libakos and perhaps Sainzelles. The northernmost site is Sainzelles, all other sites are in the south of Europe. The environments in mid-latitude Europe are, and probably were, more humid and more forested, in any case different from those in the South. Arguing on the basis of these sites in the south that there were no pigs in all of Europe, is a bridge too far.

15. Biostratigraphic Marker for the Epi-Villafranchian?

Martínez Navarro and colleagues [2] were of the opinion that the extinction of Sus strozzii at 1.8 Ma, the absence of pigs (1.8–1.2 Ma), and the re-appearance of pigs around 1.2 Ma were important biochronological markers. For a species to be an important biochronological marker, the identification should be unequivocal and the record should be abundant enough to reliably date its appearance or disappearance.

As has been demonstrated above, pigs were not abundant. The existence of the “suid gap” is still disputed. Even accepting the absence of pigs in the dated sites from Orce as representative for Europe, the onset of the “suid gap” (if it existed) is not precisely dated but, occurred in a period lasting 0.4 My. In addition, there are different opinions on which species is present at about 1.2 Ma. These facts argue against using the supposed disappearance at 1.8 Ma and re-appearance at 1.2 of Sus as biochronological markers.

16. Discussion

For many years there was little interest in the exact date of extinction of Sus strozzii. This changed when the “suid gap” hypothesis was proposed [2]. The merit of this proposal was to stimulate a discussion on the precise temporal distribution of the species of Sus of the European Pleistocene. Various papers have modified or added to this hypothesis [3,4,5].

Most of the papers, that gave range charts of the pigs from Europe, show a gap in the Early Pleistocene [10,15,18,28,29,30,31], but did not explicitly state that pigs were absent, the gaps rather represented lack of data. Other papers suggested an overlap of Sus strozzii and Sus scrofa, or at least a replacement within a particular biozone [17,19]. When the “suid gap” hypothesis was first proposed, the “gap” was not original, but earlier literature on the subject was not cited and a series of sites that date, or could date, to the time of the “suid gap” and were reported to have pig fossils, were not mentioned [2].

The “suid gap” hypothesis is a chain of arguments:

(1)

pigs are r-selected and therefore

(2a)

they are best represented by fossils of deciduous dentition,

(2b)

their fossils are abundant,

(3)

therefore, if we do not find their fossils, they were absent,

(4)

therefore, they were absent between 1.8 and 1.2 Ma,

(5)

this was because they were displaced by humans,

(6)

the absence of pigs dates the period between 1.8 and 1.2 Ma,

(7)

since pigs were absent from 1.8 Ma onward, humans were present.

Several of these arguments can be tested. Various have been tested already [6] others are tested here, in particular: because pigs are r-selected, they are best represented in the fossil record by deciduous teeth (argument 2a); and pigs are abundant in the fossil record of the Pleistocene of Europe (2b). These arguments were found to be false. In addition, it is unlikely that early humans could outcompete pigs for nuts and underground resources in all of Europe (5, 7). The fossil record is very uneven and the “suid gap” corresponds to a time when the European fossil record is not well known. This combined with the rareness of pigs, could lead to an apparent, but not real, absence (4). Accepting a “suid gap”, there is no evidence that it started earlier than about 1.4 Ma, not at 1.8 Ma (4, 6) and it does not date the presence of humans (7).

Essential to the “suid gap” hypothesis is the absence of pigs in Europe between 1.8 and 1.2 Ma. Various papers had published suid records from sites dating to the “suid gap” before this hypothetical gap was published and others did so afterwards [6], but Martínez Navarro et al. [7] challenged these records, because they could not find the fossils and by raising doubts on the identification, provenance and age of other fossils and also because they defended their hypothesis. The question of whether there were pigs during the “suid gap” or not is reduced to confidence in the identifications by Freudenthal, De Giuli, Kostopoulos and colleagues [32,81,107,108], or in Martínez Navarro et al. [7], who question those records because they do not fit their “suid gap” hypothesis.

The construction of the “suid gap” hypothesis raises questions. It seems that the essence of the hypothesis is that pigs were absent (3) because of competition with humans (5). In order to show that pigs were absent, it is argued that the apparent absence from 1.8–1.2 Ma is real. Proving absence of a species is difficult. Therefore, the approach that was taken is stating that pigs were abundant and, if we do not find them, they were not there. This supposed abundance is based on the assumption that pigs were abundant because they were r-selected. However, the link between r-selection and abundance has not been proven and it has been questioned [6]. This renders the argument of r-selection irrelevant. All the empirical data used in support of the “suid gap” hypothesis come from the PanTHERIA data base and are used to support that pigs are r-selected. They were used to construct seven graphs. In the whole publication there are no other empirical data [2]. Empirical data that could support pig abundance could come from numbers of pig fossils or pig localities compared to similar data of other species.

The other essential part of the hypothesis is that humans supposedly outcompeted the pigs, leaving them without nuts and underground resources. While it is possible, that the diets overlapped, it is necessary to show that human use of the resources was so intense that a pig population could not be sustained. This calls for some kind of quantification of the use of resources. In view of the wide range of food items taken by pigs [86], this is a hard job. No intent was done.

In this chain of arguments, argument 3 (pigs were absent) is used to arrive to argument 5 (they were displaced by humans) and then to 7 (because pigs were absent, humans were present). This looks circular.

The merits of the “suid gap” hypothesis are: having started the discussion on the temporal distribution of suid species in Europe and having mentioned a life history trait in relation to fossil pigs. It should certainly not be used to infer the ages of localities or of human dispersal into Europe. The title of this paper is a question and the answer is: no.

17. Future Directions

It is considered here that the evidence of Suidae living in Europe during the “suid gap” (1.8–1.4 Ma) outweighs the doubts that were raised. However, only the detailed publication of pig fossils from this period and their dating, will definitively settle the question.

K- and r-selection are life history traits. Most life history traits are difficult to study in the fossil record. Considering pigs, or even only the genus Sus, as a homogenous r-selected group is too simple. The living species of Suoidea show that within this group, there is an evolution in life history, nutrition and brain size. It is possible to map this evolution on a phylogenetic tree, fit information from fossils in, and even, to some extent, date the appearance of traits. Life history traits are being studied in fossil Hominidae and some of these traits have even been documented in fossil Suoidea. The study of the evolution of these features in Suoidea might serve to test hypotheses on human evolution.

The under representation of deciduous teeth and over representation of third molars is constant in the suoid fossil record. This is independent of age (Oligocene to Holocene) and human presence or not. This is not only the case for Suoidea, but seems to be a general pattern for other ungulates and possibly even for many or most mammals. The underlying reason for this is taphonomical. Documenting this pattern in more detail and identifying its causes will increase our understanding of the fossil record and may be particularly relevant for zooarchaeology.