Three study areas have been selected to undertake this study in northern Algeria (Fig. 1), where two are part of the Mitidja Plain, considered the largest sublittoral plain in Algeria according to Mutin (1977). The choice of study areas is primarily based on the presence of a sufficient number of accessible droppings for collection. Subsequently, the selection is made according to the degree of urbanisation and the intensity of anthropogenic activity. The first one (Bab Ezouar) is close to housing and is subject to a strong anthropogenic influence, the second (Zeralda) is located in a suburban area, while the third site (El Hamdania) is a forest type (natural environment).

The faeces identification is relatively simple and the risks of confusion with other mammals’ faeces are very low. Faecal samples were collected from March 2017 to October 2018, early in the morning then transported to the laboratory where they were marked (date and place of collection) after weighing and taking measurements. The analyses consisted of identifying and counting the remains of undigested prey contained in the droppings. Using a binocular magnifier, the sclerotinised pieces were separated and grouped according to their similarities, then we proceeded to the identification of each grouping. Prey fragments are determined largely by reference to the collections held in the applied zoology laboratory at the University of Bejaia (Algeria), but also thanks to a number of identification guides and specialised websites (Jones et al. 1983, Zahradnik 1984, Gaêtan 1986, Patrice and Philippe 2003, Louveaux et al. 2013). For each study site, we established the faunistic list of prey taxa consumed according to their abundances.

During the sampling period, a total of 158 faecal samples were collected, distributed across different environments. Of these, 65 samples were obtained from the urban area, 48 from natural area and 45 from suburban area.

For the examination of potential variations in the diets of Atelerix algirus, we have used the ecological composition index: the frequency of occurrence Fo %, the centesimal frequency Fc %, the Shannon diversity index (H'), the Equitability (E) and the Sorensen similarity index (Dajoz 2006, Ramade 2008). To describe the Atelerix algirus diet variations amongst the three study areas, we used the Costello graphical representation (Costello 1990), modified by Amundsen et al. (1996). It highlighted the importance of prey, their contribution to the trophic niche extent and their feeding strategy (Loua et al. 2019). The Kruskall-Wallis test was applied to test the differences between the samples. It is a non-parametric test suitable when the data do not follow a normal distribution. In the context of the study of diet, the distribution of data may be influenced by the natural variability of food resources in different environments (this is the case of our study, three groups of independent samples, etc.).

To identify the relationship between the ingested prey and the ecological characteristics of each study area, such as the level of anthropogenisation, the floristic richness (depending on the vegetation strata) and the opening level of the environment, we have applied a factorial analysis of correspondences (FAC). The results of this analysis were confirmed by the hierarchical ascending classification (HAC) which used the Euclidian distance between the variables. The test was done by using the Software XLSTAT 2021.

The highest abundance of faeces was recorded in the urbanised area with 65 faecal droppings between April and October. We noticed a total absence of droppings during April and October in the suburban area and during April in the natural area (Fig. 5). The number of droppings collected was constantly increasing from April to August, but seems to decrease from September. However, from November to March, we have not observed any afecal dropping of Atelerix algirus in the three study sites.

One hundred and sixty-one prey species were found in the Algerian hedgehog diet. The prey taxa were divided into seven classes, in which the class Insecta is the dominant and represented 97.72 % of the consumed prey (Fig. 6).

The list of prey ingested by the Algerian hedgehog in the three study sites can be found in Table 2.

Amongst the Insecta class, the Hymenoptera order is the most consumed (Hymenoptera AR = 86.90%) in the three study areas. This order is mainly represented by the Formicidae family. The species Messor barbarus is the most consumed prey; its frequencies were 83.36% in the urbanised habitat (Bab Ezouar), 77.62% in the sub-urbanised habitat (Zeralda) and 67.25% in the natural habitat (El Hamdania).

The order of Coleoptera (AR = 8.123%) took second place, mainly represented by the two families: Carabidae and Tenebrionidae. The ant Camponotus sp. 1 came in the second position after Messor barbarus with Fc = 22.15% in the suburban area and Fc = 12.81% in the natural area.

As for the urban area, the second favourite prey consumed by A. algirus was the Carabidae Acinopus picipes with Fc = 2.88%.

The third position is attributed to different species in the three study sites: the Blattelidae Ectobius sp. in the urban area with Fc = 1.14%, the Carabidae beetle Calathus sp. with Fc = 2.75% in the sub-urban area and Phyllodromica algirica with Fc = 4.62% in the natural area.

The findings show that the diversity and abundance of prey consumed are the highest in the urban area (Table 3) where we have recorded the presence of 79 prey species divided over 5574 individuals.

The parameters decreased in the natural area with 64 species and 3188 individuals. The suburban environment, on the other hand, had the lowest diversity and abundance (45 species and 1994 individuals consumed).

The difference between the samples were non-significant and confirmed by the Kruskall- Wallis test where: H (chi2): 18.24, p > 0.05.

The determination of the amplitude of the trophic niche in the three study areas showed no significant difference in the diversity and equitability index. The values of equitability tend to zero in the different types of environments (Table 3) which determine a situation of imbalance between the consumed prey and that only one species dominated the diet of A. algirus in the three environments and, thus, represented nearly the totality of the population (heterogeneous population).

The Sorensen's similarity index showed a slight similarity between the three environments (Table 4): urban and suburban areas seemed to share more similarity of consumed prey Is = 0.64; the natural area, on the other hand, showed little similarity of ingested prey (0.38 with the urban area and 0.35 with the sub-urban area).

The Costello graphical representation described the Algerian hedgehog as a generalist predator (feeding on a wide variety of prey) and inhabiting a broad ecological niche in all three environments. The ant Messor barbarus is qualified as the dominant prey. In the urban areas, the three types of prey, known as secondary, are represented by Acinopus picipes, Anisolabis maritima and Tetramorium biskrensis (Fig. 7).

In the suburban area (Fig. 8), they were represented by Camponotus sp. 1, Othiorynchus meridionalis, Sehirus sp., Aelia acuminata and Aphaenogaster sardoa.

In the natural area, the species Anisolabis maritima and Crematogaster scutellaris were secondary prey (Fig. 9). According to this representation, the other prey consumed are qualified as rare species or accidentally ingested by the Algerian hedgehog.

In the interests of evaluating the effect of biotic factors (level of anthropogenisation, number of floral taxa in the environment according to strata and openness level of the environment) on the richness of prey items consumed, we applied a factorial correspondence analysis (Fig. 10).

According to Axis F1 = 87.03%, we observed in the three study areas that the number of prey consumed depends on the density and composition of the vegetation cover particularly on the cover of the herbaceous and shrub.

The type of vegetation cover predominating the area, conditions directly the environment's degree of openness. The environment is open when the herbaceous strata is dominant (according to the F2 = 12.96%) and closed if trees are dominant.

The number of prey items consumed by hedgehogs does not depend on the anthropogenisation level, so hedgehogs adapted the same feeding strategy in different types of environments (Fig. 10).

The findings of the FCA are confirmed by the ascending hierarchical classification using the Euclidean distance to classify the variables (Fig. 11) where we observed two major groupings: the first is formed by the consumed prey and the herbaceous strata. The second is formed by the openness level and the anthropogenisation level, these results confirming those previously obtained by the FCA.

With the aim of grouping the study sites, we observed a first grouping formed by the natural area and the urban area according to affinity (they share more similarities). The suburban zone was distant from this group.

As for the variables, the number of prey consumed depends on the percentage of the herbaceous strata in the area (according to the distance between them). The openness level presented less affinity with the previous variables, which confirmed the results of the FCA.

The collection of Algerian hedgehog faeces showed that this species frequented the three different areas despite their heterogeneity (urban, suburban and natural environment). Reeve (1994) also reported that the European hedgehog (Erinaceus europaeus) tends to use a wider range of habitats ranging from forest to humanised environments.

The urban area (at 14 m altitude) was the most frequented (65 faeces samples) during the months of sampling compared to the other study sites (suburban and natural). Doumandji and Doumandji (1992) pointed out that the hedgehog does not hibernate in the coastal area at less than 50 m. However, it tends to reduce its activity, which explains the presence of droppings between April and October and their absence from November to March, because of the reduced activity in urban areas or hibernation in the other two stations (suburban: 100 m and natural: 800 m). Mouhoub-Sayah et al. (2018) recorded the highest faeces number during the pre-hibernation period because it represents a period of body reserve accumulation for hibernation.

The findings show that the Algerian hedgehog not only feeds on Insecta (97.72%), but also on Arachnida, Malacostraca, Diplopoda, Gastropoda, Chilopoda and Birds. This is confirmed by several studies (Schilling et al. 1986, Doumandji and Doumandji 1992, Mouhoub-Sayah et al. 2009, Khaldi 2014). The class of Birds recorded a single individual ingested corresponding to an egg. Axell (1956) and Boue and Chanton (1967)confirmed that A. algirus can also consume young birds and eggs on the ground.

The most consumed order of Insecta in the three study sites is the Hymenoptera with 86.90%, essentially represented by the Formicidae family.

The ant Messor barbarus is the most consumed prey by A. algirus in the three study sites with AR = 87%. This species is qualified according to the Costello diagram as the dominant prey in the diet of the hedgehog (AR > 50%).

The Kruskall-Wallis test applied to identify the difference in prey between samples was found to be insignificant which confirms that, despite the difference in habitats, A. algirus still follows the same strategy as a generalist predator with an opportunistic character feeding on Messor barbarus. This corresponds with the results of Mouhoub-Sayah et al. (2009) in the Soummam Valley in Algeria.

The Equitability index confirmed those data because it has a low value in the three study sites, which reflected a situation of imbalance (heterogeneous population) to the benefit of M. barbarus. Derdoukh et al. (2012) considered that the ant Messor barbarus is the most dominant prey in the diet of Atelerix algirus in two stations located in the Mitidja Plain.

The Coleoptera order is the second most consumed order (AR = 8.12%), represented by Acinopus picipes in the urban area. It has also been reported that the Algerian hedgehog consumed mainly Formicidae and Coleoptera, especially Carabidae (Athmani 1988, Sayah 1988). Djennoune (2019) reported that Coleoptera is the second most consumed order by A. algirus in three localities in Kabily, although it is quite uncommon. Our results support those reported in other studies conducted in Algeria (Doumandji and Doumandji 1992, Mouhoub-Sayah et al. 2009, Khaldi 2014). In contrast, most studies on the European hedgehog's diet (Erinaceus europaeus) have found that beetles were dominant in its faeces (Jones et al. 2005, Jones and Norbury 2011) in New Zealand.

The highest prey richness was recorded in the urban area (79 species). This sampling station is characterised by a mosaic landscape composed of the university’s campus buildings and open spaces where the herbaceous strata dominates (Fig. 2Table 1). Our results corroborate those of Hubert et al. (2011) who state that urban or peri-urban areas harbour hedgehogs in large numbers, about nine times more than in non-urban areas. Khidas (1998) has also noted the presence of A. algirus faeces in various open habitats in Algeria. In the natural area, in a national park, we also recorded 64 prey species of the Algerian hedgehog. This site presented a low anthropogenic activity; it is far from all types of human constructions. It is located at an altitude of 800 m; furthermore, it is characterised by a mountainous relief and a dense vegetation cover dominated by the herbaceous strata (Fig. 4Table 1). Djennoune (2019) also reported the presence of A. algirus faeces in Yakouren, which is a mountainous station located between 280 m and 1340 m altitude. Hubert (2008) revealed that stations located at high altitude with mountainous relief correspond to the availability of the hedgehogs' preferred natural food (insects mainly ants).

Reeve (1994) suggested that the hedgehog habitat depends on the existence of shelters (natural hedges and bushes), which are available in semi-open environments, such as those of El Hamdania and the university campus in urban areas.

The suburban area recorded the lowest number of faeces (45 faeces samples) and the lowest prey richness (45 species). Despite the reduced anthropogenic effect compared to the urban area, this site is characterised by the domination of two shrub and tree strata which make the habitat less open (Fig. 3). This vegetation cover harbours a few prey items that interest the hedgehog as it feeds on Arthropods that are found on the ground (depending on its shifting mode). Garcia and Puig (2014) concluded that the optimal biotope for the hedgehog is the open and semi-open environment with abundant herbaceous and shrub strata, this being confirmed in Algeria by the present study and the one of Djennoune (2019).

The presence of the hedgehog in the three different areas confirms that it uses all available habitat types. This is related to its predation strategy, which qualifies Atelerix algirus as an opportunistic generalist. Although it has food preferences, we noted, through the results of the FCA and HAC, that the richness of its prey is directly related to the dominance of the herbaceous strata in the first place because it harbours the hedgehogs' prey species.

The openness level of the environment joined the grouping in the second place because it results from the type of strata that dominates each environment. The urban environment is "open", the suburban environment and the natural environment being of semi-open type. According to literature, these two levels of openness are responsible for the food availability of the hedgehog (Garcia and Puig 2014).

The urban area, which presents a great anthropogenic activity, recorded the highest richness of prey and an important abundance of faeces than the other habitat types. It indicated that the level of anthropogenisation does not condition the activity of the Algerian hedgehog. This is confirmed by the FCA's results where the level of anthropogenisation does not belong to the grouping obtained and is also confirmed by the HAC where it comes in last place to join the parameters influencing Atelerix algirus prey’s richness.

The human activities in those areas do not have a strong influence on hedgehog populations. This is confirmed by the studies of Djennoune (2019) in Algeria. Anthropogenic activity and urbanisation do not seem to negatively affect the activity and diet of the Algerian hedgehog; instead, the urban environments appear to be conducive to hedgehogs due to higher food densities related to human occupation including natural prey, anthropogenic sources, more suitable breeding sites and reduced predation risk (Morris 1985, Micol et al. 1994Young et al. 2016, Pettet et al. 2017).