Genetic structure and expansion of golden jackals (Canis aureus) in the north-western distribution range (Croatia and eastern Italian Alps)
The golden jackal, widely distributed in Europe, Asia and Africa, is one of the less studied carnivores in the world and the genetic structure of the European populations is unknown. In the last century jackals strongly declined mainly due to human persecution, but recently they expanded again in eastern Europe. With the aim to determine the genetic structure and the origin of expanding jackals, we analyzed population samples obtained from Bulgaria, Serbia, Croatia (Dalmatia and Slavonia) and individuals sampled in north-eastern Italy. Samples were typed at the hypervariable part of the mitochondrial DNA control-region (mtDNA CR1) and at 15 canine autosomal microsatellite loci (STR), and analyzed using multivariate, Bayesian and landscape genetic methods. The mtDNA CR1 was monomorphic, showing a single haplotype shared among all the populations. The STR loci were variable, with 2–14 alleles and intermediate values of heterozygosity (Ho = 0.47; He = 0.51). Genetic diversity was significantly partitioned (θST = 0.07; P < 0.001) and the populations were partially distinct, perhaps in consequence of recent fragmentations. Jackals from Dalmatia were the most genetically differentiated. Assignment testing and gene flow analyses suggested that jackals colonizing Italy have admixed origins from Dalmatian and Slavonian populations. They are not first generation migrants, suggesting that dispersal towards north-eastern Italy is a stepping-stone process. Golden jackal and wolf colonization patterns might be different, with prevalent short-distance dispersal in jackals versus prevalent long distance dispersal in wolves. The admixed origin of jackals in the Alps ensures abundant genetic variability, which may enhance adaptive fitness and expectancy of population growth. The intersections between Dinaric–Balkan and Eastern Alps are areas of population expansion and admixture, highlighting their conservation, ecological and evolutionary values.
KeywordsAdmixture analysis Assignment testing Autosomal microsatellites Canis aureus Colonization genetics Mitochondrial DNA control-region Population structure
Climate warming and the spread of forests in mountain areas of Europe are determining deep changes in the composition of animal communities (Mawdsley et al. 2008). In the last few decades carnivore populations expanded in central and eastern Europe (Trouwborst 2010). The lowlands around the north Adriatic basin and subalpine areas of the eastern Alps have also been affected by waves of species expansion. Dinaric–Balkan populations of carnivores such as the brown bear (Ursus arctos) and the wolf (Canis lupus) expanded towards the Alps, entering in north-eastern Italy (Lapini et al. 2010; Fabbri et al. 2013), followed by lynx (Lynx lynx; Lapini et al. 1996), otter (Lutra lutra; Lapini and Bonesi 2011), raccoon dog (Nyctereutes procyonoides; Lapini 2006) and golden jackal (Canis aureus; Lapini et al. 2011). The expansion of wolves and other carnivores might follow a propagule model (Ibrahim et al. 1996), characterized by: (1) the early dispersal of vagrants, usually young males in search of new suitable territories (Valière et al. 2003; Fabbri et al. 2007, 2013); (2) the settlement of stable reproductive familial units, which may rapidly expand and saturate all suitable areas (Mech 1970). The genetic diversity of the new colonies is determined by the rates of long-range dispersal and by the consequences of founder effects (Fabbri et al. 2007). Isolated colonies might experience losses of genetic variation if population settlement and expansion are delayed by years (Vila et al. 2003). In the second phase of colonization, variable combinations of short-range dispersal, inbreeding avoidance and the turnover rates of breeding adults will determine the evolution of genetic variation (vonHoldt et al. 2010). This model well fits with wolf expansion dynamics, but the colonization genetics of other species is less known.
The golden jackal, a highly adaptable canid widespread from Africa to the Arabian peninsula, reaching central Asia, India and Indochina, is a historic invasive species in Europe (Sillero-Zubiri et al. 2004). Golden jackal populations in Europe are fragmented, particularly towards the north-west periphery of the range. The main populations occur in the Balkans, Hungary and Ukraine, and northward they reach Slovenia and Austria (Humer et al. 2007). Smaller isolated populations are distributed along the Adriatic coast of Albania, Montenegro and Croatia, and in the Black Sea coast of the Balkan peninsula (Arnold et al. 2012). Golden jackals in Europe declined until 30–40 years ago due to human persecution and overhunting and were sometimes treated as pests and eradicated (Spassov 1989). Golden jackals and wolves are competitors, and the recent waves of wolf expansion in Europe might have contributed to worsen the decline of golden jackals (Genov and Wassilev 1989; Kryštufek and Tvrtković 1990). Negative demographic trends are, however, reversing and golden jackals are currently expanding again in eastern Europe, particularly in Bulgaria (Kryštufek et al. 1997). Climate and habitat changes, and partial legal protection are favouring the expansion in areas from where the species has been absent till recent. Vagrant or reproductive individuals were recently observed in Slovenia, Austria and north-eastern Italy, probably pushed by the ongoing expansion from Bulgaria (Kryštufek et al. 1997).
Golden jackal (Canis aureus) samples analysed in this study
Country: region (acronym)
A. Vlasseva, A. Ahmed
Serbia: East Serbia (SRB)
Croatia: Slavonia (SLA)
Croatia: Dalmatia (DA)
I. Bošković, A. Galov
Italy: Friuli Venezia Giulia, Trentino Alto Adige, Veneto (IT)
Tissues, skins, teeth, hair from museum samples
Materials and methods
Sampling and DNA extraction
A total of 120 golden jackal samples were collected from Bulgaria (n = 55), eastern Serbia (close to the Bulgarian border; n = 8), Croatia (n = 50) and north-eastern Italy (n = 7; Fig. 1; Table 1). Samples from Bulgaria were obtained from animals legally shot or road-killed from 1980 to 1990 (n = 40 skins; BG-1) and in 2008–2009 (n = 15, muscle tissues collected near the city of Plovdiv and the village of Garbino c. 234 km far from Plovdiv; BG-2). Golden jackals in Croatia were legally shot or road-killed in 2008–2010 (three Dalmatian specimens collected in 2002) in two regions: continental Slavonia (n = 34) and coastal Dalmatia (n = 16). The seven samples from north-eastern Italy were collected from the regions Friuli Venezia Giulia, Veneto and Alto Adige.
Genomic DNA was extracted using a Wizard Genomic DNA Purification Kit (Promega, USA) and a Quick-gDNA™ MiniPrep kit (Zymo Research, USA). DNA was eluted in 100 μl of purified DNA/RNA-free water and stored at −20 °C until subsequent handlings.
Mitochondrial DNA control-region sequencing
We amplified 450 bp of the hypervariable left domain of the mitochondrial DNA control-region (mtDNA CR1) using PCR and sequencing primers WDLOOPL 5′-TCCCTGACACCCCTACATTC-3′, H519 5′-CGTTGCGGTCATAGGTGAG-3′ (designed on wolf mtDNA CR; Caniglia et al. 2013), in a 10 μl total PCR volume with: 2μl of 20–40 ng/μl DNA, 0.3 μM of primer mix (forward and reverse), 10x PCR buffer with 2.5 mM Mg2+ and 0.25 units of Taq polymerase (5 PRIME Inc., Gaithersburg, USA). The PCR profile was: 94 °C/2 min, 94 °C/15 s, 55 °C/15 s, 72 °C/30 s, 72 °C/5 min of final extension, for 40 cycles. The amplification products were purified using ExoSAP-IT (Affimetrix, Inc., Cleveland, Ohio USA) and sequenced in both directions using an Applied Biosystems 3130XL DNA sequencer (Life Technology). The sequences were visualized and corrected in SeqScape v.2.5 (Life Technology) and successively aligned in BioEdit (Hall 1999) with a 392 bp-long Serbian jackal sequence downloaded from GenBank (GU936689; Zachos et al. 2009). Negative controls (no DNA) were always used to check for PCR contaminations, which never occurred.
Microsatellite loci and PCR amplifications
Variability at 15 canine microsatellite loci selected to genotype the golden jackal samples used in this study
5.07 (SE = 0.74)
0.50 (SE = 0.04)
0.46 (SE = 0.03)
At a 99 % confidence interval and after Bonferroni correction, locus CPH2 showed evidence of null alleles and it did not amplify in 20 % of samples; thus it was discarded. Three loci were monomorphic (CPH3, CPH7 and FH2079), two (FH2010 and FH2054) did not consistently amplify (amplification rate ≤0.6), while the others 15 reliably amplified, were polymorphic (Ho = 0.14–0.71; PIC = 0.19–0.77; Table 2) and were used in this study.
Analysis of genetic diversity
Allele frequencies, average number of observed (Na) and private (Np) alleles, observed and expected heterozygosity (Ho, He), probability-of-identity (Waits et al. 2001) among unrelated individuals (PID) and among full sibs (PIDsibs) and AMOVA (Excoffier et al. 1992) were computed with GenAlEx 6.4 (Peakall and Smouse 2006). Fstat 188.8.131.52 (Goudet 1995) was used to compute the allelic richness (Nar), the number of alleles standardized for the smallest number of individuals in a sample (n = 15; excluding the smallest sample of seven admixed jackals that were collected in Italy). Genetix 4.03 (Belkhir et al. 1996–2001) was used to estimate FIS, θST and Nm (Weir and Cockerham 1984) and to test for Hardy–Weinberg equilibrium (HWE). Pair-wise linkage equilibrium (LE) among loci was estimated using the Guo and Thompson’s (1992) Markov chain method in Genepop 4.0 (Rousset 2008). The sequential Bonferroni correction test for multiple comparisons was used to adjust the significance levels at α = 0.05 (Rice 1989).
Population structure and landscape genetic analyses
We used Stucture 2.3.3 (Pritchard et al. 2000) to identify the optimal number of K genetically distinct clusters (assuming HWE and LE) in the samples and to assign the individuals to the clusters. We applied the admixture and correlatedallele frequencies (F) models (Falush et al. 2003), with K from 1 to 10, running five replicates each of 4 × 105 iterations following a burn-in of 4 × 104 iterations. We also used the locprior model, assuming that sampling locations are informative priors to assist the clustering procedure (Hubisz et al. 2009). To evaluate eventual substructure in the Bulgarian samples we considered BG-1 and BG-2 groups as two distinctive locpriors. The statistic ΔK was used to identify the major increase in the posterior probability, Ln P(D), between each successive K (Evanno et al. 2005). At optimal K we assessed the average proportion of membership (Qi) of the sampled populations and the individual proportion membership (qi) to the clusters, using an arbitrary threshold qi = 0.80 (which has been used in other carnivore studies: Oliveira et al. 2008; Randi 2008; vonHoldt et al. 2010). The spatial locations of genetic clusters were reconstructed by landscape genetic analyses in Geneland 4.0.3 (Guillot et al. 2005), with 105 MCMC iterations, (thinning = 103 and post-process burn-in = 102 iterations), the correlatedallele frequency model and coordinates uncertainty = 0.01 corresponding to 1.6 km (radius of a territory of 8 km2; Taryannikov 1977). In addition we used a spatial principal component analysis in Adegenet (sPCA, Jombart et al. 2008), which describes the spatial pattern of population structuring independently on any assumption on population genetic equilibrium. The geographic coordinates of samples collected in Bulgaria from 1980 to 1990 were not available, thus these samples were not used in Geneland and Adegenet analyses. Current rates of gene flow between the populations were estimated by Bayesass (Wilson and Rannala 2003), with 3 × 106 iterations, a thinning interval of 2000 and a burn-in of 105. We compared the results obtained by Bayesian methodologies with Paetkau et al.’s (2004) frequency-based assignment test implemented in GenAlEx.
We obtained a fragment of 450 bp of the hypervariable domain of the mtDNA CR1 from 120 golden jackals collected in Bulgaria, Croatia, Serbia and Italy. The aligned sequences were identical: all jackals shared the same haplotype and, consequently, haplotype and nucleotide variability were zero. This unique haplotype was identical to the one described in Serbian jackals (Zachos et al. 2009; GenBank accession number KF588364). In contrast, 15 of the 21 STRs were polymorphic. We obtained complete multilocus genotypes at the 15 polymorphic STRs in 114 samples, whereas six individuals (5 %) showed from 1 to 5 loci with missing data: four at one locus, one at two loci and one at five loci. The number of alleles per locus varied from Na = 2 (in locus FH2088) to Na = 14 (FH2137); the observed heterozygosity varied from Ho = 0.14 (CPH12) to Ho = 0.71 (FH2137; Table 2). Overall there were 8/15 (53 %) loci not in HWE (P < 0.05) due to presence of genetic substructure (see results below), with average FIS = 0.11, significantly different from 0 (P < 0.001). The loci were not in linkage disequilibrium (at P = 0.05 after Bonferroni correction).
Population structure and landscape genetic analyses
Average proportion of membership (Qi) computed from five replicated runs of Structure of golden jackal genotypes grouped in five predefined populations (the samples from Bulgaria are in the same group; sample size in parenthesis), and assigned to three (K = 3) clusters (90 % credibility intervals in parenthesis)
Only jackals from Dalmatia clustered separately from the other samples, independently of the model, with average proportion of membership QIII > 0.90 and individual proportion of membership qi = 0.82–0.96 (admixture model) and 0.96–0.99 (locprior model). The other populations were not assigned to any single cluster with the admixture model (Table 3), whereas a sharper sub-structuring was detected using the locprior model. The Slavonian and Serbian jackals were assigned to same genetic cluster with QII = 0.90 and 0.82, respectively, while the samples from Bulgaria showed signals of admixture (Table 3). Independently of the model, golden jackals sampled in Italy were genetically admixed (Fig. 2; Table 6). With the admixture model, three samples (1216, 1219 and 1220) were assigned to the Dalmatian cluster with qi = 0.96, 0.79 and 0.90, respectively. The other four samples showed genetic components shared with the Dalmatian and Slavonian jackal clusters (Table 6). The assignments of the Italian samples with the locprior model were weaker (Table 6).
Genetic variation in the subpopulations
Genetic variability in golden jackals sampled in Bulgaria 1980–1990 (BG-1), Bulgaria 2008–2009 (BG-2), Slavonia (SLA) plus Serbia (SRB), Dalmatia (DA), and in north-eastern Italy (IT), and genotyped at 15 autosomal STRs (N = sample size)
Gene flow and origin of jackals in Italy
a Matrix of inter-population θST values (lower left) and the corresponding estimates of historical gene flow (Nm; upper right) among the sampled golden jackal populations; in parenthesis probability values based on 999 permutations. b Mean (standard deviation) posterior probabilities of migration rates m into each jackal population computed in Bayesass (Wilson and Rannala 2003). The populations from which individuals were sampled are listed in the rows; the populations from which they migrated are listed in the columns. Values along the diagonal are the proportions of individuals originated from the source population (not migrants). Es: 73.8 % individuals sampled in BG are from BG (source population), 24.6 % are from SLA, 0.6 % from DA, etc.
Identification (ID), sampling year and locality of golden jackals collected in north-eastern Italy (see: Fig. 1)
BASS Time 1
BASS Time 2
Despite its large range distribution, the golden jackal has been little investigated and few genetic studies have been published so far. Zachos et al. (2009) genotyped 121 individuals from Serbia using eight STRs, finding a very low genetic variability. Cohen et al. (2013) analysed 88 jackals from Israel using 14 STRs, and found a high level of genetic diversity and no evidence of bottleneck, even if the species was near-completely eradicated in this region. In our study the golden jackal populations sampled in Bulgaria, Serbia and Croatia share the same mtDNA CR1 haplotype, but they have polymorphic STRs with 2–14 alleles and intermediate values of heterozygosity (Ho = 0.47; He = 0.51). Although more empirical data are needed, these first results may suggest that Europe was colonized a few centuries ago by small numbers of founders, which carried a limited portion of the total genetic diversity of the southern golden jackal source populations.
The monomorphic mtDNA CR1 sequence consistently found in this study and in Zachos et al. (2009) suggests a scenario of historical population contraction to restricted refuge areas, followed by recent re-expansion waves and rapid demographic expansion. The number of sampled populations and the empirical evidence available so far are not enough to allow reconstructing a reliable global phylogeographic framework for the golden jackal in Eurasia. Thus, at the moment it is not possible to further speculate on the phylogeographical relationships of extant golden jackal populations, and identify the causes of the observed mtDNA monomorphism. It is well known that the control-region includes the most variable sequences in the mtDNA genome of the vertebrates (Simon 1991). There are some described exceptions, and among them, notably, a carnivore, the otter (L. lutra) that showed very conserved mtDNA CR sequences in all the populations studied so far (Mucci et al. 2010). The two competing hypotheses, that the golden jackal mtDNA CR monomorphism derives: (1) from an exceptionally conserved molecular dynamics, or (2) from a recent origin of the extant European jackal populations, may be tested by sequencing other mtDNA and nuclear DNA genes (e.g., polymorphic genes from the major histocompatibility complex, or olfactory receptors; Aguilar et al. 2004; Quignon et al. 2012) and comparing their mutation rates, or by genotyping golden jackal samples collected throughout the entire distribution range in Eurasia and Africa.
The STRs allele distributions in the sampled golden jackal populations were significantly different (θST = 0.07; P = 0.001), yielding consistent pictures of genetic structuring, using multivariate, Bayesian or landscape population genetic methods. These three statistical methods are based on different assumptions. In particular, Structure assumes that an admixed sample of unknown subpopulations can be split in distinct genetic clusters maximizing both HWE and LE within each of them (Pritchard et al. 2000). Obviously, empirical multilocus datasets (including our jackals) might not include subpopulations in genetic equilibrium, because of recent founder effects, ongoing admixture or biased sampling. The consequences of assumption violations on Structure results are, however, poorly known (Kaeuffer et al. 2007; Kalinowski 2011), and it is wise to compare the results obtained with different methods. Landscape genetic models assume that spatial proximity should be a priori related to genetic similarity, and that the use of spatial information could improve clustering in populations that, as in the golden jackal case-study, are not deeply divergent (θST < 0.10; Guillot 2008).
Bayesian and landscape genetics models showed a weak population subdivision, but golden jackals sampled in Dalmatia were clearly separated from all the other samples. In Structure jackals from Dalmatia clustered separately from the other samples assuming K = 3. Jackals from Slavonia and Serbia were assigned to the same genetic cluster, while samples from Bulgaria remained distinct, showing strong signals of admixture. The spatial genetic analyses (Structure with the locprior model; sPCA) confirmed these results, showing a major subdivision between jackals from Dalmatia versus all the other sampled groups. Geneland confirmed this pattern, showing sharp subdivisions among jackals sampled in Dalmatia, Slavonia–Serbia and Bulgaria. These results are in agreement with the available historical information. The golden jackal is a historical invasive species in Europe, and its permanent presence has been reported at least since the Middle Ages in southern Dalmatia (Kryštufek and Tvrtković 1990). Permanent populations in northern Dalmatia were established much later, during the 20th century. Further north-western expansions towards Istria were more recent (in the 1980s; Kryštufek and Tvrtković 1990). The coastal golden jackal populations might have originated centuries ago from unknown source populations, and might have survived in isolation for hundreds of generations. The empirical genetic evidence support this hypothesis. In fact, the 15 selected STRs were polymorphic in all the sampled golden jackal populations, except for locus CPH12 that was monomorphic in the Dalmatian samples. Moreover, allele richness and heterozygosity were smaller in Dalmatia than in the other populations. Reduced genetic diversity suggests that jackals in Dalmatia might have experienced a demographic bottleneck in the past (a test designed to detect recent bottlenecks did not yield significant results; Cornuet and Luikart 1996) and have survived in isolation for centuries. Once again, historical information and genetic evidences are in agreement. The expansion of jackals in Slavonia, in the continental eastern part of Croatia, occurred only from 1903 onwards, and the more recent expansion wave occurring in the last 15 years is most probably due to immigration of jackals from Bulgaria, Romania and Serbia (Giannatos 2004).
This pattern of population clustering yields a reference framework for efficient assignments of individuals to the populations. Although only jackals from Dalmatia are always consistently assigned to their population of origin, at least a portion (71 %) of the Slavonian–Serbian samples was correctly assigned using genetic information only. All three geographic groups (Dalmatia, Slavonia–Serbia and Bulgaria) were identified and the individuals were correctly assigned by integrating the geographical information in Geneland, which also assigned the admixed Italian samples to a fourth group (Fig. 1). These results indicate that weakly differentiated populations (θST < 0.10) can be identified and individuals assigned if multilocus genetic and geographical information are integrated in landscape genetics models. The results obtained from landscape analyses allowed drafting a preliminary reconstruction of jackal expansion in eastern Europe.
Jackals in Bulgaria, Croatia and Slavonia–Serbia showed low rates of both historical and current gene flow (Table 5), meaning that they were and are substantially isolated from one other. The only exception is the jackal population in Bulgaria, which appears strongly admixed and historically connected with jackals in Slavonia–Serbia (Nm = 15.5). These findings are in agreement with demographic data suggesting that the expansion of the more abundant and widespread jackal populations in Romania and Bulgaria increased local admixtures and pushed migration towards north-western areas. Current rates of gene flow estimated in this study, though they should be evaluated with caution because of partial differentiation among populations (Faubet et al. 2007), suggest that Slavonia is the main source of jackals migrating both eastwards (towards Bulgaria) and westwards (towards Dalmatia and Italy). Therefore, as the reliability of gene flow estimates based on Fst or assignment procedures is conditioned by a number of assumptions and sampling design (Whitlock and Maccauley 1999), this hypothetical scenario should be tested indirectly by genotyping additional samples, or directly by reconstructing the dispersal movements of radio-tagged jackals. The historical invasion of golden jackals in Europe could have been sustained by anthropogenic habitat changes, which may favour opportunistic mesocarnivores, and by global climate changes, which may favour the expansion of species of southern origins. These dynamics could modify the ecological interaction in prey–predator communities in southern Europe, with implications for conservation of both carnivores and their prey.
Population structure and assignment tests concordantly indicated that jackals colonizing Italy have admixed genotypes that originated from the Dalmatian and Slavonian populations (Figs. 1, 2, 3). Admixed origins, further supported by the absence of any private allele, and evidence that the genotypes sampled in Italy are not first generation migrants but might have originated one generation back, suggest that jackal dispersal from Dalmatia and Slavonia towards north-eastern Italy is a stepping-stone process. Jackals migrating from their source populations probably met and mated in some (still un-sampled) intermediate areas. In this perspective, golden jackal and wolf colonization patterns might be significantly different, due to the prevalent short-distance dispersal with intermediate admixture in jackals versus prevalent long distance dispersal in wolves. This hypothesis might be tested with additional genetic, demographic and behavioural data. The admixed origin ensures abundant genetic diversity in the golden jackal colonizers, which may enhance their adaptive fitness and expectancy of population growth. These findings and the recent expansion of wolf (Fabbri et al. 2013), brown bear, lynx and otter (Lapini et al. 1996; Lapini and Bonesi 2011) confirm that the intersections between the Dinaric–Balkan and Italian Alps are areas of population expansion and admixture, highlighting their conservation, ecological and evolutionary values. Monitoring the ongoing jackal colonization process in the Dinaric–Italian Alps might take advantage of improved non-invasive genetic sampling and molecular identification procedures (Waits and Paetkau 2005). Jackals colonizing Italy show enough genetic diversity and low probability-of-identity to allow reliable individual genotype identifications also from DNA extracted from non-invasive samples.
We warmly thank everybody who made it possible to realize this research project, and that contributed to obtain samples used in this study. This project was supported by ISPRA, by the Italian Ministry of Environment, Direction of Nature Protection, and it was partly supported by the European Social Fund through the Ministry of Education, Youth and Science of Bulgaria (BG051PO001-3.3.04/41). We are particularly grateful to Aritz Ruiz-Gonzalez for his useful suggestions and help in landscape genetic analysis and for his comments on a preliminary version of this manuscript.
- Banea OC, Krofel M, Červinka J, Gargarea P, Szabó L (2012) New records, first estimates of densities and questions of applied ecology for jackals in Danube Delta Biosphere Reserve and hunting terrains from Romania. Acta Zool Bulg 64:353–366Google Scholar
- Belkhir K, Borsa P, Chikhi L, Raufaste N, Bonhomme F (1996–2004) Genetix 4.05, logiciel sous Windows TM pour la génétique des Populations. Laboratoire Génome, Populations, Interactions, CNRS UMR 5000, Université de Montpellier II, Montpellier, France Google Scholar
- Evanno G, Regnaut S, Goudet J (2005) Detecting the number of clusters of individuals using the software structure: a simulation study. Mol Ecol 14:611–2620Google Scholar
- Fabbri E, Miquel C, Luchini V, Santini A, Caniglia R, Duchamp C, Weber J-M, Lequette B, Marucco F, Boitani L, Fumagalli L, Taberlet P, Randi E (2007) From the Apennines to the Alps: colonization genetics of the naturally expanding Italian wolf (Canis lupus) population. Mol Ecol 16:1661–1671PubMedCrossRefGoogle Scholar
- Fabbri E, Caniglia R, Kusak J, Galov A, Gomerčić T, Arbanasić H, Huber D, Randi E (2013) Genetic structure of expanding wolf (Canis lupus) populations in Italy and Croatia, and the early steps of the recolonization of the Eastern Alps. Mamm Biol (in press)Google Scholar
- Genov P, Wassilev S (1989) Der Schakal (Canis aureus L.) in Bulgarien. Ein Beitrag zu seiner Verbreitung und Biologie. Z Jagdwiss 35:145–150Google Scholar
- Giannatos G (2004) Conservation action plan for the golden jackal Canis aureus L. in Greece. WWF Greece, p 47Google Scholar
- Goudet J (1995) Fstat (Version 1.2): a computer program to calculate F-statistics. J Hered 86:485–486Google Scholar
- Hall TA (1999) BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/97/NT. Nucleic Acids Symp Ser 41:95–98Google Scholar
- Humer A, Heltai M, Murariu D, Spassov N, Häcklander K (2007) Current status and distribution of golden jackals (Canis aureus) in Europe. In: Sjöberg K, Tuulikki R (eds) Abstracts XXVII Congress IUGB, Uppsala, Sweden. Swedish University of Agricultural Sciences, Umeå, p 272Google Scholar
- Kryštufek B, Tvrtković N (1990) Range expansion by Dalmatian jackal population in the 20th century (Canis aureus L. 1758). Folia Zool 39:291–296Google Scholar
- Lapini L (2006) Il cane viverrino Nyctereutes procyonoides ussuriensis Matschie, 1908 in Italia: segnalazioni 1980–2005 (Mammalia: Carnivora: Canidae). Boll Mus Civ St Nat Venezia 57:235–239Google Scholar
- Lapini L, Bonesi L (2011) Evidence of a natural recovery of the Eurasian otter in northeast Italy. In: Proceedings of the 29th European Mustelid Colloquium, Southampton, UKGoogle Scholar
- Lapini L, Dall’Asta A, Dublo L, Spoto M, Vernier E (1996) Materials for the theriofauna in north-eastern Italy (Mammalia, Friuli-Venezia Giulia) Gortania. In: Proceeding of the museum of the natural history of friuli 17, pp 149–248 (in Italian)Google Scholar
- Lapini L, Molinari P, Dorigo L, Are G, Beraldo P (2009) Reproduction of the golden jackal (Canis aureus moreoticus I. Geoffroy Saint Hilaire, 1835) in Julian pre-Alps, with new data on its range-expansion in the high-adriatic Hinterland (Mammalia, Carnivora, Canidae). Boll Mus Civ St Nat Venezia 60:169–186Google Scholar
- Lapini L, Brugnoli S, Krofel M, Kranz A, Molinari P (2010) A grey wolf (Canis lupus Linné, 1758) from Fiemme Valley (Mammalia: Canidae: north-eastern Italy). Boll Mus Civ St Nat Venezia 61:117–129Google Scholar
- Lapini L, Conte D, Zupan M, Kozlan L (2011) Italian jackals 1984–2011. An updated review (Canis aureus: Carnivora, Canidae). Boll Mus Civ St Nat Venezia 62:219–232Google Scholar
- Mech LD (1970) The wolf: the ecology and behavior of an endangered species. Natural History Press, Garden City, NYGoogle Scholar
- Rice WR (1989) Analyzing tables of statistical tests. Evolution 43:223–225Google Scholar
- Selanec J, Lauš B, Sindičić M (2011) Golden jackal (Canis aureus) distribution in Croatia. In: Abstract of VI European Congress of mammalogy, Paris, France, p 60Google Scholar
- Sillero-Zubiri C, Hoffmann M, Macdonald DW (2004) Canids: foxes, wolves, jackals and dogs: status survey and conservation action plan, 2nd edn. IUCN Canid Specialist Group, CambridgeGoogle Scholar
- Simon C (1991) Molecular systematics at the species boundary: exploiting conserved and variable regions of the mitochondrial genome of animals via direct sequencing from amplified DNA. In: Hewitt GM, Johnston AWB, Young JPW (eds) Molecular techniques in taxonomy. Springer, Berlin, pp 33–71CrossRefGoogle Scholar
- Spassov N (1989) The position of jackals in the Canis genus and life-history of the golden jackal (Canis aureus L.) in Bulgaria and on Balkans. Hist Nat Bulg 1:44–56Google Scholar
- Taryannikov VI (1977) On the population structure and dynamics of Canis aureus aureus in the Syrdarja river valley. Zool Z Mosca 71:1423–1425Google Scholar
- Zachos FE, Cirovic D, Kirschning J, Otto M, Hartl GB, Petersen B, Honnen A (2009) Genetic variability, differentiation, and founder effect in golden jackals (Canis aureus) from Serbia as revealed by mitochondrial DNA and nuclear microsatellite loci. Biochem Genet 47:241–250PubMedCrossRefGoogle Scholar