Study areas
Fieldwork was carried out in the northwest of the Iberian Peninsula: Galicia, Zamora and Valladolid regions (Spain) (Fig. 1). In Galicia, the study location is a protected area of 5722 ha called Montes do Invernadeiro Natural Park. This area falls within an altitudinal range of 803–1707 m with a series of low mountains and deep valleys. The vegetation mainly consists in scrubland composed by heather (Erica australis), prickled broom (Pterospartum tridentatum) and sandling (Halimium lasianthum). There are also deciduous forests in the valleys and along watercourses, characterised by oak (Quercus robur), birch (Betula celtiberica) and holly (Ilex aquifolium). Replanted Scot pine (Pinus sylvestris) forests are also common in the area. This region, especially the natural park, has a high density of wild ungulate prey such as roe deer (Capreolus capreolus), red deer (Cervus elaphus) and wild boars (Sus scrofa).
Study area in Zamora comprises the “Sierra de la Culebra” Regional Hunting Reserve. The Sierra de la Culebra is a mountain chain ranging from 800 to 1243 m of altitude. The vegetation is characterised by scrubland composed by heather (Erica australis). While natural forests are scarce and mostly found in the valleys, replanted Scot pine (Pinus sylvestris) forests are abundant in this reserve. As in Galicia, wolf preys primarily are roe deer (Capreolus capreolus), red deer (Cervus elaphus) and wild boar (Sus scrofa).
In Valladolid, the study region is dedicated to agriculture and livestock, being very flat and with very low tree density. There are no red deer in the area, and wild boars and roe deer are scarce. On the contrary, lagomorphs such as European wild rabbits (Oryctolagus cuniculus) and Iberian hares (Lepus granatensis) are abundant. There are also domestic pig (Sus scrofa domesticus) farms in the study area where wolves can scavenge.
Wolf group detection and prey availability
Breeding wolf groups were spotted during May-June by the increase in faecal marking behaviour, which is more intense during breeding season (Barja et al. 2005). Faecal marking is carried out by the dominant pair and only in reproductively active groups (Barja et al. 2008). Wolves use conspicuous substrates to deposit faeces to increase the effectiveness of the signal (Barja 2009b), this way wolves know which territories are occupied by other neighbouring groups. Later, breeding groups were confirmed by camera trapping in Galicia and by sightings in Zamora and Valladolid. A total of nine rendezvous sites (zones where pups are left from July to September while adults hunt, Mech and Boitani 2006) belonging to nine Iberian wolf breeding groups were located: four in Galicia (Invernadeiro Natural Park, Saguñedo, Castrelo and Baldriz), four in Zamora (Boya, Ferreras, Mahide and Villardeciervos) and one in Valladolid.
To estimate wild ungulate prey availability (roe deer, red deer and wild boar) in Galicia, we set 12 camera trapping stations for one month in each one of the four group rendezvous site locations. We corrected the total number of photographic events between the number of days that cameras were active to set comparable conditions. In the case of Zamora and Valladolid, to know wild prey relative abundance we request to the Administration and the Hunting Reserve, respectively, the official census of wild prey. We also set a correction index for the census by dividing the abundance of wild ungulates by the total area (ha) occupied by each rendezvous site.
Faecal sample collection
Scats were collected during breeding months (June-September). For this, we surveyed transects along roads and firebreaks near the rendezvous sites walking and with a vehicle (10 km/h) for 15 days each month. Transects were done early in the morning, when the probability to detect fresh faeces is higher due to wolf nocturnal and crepuscular habits (Barja et al. 2008; Barja et al. 2018). Fresh scats were discriminated from old ones by the strong smell, a layer of mucus, and no signs of dehydration (Barja 2008; 2009a; Martín et al. 2010). From each scat we collected two subsamples: one for genetic analyses and another for nitrogen contents. Samples for genetic analyses were preserved in ethanol at -20ºC whereas samples for nitrogen content were frozen at -20ºC until laboratory analyses. Moreover, from each fresh scat detected, we collected hair and bone samples to conduct dietary analyses.
We also register if the scat was a scent mark or not, considering as scent marks only faeces deposited in conspicuous or elevated substrates or crossroads (Barja 2009b). In wolf groups only the dominant pair exhibit this marking behaviour (Barja et al. 2004; Barja et al. 2008; Barja 2009b), hence, we considered that collected faeces with a marking function belonged to dominant wolves (Barja et al. 2008).
Genetic analysis
To ensure that faecal samples belonged to Iberian wolves and not to other sympatric carnivores, we performed a genetic identification on the faecal samples collected in the field by sequencing mitochondrial DNA (mtDNA). We took a subsample of each faecal sample, placed it in tubes filled with ethanol (96%) and stored it at -20ªC. For the extraction of the DNA, we used an extraction kit consisting in silica membranes and adapted to non-invasive samples (QIAamp DNA Stool Mini Kit, Qiagen). To identify the species origin of the samples, we sequenced a 440 bp fragment of the mitochondrial DNA control region following Vilà et al. 1999 methodology. We used the PCR (Polymerase Chain Reaction) technique and the universal primers Thr-L 15926 and DL-H 16340 for the amplification of the DNA. Then, we used gel electrophoresis to verify the success of the DNA amplification. In order to eliminate the primers and the excess of deoxynucleotides, we applied the alkaline phosphatase and exonuclease I (ExoSAP-IT) method for the cleaning and purification of the amplified product. Finally, the sequencing on this cleaned PCR product was conducted by the application of the commercial kit dRhodamine Terminator Cycle Sequencing Ready Reaction (Applied Biosystems) and an automatic sequencer ABI PRISM Model 3130 (Applied Biosystems). For the species identification, we compared the sequences obtained with reference sequences of dogs and wolves obtained in previous studies (Vilà et al. 1999; Randi et al. 2000; Pilot et al. 2010) and reference sequences deposited in the GenBank databases (http://www.ncbi.nlm.nih.gov/) using the BLAST 2.0 algorithm (http://www.ncbi.nlm.nih.gov/BLAST/).
We followed the method described in Seddon (2005) for the sex determination of the samples. By using the PCR technique, we amplified two specific canine markers: the DBX intron6 (249 bp), which identifies the X chromosome in females and males, and the DBY intron7 (118 bp) for the Y chromosome in males. To verify the success of the DNA amplification, we conducted an electrophoretic migration of the amplified product in 1.5% agarose gels. Males were identified by the presence of two bands corresponding to the X and Y chromosomes, while females only presented the band of the chromosome X. All samples were processed in duplicate to cope with the low quantity and quality of DNA extracted from the faecal samples. When bands were faint or fuzzy and thus the identification by agarose gel was doubtful, samples were genotyped with two replicates using an automatic sequencer (ABI PRISM 3130, Applied Biosystems). We used the program GENEMAPPER version 4.0 (Applied Biosystems) to detect the fragments corresponding to the X and Y chromosomes.
Diet analysis
Wolf diet was determined by identifying guard hairs as well as bone remains in the scats. Since cuticle patterns vary between species (Teerink 1991; Barja et al. 2021), we prepared cuticle slides using hair spray as medium (Barja 2009a). Then, we used a microscope (Olympus 400X) to compare the cuticle patterns found in the samples with those in reference manuals (Teerink 1991; Barja et al. 2021) and with reference hairs collected in the study area (Barja 2009a; Barja et al. 2021). Bone remains were identified using dichotomic keys and by comparing with a reference collection.
Elemental nitrogen analysis
Since total nitrogen content of faeces seems to be a good indicator of protein ingestion and hence, diet quality (Aldezabal et al. 1993; Robbins et al. 2005; Baldwin and Bender 2009; Navarro-Castilla et al. 2023), we analysed nitrogen contents of wolf faeces to evaluate nutritional condition (protein intake) of individuals.
Frozen faecal samples were dried in the laboratory oven at 90ºC until they exhibited a constant weight, which took 24 hours. Following, using liquid nitrogen, we pulverized the samples in a mortar and 1g of each pulverized sample was stored and later analysed at the Research Support Central Services (SCAI - University of Málaga, Spain). Total faecal nitrogen was determined by carrying out the elementary chemical analysis on a PERKIN-ELMER 2400 CHN elemental analyser, using the classical Pregl-Dumas method according to Sergiel et al. 2020. Faecal nitrogen content is presented as g N/100 g dry faeces.
Data analysis
The composition of the diet was expressed in terms of frequency of occurrence (the total number of times that each prey species appeared in faecal samples) and the percentage of consumed biomass. Since the energy provided by each prey is different depending on its weight, the consumed biomass of each prey species was estimated by multiplying its frequency of occurrence by that prey mean weight, considering both adult and juvenile weights (C. capreolus 15.8 kg; C. elaphus 57.5 kg; S. scrofa 48.5 kg; C. aegagrus 15.7 kg; O. aries 16.8 kg; E. africanus asinus 100.0 kg; S.scrofa domesticus 80.0 kg; O. cuniculus 1.2 kg) (Urios 1995; Llaneza et al. 1996; Blanco 1998; Mateos-Quesada, 2002; Soffiantini et al. 2006; Barja 2009a; Meriggi et al. 2015). When species identification was not possible, consumed biomass was estimated using the mean weight of the corresponding member species of that group (Unidentified ungulate: 40.6 kg, mean between C. capreolus, C. elaphus and S. scrofa mean weights).
To analyse the relationship between prey consumption and wild ungulate availability in each wolf breeding group, we recorded the total number of times that each prey species appeared in faecal samples (ObsF). Since the number of scats collected in each group was different, ObsF were corrected using the following equation:
(\({ObsF}^{*}=ObsF*{I}_{c}\))
where ObsF* is the corrected frequency and Ic is the correction index:
( \({I}_{c}=\frac{{N}_{m}}{N}\))
Ic index was calculated by dividing the number of faecal samples collected in each breeding group (Nm) by the mean number of faecal samples collected in all groups (N).
To estimate wild ungulate prey availability, we calculated the expected frequencies:
\(\%Esp=\frac{{D}_{i}·100}{{D}_{t}}\) where Di corresponded to each prey species availability in each group and Dt the total ungulate availability in each group. Expected frequencies (ExpF) were calculated as: \(ExpF=\frac{ Ob{s. F}^{*}·\%Exp}{100}\)
Jacob's (1974) prey selection index was used to calculate wolf ungulate preferences: \(D=\frac{r-p}{r+p-2pr}\) where r is the contribution of each ungulate species in relation to the total number of prey and p is the abundance of that prey in that study area. D can take values between from − 1 to + 1, -1 implies a negative selection, 0 no selection and + 1 positive selection.
Moreover, we used Levin’s (1968) index (L) to estimate trophic niche breadth:
\(B=\frac{ 1}{{Pi}^{2}}\) where Pi is the contribution of each prey to total biomass ingested in each wolf group. B values next to 1 indicate a highly specialised diet whereas larger values indicate an opportunistic trophic behaviour.
To analyse diet similarity, we used Pianka (1973) index:
\({\alpha }_{gz}=\left({P}_{g}·{P}_{z}\right)·\left[\right({P}_{g}{)}^{2}·({P}_{z}{)}^{2}{]}^{-\text{0,5}}\) where αpz would be the similarity between wolf breeding groups in Galicia and Zamora, Pg the contribution of one prey species to the total biomass ingested in Galicia and Pv the contribution of one prey species to the total biomass ingested in Zamora. This index was calculated by comparing all study areas between them. Values close to 0 indicate the minimum niche overlap.
To compare wild ungulate abundance within zones, we used a t-test. Since data was not normal distributed, we used non-parametric Chi-square (χ2) tests to check the independence between the observed and expected prey presence in diet depending on the wolf breeding group, age and social status. To analyse differences in the type of prey consumed between different groups (zone, wolf breeding group, social status, sex and age) we used contingency table analysis. We used Pearson χ2 when the table had less than 20% of the expected frequencies > 5. In contingency tables where more than 20% of the expected frequencies were < 5, the Monte Carlo’s exact test was used. In 2x2 tables where df = 1, we used Yates’s continuity correction. In 2x2 tables we used Fisher exact test and χ2 de Pearson for the rest of the cases.
To analyse differences in total faecal nitrogen (%) between breeding groups, sexes, dominant and subordinate individuals we performed non-parametric Kruskal-Wallis and Mann-Whitney tests because data did not fit normal distribution, not even transformed. We only had data of faecal nitrogen contents for Galicia and Valladolid wolf breeding groups because we did not receive any funding to carry out the analysis in Zamora’s wolf groups.
Results were considered significant at α < 0.05. Data are represented as mean ± standard error (SE). The software used to perform the statistical analysis was SPSS 23.0 for Windows (SPSS Inc, Chicago, IL, USA).
Animal Ethics
The study methodology was strictly non-invasive and this research was performed in compliance with all applicable laws and rules set forth by the Spanish Government.