Spatial and Temporal Monitoring of North African Turtle Doves Streptopelia turtur arenicola (Hartert, EJO, 1894): First Migrants Arrive Early and Select Nesting Trees next to Foraging Resources while Second Breeders’ Wave Breed around Earlier Nests

,


Introduction
Te Turtle Dove Streptopelia turtur (Linnaeus, 1758) is a long-distance migrant Columbidae that migrates between Europe and Asia as breeding habitats (Mars-September) and Sub-Saharan Africa as wintering grounds between October and February [1][2][3]. In Europe, Streptopelia turtur is mentioned as a breeder in Spain, France, the UK, and Germany in the west [1,4,5], as well as in Turkey, Greece, and Bulgaria in the East [6,7]. In Africa, the species is recorded in North Africa as breeding migrant [8][9][10], while in Sub-Sahara it is considered as winterer [2]. On the other hand, Turtle dove is the remarkable example of a European long-distance migrant bird that has sufered a rapid and severe decline across its western European range (−78% in Great Britain from 1980 to 2020 as well as −70% in the Iberian peninsula, mainly in Spain from 1980 to 2017) [11]. Terefore, it has been ranked as "Vulnerable" globally and "Near Treatened" within the Northern slope of the Mediterranean basin following recent evaluation [12,13]. Potential causes responsible for the species" deterioration include deprivation of breeding sites [14], scarcity in food availability due to intensifcation of farming activities [15][16][17], unsustainable hunting policies [18], and variation in ecological conditions throughout the migration fyways [16,19].
In Morocco, studies of North African Turtle Dove S. t. arenicola were mostly conducted in farmlands, principally olive, orange, and apple orchards [31,33,47]. Tese studies have detailed breeding biology, reproductive success, and menacing factors. However, the selection of nesting trees inside breeding orchards was neglected in Europe [41,48] and North Africa [9,30,47,49]. Tis element is of great importance in clarifying patterns of nesting site selection with regards to orchard characteristics, tree heights, and disturbing factors [49]. If we consider the vulnerable status and low population densities, an understanding of such elements is suggested to improve the management measures in agricultural habitats mostly colonized by this declining game species.
Tis study aimed to (i) map the microdistribution of nests built by North African Turtle Doves inside orange orchards, (ii) analyse breeding parameters counting chronology and success with regard to distribution in monitored orchards, and (iii) compare breeding parameters between earlier and later clutches. Tese elements are suggested to fll the gap concerning the zonation and nesting strategies of Turtle Doves at selected breeding sites [30].

Study Area. Fieldwork was conducted in the Beni
Mellal-Khenifra region, located in the center of Morocco ( Figure 1). Te study area is dominated by various climatic stages linked to altitudinal zonation, from the plains (400 m altitude) to the mountains (up to 1000 m altitude), and this induces a spatial variability of precipitation and temperature. Te rainfall regime in the mountains (northern slopes of the Middle and High Atlas) is Mediterranean with oceanic infuence, with annual precipitation between 550 mm and 700 mm in Azilal and up to 1000 mm in the High Atlas. In contrast, precipitation in the Beni Mellal and Tadla plains is low (around 436 mm). Equally, temperatures vary from 1.1°C in January to 35.7°C in August. However, these climatic conditions are subject to strong interannual variability [50].
To monitor the migratory Doves, we selected one orange grove in Abou Khayma El Bazzaza village, located in the north of Beni Mellal (Figure 1). Te grove is about four hectares with 1,182 trees of Valencia late (Citrus sinensis), placed in lines and separated by 8 meters between trees. Cereals and legumes surround the grove from the south and east, while the west and north are limited by other orange groves. Te irrigation system is installed underground, and the water is opened weekly.

Breeding Chronology.
Monitoring of the area started at the beginning of March (2016), considered the date of arrival for migrant Doves in Morocco and Northwest Africa [8,31]. We examined the dates of frst arrivals to the region, and after breeding season, we noted the date where the last Doves were observed in the area (last departure dates). After the arrival of the frst individuals, we conducted visits to the study orchard weekly to survey breeding activities (during the beginning of April, the orchard was visited twice to identify the frst nesting attempts, while from May on, visits were reduced to once per week due to lower nesting activities). Breeding chronology, counting the construction of nests, eggs laying, hatching, and fedging of chicks, were monitored from the frst week of March to the last week of September. We noted the evolution of nests, eggs, and chicks, as well as the failed ones. Failure factors were noted per visit and nest for each breeding stage.

Spatial Distribution.
To evaluate the spatial distribution of breeders inside the study orchard, we noted nesting trees with specifc numbers and dates. To be more accurate, we used lines and rows of citrus trees (Valencia late) as coordinates, as explained in Figure 1. For each nest found, we sanded the tree trunk and marked the nest number using sandpaper and a permanent marker. Tis method has helped to fnd the position of the nest during the ulterior visits (surveillance of nests from the construction to the success or failure of chicks) easily. We used a selfe stick to photograph the content of the nest and a piece of paper containing the plan of the orange grove to locate each nest with a symbol representing its content. Once the nest is located, it is monitored during each visit, from construction to fedging or loss of clutches. Further, we noted the distance of nests to the cereals and legumes, to the central zone of the orchard (epicenter) and to the periphery of orchards (periphery next to cereals and periphery next to other orange orchards, as clarifed in Figure 1) based on the distance separating nesting-citrus-trees and the targeted zone.

Statistics.
Breeding season was divided into frst and second phases based on a long break in nest construction and egg-laying (hatching and fedging were neglected since the failure factors are susceptible to impacting them). Reproductive rates counting nests (occupied nests/constructed nests), eggs (hatched eggs/laid eggs), and chicks (chicks that survived/hatched chicks) were calculated for the entire season from the frst nest to the last chicks and for each breeding phase. We chequered for normality and homogeneity of variance for all breeding parameters (variables) via the Kolmogorov-Smirnov test. We compared breeding success rates and failure factors (predation, desertion, destruction, and unhatched) among breeding stages (nesting, laying, hatching, and fedging) using ANOVA One way test. Further, we compared breeding parameters and failure factors between the frst and second breeding phases using the T test.
To evaluate the correspondence between breeding stages, nesting (number of nests), laying (number of eggs), hatching (occurred chicks), and fedging (fedged chicks), considered as response variables, and periods-4 weeks per month (April to September), considered as factors, we used principal For spatial distribution, we created the graphical planning of the orchard via Adobe Illustrator 26.0.3 (2022) software [51] (based on map extracted from Goggle Earth). Nests were placed on the nesting-trees in the graphical planning, based on data collected from the feld (the lines and rows of nesting-citrus-trees were used as coordinates). However, the nests were divided into three periods: (i) frst nesting wave (the frst breeders to colonize the orchard after arrival dates); (ii) frst breeding phase (frst massive nesting stage after the colonization of breeding orchard); and (iii) second breeding phase (second massive nest construction after the long break of the frst phase). However, we reinforced the graphical mapping with detrended correspondence analysis (DCA) to demonstrate statistically the distribution of nests inside the orchard as realized currently by Mansouri et al. [8] and Squalli [52] for passerine birds and Columbidae. In our case, nesting sites in orchard counting central zone (the epicenter of the orchard), peripheryorchard (the marginal zones surrounded by other orange orchards), and periphery-cereals (marginal zones surrounded by cereals and legumes (potential foraging resources)) were considered as response variables, while distances of nests (148 nests) toward central zone, peripheryorchard, periphery-cereals, and cereals were considered as dependent variables. For graphical plot, only eigenvalues superior to 1.0 and an axis with a percentage of variance >50% were selected. Tese methods are commonly used to assess the ecological requirements of birds, including dove species [7,12]. All statistical analyses were executed using SPSS 18 [53], while graphs were created by GraphPad Prism 8.3.0 [54]. Results were given as percentage for success rates and sample size and as the mean ± SD for breeding parameters.

Arrival and Departure.
In Beni Mellal province, the frst birds of S. t. arenicola were observed during the third week of March. Tese birds arrived solitary (single and pairs) during the last days of March, while in April, arrivals were in groups of few hundreds and mostly observed in electrical lines in the vicinity of roads principally in rural areas. After breeding season, migratory birds (adults and subadults), were gathered in foraging sites near water resources counting rivers, dams, and irrigation tunnels. Te last migrants were seen at 10 October, which marked the latest departure date in the Beni Mellal area.

Chronology of Breeding Activities.
Breeding chronology of S. t. arenicola in Beni Mellal is summarized in Figure 2. Construction of nests started during the second week of April, and nesting activities continued into the frst week of August. Nesting activities were divided into the following two phases: the frst phase from the frst week of April to the frst week of June with a peak of nesting during the second week of May, and the second phase from the second week of June to the second week of August with a peak during the third week of June. Laying of eggs started during the third week of April, and egg-putting activities sustained to the frst week of August. Te laying phases were divided into two phases; the frst one from the second week of April to the second week of June with a peak of egg-positioning during the third week of May, and the second phase from the third week of June to the second week of August with a peak during the third week of June. Te number of eggs was signifcantly superior during the frst laying stage. Te occurrence of chicks (hatching of eggs) started during the frst week of May, and hatching activities sustained until the third week of August.
Te hatching and fedging activities showed two peaks in the fourth week of May and the third week of June, respectively, and then fuctuations continued until the end of the breeding period. Both phases seem to be unclear in the curves, and this can be explained by modifcations that occur in each nest during the incubation and rearing period, i.e., predation of eggs or broods, desertion due to disturbance, destruction of nests, etc.
Multivariate analysis (PCA) of optimum periods for breeding activities is summarized in Figure 3. Nesting and laying activities were concentrated principally during the frst three weeks of May and the last two weeks of April. Hatching was concentrated between the fourth week of May and the second week of July, while fedging was mainly recorded between the fourth week of July and the frst week of September. Tese periods marked the optimal breeding times for migrant Doves.

Spatial Distribution of Nests. Te distribution of nests inside the breeding orchard is summarized in Figures 4 and 5.
First nests (the frst nests after prenuptial migration between second and third week of April) were placed in the periphery of the orchard surrounded by cereal farms. During the frst nesting phase (after the installation of the earliest nests), breeders occupied the nesting trees without any oriented selection (nests were documented in the entire orchard). During the second phase, nests were constructed in gregarious forms next to nests of the frst breeding phase. Further, three support trees used during the frst phase were reoccupied during the second breeding phase. Table 1. Among the 148 monitored nests, only 9.45% were deserted during the nesting stage. Among 134 occupied nests (eggs), 34.32% did not achieve the hatching stage. Among the 261 counted eggs, only 153 (58.62%) have succeeded to achieve the hatching stage, while 95 were predated, deserted, destructed, and unhatched. Te loss rate in the fedging stage was limited; among recorded 166 chicks, 153 fedged successfully with a loss rate of 7.83%.

Reproductive Rates. Te total reproductive success of the North African Turtle Dove at Beni Mellal is summarized in
Failure factors were variable (DF � 2, F � 33.960, P < 0.001). Nest desertion caused the highest loss of S.t. arenicola clutches (14 nests, 42 eggs, and 8 chicks), followed by predation (35 eggs and 4 chicks), and destruction (10 eggs). Nest desertion was caused principally by anthropic activities counting, tree pruning, fruit harvesting, irrigation, pesticide use, and hunting, which were applied in coincidence with the breeding activities. Te main predators observed in the monitored orchard were reptiles mostly Montpellier snake Malpolon monspessulanus (Hermann, 1804) and Horseshoe whip snake Hemorrhois hippocrepis (Linnaeus, 1758) documented on nesting-trees, as well as raptors counting, Common kestrel Falco tinnunculus, Peregrine falcon Falco peregrinus, Black-winged kite Elanus caeruleus, and Barn owl Tyto alba. Te comparison of reproductive success parameters and failure factors between breeding phases is summarized in Table 2. For breeding success, only fedging rates difered between the two breeding phases, while laying, hatching, and nesting were similar. Equally, failure factors (desertion, predation, and destruction) were similar between the frst and second breeding phases.

Discussion
Tis study highlighted the temporal and spatial microdistribution of breeding North African Turtle Doves Streptopelia turtur arenicola at Beni Mellal's irrigated perimeter (Morocco). Our central objectives were to provide detailed data on the chronology of breeding activities and the distribution of nests inside the breeding orchard. We obtained the frst data mapping distribution of nests in an occupied orchard and the time evolution of breeding activities. Tese fndings are of great importance for mapping zones of high breeding rates and then orienting well-adapted conservation actions to protect this threatened game in Morocco and the entire southern slope of the Mediterranean basin.
Our study documented the breeding activities of Doves, which confrms the vital importance of Beni Mellal area as a breeding and stopover area for Moroccan (breeders) and European (migrants) populations of Turtle Doves, respectively [8,32,46]. In our case, during spring, the frst Doves were witnessed in the Beni Mellal province during the third week of March, which is in agreement with results cited by Vaurie [20] in the same zone (last decade of March). However, these dates are earlier when compared with European migratory Doves (S. t. turtur) observed on 25 April, in the Beni Mellal area and Moroccan breeders (S. t. arenicola) in high altitude zones [32]. In high altitudes, the low temperatures and high precipitations push Doves to delay their entire breeding chronology counting arrival dates to avoid the abortion of their clutches [9]. In contrast, departure dates (mid-October) were similar between Beni Mellal and other highlands in Morocco [9], in which Doves migrate on 13 th October. Further, [8] recorded currently International Journal of Zoology many wintering Doves in Morocco, and this is suggested to modify the phenological status of the species in entire North Africa.
Our study revealed that the breeding season of S.t. arenicola in Beni Mellal province is divided into earlier and tardive clutches, which is in contradiction with previous studies conducted in other Moroccan [30,32,33,49] and Algerian regions [55] on the southern slope of the Mediterranean, as well as with prior studies conducted in Spain [56] and Britain [14,34] on the European side. Te abundance of cereals and other cultivated seeds around breeding orchards is strongly suggested to encourage the two breeding clutches in S. t. arenicola at Beni Mellal [17]. Further, in Beni Mellal, breeding activities started with the nest initiation recorded in the second week of April, followed by laying of eggs during the third week of April, hatching during frst week of May, and fedging during fourth week of May. Similar results were recorded in the adjacent areas, in Tadla and Midelt which are only 30 and 180 km from Beni Mellal, respectively [9,32]. However, these breeding dates are markedly late when compared with sublatitudinal breeding habitats in Taroudant located 400 km to the South of Morocco [33], where breeding activities started in March, and earlier when compared with Northern breeding habitats in Spain [56] and Britain [34] where breeding season starts in mid-April. Tese diferences may indicate a suggested efect of latitudinal gradient as recorded in many Western Palearctic birds counting the European Turtle doves Streptopelia turtur turtur that take nearly 17 days between North African stopovers (as low latitudinal limits) and European breeding grounds (as higher latitudinal limits in the Northern hemisphere) [41,41]. However, more investigations are needed to confrm this issue.
Tis study highlighted the distribution of nests inside breeding orchards, which is the frst of its kind. First nesting sites were selected in marginal trees placed near cereals and other cultivated seeds, and this indicates the crucial role of foraging resources in the selection of breeding sites and habitats [9,17]. Te selection of breeding trees near foraging seeds is suggested to support breeding pairs and their nestlings during the breeding season as mentioned currently by Mansouri et al. [9]. Nests of the second breeding phase were constructed in gregarious forms and nesting trees were selected next to nests of the frst breeding phase. As a potential explanation, we suggest that the frst breeders prospect the selected orchards for the security of nests and availability of forage via their earlier nests, while during the second wave of breeders (second breeding phase), Doves nest intensively near the trees selected by the prospectors (frst wave of breeding phase) based on the security and food availability ofered for frst nests. However, this issue needs a specifc investigation, and its results are suggested to classify nesting-nucleus where the density of nests is higher as the case of many gregarious and social birds counting greater famingos (Phoenicopterus roseus) [57,58] and the Eurasian Coot Fulica atra [59] that colonise secure sites and other nesting sites of lower density.
According to our annual breeding success evaluation, the breeding success rates of Turtle doves at Beni Mellal province were medium during all breeding phases. In total,   92.17% of chicks have survived from nesting to fedging. Tese results are highly close to those cited in apple orchards at Midelt [32], in olives at Taroudant [33], and in palms at Biskra in Algeria [60]. Despite the availability of food resources around orchard and water inside it, Turtle dove clutches were regularly disturbed.

Conclusion
Tis study ofers new data on the temporal and spatial distribution of the North African Turtle dove subspecies (S.    In summary, frst breeders arrived early to nesting sites and the breeding season was divided into the frst phase from the frst week of April to the frst week of June (earlier), and the second breeding phase between the second week of June and the second week of August. Te frst breeders prospect breeding orchards and select nesting trees close to foraging resources, while during the second breeding phase nests were selected in gregarious forms close to those of the frst phase. However, despite the abundance of foraging resources and breeding requirements, breeding success was lower due to human disturbance, natural enemies, and abortion of eggs. Tese data are of great importance for comparative research concerning the microdistribution of the vulnerable Doves in breeding sites, as well as for conservation actions at orchard-scale via the reduction of human activities in most populated areas of breeding groves. Further, the installation of the second breeding phase between June and September needs more investigations, principally the impact of hunting activities from July to September, on breeders that realize sorties between nests and foraging habitats.

Data Availability
Te data used to support the fndings of this study are included in the article.

Conflicts of Interest
Te authors declare they have no conficts of interests.