Aim: Migratory animals regularly move between often distant breeding and non-breeding ranges. Knowledge about how these ranges are linked by movements of individuals from different populations is crucial for unravelling temporal variability in population spatial structuring and for identifying environmental drivers of population dynamics acting at different spatio-temporal scales. We performed a large-scale individual-based migration tracking study of the lesser kestrel (Falco naumanni), an Afro-Palearctic migratory raptor, to determine the patterns of migratory connectivity of European breeding populations. Location: Europe, Africa. Methods: Migration data were recorded using different devices (geolocators, Argos PTTs, GPS loggers) from 87 individuals breeding in the three core European populations, located in the Iberian, Italian and Balkan peninsulas. We estimated connectivity by the Mantel correlation coefficient (rM), and computed both the degree of separation between the non-breeding areas of individuals from the same population (i.e. the population spread) and the relative size of the non-breeding range (i.e. the non-breeding range spread). Results: European lesser kestrels migrated on a broad-front across the Mediterranean Sea and Sahara Desert, with different populations using different routes. Iberian birds migrated to western Sahel (Senegal, Mauritania, western Mali), Balkan birds migrated chiefly to central-eastern Sahel (Niger, Nigeria, Chad), whereas Italian ones spread from eastern Mali to Nigeria. Spatial differentiation of non-breeding areas led to a strong migratory connectivity (rM = 0.58), associated with a relatively high population (637 km) and non-breeding range (1149 km) spread. Main conclusions: Our comprehensive analysis of the non-breeding distribution of European lesser kestrel populations revealed a strong migratory connectivity, a rare occurrence in long-distance avian migrants. The geographic conformation of the species' breeding and non-breeding ranges, together with broad-front migration across ecological barriers, promoted the differentiation of migratory routes and non-breeding areas. Strong connectivity could then arise because of both high population spread and broad non-breeding range.

The consequences of converting tropical rainforest to oil palm are well-documented, but the impacts of oil palm conversion on natural savanna landscapes remain little-known. Natural savannas in South America have been identified as fertile grounds for future oil palm expansion, partly due to perceived low biodiversity impacts relative to forest systems. We quantify the impacts of oil palm conversion for bird communities inhabiting natural savannas in the Colombian Llanos. Bird species richness and abundance were significantly reduced at sampling points in oil palm relative to adjacent natural savannas, with marked concomitant shifts in community composition. Aquatic, forest and grassland specialists all showed significantly lower abundances in oil palm habitat relative to savanna, as did migratory species and carnivorous/insectivorous dietary guilds. In both habitats, point-level species richness and total bird abundance increased with proximity to remnant forest patches. Within savanna, total bird abundance also decreased with proximity to adjacent oil palm, though species richness was not significantly affected. Within oil palm, point-level abundance increased with proximity to savanna, suggesting that communities in both habitats may be impacted by reciprocal edge or spill-over effects. Point-level abundance in oil palm also increased significantly with the presence of an invasive vine ‘kudzu’, a species introduced in some plantations as a soil protection measure, although species richness was not affected. Our work underlines the need for careful monitoring of further large-scale agro-industrial conversion in the Llanos. In particular, we suggest greater efforts to direct oil palm expansion towards already degraded lands (e.g. improved grassland areas currently used for intensive cattle grazing) to ensure remaining natural savannas are spared

External factors such as geography and weather strongly affect bird migration influencing daily travel schedules and flight speeds. For strictly thermal-soaring migrants, weather explains most seasonal and regional differences in speed. Flight generalists, which alternate between soaring and flapping flight, are expected to be less dependent on weather, and daily travel schedules are likely to be strongly influenced by geography and internal factors such as sex. We GPS-tracked the migration of 70 lesser kestrels (Falco naumanni) to estimate the relative importance of external factors (wind, geography), internal factors (sex) and season, and the extent to which they explain variation in travel speed, distance, and duration. Our results show that geography and tailwind are important factors in explaining variation in daily travel schedules and speeds. We found that wind explained most of the seasonal differences in travel speed. In both seasons, lesser kestrels sprinted across ecological barriers and frequently migrated during the day and night. Conversely, they travelled at a slower pace and mainly during the day over non-barriers. Our results highlighted that external factors far outweighed internal factors and season in explaining variation in migratory behaviour of a flight generalist, despite its ability to switch between flight modes.

Migratory birds often spend a large proportion of their annual cycle in their non-breeding areas. While some species are sedentary after arriving in their non-breeding areas, others engage in itinerary or nomadic movements. Using GPS-tracking we aimed to investigate the little known non-breeding movements of the lesser kestrel Falco namumanni in sub-Saharan Africa. We segment non-breeding tracks (n = 78 tracks by 54 individuals) into staging events (131± 25 days), itinerant movements between staging sites (11 ± 10 days), and non-directed exploratory movements (6 ± 5 days). We then describe timing and directionality of itinerary movements by male and female kestrels, and explore shifts in habitat use throughout the non-breeding season. Regardless of sex, lesser kestrels spent on average 89% of the non-breeding season at 2-3 staging sites in West Africa. Upon arrival at the end of September, kestrels used disparate sites throughout the Sahel. By December, however, they congregated into two clearly delineated areas in Senegal and along the Malian-Mauritanian border. The birds stayed longer and showed greater daily activity in the latter areas, which were situated closer to wetlands compared to their first and intermediate ones. While 20 of 24 individuals tracked across multiple annual cycles showed consistent staging sites, a few switched between the Senegal and Mauritanian/Malian staging clusters. These results show that rather than tracking suitable climatic conditions southward, lesser kestrels survive the West African dry season by moving into wetland areas within the Sahelian climatic belt. Our findings match reports of lesser kestrel super-roosts in West Africa and stress the conservation value of wetlands along the Senegal River and the Inner Niger Delta in West Africa for the Spanish lesser kestrel population. These areas host 81% of lesser kestrels during 20% of their annual cycle.

European ringing records reveal that the Common Kestrel (Falco tinnunculus) shows a continuum of migratory behavior, ranging from long- and medium-distance migratory populations in northern and central Europe to resident ones in southern and western Europe, with partially migratory populations in between. In the southern half of Spain, the species is described as resident. With high resolution solar GPS data loggers on 17 adult Common Kestrels (seven males and 10 females) in two study areas in southwestern Spain, we documented four cases of long-range movements between 2018 and 2019. Three females engaged in these journeys during the nonbreeding season, but there was no relationship between the occurrence of these events and the sex of tagged kestrels. We found a strong individual variation in duration, direction, distance, and timing of travel. Our results suggest that the Common Kestrel behaves as a partial migrant in southwestern Spain. The observed variation in movement patterns reflects the flexibility documented in dispersive migrants. To our knowledge, this is the first study to report Common Kestrel seasonal movements using high resolution movement data.