INTRODUCTION

Long-distance migratory birds fly hundreds to thousands of kilometers across continents and often undertake nonstop flights that cross large ecological barriers and require immense energy. To power such energetically expensive flights during migration, many migratory bird species break up their journeys by stopping over at select sites to rest and refuel (Alerstam et al. 2003, Hedenström 2008). The decision to stop over and the time spent at such sites are influenced by multiple factors including food resources and environmental conditions (Jenni and Schaub 2003, Ottich and Dierschke 2003, Dossman et al., 2016), and also factors intrinsic to the migrant itself such as physiological fitness and endogenous time programs (Goymann et al. 2010, Merlin and Liedvogel 2019, Bounas et al. 2023). The stopover sites, therefore, are critical for migratory birds for ensuring optimal migration because the quality of such sites dictates the overall speed of migration and subsequent survival and reproduction (Newton 2006, Alerstam 2011, Duijns et al. 2017, Gómez et al. 2017).

For many migratory raptors, particularly obligate soaring birds (e.g., vultures, eagles, kites, and large hawks), which are dependent on thermal updrafts, the flights over open oceans and seas is challenging (Alerstam 2001, Bildstein 2006). Therefore, in order to avoid such crossings, these birds undertake detours to cross along narrow land-bridges or over straits, or they adopt an island-hopping strategy (e.g., Short-toed Snake-Eagle [Circaetus gallicus]; Mellone et al. 2011, Panuccio et al. 2012; Oriental Honey-Buzzard [Pernis ptilorhynchus]; Yamaguchi et al. 2008). Conversely, migrating raptors that alternate between soaring and flapping flight (e.g., Ospreys [Pandion haliaetus], honey-buzzards, harriers, small hawks, and falcons) perform sea crossings of hundreds of kilometers (Agostini et al. 2005, Bildstein 2006), and in some exceptional cases thousands of kilometers, such as Eleonora's Falcon (Falco eleonorae; Gschweng et al. 2008) and Amur Falcon (Falco amurensis; Dixon et al. 2011, Meyburg et al. 2017). Therefore, for birds crossing large ecological barriers, building sufficient energy reserves at stopover sites is crucial (Gschweng et al. 2008, Javed et al. 2012, Monti et al. 2018).

Geographical constraints often cause populations of migratory birds to converge and concentrate in a few high-quality, resource-rich stopover sites to refuel; thereby such sites act as ecological bottlenecks (Bayly et al. 2016, Cohen et al. 2017, Gómez et al. 2018). The qualities and constraints at these areas may enhance positive or negative carry-over effects, significantly influencing individual fitness and population dynamics (Webster and Marra 2005, Boulet and Norris 2006, Betini et al. 2015). Certain populations of birds during their migration are relatively highly localized to few sites for stopping over, making them vulnerable to habitat changes at local scales, and hunting or poaching. Migratory birds may shift their stopover locations in response to a reduction of stopover habitat quality (Ruffs [Calidris pugnax]; Verkuil et al. 2012), or to human disturbance in the form of hunting (Snow Geese [Anser caerulescens]; Béchet et al. 2003, 2004), and may respond to favourable food availability by prolonging their stopover stay (Great Knots [Calidris tenuirostris]; Chan et al. 2019, and Bar-tailed Godwits [Limosa lapponica]; Conklin et al. 2021 in the Yellow Sea region of China).

Many migratory birds, particularly from the northern Asian breeding grounds, fly along the Central Asian Flyway and enter the Indian subcontinent, either as passage migrants that stop in India to feed but continue on to other locations or as wintering migrants that stay for the entire season. The Amur Falcon is a passage migrant in India, traveling from its breeding grounds in eastern Russia and northeastern Mongolia to northern and eastern China (Ferguson-Lees and Christie 2001), then stopping over in large congregations in Northeast India for 2–3 wk in October and November each year (Ali and Ripley 1978, Bildstein 2006, Naoroji 2006, Kumar 2021, Kaur et al. 2022) while en route to its nonbreeding grounds in southern Africa (Bildstein 2006). A small raptor with males averaging 135 g (range = 97–155 g) and females averaging 148 g (range = 111–188 g; Jenkins 2005), the Amur Falcon is a transcontinental migrant that undertakes a nonstop flight of nearly 6000 km from its stopover sites in Northeast India to the east coast of Africa, a trip that includes the longest overwater flight of any raptor (Kumar 2021). This implies that stopping over at sites in Northeast India is critical for the successful migration of Amur Falcons.

Like other small raptors, Amur Falcons are insectivorous (  Supplemental Material Fig. S1 (JRR-23-49_Supplemental_Material_Fig.S1.jpg)) but occasionally feed on rodents and small birds. Diet data from their nonbreeding grounds in southern Africa were obtained primarily through examination of regurgitated pellets, and insects (Coleoptera, Orthoptera, Isoptera, and Solufigae) were the main prey identified (Kopij 2009, Pietersen and Symes 2010, Alexander and Symes 2016). In India, only anecdotal reports exist of Amur Falcons foraging on swarms of insects such as winged termites and ants (Ali and Ripley 1978, Naoroji 2006). Thus, to determine what drives the large congregations of Amur Falcons during autumn passage in Northeast India, we undertook a detailed diet study through the examination of regurgitated pellets to identify their prey.

METHODS

Study Sites. We carried out this study across two seasons in October–November of 2017 and 2018, during Amur Falcons' stopovers at three major stopover sites in Nagaland State: Pangti, Yaongyimchen, and Hakhizhe in Wokha, Longleng, and Dimapur districts, respectively. Additionally, in 2019, we sampled at two other major sites: Umrangso in Dima Hasao district of Assam and Puching in Tamenglong district of Manipur, both of which adjoin Nagaland (Fig. 1).

Figure 1.

Map showing Amur Falcons' major and minor stopover sites in Nagaland State and neighboring Assam and Manipur States in Northeast India. Major stopover sites are where large congregations of Amur Falcons were observed and were sampled for regurgitated pellets for the diet study.

The study sites, which are located south of the Brahmaputra floodplains, are part of the NorthEast Biogeographic Zone of India (Rodgers et al. 2002), and fall in the Indo-Burma biodiversity hotspot (Myers et al. 2000). The stopover sites, characterized by hilly terrain (300–600 masl), are primarily situated in the Naga Hills of Nagaland and Manipur States. The sites used in this study (Pangti, Yaongyimchen, Hakhizhe, Puching, and Umrangso) are considered major stopover sites, wherein we observed large congregations of falcons (>100,000 per major stopover site) during the October–November study period (Fig. 1).

The stopover sites were primarily near rivers, and notable among them were the Pangti and Umrangso sites, which were located near large reservoirs of the Doyang and Kopili Rivers, respectively. The vegetation at the stopover sites mainly consisted of bamboo forests and secondary forests, a consequence of historical slash-and-burn agricultural practices. However, specific vegetation types varied across sites; for example, at Hakhizhe the falcons roosted on teak (Tectona gran-dis) plantations, whereas at Pangti, they preferred primary forests along with beechwood (Gmelina arborea) plantations, and at Umrangso they used pine (Pinus kesiya) plantations.

Pellet Collection and Examination. At the Amur Falcon stopover sites, we searched the forest floor for freshly regurgitated whole pellets in the morning soon after the falcons left the roost, and we collected the pellets in individual paper bags. We collected pellets at each of the sites within a 3-d period every 15 d across the 2 mo. We then labeled the pellet bags and sun-dried them to remove any moisture. Additionally, to help identify prey remains in the pellets, we opportunistically collected insects in and around the stopover sites for use as reference.

In the lab, we placed individual pellets in a petri dish, and gently segregated them to look for any visible prey remains. We then examined the pellet under a dissecting microscope (10–40× optical zoom), and counted the individual prey items using the minimum number of individuals approach following Marti et al. (2007). We identified prey remains to their taxonomic order, and then to the lowest taxonomic level possible, based on the reference insect samples collected and relevant field guides (Roonwal and Chhotani 1989, Chhotani 1997, Bose 1999, Krishna et al. 2013, Richards and Davies 2013). The remains that could not be identified were classified as “unknown.”

We followed Alexander and Symes (2016) to calculate the frequency of occurrence (%) of each prey item in a pellet, which is the number of items of each taxon divided by the total number of all items identified in that pellet. We then averaged (±SD) the percentage for each taxon per pellet across all pellets sampled from each study site.

For the most commonly occurring prey item recorded in the pellets we performed statistical tests to examine differences in their frequency of occurrence across years and across sites using the nonparametric Mann-Whitney U-test, and the Kruskal and Wallis test followed with a post-hoc Dunn's Ztest. All statistical analyses were performed in program R (R Core Team 2021).

Because small raptors such as falcons generally regurgitate one to two pellets each morning (Bond 1936, Duke et al. 1976), we also estimated the termite biomass and numbers of termites for one and two regurgitated pellets using the average number of termite heads counted in 200 randomly selected pellets, for an average 15-d stopover duration (Kumar 2021, Kaur et al. 2022) of Amur Falcons in Northeast India.

RESULTS

Pellet Collection and Examination. We examined 1200 pellets collected from the three major stopover sites in Nagaland in 2017 and 2018 (200 pellets/site/yr). Overall, we found pellets comprised only insect prey belonging to five taxonomic orders: Orthoptera, Isoptera, Hemiptera, Coleoptera, and Hymenoptera. Remains of Isoptera (termites) were observed in 1107 pellets (92%) examined, and in 66% of these, it was the only prey item recorded. Across 2 yr of the study and across sites, the average percentage of termites in the pellets was 88% (Table 1). Further, based on the main identifying characteristics for termite families, genera, and species (in this case presence of a tooth on the left mandible and tongue-shaped labrum with a pointed tip, along with maximum head width; Krishna et al. 2013), we identified two termite species (Odontotermes feae and O. horni; the fungus-growing subfamily Macrotermitinae, family Termitidae) in the diet of Amur Falcons.

Table 1.

Average percent of prey items in Amur Falcon regurgitated pellets across 2 yr (2017 and 2018) and at three major stopover sites (H = Hakhizhe, P = Pangti, and Y = Yaongyimchen) in Nagaland, Northeast India.

Considering all sites together, the average percentages of termites in pellets in 2017 (89% ± 26.37) and 2018 (87% ± 30.97) did not differ significantly (U = 175,248, P = 0.35). However, it was significantly different across the three stopover sites in 2017 (χ² = 74.97, df = 2, P < 0.05) and in 2018 (χ² = 142.36, df = 2, P < 0.05; Fig. 2). Further, the post-hoc Dunn's test revealed that the percentage of termites in the diet at Hakhizhe differed significantly from that at Pangti (Dunn's test, P < 0.05) and at Yaongyimchen (Dunn's test, P < 0.05). The Yaongyimchen site had the highest averaged percentage of termites (98%) across years.

Figure 2.

Average percent of Isoptera in the diet of Amur Falcons across the three stopover sites and across 2 yr.

At Hakhizhe, the average percentage of termites in the pellets was 83% ± 28 in 2017 but only 63% ± 44 in 2018 (U = 23,128, P < 0.05). At Pangti (85% ± 33 in 2017 and 99% ± 25 in 2018) and at Yaongyimchen (98% ± 6 in 2017 and 98% ± 8 in 2018) the percentages of termites in the pellets did not differ between years (U = 37,630, P = 0.11 and U = 19,484, P = 0.52, respectively).

Like the sites in Nagaland, at Umrangso in Assam, Isoptera was the most dominant prey remain, with 73% frequency of occurrence followed by Hemiptera (19%), Coleoptera (7%), and Orthoptera (<1%). Similarly, at Puching site in Manipur, Isoptera again made up the majority diet (64%), followed by Hemiptera (30%), Coleoptera (4%), and Orthoptera (1%). However, unlike in Nagaland, Hymenoptera was absent in the pellets collected from both Umrangso and Puching.

Of 200 randomly selected pellets (from the 1200 total study pellets), 60% contained an average of 64 termite heads (range: 14–177, n = 120). We weighed 50 alates and used their average mass (0.07 g) to calculate the mass of 64 alates in one pellet (4.48 g). We then calculated the mass of termites that one falcon could consume over a 15-d period; if the falcon produced one pellet/d the mass consumed would be 67.2 g, and if the falcon produced two pellets/d the mass consumed would be 134.4 g. Consequently, for a 15-d stopover in Northeast India, a million Amur Falcons could consume somewhere between 67 and 134 metric tons of termite biomass. This equated to about 0.96 to 1.92 billion termites (Table 2).

Table 2.

Termite alate biomass and number of alates consumed by Amur Falcons that stopped over for an average of 15 d at three sites in Nagaland, Northeast India.

DISCUSSION

Birds on migration are reported to meet their energetic demands by making stopovers to replenish energy reserves to power their onward flights (Alerstam et al. 2003, Newton 2006, Hedenström 2008). Stopping over at resource-rich sites and those that are predictable in terms of food availability is therefore likely an important strategy adopted by migrating birds. This appears to be the case with Amur Falcons on migration, with entire populations numbering in the hundreds of thousands on their southbound passage. They funnel into Northeast India to stop over for about 2 wk or more, and as a result the region, particularly Nagaland is now popularly known as the “falcon capital of the world” (Williams 2013). The findings of our study clearly indicate that Amur Falcons stop over at select sites in Northeast India not only to rest, but more importantly, to refuel—a strategy adopted by the falcons before undertaking their nonstop leg of migration ahead.

Many long-distance migratory raptors adopt a behavioral strategy by synchronizing their arrival at stopover sites to coincide with the availability of rich and plentiful food sources at stopover sites (Gschweng et al. 2008, Symes and Woodborne 2010, Javed et al. 2012). This adaptation allows the migratory birds to undergo a state of hyperphagia to prepare for the migratory flights ahead (Guglielmo 2018). During the surveys to locate stopover sites and to conduct conservation awareness campaigns across Nagaland, we learned from the communities that falcons in the region are called loi meaning “insect-eater,” feeding specifically on alhu or winged-termites (alates). Local people who were once falcon hunters in the region noted that falcons' meat lacked fat at the start of the stopover season. As a result, they preferred to hunt falcons toward the season's end when the birds had accumulated sufficient fat from intensive feeding on insects. This along with our observations on falcons' foraging suggests premigratory fattening on the autumn stopover in Northeast India, and thereby, implicates these stopover sites as critical for the falcons' successful onward migration. Success of the migration is often dependent on stopover sites located adjacent to or just before large ecological barriers (Delingat et al. 2008, Bonter et al. 2009, Bayly et al. 2012, Gómez et al. 2017).

The dietary analysis in our study clearly showed Amur Falcons to be highly insectivorous during their stopover in Northeast India. Outside of their breeding areas, small raptors often show a dietary shift, switching to an insectivorous diet in response to locally abundant available prey (Kok et al. 2000, Palatitz et al. 2009, Bouwman et al. 2012, Samraoui et al. 2022). Similarly, Amur Falcons at their nonbreeding grounds in southern Africa feed largely on Coleoptera, Orthoptera, Isoptera, and Solufigae, and occasionally on rodents and small birds (Kopij 2009, Pietersen and Symes 2010, Alexander and Symes 2016). In Northeast India, the diet of Amur Falcons comprised mostly Isoptera (termites), which fuel the birds' onward migration. Further, it is generally believed that Amur Falcons benefit by preying on co-migrating dragonflies (globe skimmer [Pantala flavescens]) during oceanic crossings (Anderson 2009). However, considering the optimal migration (Alerstam 2011) and the falcons' behavioral strategy of building fuel reserves at stopover sites in Northeast India, we suggest that Amur Falcons may not need to forage much during their migration.

Termites are generally known for their high nutritional properties as they are particularly rich in easily digestible proteins (Kinyuru et al. 2010), contain high amounts of fat (Redford and Dorea 1984, Rumpold and Schlüter 2013), and form an important source of food for many species. In Nagaland, the termite species O. feae and O. horni emerge en masse in October and November with the retreat of the southwest monsoon, an event that coincides with the arrival of Amur Falcons. These two species of termites are members of the family Termitidae, reported to swarm only during a restricted season (Harris 1971), generally in rainy conditions (Mitchell 2008). Roonwal and Verma (1991) reported that O. feae swarm once a year in India, varying geographically from early June to the end of November. During the autumn stopover, we recorded intense swarming episodes of O. feae and O. horni across Northeast India that only occurred during low-light conditions, between 1530 and 1700 H for a maximum of 15–20 min. On days of such swarming episodes, we observed Amur Falcons departing from their roost together in large numbers, followed by milling above the area. Then, as the termites emerged in mass numbers, we witnessed falcons engaging in frenzied feeding. On days of incessant rainfall with no insect emergence, Amur Falcons mostly rested in trees or on powerlines close to their roost sites throughout the day. The absence of such mass termite emergence in April and May in Northeast India may explain why no similar mass gathering of falcons occurs during spring migration.

The variation in percentage occurrence of termites across study sites likely results from spatial and temporal shifts in termite emergence, in addition to availability of other insect prey. Hakhizhe, predominantly a lowland, is part of Brahmaputra floodplains, with wet paddy cultivation, whereas Umrangso is close to open grasslands; thus, these two sites offer high abundance of other insect groups such as Hemiptera, Coleoptera, and Orthoptera, as observed in the dietary analysis. Dietary studies of Amur Falcons during their 5–6-mo nonbreeding season at sites in southern Africa (Kopij 2009, Pietersen and Symes 2010, Alexander and Symes 2016) showed that termites constitute a small part of their diet, with Coleoptera and Orthoptera being primary prey; this diet also varies seasonally and is influenced by rainfall. However, in Northeast India, where the stopover is rather short, termites appear more important than other insect prey.

The low occurrence or absence of other prey items in the diet of Amur Falcons may be attributed to their opportunistic predation strategy. For example, we observed Amur Falcons unsuccessfully chase a White-throated Fantail (Rhipidura albicollis) and feed in a frenzy on mayflies (Ephemeroptera) at Puching. In addition, the complete digestion of certain taxa, specifically soft-bodied prey such as mayflies, can also result in low detection in pellets. Pellet analysis is useful because (a) it is relatively easy to determine the raptors' diet with minimal disturbance (Rosenberg and Cooper 1990, Redpath et al. 2001), (b) remains can be preserved to identify later (Marti 1987), and (c) incorporation of spatial and temporal sampling is possible and can be used to determine variability and temporal shifts in diet (Bakaloudis et al. 2012). However, pellet examination has its biases because it may underestimate the contribution of small vertebrates such as Passeriformes, Rodentia, and Soricomorpha; and small, soft-bodied prey such as Lepidoptera and Diptera (Kopij 2009, Alexander and Symes 2016).

Amur Falcons' tracking of food resources and ability to feed on a wide range of invertebrates across their distributional range likely play a crucial role in maintaining ecological balance. By foraging extensively on swarms of termites in Northeast India, Amur Falcons contribute to controlling termite populations and thereby provide an important ecosystem service. Our estimate of consumption of approximately one to two billion termite alates (∼67–134 metric tons of biomass) by a million Amur Falcons (Williams 2013, Kumar 2021) stopping over for at least 15 d in Northeast India indicates the value of these insectivorous raptors in the region. This further underscores the importance of continuing and strengthening the conservation efforts on behalf of Amur Falcons that were initiated after the release of the Conservation India (2012) report on mass-scale harvesting of falcons in the region. Our results and the study by Bouwman et al. (2012) in southern Africa highlight that a change in the number of Amur Falcons across its distributional range can cause cascading effects with respect to insect and pest populations.

Long-distance migrants have more structured migration schedules, and this inflexibility make them vulnerable in a rapidly changing environment (Klaassen et al. 2012). Climate change influences the phenology of many species, resulting in fluctuations in seasonal resource availability, which thereby can affect the migratory cycles of birds (Baker et al. 2004, Both et al. 2006, 2010, Møller et al. 2008, Saino et al. 2011). Several studies analyzing long-term climate data have highlighted a significant decrease in monsoon rainfall in Northeast India (Kumar et al. 2010, Jain and Kumar 2012), along with a significant increase in temperature (Mondal et al. 2015). Further, the impact of indigenous and historical jhum (shifting) or slash-and-burn agriculture practices in Northeast India on termite populations remains unknown. Any phenological change or fluctuation in the prey base at stopover sites will in turn affect the cycle of Amur Falcon migratory behavior in the region. This is also critical when migratory populations rely on a small number of sites to acquire energy (Bayly et al. 2013, McKinnon et al. 2013). This was evident during our study, in which we observed that over the years Amur Falcons shifted their stopover sites in the region, possibly due to fluctuating local resources, or to human disturbance. Our results highlight that Amur Falcons during their transcontinental migration rely on specific sites to stop over and replenish their energy before undertaking the 3500-km long Arabian Sea crossing. This has important implications for strengthening the conservation efforts across all stopover sites in Northeast India, suggesting that the specific needs of Amur Falcon populations passing through the region may be highly localized.

SUPPLEMENTAL MATERIAL (available online).   Figure S1 (JRR-23-49_Supplemental_Material_Fig.S1.jpg): Amur Falcon getting ready to capture a flying termite. Image credit: Saikat Banik.