Apparent absence of avian malaria and malaria-like parasites in northern blue-footed boobies breeding on Isla Isabel

Haemosporidian parasites are common in birds but are seldom reported in seabirds. The absence of vectors or genetic resistance to infection have been proposed to explain this pattern. However, screening of blood parasites in many seabirds has been done only by visual inspection of blood smears, which can miss low-intensity infections, and molecular detection of blood parasites must be supported by detection in blood smears to confirm the presence of haemosporidians and avoid false positive cases. Here, we tested for the presence of blood parasites of the genera Plasmodium, Haemoproteus and Leucocytozoon, combining inspection of blood smears and PCR-based detection methods in a highly philopatric colony of blue-footed boobies (Sula nebouxii) in the Tropical North Pacific. Our results indicate that adults in this colony are likely free of these blood parasites, probably due to unsuitable conditions for insect vectors in booby breeding sites, although potential genetic resistance of blue-footed boobies to infection deserves examination. Apparent absence of blood parasites in Isla Isabel boobies indirectly adds to the growing evidence of variation in parasite infections among avian host species that coexist locally.

, has been estimated to be 16% (n = 251 frigatebirds 19 ), reinforcing the idea that there are suitable insect vectors of haemoparasites on the island.

Results
No blood parasites were visually detected in 64 blood smears from 33 male and 31 female blue-footed boobies of ages 7 to 13 years that were caring for a clutch or brood, and none of the 64 samples tested molecularly showed evidence of parasite DNA amplification.

Discussion
Failure to find haemosporidian parasites of the genera Plasmodium, Haemoproteus and Leucocytozoon, using microscopic screening of blood smears and a broadly used PCR-based detection method suggests that breeding male and female blue-footed boobies on Isla Isabel are likely free of these blood parasites. Our results are consistent with Clark and Swinehart's (1969) 20 failure to find these parasites in blood smears from 19 blue-footed boobies of Mexico (the number of samples per site is not specified, but the samples were taken on different islands, including Isla Isabel, Todos Santos, Cedros, Cabo San Lucas, San José del Cabo, Cerralvo, Isla Partida, Espíritu Santo, Islas Marietas, Socorro, San Benito and San Martin). These findings are further evidence of the scarcity of blood parasites in seabirds, which is often attributed to unsuitable conditions for vectors in seabird habitats 5,21 .
Haemoproteus parasites were detected in Isla Isabel frigatebirds in 1999-2001 (prevalence ranging from 16 to 54% 19,22 ). Assuming they were still present in 2019, lack of blood parasites in our sample implies that presence of parasite infections varies among Isla Isabel's seabird species.
The difference between boobies and frigatebirds in the prevalence of blood parasites could be due to spatial variation in the abundance and activity of vectors within the island 23,24 . Infected frigatebirds were sampled in www.nature.com/scientificreports/ the southwestern end of the island, in a shady, vegetated area with large trees 40 m from a concrete building where rainwater accumulates, and 150 m from the only pool of alkaline water on the island 19,25 . This area is protected from the wind and suitable for development and proliferation of some potential vectors 26 . Boobies breed mostly ~ 1.0-1.5 km away, at the wind-exposed northeastern end of the island, where we sampled them. The latter area is mainly covered by stunted garlic pear trees recurrently damaged by hurricanes 27,28 , and the hot and dry conditions there may limit the abundance and activity of potential vectors of blood parasites 6,23 , explaining the absence of blood parasites in our sample. In addition, the difference between boobies and frigatebirds in the prevalence of Haemoproteus parasites could be due to differences in the sampling years if parasites found two decades ago are no longer found in either species now. Alternatively, the absence of blood parasites in blue-footed boobies despite the presence of Haemoproteus parasites in frigatebirds, could be due to strong specificity between the parasite or the vector and their vertebrate host 29,30 . The blood parasites detected in Isla Isabel frigatebirds belong to the morphospecies Haemoproteus iwa 19 , which is highly prevalent on frigatebirds and potentially specializes in parasitizing them 18,22,30 . H. iwa is transmitted by louse flies (Hippoboscidae) 30 . These blood-sucking insects dwell mainly on their vertebrate hosts and present higher host specificity than mosquitoes, Culicoides and black flies. However, specificity of louse flies may be low and they could potentially transmit Haemoproteus parasites to Isla Isabel boobies, since louse flies are often reported on boobies 31,32 and have been anecdotally observed during the manipulation of blue-footed boobies on Isla Isabel (authors per. obs. 33 ). Then, strong host specificity of the parasite not the vector could preclude transmission of Haemoproteus parasites to Isla Isabel boobies.
Haemosporidians have earlier been detected in southern populations of blue-footed boobies 14,18 but these observations are under debate 34 and these boobies could be free of blood parasites along their geographic range. Galapagos boobies were found to be infected by Haemoproteus parasites using PCR 14 34 . Further confirmatory PCR essays of boobies' samples tested negative for avian haemosporidians and potential cross-contaminations could not be ruled out since blood smears were not collected in parallel with blood for molecular screening, which is required to confirm the presence of haemosporidians and discard such false positive cases 9,34 . Lee-Cruz et al. ' s detection (2016) 14 of Haemoproteus spp. by PCR of individuals with negative blood smears was attributed to low levels of parasitaemia or detection of free DNA rather than viable parasites, but the discrepancy could be an indicator of potential PCR contamination 34 . Only a parasite tentatively identified as Leucocytozoon spp. was detected using blood smear screening 14 . Thus, further analyses of blue-footed boobies of the Galapagos and further south (e.g., Peru) are needed to confirm infection by avian haemosporidians.
The absence of blood parasites in boobies despite previous detection of potential vectors in the area could also be explained by overly short exposure to parasitic infections, physiological incompatibility with haemoparasites that prevents them from completing their life cycles, or high immunological capacities 8,9 . Moreover, genetic differences between blue-footed booby populations 35 could cause differences in immune capacity and prevalence of blood parasites. Studies comparing immunocompetence among populations of the same species are still scarce, but data for fish 36 , birds and mammals 37,38 suggest that this possibility deserves examination.
Importantly, high and rapid mortality after blood parasite infection may occur in naïve avian host populations 39,40 , drastically reducing the prevalence of parasites in sampled birds. However, it is unlikely that this explains the apparent absence of blood parasites in the Isla Isabel boobies, since haemosporidians were detected on the island twenty years earlier 19 . Similarly, it is unlikely that a sample of 64 birds in a single year yielded a biased estimate of blood parasite prevalence in Isla Isabel boobies. Firstly, sample sizes above 15 individuals are expected to produce robust prevalence estimates 41 . Secondly, although environmentally driven inter-annual variation in the prevalence of blood parasites is common 42,43 , antagonistic interactions between bird hosts and blood parasites tend to be stable over time 44 . Nevertheless, additional screening of birds in other years is desirable to fully confirm the absence of blood parasites in this booby colony.
In conclusion, we failed to find evidence of blood parasites in one of the largest colonies of blue-footed boobies of the North Pacific coast, after microscopic examination of blood smears and state-of-the-art molecular analysis for detection of avian blood parasites 9,45 . Apparent absence of blood parasites in Isla Isabel boobies indirectly adds to the growing evidence of variation in parasite infections among avian host species that coexist locally 23,46 , and highlights the relevance of performing evaluations of the prevalence of blood parasites in different populations of widespread host species 21 .

Methods
Study site. Isla Isabel is an 82-ha volcanic island 28 km off the west coast of Mexico, in the Eastern Tropical North Pacific (21°52′ N, 105°54′ W). The island is mainly covered by deciduous dry forest of Crataeva tapia trees, Euphorbia schlechtendalli bushes, and coastal grasslands. The climate is sub-humid tropical with rains in June-November (the hurricane season). In the rainy season, water is collected in three endorheic basins in the center and north of the island. There is also a shallow pool of alkaline water of approximately 50 m in diameter in the south of the island that is the result of rainwater runoff and the entry of seawater during storms 27  www.nature.com/scientificreports/ metal ring (since 1989, tens of thousands of fledglings and adults have been banded on Isla Isabel 13 ) and sexed all captured boobies by voice (females grunt, males whistle). Our sample included 28 breeding boobies banded as fledglings aged from 7 to 12 years in females and 8 to 13 years in males. We captured adults that had been incubating a clutch (17 males and 16 females) or caring for a brood (16 males and 15 females) for at least 15 days, to minimize the risk of adults abandoning their clutches and milksnakes preying on chicks. See Drummond et al. 48 for further details on the field procedures used. Approximately 1.5-2.0 ml of blood were taken from each adult's brachial vein. A drop of blood was used for blood smear preparation and 200 µl were split in two aliquots (100 µl each) and stored in 96% ethanol for molecular screening of blood parasites. The remaining blood was centrifuged and stored for future research. Blood smears were fixed in 96% ethanol and subsequently stained with Giemsa. Manipulation of booby adults took less than 10 min and bleeding stopped before release at the site of capture. All adults resumed nest attendance 5-10 min after release. Data collection and blood sampling protocols comply with the current laws and ethical standards of animal research in Mexico (NOM-059-SEMARNAT-2010) and were revised and approved by the Secretaría de Medio Ambiente y Recursos Naturales (SEMARNAT; permit number SGPA/DGVS/01216617). We confirm that all methods are reported in accordance with ARRIVE guidelines 2.0 (https:// arriv eguid elines. org).
Parasite identification. Blood smears were scanned for the presence of blood parasites using a light microscope Nikon Eclipse Ti-Arcturus XT of Applied Biosystems. Half of each blood smear was scanned at 400× magnification in search of larger parasites, including Leucocytozoon, during 30-45 min. The other half of each smear was scanned at 1000× magnification in search of Haemoproteus, Plasmodium and Leucocytozoon parasites in up to > 10,000 erythrocytes per smear (see 49 ) during 30 min.
DNA from blood samples was extracted using the Maxwell ® 16 LEV system Research (Promega, Madison, WI) 50 . Samples were run in 0.8% agarose gels to check DNA integrity. To detect and identify avian parasites of the genera Plasmodium, Haemoproteus and Leucocytozoon, we carried out two nested PCRs following Hellgren et al. 51

Data availability
All data generated during this study are included in this published article [and its supplementary information files].