Study Areas

I collected food materials at active nests and roosts of owls in the Dominican Republic incidentally and at irregular intervals from December 1975 through December 2004 (Table 1). During those periods, I collected data once every two to three weeks, from five localities (four habitats) for Ashy-faced Owls and from nine localities (three habitats) for Barn Owls (Table 1). Descriptions of the collection areas are presented in Wiley and Wiley (1981), Snyder et al. (1987), Wiley (1998), Keith et al. (2003), and Latta et al. (2006).

Table 1

Localities, habitats, and year of collection for prey remains and regurgitated pellets gathered from nests and roosts of Ashy-faced Owl (Tyto glaucops) and Barn Owl (T. alba) in the Dominican Republic, 1975–2004.

Prey Material Collection and Preparation

I gathered prey remains and regurgitated pellets at active nests and roosts of owls. Regurgitated pellets and uneaten prey at nests and roosts provide an accurate representation of the diet of owls (Marti et al. 2007), and are an accepted means of determining feeding ecology of Barn Owl (Taylor 1994). Although all pellet materials were collected and analyzed, here I include only those prey remains gathered between the first and last visits to the locality during the observation period; i.e., older prey items accumulated before my work began at a site are not included in analyses to minimize the possibility that species other than the owls were involved in those prey captures. Pellets and prey remains were placed in plastic bags tagged with the date and site of collection, and protected from pests by naphthalene crystals. Collected materials were later identified in the laboratory, using a dissecting microscope when necessary. Whole pellets were dissected individually, whereas partial pellets were grouped together by date and locality. Mandibles, skulls, and femurs of mammals; bones and feathers of birds; bones and skin of reptiles and amphibians; and invertebrate parts were separated and identified. If mammal skulls were too damaged or absent, I counted pairs of mandibles or numbers of atlas vertebrae. Identifiable invertebrates were found within pellets and as food remains at nests and roosts, but I did not attempt to quantify these remains because of the degraded nature of much of the materials. Voucher specimens were examined in the Museo Nacional de Historia Natural de Santo Domingo (MNHNSD) and the private collection of Albert Schwartz. When possible, prey body masses were obtained from fresh materials. Other sources of prey masses included data on specimen labels in the MNHNSD and other collections, and published data in Silva Taboada (1979), Dunning (1993), and Arendt et al. (2004). Body masses of a few species of smaller frogs and lizards for which I could find no mass data were estimated by weighing similar-sized species. For size-dimorphic species, body mass was estimated by averaging mean male and female weights.

Nomenclature follows American Ornithologists' Union (1998 and supplements) for birds, Hedges (2008) for reptiles and amphibians, and Wilson and Cole (2000) for mammals.

Analysis

Data are presented as the minimum number of individuals and percent frequency of occurrence (frequency), which is the proportion of the total number of prey individuals. The data obtained were used to assess the locality-related diversity of the prey fauna represented in the owl pellets and prey remains. To assess and compare diversity in the diets of the owls, I used four indices in the program Species Diversity and Richness IV (Seaby and Henderson 2006): (1) species richness (R), (2) Shannon's index (H′), (3) Simpson's index (D), and (4) Pielou's evenness (J) estimator (Magurran 1988). Hutcheson's method was used for testing significance in diversity of the two owl species' assemblages of prey (Hutcheson 1970), and Solow's Randomization Test was used to assess differences in diversity index values (Seaby and Henderson 2006). I measured niche overlap using Pielou's (1977) equation for α. Statistical computations were performed using the software package Minitab (2000), with an alpha value for statistical significance set at 0.05.

Prey Frequency

Mammals predominated in the identified prey for both species, but Ashy-faced Owls ate a substantially smaller proportion (52.0%) of mammals than did Barn Owls (76.7%; Mann-Whitney test U  =  381.0, P < 0.05; Table 2). House rats (Rattus rattus) predominated in the prey of both Ashy-faced (38.1% of total items identified) and Barn (57.4%) owls, with house mice (Mus musculus) making up a moderate proportion of the remaining prey (2.8% for Ashy-faced Owl, 6.7% for Barn Owl; Table 2). Brown rat (Rattus norvegicus), a species mainly associated with urban habitat in the Dominican Republic, formed a small proportion (0.4%) of the Barn Owl's diet. In total, rodents made up 40.9% of the identified items for the Ashy-faced Owl, and 64.5% for the Barn Owl (U  =  9.0, P > 0.05). Thirteen species of bats were detected in prey remains and pellets of the Ashy-faced Owl, whereas 16 species were found in the diet of Barn Owl. All bat species combined made up a total of 11.1% of the material examined for Ashy-faced Owl, and 12.2% for Barn Owl (U  =  282.5, P > 0.05).

The second most frequent prey class for both owl species was birds, making up 28.8% of the material examined for Ashy-faced Owls, and 12.3% for Barn Owls (U  =  8037.5, P < 0.001; Table 2). I found 78 avian species among the material for Ashy-faced Owl, and 52 species for Barn Owl. Among birds taken by Ashy-faced Owls, passerines (n  =  41 species, 739 individuals) were more frequent than non-passerines (n  =  37 species, 364 individuals). In contrast, more species of nonpasserines (n  =  29) than passerines (23) were recorded for Barn Owls, although its diet consisted of more individuals of passerines (349) than nonpasserines (203).

Reptiles and amphibians composed 19.2% of the prey material examined for Ashy-faced Owl and 11.1% for Barn Owl (U  =  1737.5, P < 0.05; Table 2). Amphibians made up a higher proportion of the materials examined for Ashy-faced Owl (11.5%) compared to that of Barn Owl (4.6%; U  =  164.5, P < 0.05). The same frequency pattern was true for reptiles, with Ashy-faced Owl material composed of more reptiles (7.7%) than found in Barn Owl material (6.5%; U  =  837.5, P < 0.01). Ashy-faced Owls preyed on 15 species (210 individuals) of lizards, most of which were anoles (Anolis: seven species, 121 individuals), seven species of snakes (85 individuals), and one species of amphisbaenid (1 individual; Table 2). Barn Owl prey included 24 species (215 individuals) of lizards, seven species (74 individuals) of snakes, and one amphisbaenid (3 individuals). Among amphibians in the diet of Ashy-faced Owl, more species (six species; 117 individuals) of Eleutherodactylus frogs than other genera were represented, but Hispaniolan laughing treefrog (Osteopilus dominicensis; 212 individuals) accounted for nearly half of the amphibian prey. Similarly, Barn Owl material contained more species of Eleutherodactylus frogs (six species; 65 individuals) than other genera, but more individuals (118) of Hispaniolan laughing treefrog were found. A total of nine species of amphibians was found among the Ashy-faced Owl material, and 11 species among that for Barn Owl.

Tarantulas, scorpions, beetles, and orthopterans predominated in invertebrate material. I did not determine numbers of invertebrates in the samples, but 29% of examined Ashy-faced Owl pellets and 35% of Barn Owl pellets contained invertebrate remains.

Prey Biomass

Ashy-faced Owl prey masses differed among classes from those of Barn Owl (χ²  =  52898.0, df  =  3, P < 0.001). Mammals formed the bulk of prey biomass for both species, but in the diet of Ashy-faced Owl, mammals formed a smaller proportion (73.9%) of biomass than in the Barn Owl's diet (90.7%; U  =  363.5, P < 0.05; Table 2). House rats predominated in the prey of both Ashy-faced (70.8% of total biomass) and Barn (86.6%) owls, with house mice making up a modest proportion of the remaining prey identified (0.5% for Ashy-faced Owl, 1.0% for Barn Owl; Table 2). Brown rats formed a small proportion of the rodents found in the Barn Owl's diet (0.9% of mass). In total, rodents composed 71.3% of the biomass for Ashy-faced Owls, and 88.4% for Barn Owls (U  =  8.0, P > 0.05). All bat species combined made up a total of 2.6% of the prey biomass for Ashy-faced Owls, and 2.2% for Barn Owls (U  =  271.0, P > 0.05).

Birds were second in importance, by total biomass, in materials for both owl species, with that class making up 14.8% of the mass for Ashy-faced Owl, and 5.1% for Barn Owl (U  =  7792.5, P < 0.001; Table 2). The total prey mass of reptiles and amphibians combined was 11.3% for Ashy-faced Owl and 4.3% for Barn Owl (U  =  1807.5, P  =  0.053). Reptiles formed 8.5% of total prey mass for Ashy-faced Owl and 3.2% for Barn Owl (U  =  892.0, P < 0.05), and amphibians comprised 2.8% for Ashy-faced Owl and 1.1% for Barn Owl (U  =  170.0, P > 0.05; Table 2).

Prey Species Richness and Diversity

The numbers of vertebrate species used by the two species differed slightly, with the Ashy-faced Owl preying on 125 species, and the Barn Owl using 114 species. The two owl species captured 92 species in common (14 mammals [2 rodents, 12 bats], 48 birds, 7 amphibians, 23 reptiles), whereas the samples for Ashy-faced Owl included 33 unique species (1 bat, 30 birds, 2 amphibians) and those for Barn Owl included 22 distinct species (5 mammals [1 rodent, 4 bats], 4 birds, 4 amphibians, 9 reptiles). A Chao and Lee Richness Estimator yielded a species richness (R) estimate of 147.2 species (two samples), which equaled the number of collective prey species (n  =  147) observed among the two owl species' pellets and remains.

Indices for the diversity of the owls' prey revealed that the diet of the Ashy-faced Owl was more diverse than that of the Barn Owl: i.e., H′_{Ashy-faced Owl}  =  3.044 and H′_{Barn Owl}  =  2.214; P < 0.05. Similarly, Simpson's index showed the prey species of Ashy-faced Owl to be more diverse than those in the diet of Barn Owl: D_{Ashy-faced Owl}  =  6.316 and D_{Barn Owl}  =  2.933 (P < 0.05). Prey evenness, the pattern of distribution of the prey species, as determined by Pielou's index, indicated that Ashy-faced Owls had more diverse prey than Barn Owls: J_{Ashy-faced Owl}  =  0.6101 and J_{Barn Owl}  =  0.4437 (P < 0.05).

Another measure of prey diversity is the minimum number of species needed to form ca. 80% of all items in the predator's diet. By prey frequency, 19 prey species formed 80% of the items in the diet of Ashy-faced Owl, whereas nine species formed that level of prey items for Barn Owl. In terms of prey biomass, only eight species made up >80% of the diet of Ashy-faced Owls and a single species made up that amount for Barn Owls.

Prey Niche Overlap

An analysis of prey niche revealed moderate overlap (α₂₁  =  0.60) for food resources between Ashy-faced and Barn owls. I collected prey remains and pellets from four habitats for the Ashy-faced Owl and three for the Barn Owl (Tables 1, 3). I found no differences among prey proportions in materials collected among the habitats for the Ashy-faced Owl (χ²  =  2.90, df  =  9, P > 0.05) or those for the Barn Owl (χ²  =  3.704, df  =  6, P > 0.05); i.e., there were consistent relative proportions of mammals, birds, amphibians, and reptiles within each owl species' prey. Prey remains and pellets were collected from two habitats (dry scrub/coastal woodland and montane broadleaf evergreen forest) shared by Ashy-faced and Barn owls. In both habitats, proportions of prey classes differed significantly between owl species (for dry scrub/coastal woodland, χ²  =  19.010, df  =  3, P < 0.001; for montane broadleaf evergreen forest, χ²  =  12.639, df  =  3, P < 0.01). In the two habitats where I collected prey remains and pellets for both species, the niche overlap index (α₂₁) was 0.57 for montane broadleaf evergreen forest, and 0.63 for dry scrub/coastal woodland.

Prey Size

Size of the two owl species' prey ranged from 0.6 g (tuck-wheep frog [Eleutherodactylus abbottii]) to 1200 g (blunt-headed green treesnake [Uromacer catesbyi]) in Ashy-faced Owl, and 0.6 g (tuck-wheep frog) to 650 g (Hispaniolan desert boa [Epicrates fordii]) in Barn Owl. Prey species size did not differ between the Ashy-faced Owl (mean  =  79.0 ± 169.4 g, n  =  125, range  =  0.6–1200 g) and Barn Owl (mean  =  64.53 ± 101.65 g, n  =  114 species, range  =  0.6–650 g; U  =  14013.0, P > 0.05).

Ashy-faced Owl Diet

The Ashy-faced Owl samples I examined were substantially lower in mammalian prey (52.0%) than those reported from the two fully quantified sites studied by Wetmore and Swales (1931). Materials collected at two of four localities in Haiti and the Dominican Republic were characterized qualitatively for Rattus spp. as having “a large number” and “a great quantity,” respectively. At the other two sites, where mammal remains were quantified, mammals formed 67.4% and 94.1%, respectively, of the Ashy-faced Owl prey items. Rodents composed 48.6% and 91.5%, whereas bats formed 18.8% and 2.6%, respectively, of the total number of prey (Wetmore and Swales 1931).

Wetmore and Swales' (1931) report of 18.8% bats at one of their sites, and my tally of 11.1% bats for Ashy-faced Owl, are similar to the proportion of bats in Barn Owl prey in Hispaniola (12.2%; present study), but higher than the proportion of bats among Barn Owl prey reported elsewhere in the Caribbean Islands (1.7–2.0%; Buden 1974, Arredondo Antúnez and Chirino Flores 2002).

Wetmore and Swales (1931) found birds to be common among Ashy-faced Owl food remains (Table 4). Because mammal remains were not quantified at two of the sites, the proportion of birds reported among prey of the Ashy-faced Owl by Wetmore and Swales (1931) is likely an overestimate. At the two fully quantified sites, birds represented 23.6% (n  =  144) and 5.9% (n  =  153), respectively, of remains. Wetmore and Swales (1931) recorded 30 bird species (including 3 identified only to genus), of which Narrow-billed Tody (Todus angustirostris) was the only species that was not present in the materials I collected. As in the Ashy-faced Owl prey remains I examined, the Black-crowned Palm-Tanager (Phaenicophilus palmarum) was the most frequent bird species identified by Wetmore and Swales (1931). In their samples, passerines were equal to nonpasserines in representative species (15 each), but the former comprised 68% of bird prey individuals; these proportions are similar to those I found for Ashy-faced Owl. Birds were better represented in Ashy-faced Owl diet (26.5%, Wetmore and Swales 1931; 28.8%, this study) than in the diet of Hispaniolan Barn Owls (12.3%).

Table 4

Prey of Ashy-faced Owl (Tyto glaucops), as published in Wetmore and Swales (1931). Data from regurgitated pellets and prey remains collected by G.S. Miller, Jr. and H.W. Krieger at four localities in Hispaniola (Haiti: Diquini, near Port-au-Prince; L'Acul; and Trujín, La Salle, Morne La Viste; and Dominican Republic: San Lorenzo Bay). The quantitative data represent prey collected at L'Acul and Trujín, where rodent numbers were tallied.

No reptiles or amphibians were found among the Ashy-faced Owl prey remains for one site reported by Wetmore and Swales (1931), whereas amphibians and reptiles formed 9.3% of the prey at the second of the two sites where remains were quantified (Wetmore and Swales 1931). In my study, Ashy-faced Owls used amphibians more frequently (11.5%) than did Barn Owls in Hispaniola (4.5%) or other reported West Indian Barn Owl populations (0–7.4%; Arredondo Antúnez and Chirino Flores 2002, Vilató Viamontes et al. 2002).

Barn Owl Diet

Diet data have been widely reported for Barn Owls through much of their cosmopolitan range, although most reports are for continental populations (see compilation by Taylor 1994). Of 52 studies listed by Taylor (1994), only two were for island (Galapagos and Canary islands) populations. Prey recorded in 49 continental Old and New World, and Australian diet studies consisted primarily of small mammals (range 77.7–100% of all items), whereas island populations showed lower proportions of mammal prey; e.g., Canary Islands: 74.5% and Galapagos Islands: 44.9% (Taylor 1994).

Overall, my observations of Barn Owl prey proportions among vertebrate classes were similar to those reported in the Bahamas and West Indies by other authors. In contrast to Bond (1977), who noted that rodents, mainly rats, compose only a minor part of the Barn Owl's diet in the Antilles and Bahamas, rodents have been reported as the primary diet items of the owl by the several authors who have conducted detailed studies of the owl's food habits in the region (e.g., Johnston 1974, Hernández Muñoz 2001, Arredondo Antúnez and Chirino Flores 2002), especially introduced Old World rodents (subfamily Murinae; Buden 1974, Debrot et al. 2001). Through competition, these exotic rodents probably aided in the decline and extinction of the original prey species; in the West Indies these were likely small native rodents (e.g., hutias [familly Capromyidae]; rice rats [Ozyzomys spp.]) and insectivore mammals (e.g., Nesophontes [family Nesophontidae]; Woods 1989, Wing 2001, Silva Taboada et al. 2007). With the exception of bats, introduced rodents were the only mammalian prey in pellets and prey remains of Hispaniolan Barn Owls. However, I found fewer rodents in Hispaniolan Barn Owl diet than reported by investigators for other islands in the region (64.5% vs. 79.9–93.5%, respectively). Although house rats were far more common than house mice in the diet of Hispaniolan Barn Owls in my study, Rutten (1934) and Arredondo Antúnez and Chirino Flores (2002) reported that house mice far outnumbered house rats among Barn Owl prey in Cuba. I found a ratio of 152 house rats to one brown rat in Barn Owl prey in Hispaniola, which was similar to Buden's (1974) ratio of 116 house rats to one brown rat in Barn Owl prey in the southern Bahamas. Ecological differences between the two rat species explain the disparity; i.e., brown rat is mainly an urban species, whereas house rat occurs mostly in rural habitats.

Bats generally form a low proportion of prey among continental populations of Barn Owls (Taylor 1994) and most West Indian Barn Owls (e.g., Buden 1974, Hernández Muñoz 2001, Arredondo Antúnez and Chirino Flores 2002). My observation of the proportion of bats (12.2%) in prey of Hispaniolan Barn Owls was substantially greater than that reported in other Barn Owl diets in the region (1.7–2.0%; Buden 1974, Arredondo Antúnez and Chirino Flores 2002).

Bond (1977) suggested that Barn Owls capture more land birds in the Antilles and Bahamas than all other native predators combined. Several others have reported birds as important components of Barn Owl diets in the region (e.g., Rutten 1934, Johnston 1974, Suárez 1998). The proportion of Barn Owl avian prey I observed in Hispaniola (12.3%) was between the extremes of reported proportions (2.6–ca. 40%; Arredondo Antúnez and Chirino Flores 2002, Johnston 1974) among populations in the Bahamas and Caribbean islands.

The proportion of reptiles and amphibians (11.1%) in Barn Owl samples from Hispaniola was slightly greater than the extremes reported (0.8–7.4%; Buden 1974, Arredondo Antúnez and Chirino Flores 2002) in the region.

Few studies have quantified the importance of invertebrates in the diet of Barn Owl in the Bahamas and Caribbean islands (i.e., 3.4% of total items in Curaçao [Debrot et al. 2001], and 3.8% in Cuba [Arredondo Antúnez and Chirino Flores 2002]). Where they were not quantified, invertebrates have been reported as minor items in the diet of Barn Owls in the Caribbean Islands by several authors (Buden 1974, Hernández Muñoz 2001, Vilató Viamontes et al. 2002).

Prey Characteristics for Ashy-faced and Barn Owls

I found considerable difference between Ashy-faced and Barn owl diets in Hispaniola. All prey diversity indices suggested that the Ashy-faced Owl has a broader feeding niche than does the Barn Owl, though both species showed remarkable breadth in prey species used. In addition, the fact that fewer species made up the majority (frequency and biomass) of prey for Barn Owls compared with Ashy-faced Owls further confirms the narrower prey niche of the Barn Owl. Taylor (1994) noted that most barn owl diets tend to be dominated by a relatively small number of species and to have a long tail of species that are taken regularly or sporadically but which, in total, contribute only small amounts. In the Hispaniolan populations, these characteristics were true of both species, but particularly the Barn Owl.

Part of the difference between prey proportions in the two owl species is explained by the greater number of bird species used as prey by the Ashy-faced Owl compared with the Barn Owl (78 vs. 52). Also, whereas Ashy-faced Owl prey included far more unique bird species (30) than Barn Owl, Barn Owl prey consisted of more unique reptiles (nine species) and amphibians (four species).

My results, like those of most pellet analyses, are likely to be biased toward the sturdier mammals, and against the more delicate birds, reptiles, and amphibians, and, certainly, invertebrates. Nevertheless, it is unlikely that the two Tyto species have substantial differences in the mechanisms for the intake and processing of prey, so the diets presented here are at least relatively representative as comparisons of the feeding ecology between the two species.

I observed no diurnal hunting activity by either owl species, although they commonly foraged in low light of dawn and dusk. Therefore, it is remarkable that so many diurnal species were captured by both Tyto species. Birds are most likely snatched from their overnight roosts, where they are exposed to such depredations. Most lizard species are diurnal, and generally do not occur commonly among Barn Owl prey (Taylor 1994). In addition, several species of diurnal Hispaniolan lizards have subterranean overnight retreats, making them even less likely to be available as nocturnal owl prey; e.g., Ameiva enter their underground retreats in the mid- to late afternoon and do not emerge until midmorning of the next day. On the other hand, some Hispaniolan reptiles (e.g., boas) are primarily nocturnal. It is also notable that the Hispaniolan dwarf amphisbaena (Amphisbaena manni) and Hispaniolan common blindsnake (Typhlops pusillus) were among the prey captured by both Ashy-faced and Barn owls in Hispaniola, given the fossorial habits of these reptiles.

That bats formed a substantial proportion of prey items for both owl species in the Dominican Republic is not surprising because both Ashy-faced and Barn owls nest and roost in caves and sinkholes where several of the bat species (e.g., Jamaican fruit-eating bat [Artibeus jamaicensis], big brown bat [Eptesicus fuscus]) occur (Silva Taboada 1979). I observed Ashy-faced and Barn owls capture bats in flight, but only at sinkholes and cave entrances, where bats exited or entered their retreats in large numbers. The presence of remains of bats that do not live in caves (e.g., tree-foliage-roosting Cuban fig-eating bat [Phyllops falcatus]), however, raises other questions of availability and capture techniques used by the owls.

Prey Size

The small size of some prey species of the two owl species is remarkable, with some items as small as 0.6 g in mass. The smallest (<10 g) items were rare in the diet of both species in my samples, forming 5.0% (<0.1% of biomass) of items for Ashy-faced Owl and 4.3% (0.1% of biomass) for Barn Owl. Bird masses in Ashy-faced Owl prey samples ranged from 4 to 1150 g, and 4 to 300 g in Hispaniolan Barn Owl samples. Small birds have been reported among prey remains for Ashy-faced Owl (ca. 6–9 g, Wetmore and Swales 1931) and Barn Owls in the Antilles and Bahamas (ca. 4–6 g, Bond 1977). The large size of some prey items taken by both Tyto species in the Dominican Republic is also extraordinary, with estimated weights of some prey substantially greater than owl masses; i.e., Ashy-faced Owl, 260–534 g (Weick 2006); Barn Owl in the West Indies, 455–575 g (Arendt et al. 2004). Items averaging >500 g made up only 0.4% (2.6% of biomass) of total items for Ashy-faced Owl, and 0.3% (1.4% biomass) for Barn Owl in this study.

Future Research

It will be interesting to study the development of coexistence between the two Tyto species in Hispaniola to determine if further niche separation develops. An important step in establishing the character of the two species' niches would be to ascertain the nature of competition for shared resources. My study did not address prey populations, so it is still unknown whether the two owl species compete for limited resources. My unquantified estimate of the numbers of the most frequently shared prey species was that they are common and available (i.e., rats, mice, and other high-ranked prey species appeared to be common in my study localities).