The influence of clay lick proximity on parrot flight dynamics in south-eastern Peru
Steven Ward
Dissertation submitted in part fulfilment of requirements for the degree of Master of Science in Conservation Biology, November 2007.
Abstract
Geophagy in parrots has been recorded in many different parts of the world but most questions concerning its true importance remain unanswered. In order to begin to address some of these questions this study examined the influence of clay lick proximity on volume and patterns of flight in a range of parrot species in lowland Peru. Volumes of flights of geophagous species were expected to be higher near clay licks, but this was found in only four out of the eleven species that regularly fed at the clay licks in the study area. Diurnal flight activity patterns differed with proximity to the clay licks in four species; a disproportionately larger number of flights were recorded near the clay licks at times of usage. Community composition (the proportion of all group records made up by each species) differed with proximity to the clay licks but this was considered more likely a response to habitat differences between sites than due to an increased dominance of geophagous species near the clay licks. Patterns of movement were non random in most frequently recorded species at all sites. Flight directions far from the clay licks appeared to reflect movements between feeding and roosting sites; near the clay licks all non random flight patterns were associated with clay lick visitation. The ability to detect movements associated with the clay licks appeared to decline with distance from them. The results question the significance of the clay licks as attractors of parrots when compared to other resources in the landscape. They may also support the notion that geophagy is a local phenomenon practiced only by those parrots with accessible clay lick resources.
1. Introduction
Geophagy, the intentional consumption of soil, has been documented in many parts of the world for many different animal taxa and is best known among mammals, most notably ungulates and primates, including humans (Jones and Hanson, 1985; Abrahams and Parsons, 1996; Krishnamani and Mahaney, 2000). Invertebrates and reptiles have also been reported consuming soil (Sokol, 1971; Arms et al., 1974). Examples among birds include parrots, cockatoos, pigeons and cracids (Diamond et al., 1999; Burger and Gochfeld, 2003; Downs, 2006; Symes et al., 2006). Parrots arguably provide the most striking displays of avian geophagy, congregating daily in their hundreds at clay lick sites across south-eastern Peru, creating a wildlife spectacle that helps drive the ecotourism trade in the area (Emmons and Stark, 1979; Beissinger and Snyder, 1992; Gilardi et al., 1999; Brightsmith, 2004; Brightsmith and Aramburu, 2004).
The functions of geophagy in animals have been variously proposed as: mineral supplementation (Jones and Hanson, 1985; Sanders and Jarvis, 2000; Brightsmith and Aramburu, 2004), adsorption of dietary toxins (Diamond et al., 1999; Gilardi et al., 1999; Mahaney et al., 1999), mechanical aid to digestion (Best and Gionfriddo, 1991), pH buffering (Kreulen, 1985), and gastrointestinal cytoprotection (Mahaney et al., 1996). Soils chosen by parrots in lowland Peru have been found to reduce the toxicity of some plant secondary compounds and to have a higher cation exchange capacity, both suggesting that adsorption of dietary toxins is the primary role of geophagy in these birds (Gilardi et al., 1999). A similar study in a proximate location in Peru found that soils preferred by parrots had higher concentrations of exchangeable sodium but not a higher cation exchange capacity or clay content, supporting an alternative hypothesis that salts are driving soil selection (Brightsmith and Aramburu, 2004). Despite these contrasting conclusions, geophagy in these parrots may serve multiple functions including mineral supplementation and adsorption of toxins, and soil selection in a particular area may be influenced by local conditions (Brightsmith and Aramburu, 2004). Other studies on parrot geophagy from different continents have also proposed either one or both of these as likely functions (Diamond et al., 1999; Downs, 2006; Symes et al., 2006).
The fact that such a diversity and volume of parrots visit exposed clay lick sites in south-eastern Peru on a regular basis, increasing their risk of being predated, suggests that the consumption of soil is important (Burger and Gochfeld, 2003). Yet most major questions concerning the true importance of geophagy in parrots remain unanswered. Clay lick resources are apparently patchily distributed, so ease of access to individual birds is assumed to differ considerably in different locations. Therefore, is geophagy universal among individuals of a species, or does it only occur in parrots that have accessible clay licks? If the latter is true, does clay lick presence allow an increased carrying capacity of certain species in the surrounding area (Brightsmith, 2004), or provide advantages in terms of increased breeding output and increased dietary breadth? During periods of local food shortage, the ability to eat chemically well-defended seeds and unripe fruits could reduce the risk of population crashes and lessen the need for seasonal movements (e.g. Renton, 2001, 2002; Karubian et al., 2005). If geophagy is practised by all or most individuals of a given species then is the distribution of that species linked to clay lick distribution (Diamond et al., 1999)? Finally, it is not yet known how often individual parrots consume soil (Brightsmith, 2004), or over what area of forest clay licks exert an influence. If they attract parrots from large distances then this has important implications for their protection and use for ecotourism purposes.
One way to begin to address some of these questions is to compare aspects of parrot biology between areas with and without clay licks. This study aims to examine the influence of clay lick proximity on volume and patterns of flight in a range of parrot species in lowland Peru. Two sites were selected: the first, Posada Amazonas, has two known clay licks within 1 km of the lodge (Fig. 4); the second, Picaflor Research Centre, is situated more than 6 km from the nearest known parrot-visited clay lick and 11 km from the clay licks at Posada Amazonas (Fig. 3). The Tambopata River in south-eastern Peru, on which the sites are located, has for many years been frequently used by boats associated with ecotourism and therefore, together with research activities in the area, it is assumed with confidence that all major clay licks used by parrots are known and mapped.
It was expected that the site near to clay licks (Posada) would have, in comparison with the site far from clay licks (Picaflor), the following:
1. A greater volume of flights by parrot species that regularly use clay licks
2. A different community make up of parrots (in flight), more dominated by species associated with clay licks.
3. A detectable movement of parrots towards clay licks at times of known clay lick usage.
4. Detectable differences in diurnal flight activity patterns that could be directly attributed to movements associated with clay licks.
2. Methods
2.1. Study sites
The two lodges around which the study was based are located along the Tambopata River, in the buffer zone of the Tambopata National Reserve (275 000 ha) in south-eastern Peru. The area receives c. 2800 mm of rain per year and the dry season extends from April to October (Pearson and Derr, 1986). Posada Amazonas (12˚48’S, 69˚18’W) is situated 300 m from the southern bank of the Tambopata River, 26 km southwest of Puerto Maldonado, within a 4000 ha ecological reserve owned by the Native Community of Infierno (Fig. 4). It is surrounded by floodplain and terra firme forests. Within 1 km of the lodge, on the south bank of the river there are a series of near vertical exposed riverbank sections that are used as clay lick sites (Colpa Hermosa; Fig. 4). There is a further small clay lick, formed by the erosion of the bank of a stream, 100 m inland of the river, 1 km upstream of Colpa Hermosa (Colpita; Fig. 4). Picaflor Research Centre (12˚49’S, 69˚23’W) lies on the northern bank of the Tambopata River, 33 km southwest of Puerto Maldonado (Fig. 3). It is surrounded to the north by terra firme forest, interspersed with many patches of bamboo, and to the south (across the river) by floodplain forest. The nearest known clay lick site used by parrots is 6 km to the southwest (Colpa El Gato; Fig. 3) and the clay licks at Posada Amazonas are 11 km to the east.
2.2. Parrot surveys
Parrots were surveyed between 14th June and 7th August 2006, a period within the dry season when parrot usage of clay licks in the area is known to be relatively low (Brightsmith, 2004). Three survey sites were established at each lodge, chosen primarily for their unrestricted view of a large area of sky: at Posada these consisted of a canopy tower within the forest and two river overlook sites (Fig. 4); at Picaflor three river overlook sites presented the best viewing areas (Fig. 3). Visual and vocal identification of all parrot species known to occur in the area was practiced for 10 days prior to beginning the study. Parrot surveys lasted 2.5 hours and all sites were visited 7 times in each of 3 periods: 05:45 – 08:15, 08:45 – 11:15 and 15:00 – 17:30. Sites were not visited more than once daily, were visited in different periods on consecutive days and surveying at each lodge was alternated approximately fortnightly. During each survey, the species, group size, time and flight bearing were recorded for every parrot group seen in flight. In the second half of the study, during the last 10 days surveying at each lodge, to measure the volume of flight activity it was also recorded whether each parrot group flew over a designated area at each site. These areas were created by choosing two points on the opposite bank of the river that were 90˚ apart from the observer and then connecting these points and the observer to form a triangle. The width of the river at each site (measured using a range finder) allowed the area of each triangle to be calculated. Delineating such an area accurately from the canopy tower at Posada (Site 2) was not possible.
2.3. Clay lick monitoring
As part of an ongoing project, both Colpa Hermosa and the Colpita at Posada were monitored from established hides for around 10 days per month. Observers arrived at the hides before sunrise, before the first birds arrived at the licks, and monitored the licks continuously until late afternoon (around 17:30). Every 5 minutes the number of birds of each species feeding on the licks was recorded with the aid of binoculars. From the hides observers could distinguish all common parrot species known to occur in the area. During the months in which flight activity surveys were conducted, June, July and August 2006, 10 parrot species were recorded feeding at Colpa Hermosa and 9 species were recorded feeding at the Colpita (Table 1).
2.4. Data analysis
Flight frequency (the number of groups or individuals flying per hectare per hour) was compared between four sites for each species using Kruskal-Wallis non-parametric analysis of variance. Dunn’s multiple comparison tests were then used to
TABLE 1. Clay lick usage dynamics of fifteen parrot species at sites near Posada Amazonas lodge, Tambopata, Peru, during June, July and August 2006a
Colpa Hermosa(30 survey days) / Colpita
(25 survey days)
No. days when feeding occurred / Peak feeding timeb / Max count (mean ± S.D)c / No. days when feeding occurred / Peak feeding timeb / Max count (mean ± S.D)c
Black-capped Parakeet
Pyrrhura rupicola / 18 (60%) / 10:00* / 6.7 ± 4.0 / 5 (20%) / 09:00 / 7.2 ± 4.7
Blue-headed Macaw
Primolius couloni / 0 / - / - / 0 / - / -
Blue-headed Parrot
Pionus menstruus / 23 (77%) / 07:00 / 20.6 ± 9.1 / 13 (52%) / 07:00 / 6.2 ± 6.5
Blue-and-yellow Macaw
Ara ararauna / 0 / - / - / 0 / - / -
Chestnut-fronted Macaw
Ara severa / 19 (63%) / 07:00 / 11.5 ± 7.5 / 1 (4%) / 07:30 / 11
Cobalt-winged Parakeet
Brotogeris cyanoptera / 0 / - / - / 9 (36%) / 15:30 / 126 ± 78.5
Dusky-headed Parakeet
Aratinga weddellii / 25 (83%) / 07:00 / 36.8 ± 21.1 / 11 (44%) / 07:30 / 19.7 ± 23.8
Mealy Parrot
Amazona farinosa / 11 (37%) / 07:00 / 4.1 ± 3.5 / 6 (24%) / 08:00 / 16.5 ± 23.5
Orange-cheeked Parrot
Pionopsitta barrabandi / 19 (63%) / 07:00 / 5.9 ± 5.2 / 14 (56%) / 08:00 / 10.5 ± 11.1
Red-bellied Macaw
Orthopsittaca manilata / 0 / - / - / 0 / - / -
Red-and-green Macaw
Ara chloroptera / 30 (100%) / 15:30* / 8.0 ± 4.2 / 0 / - / -
Scarlet Macaw
Ara macao / 2 (7%) / 10:00 / 1.5 ± 0.7 / 0 / - / -
White-bellied Parrot
Pionites leucogaster / 0 / - / - / 0 / - / -
White-eyed Parakeet
Aratinga leucopthalamus / 2 (7%) / 07:00 / 4.5 ± 0.7 / 7 (28%) / 07:00 / 5.3 ± 3.7
Yellow-crowned Parrot
Amazona ochrocephala / 20 (67%) / 07:00 / 8.5 ± 3.6 / 1 (4%) / 08:00 / 4
a data courtesy of Alan Lee
b the half hour period when the most feeding occurred (times represent the midway point in each half hour period, for example 07:00 = 06:45 – 07:14), * but feeding occurred throughout the day
c maximum number of birds feeding simultaneously on the clay lick per day (days when feeding did not occur were omitted from mean calculations)
show significant between-site differences. It was not possible to delineate an accurate area at Posada Site 2 so this site at both lodges was removed from these analyses. Late morning (08:45 – 11:15) records were also removed due to reduced flight activity during this period. Community composition (the proportion of all group records made up by each species at each site) was compared between sites and lodges with a hierarchical cluster analysis using average linkage between-groups and a squared Euclidean distance interval.