Donald Brightsmith
Duke University Department of Biology, Durham NC, USA, djb4@duke.edu
Corresponding author
Kyle Van Houtan
The Center for Conservation Biology at Stanford University, Stanford CA, USA
Geophagy, or the eating of dirt, is a relatively common phenomenon among herbivores (Gilardi et al 1999 and references therein). One of the most visually impressive manifestations of this behavior is seen in the lowland forests of the Southwestern Amazon Basin where hundreds of individuals of a dozen or more species of parrots, parakeets and macaws gather to take clay from exposed riverbanks (Munn 1994). To date the studies of parrot geophagy have focused on the physiological question of why birds eat clay (Diamond 1999, Emmons and Stark 1979, Gilardi et al 1999), or the potential of licks as ecotourism attractions (Munn 1991, Munn 1992) and left the behavioral and ecological aspects unexplored.
This work documents the abundance and diversity of birds using a large clay lick near the Tambopata Research Center in southeastern Peru during the rainy season. It also shows the effects of weather on clay lick use. The information presented here will form the basis for future studies looking at annual cycles of clay lick use and provide baseline information useful in plotting long term changes in clay lick use. These data will then be used to ask focused questions about the impact of foraging ecology and ecotourism in clay lick use by parrots and other birds.
This study was conducted from 11 January - 4 April 2000 at the large clay lick adjacent to Tambopata Research Center on the right bank of the Tambopata River, Department of Madre de Díos, Peru. The lick was observed from a distance of about 100-150m without the use of a blind. No blind was used because the frequent rise and fall of the river destroyed blinds within a few days of their construction and because birds do not show any visible reaction to people located at this distance. This clay lick is frequently observed by guided groups of tourists from this distance. On many occasions tourist groups were present during early morning data collection. The possible impacts of these tourists will be analyzed in other papers and are not considered further. Observers and tourists arrived at dawn, at least 15 – 30 minutes before the parrots began staging to descend to the lick. The few exceptions to this were usually caused by rain, which delayed the arrival of birds as well as the observers. The lick was observed using 7-10 power binoculars and a Bushnell Spacemaster spotting scope with a 15-40X zoom eyepiece. The optical equipment used did not allow rapid and reliable differentiation between Chestnut-fronted Macaw (A. severa) and Red-bellied Macaw (Orthopsittaca manilata). These two small green macaws are easily identifiable in flight, but at rest on the lick they are extremely difficult to separate rapidly. As a result these two species are lumped for all of the analyses in this paper.
Every five minutes the number of birds on the clay lick was counted and the weather was noted. Weather was scored as follows: 1) Sun, sunshine visible from the observation point, 2) Cloudy, no sun visible from the observation point, 3) Rain, rain falling on the observation point. The presence or absence of fog was also recorded. For the analysis presented here, the data were divided in to three time periods: early morning (dawn-07:30), late morning (7:30-12:00) and afternoon (12:00-17:30). These three groupings were logical because each morning without rain there was one strong pulse of activity dominated by small macaws and parrots that ended in all but one case by 07:30. There was often one or more morning and afternoon visits by large macaws. For each time period on each day the maximum number of birds seen simultaneously on the lick was calculated and used for the analysis. In addition a single composite weather value was calculated for each time period as follows: Sun >80% of time sun visible; Sun/Clouds >20% of time sunny and >20% cloudy; Cloud/Rain >20% time cloudy and > 20% time raining; Rain >80% time raining.
This report deals only with species in the family Psittacidae. Other species including pigeons (Columbidae) and Guans (Cracidae) were recorded but they are not considered here. The total number of psitacids for each time period per day was calculated for the following: each individual species, all large macaws combined (Ara chloroptera, A. ararauna, and A. macao), and all psitacids combined. For each complete day that the lick was observed, the maximum number recorded in the early AM, late AM and PM were summed for each species to get an index of the total number of birds that visited the lick. This index definitely underestimates the numbers of birds that use the lick in a single day because the members of a single species do not all descend to the lick simultaneously. In fact less than half of the total number of any one species may be on the lick at any one time.
Data were collected for a total of 409 hours on 59 different days. Observations were completed for 56 early mornings, 34 late mornings, 29 afternoons and 27 full days. Fourteen different species of psittacids were seen eating clay during these observations (Table 1). Other birds seen eating clay included Ruddy Pigeon (Columba subvinacea), Plumbeous Pigeon (C. plumbea) and Pale-vented Pigeons (C. cayennensis), Speckled Chachalaca (Ortalis guttata), Common Piping-Guan (Aburria pipile), Spix’s Guan (Penelope jacquacu), and Orange-breasted Falcon (Falco deiroleucus). These non-parrot species will not be discussed further in this report. All 14 psittacid species were recorded during the early morning observations while only 10 and 6 species were recorded during the late morning and afternoon observations respectively (Tables 1-4). The early mornings were dominated by a variety of small macaws, parrots and parakeets, while large macaws dominated the late mornings and afternoons. On average 7.4 species visited the lick daily (sd=2.62, n=27). The average species diversity was significantly higher during the early morning (5.52 species, sd=3.11, n=56), than in the late morning (3.41 species, sd=1.8, n=34, p<0.05) and afternoon (2.90 species, sd=1.08, n=29, p<0.05), but late morning and afternoon did not differ significantly in their diversity (Tukey-Kramer multiple comparisons).
On average 236 birds visited the clay lick daily (sd=60, n=27 full days, Table 4). Most birds visited during the early morning (203, sd=132, n=51, Table 1) and significantly fewer visited during the late morning (20.8, sd=23.9, n=34, p<0.05, Tukey-Kramer, Table 2) or afternoon (40.4, sd=40.5, p<0.05, Tukey-Kramer, Table 3). There was no significant difference in the total number of birds between the late morning and afternoon.
Mealy Parrot (Amazona farinosa) was the most common species at the lick with over 300 seen during some early mornings and nearly 100 seen on average (Table 1, 4). Five species of macaws including Ara chloroptera, A. ararauna, A. macao, A. severa and Orthopsittaca manilata were common at the lick. They were all seen most days and usually in fairly good numbers (Tables 1 -4). The large macaws (A. chloroptera, A. ararauna and A. macao) were much more common during the late morning and afternoon than during the early morning. In fact A. chloroptera was rarely seen at the lick before 07:30 (6 of 56 mornings). In contrast the “green macaws” A. severa and Orthopsittaca manilata were common early morning visitors and would only occasionally appear in small numbers in the late morning or afternoon (Tables 1-4). A sixth species of macaw Propyrrhura couloni was recorded on the lick only once during data collection. This uncommon endemic is resident in the area but is rarely seen at the lick.
A detailed analysis of the daily temporal sequence of clay lick use by the different parrot species will be presented in future publications.
Observations were conducted during five early mornings during which it rained over 80% of the time. During these observations no psittacids visited the lick. Observations also indicate that heavy rains in the early morning cause birds to fly en masse from the area of the lick (personal observation). No observations were conducted during late mornings or afternoons during which rain fell for over 80% of the time. Mornings with sun (weather class sunny and sun/cloud) had larger numbers of birds and a higher diversity than cloudy and cloudy/rainy mornings (ANOVA all p-values <0.05). This trend was significant for the total number of large macaws, total numbers of green macaws, Mealy Parrots, Blue-headed Parrots, White-eyed Parakeets and White-bellied Parrots and the total number of psitacids combined.
Mornings with fog had a significantly smaller total number of species, less total individuals, less total large macaws, and significantly less Mealy Parrots. No other species showed a significant response to the presence of fog, but data were collected on only 4 foggy mornings.
As the study progressed there was a significant decline in the total number of individual psittacids visiting the clay lick (Figure 2, Table 4, ANOVA p<0.03 data). When the different parts of the day were analyzed separately, the number of birds arriving in the early morning (p<0.001) and late morning (p<0.001) was less in April than it was in January. However when the afternoon data were analyzed separately the decline was not significant (p=0.15). Looking at individual species, this decline over the course of the study was significant for the Mealy Parrot, Dusky-headed Parakeet, and all three species of large macaws (Tables 1, 4, Figure 1, 3). However, most species had no strong trend in their numbers, and some, like Chestnut-fronted Macaws, Red-bellied Macaws and Blue-headed Parrots, showed weakly increasing trends (Tables 1, 4, Figures 3, 4).
Dozens of clay licks used by parrots are known from the lowlands of the southwestern Amazon basin (Munn 1992). Fourteen parrot species and 7 other bird species were recorded during data collection. An additional three parrot species are known to use this lick: Dusky-billed Parrotlet (Forpus sclateri), Amazonian Parrotlet (Nannopsittaca dachilleae) and Tui Parakeet (Brotogeris sanctithomae). This puts the total known parrot diversity for this lick at 17 species. This represents the first published diversity estimate for any parrot clay lick. This diversity is of interest because it is known that not all licks are used by all species present in an area. For example the clay lick on the lower Tambopata River on the lands of the Native Community of Infierno is visited regularly by Green-winged Macaws, rarely by Scarlet Macaw and never by Blue-and-Yellow Macaws. The lick on the Manu River near Cocha Cashu Biological Station is visited regularly by Green-winged and Scarlet Macaws but never by Blue-and-Yellow Macaws. Finally the lick on the Madre de Dios River near Blanquillo is visited regularly by Green-winged Macaws, rarely or never by Scarlet Macaws and never by Blue-and-Yellow Macaws (personal observation and discussions with guides and researchers). In all three of these locations the three species of large macaws occur. Rigorously documenting species at each lick is the first step towards determining the physical and chemical factors of the licks that attract the different species.
The numbers reported here for any one species do not represent the number of birds in the area or even the number of birds that took clay. These numbers are only the maximum number seen during a count. The counts were instantaneous scans done every five minutes. In most instances the maximum number of birds on the lick did not coincide with the timing of the censuses. So this index is nearly always an underestimate of the maximum number of birds that took clay at one time. In addition during the peak of early morning clay lick activity there are always large numbers of birds flying and perched in the trees surrounding the lick. Observations suggest that many birds arrive in the trees surrounding the lick but do not descend to ingest clay. For this reason the numbers recorded here are much less than the total number of birds in the area surrounding TRC.
Local guides and researchers have known for some time that weather affects the use of clay licks (Munn 1992). The data presented here document for the first time that psittacids do not visit clay licks during the rain. Even when rainy days are eliminated, weather is still has a significant impact on lick use with more birds coming during sunny and partly sunny conditions than cloudy conditions. The presence of fog reduces the number of birds visiting the lick. These two findings show that future comparisons between different clay licks or comparisons within clay licks between seasons and between years must take into account weather. Failure to account for weather differences could lead to misinterpretations of apparent trends in clay lick use (especially in comparisons of rainy and dry season).
Seasonal changes in clay lick use by macaws and parrots have been known for some time but these trends have never been rigorously documented (Munn 1994). Despite the fact that data were collected during 3 months all within the rainy season, Mealy Parrots and all three large macaws showed significant declines with time. These declines are consistent Munn’s (1994) contention that large macaw use of the licks peaks in August and September and drops to an annual low during May and June. During this same time period the Chestnut-fronted Macaws, Red-bellied Macaws and Blue-headed parrots showed a non-significant but increasing trend showing that not all species may have the same seasonal fluctuations in lick use.
One large question that remains unanswered is where are the birds when they are not at the lick. At present it is unknown if the birds undergo seasonal movements to other areas or if they remain in the area but simply do not visit the clay lick. Answering this question at TRC will require extensive parrot censuses away from the lick throughout the year. Information on this issue may be easier to obtain at Posada Amazonas, another clay lick along the lower Tambopata River. Here anecdotal observations suggest that 100-200 Mealy Parrots remain in the area year round but that it is only during the dry season that they regularly descend in numbers to take clay. Until these observations can be confirmed with data, the question remains open. However if the seasonal trends in clay lick use are not due to migrations of parrots, the differences in seasonal trends would suggest corresponding differences in the foraging ecology of the species involved. As future work continues to document the annual cyclic changes in clay lick use it will suggest new avenues of research that will help us better understand the ecological role clay licks play in determining the foraging niche of different parrot species.
Thanks to my assistants without whom none of this work would have been possible. Bob Wilkerson, Ryan Jones, R. Chris Egan, Leigh Detweiler, Alex Smith, Marc Dragiewicz, Andrés Reatague, Verena Matsufuji, and Fabiana Huaman. Thanks also to the Peruvian Government office of INRENA for permission to carry out this study. We would also like to thank Rainforest Expeditions for their support of this work, especially Kurt Holle, Eduardo Nycander, Mario Napravnik, Patricia Deza, Patricia Herrera, Eliott Caviedes, Maximo Gonzales and Ivette Mendoza.
Diamond, J. M. (1999). Dirty eating for healthy living. Nature 400,8 July 1999.
Emmons, L. and N. M. Stark (1979). Elemental composition of a natural mineral lick in Amazonia. Biotropica 11: 311-313.
Gilardi, J. D., S. S. Duffey, et al. (1999). Biochemical functions of geophagy in parrots: detoxification of dietary toxins and cytoprotective effects. Journal of Chemical Ecology 25: 897-922.
Munn, C. A. (1992). Macaw biology and ecotourism, or when a bird in the bush is worth two in the hand. In New World Parrots in Crisis. S. R. Beissinger and N. F. R. Zinder (editors). Washington, Smithsonian Institution Press: 47-72.
Munn, C. A. (1994). Macaws winged rainbows. National Geographic 185(1): 118-140.
Munn, C. A., D. Blanco, E. Nycander, and D. Ricalde. (1991). Prospects for sustainable use of large macaws in southeastern Peru. The First Mesoamerican Workshop on the Conservation and Management of Macaws, Tegucigalpa, Honduras, Center for the Study of Tropical Birds, Inc.
Table 1: Psitacids seen at the clay lick near Tambopata Research Center, Madre de Diós, Peru between dawn and 07:30. Data were collected on 56 different mornings. The means and standard deviations for each species include only days in which that species was seen. n= the number of days on which each species was seen. A “-” or “+” under Trend indicates the number of individuals at the lick declined or increased as the study progressed. Asterisks indicate statistical significance for ANOVA tests. Only species seen 10 or more times were tested for trends.
Early Morning Totals |
Mean |
stdev |
n |
Trend |
|
Number of species |
5.52 |
3.11 |
56 |
- |
|
Number of individuals |
184.91 |
139.12 |
56 |
-**** |
|
Large macaws |
8.36 |
4.54 |
33 |
- |
|
Red-and-green Macaw (Ara chloroptera) |
2.00 |
1.55 |
6 |
|
|
Blue-and-yellow Macaw (A. ararauna) |
6.03 |
3.43 |
31 |
- |
|
Scarlet Macaw (A. macao) |
2.85 |
1.63 |
27 |
- |
|
Green macawsa |
56.83 |
34.71 |
47 |
+ |
|
Mealy Parrot (Amazona farinosa) |
97.80 |
92.22 |
44 |
-**** |
|
Yellow-crowned Parrot (A. ochrocephala) |
7.48 |
6.50 |
29 |
0 |
|
Blue-headed Parrot (Pionus menstruus) |
26.24 |
23.97 |
42 |
+ |
|
Orange-cheeked Parrot (Pionopsitta barrabandi) |
7.71 |
7.46 |
24 |
+ |
|
Dusky-headed Parakeet (Aratinga weddellii) |
33.76 |
22.50 |
29 |
-* |
|
Cobalt-winged Parakeet (Brotogeris cyanoptera) |
3.00 |
1.41 |
2 |
- |
|
White-eyed Parakeet (Aratinga leucopthalmus) |
25.13 |
21.05 |
24 |
|
|
White-bellied Parrot (Pionites leucogaster) |
3.67 |
1.53 |
3 |
|
|
Blue-headed Macaw (Propyrrhura couloni) |
2.00 |
- |
1 |
aGreen macaws contains the combined data for Chestnut-fronted Macaw (Ara severa) and Red-bellied Macaw (Orthopsittaca manilata) since these could not be easily distinguished while perched on the lick.
* Trend marginally significant at p<0.1
**** Trend significant at p<0.0001
Table 2: Psitacids seen at the clay lick near Tambopata Research Center, Madre de Diós, Peru between 07:30 and 12:00. Data were collected on 34 different mornings. The means and standard deviations for each species include only days in which that species was seen. n= the number of days on which each species was seen. A “-” or “+” under Trend indicates the number of individuals at the lick declined or increased as the study progressed. Asterisks indicate statistical significance for ANOVA tests. Only species seen 10 or more times were tested for trends.
Late Morning Totals |
Mean |
stdev |
n |
Trend |
|
Number species |
2.47 |
1.80 |
34 |
-** |
|
Number Individuals |
26.76 |
23.87 |
34 |
-**** |
|
Large macaws |
34.21 |
18.01 |
24 |
-**** |
|
Red-and-green Macaw |
14.26 |
10.06 |
23 |
-**** |
|
Blue-and-yellow Macaw |
13.22 |
6.13 |
23 |
-*** |
|
Scarlet Macaw |
8.59 |
3.86 |
22 |
-**** |
|
Green macawsa |
2.75 |
1.50 |
4 |
|
|
Mealy Parrot |
5.00 |
5.66 |
2 |
|
|
Blue-headed Parrot |
10.67 |
9.61 |
3 |
|
|
Orange-cheeked Parrot |
4.00 |
- |
1 |
|
|
Dusky-headed Parakeet |
2.00 |
- |
1 |
|
|
White-bellied Parrot |
6.00 |
1.22 |
5 |
aGreen macaws contains the combined data for Chestnut-fronted Macaw (Ara severa) and Red-bellied Macaw (Orthopsittaca manilata) since these could not be easily distinguished while perched on the lick.
** Trend significant p<0.05
**** Trend significant at p<0.0001
Table 3: Psitacids seen at the clay lick near Tambopata Research Center, Madre de Diós, Peru between 12:30 and dusk. Data were collected on 29 different afternoons. The means and standard deviations for each species include only days in which that species was seen. n= the number of days on which each species was seen. A “-” or “+” under Trend indicates the number of individuals at the lick declined or increased as the study progressed. Asterisks indicate statistical significance for ANOVA tests. Only species seen 10 or more times were tested for trends.
Afternoon Totals |
Mean |
stdev |
n |
Trend |
|
Number species |
2.90 |
1.08 |
29 |
-*** |
|
Number individuals |
40.38 |
40.52 |
29 |
- |
|
Large macaws |
33.15 |
15.60 |
26 |
-* |
|
Red-and-green Macaw |
19.84 |
11.22 |
25 |
-* |
|
Blue-and-yellow Macaw |
6.60 |
4.83 |
25 |
- |
|
Scarlet Macaw |
7.73 |
4.06 |
26 |
-** |
|
Cobalt-winged Parakeet |
93.33 |
36.86 |
3 |
* Trend marginally significant at p<0.1
**Trend significant at p<0.01
***Trend significant at p<0.001
Table 4: Psitacids seen at the clay lick near Tambopata Research Center, Madre de Diós, Peru during 27 continuous dawn to dusk observations. The means and standard deviations for each species include only days in which that species was seen. n= the number of days on which each species was seen. A “-” or “+” under Trend indicates the number of individuals at the lick declined or increased as the study progressed. Asterisks indicate statistical significance for ANOVA tests. Only species seen 10 or more times were tested for trends.
Full Day Totals |
Mean |
stdev |
n |
Trend |
|
Number species |
7.41 |
2.62 |
27 |
0 |
|
Number individuals |
235.63 |
159.78 |
27 |
-** |
|
Large macaws |
60.33 |
34.93 |
26 |
-*** |
|
Red-and-green Macaw (Ara chloroptera) |
27.63 |
19.29 |
26 |
-** |
|
Blue-and-yellow Macaw (A. ararauna) |
18.41 |
10.49 |
26 |
-*** |
|
Scarlet Macaw (A. macao) |
14.30 |
7.86 |
26 |
-** |
|
Green macawsa |
53.04 |
39.64 |
23 |
+ |
|
Mealy Parrot (Amazona farinosa) |
80.62 |
68.53 |
21 |
-*** |
|
Yellow-crowned Parrot (A. ochrocephala) |
7.25 |
3.93 |
12 |
- |
|
Blue-headed Parrot (Pionus menstruus) |
33.32 |
32.71 |
19 |
+ |
|
Orange-cheeked Parrot (Pionopsitta barrabandi) |
6.33 |
4.10 |
12 |
+ |
|
Dusky-headed Parakeet (Aratinga weddellii) |
33.35 |
21.56 |
17 |
- |
|
Cobalt-winged Parakeet (Brotogeris cyanoptera) |
51.50 |
62.88 |
4 |
|
|
White-eyed Parakeet (Aratinga leucopthalmus) |
25.11 |
29.57 |
9 |
|
|
White-bellied Parrot (Pionites leucogaster) |
5.40 |
1.34 |
5 |
aGreen macaws contains the combined data for Chestnut-fronted Macaw (Ara severa) and Red-bellied Macaw (Orthopsittaca manilata) since these could not be easily distinguished while perched on the lick.
**Trend significant at p<0.01
***Trend significant at p<0.001
Figure 1: The total number of Mealy Parrots recorded eating clay at the lick adjacent to Tambopata Research Center during 56 mornings (dawn – 07:30) of observation. The decreasing trend is significant (p<0.0001, ANOVA).
Figure 2: Total number of psitacids recorded at the clay lick adjacent to Tambopata Research Center over 27 days of continuous dawn to dusk observation. The declining trend is significant at p<0.01, ANOVA).
Figure 3: The numbers of large macaws recorded at the clay lick adjacent to Tambopata Research Center during 27 days of dawn to dusk observation spread over 2 months. Data are presented for Red-and-green Macaw (Ara chloroptera), Blue-and-yellow Macaw (A. ararauna) and Scarlet Macaw (A. macao) combined. The declining trends for each species individually and all species combined are significant at p<0.01, ANOVA.
Figure 4: The total number of Chestnut-fronted and Red-bellied Macaws (combined) recorded eating clay at the lick adjacent to Tambopata Research Center during 56 mornings (dawn – 07:30) of observation. The data for these two species are combined because they could not be readily differentiated while perched on the lick. The increasing trend is not significant (p>0.5, ANOVA).
Figure 5: The total number of Blue-headed Parrots recorded eating clay
at the lick adjacent to Tambopata Research Center during 56 mornings (dawn
– 07:30) of observation. The increasing trend is not significant (p>0.8,
ANOVA).