Glasgow University Exploration Society

Trinidad Expedition 2001

Final report

Design of t-shirt sold for fundraising for 2001 expedition.

Edited and compiled by Joanna M Smith

Contact: frojo2525@hotmail.com

The beauty and genius of a work of art may be reconceived, though its first material expression be destroyed; a vanished harmony may yet again inspire the composer; but when the last individual of a race of living things breathes no more, another heaven and another earth must pass before such a one can be again.

William Beebe

TABLE OF CONTENTS

CONTENTS	START PAGE
Acknowledgements …..………………………………………..…………………		7
Glasgow University Expeditions to Trinidad…..…………………………..		9
Reports of the Bird Group 2001 …..………………………………………….		11
Investigations into the effects of exploitation and disturbance on rainforest bird populations in Trinidad 2001 – an overview …..……………………………………..		13
The effects of disturbance on rainforest birds …..……………………………………		17
Hummingbird populations relative to hist plant availability …..……………………..		31
Avian species diversity in three types of Tropical rainforest …..……………………..		43
The role of sunning behaviour in temperature regulation in Black Vultures, Coragyps atratus ……………………………………………………………………..		55
Reports of the Turtle Group 2001 …..………………………………………..		57
Insect infestation of Leatherback turtle nests …..……………………………………..		59
Censusuing public opinion on tourism development …..……………………………..		71
Reports of the Tree Frog Group 2001…..…………………………………...		73
Mating behaviour in the frog Phyllomedusa trinitatis…..…………………………….		75
Adhesion in Trinidadian tree frogs…..………………………………………………..		89
Fauna in Bromeliads…..………………………………………..……………………..		101
Appendices
I: Accommodation …..………………………………………..………………………		105
II: Personnel …..………………………………………..…………………………….		106
III: Finances …..………………………………………..……………………………		107
IV: NGS Kids Camp …..………………………………………..……………………		109
V: Field trips …..………………………………………..……………………………		111
VI: Turtle group questionnaire …..………………………………………..…………		113

ACKNOWLEDGEMENTS

Sponsorship: Firstly we would like to extend our appreciation to all those that contributed funding to the Expedition, the Albert Reckitt Trust, the Blodwen Lloyd Binns Trust, The British Ecological Society, The Carnegie Trust, The Dennis Curry Charitable Trust, The Royal Geographic Society, Tescos Ltd., The University Court and The Walter Guinness Charitable Trust. Individual members of the Expedition would also like to thank the following: Glasgow University Postgraduate Scholarships (Jennifer Ward) NERC & Blodwen Lloyd Binns (Joanna Smith) The Explorers Club (Dan Thornham) and The Carnegie Trust for Scotland (Stewart White). For a full breakdown of finances see Appendix III.

In Glasgow: Cheers to Roger Downie for all his help. Glasgow University Finance Department for helping arrange our insurance so efficiently. We would also like to thank all the local businesses and companies that generously provided raffle prizes, especially the Student Travel Association, Scottish Football Association, Los Borrachos, The Stand Comedy Club, The Rainbow Rooms, Tomatin Breweries. Thanks also to Bar-One Ltd. for cutting us a deal on those t-shirts! And last but certainly not least…. thanks to all the long-suffering friends and families that were coerced into buying t-shirts, tickets, toasties and too much tipple in the name of sending us to Trinidad.

In Trinidad: Special thanks to old friends in Trinidad who helped everything run smoothly and who were always there; Tony Dougdeen, Daria St-Hill and Ryan Flemming. At Simla, a huge thank you to Ronnie. Thanks to everyone at Trinidad Governmental Department of Forestry and Wildlife. Special thanks to Nadra Nathai-Gayan, David and Wendy Boodoo. Thanks to all at the University of the West Indies: As usual UWI was unstinting in its support for the expedition from Glasgow, thanks especially to Professor Peter Bacon, Dr Mary Alkins-Koo, Dr Mike Oatham and Karen Duhn. Thanks to everyone at Petrotrin for their friendship and help (especially Roxana for putting up with Nev!). Thanks to all at the Asa Wright Centre, cheers to Painter and Pamela in particular. Thanks to Linda too for helping tidy up after us messy bunch and keeping us entertained. Thanks to Nature Seekers, Marissa, Avanash, Sprang, Dennis Sammy and Anderson Innes for all their help. Thanks to Wynda Chandler and Maria, Christopher and Joel of NGS Kids camp for allowing us the opportunity to get involved. Thanks for nights of hospitality are due to Mister Mac, Cynthia, Dijean and Ali Supersad and Bumpy, Scotty & Walker for teaching us riddim!

GLASGOW UNIVERSITY EXPLORATION SOCIETY EXPEDITIONS TO TRINIDAD: AN INTRODUCTION.

The neighbouring islands of Trinidad and Tobago are the most southerly members of the Caribbean Island chain situated just off the coast of Northeast Venezuela. These contrasting islands have a long history of colonial involvement (being successively owned by Spain, France and most recently Britain) and are now a single nation state within the Commonwealth. Trinidad, the larger of the two islands was known as Lere or ‘land of the hummingbirds' to the native Amerindians.

Once a part of the South American mainland, its flora, fauna and climate closely resemble that of nearby Venezuela. The combination of plains and hill regions, together with a warm, damp climate, made Trinidad suitable for sugar, cocoa, coffee, and rice plantations. The economy of Trinidad today still relies substantially on industrial trade overseas.

This diversity in geography and land use in Trinidad, coupled with its closeness in species diversity to South America, makes the island a rich source for biological research as well as for industry, especially where studies are concerned with conservation. The rich variety of flora and fauna throughout these islands has unfortunately been greatly disturbed by colonial plantations and modern development. However there are still many largely undisturbed habitats available for study including thick swathes of primary rainforest in the south of the island, rain and cloud forest in the Northern Range mountains, beaches used by five species of turtle for egg laying and several large coastal mangrove swamps. Comparative studies of areas in respect to land use can therefore be beneficial to understanding future management strategies for areas that are still pristine at present. As each of these habitats is complex and specialised and supports a number of rare and indigenous species of plants and animals, scientific research is very important. Unfortunately, resources for research within Trinidad are limited at present but a number of local conservation groups are starting to make an impact on public awareness. The Governmental Wildlife Division has been keen to encourage the presence of foreign study groups to aid in the collection of vital data.

The Glasgow University Exploration Society reformed in 1988 after a lapse of around 10 years. It mounted its first expedition to Trinidad and Tobago in June 1989 and since then has gone from strength to strength with more and more students striving to become involved and to gain from the opportunities it offers. Many expeditions have been run each year to destinations such as Cyprus, Patagonia, Ecuador and Iceland with subject matter ranging from zoology and conservation issues to geology and rock formations.

Previous expeditions to Trinidad and Tobago have been run in 1998, 91, 93, 94, 96, 98, 99, and 2000. This year’s expedition was therefore the University’s ninth to the country. We strongly believe that there are considerable benefits in this continuity. As a recurring expedition we strive for good relations and continuity and were therefore fortunate in having had expedition members with previous experience and contacts within Trinidad and Tobago. This helps by returning a constant level of understanding to the country enabling our Trinidad friends and acquaintances to relate to people and not just a name. Additionally, having a large number of first time visitors brings a fresh and enthusiastic perspective to every aspect of biological study undertaken. This is a vital asset, especially in any long-term well-established study and was prevalent in abundance in this years ‘new’ students, who pulled together well as a team in some difficult circumstances, including Jimmy’s dengue and Rach’s broken arm! A full list of members (all of whom made it in one piece) can be found in Appendix II.

For every expedition run to Trinidad reports have been published and circulated to all sponsors and collaborating organisations in Trinidad and Tobago. In addition, many research papers have been published in refereed journals on the entomological and herpetological work these expeditions have carried out. Its not all about work and data collection and we took part in a number of interesting field trips to various other nature centres and music festivals around the island, details of which can be seen in Appendix V.

REPORTS OF THE BIRD GROUP 2001

Bird group at Simla: L to R back row: Gemma Doran, Carrie Finlayson, Kirsty MacNaughton, Dr. Stewart White. Foreground: Neville Broadis + photographer: Jen Ward.

The main body of work carried out by the Bird Group 2001 comprised of Avifaunal Surveys of the Northern Range and the Southern Rainforest, follow-up studies complementing and adding to a database of information collected through the years by expeditions from both Glasgow and Dundee Universities. Additionally, several final year students working within the group have taken on the task of analysing elements of the data collected in the course of the main surveying work and produced ‘mini-studies’ to examine the data with respect to more specific lines of enquiry. Jennifer Ward, a second year Postgraduate student conducting research into the natural sunning behaviour of black vultures, also joined the group for a time.

The bird group would like to give their particular thanks to Ronnie Hernandez, Mac Supersad (and grandsons) in Verdant Vale, Mike & Doreen Oatham, Floyd Hayes at the University of the West Indies and everyone at Petrotrin. Jen would also like to offer her personal thanks to all the members of the turtle group who helped her whilst working at Matura and to Marissa Ramjattan and Abiraj “Sprang” Rambaran for all their assistance and to all the “Natureseekers” turtle rangers, Matura Beach, for their help and advice.

Investigations into the effects of exploitation and disturbance on rainforest bird populations in Trinidad 2001 – An overview.

Dr. Stewart A. White

Tangara mexicana (Simla) (Photo: J.Smith)

Young Black Hawk caught at Morne Bleu

(Photo: C. Finlayson)

INTRODUCTION

The study conducted in the summer of 2001 in Trinidad, West Indies had two main aims. The first part of the study looked at the effect of timber exploitation on bird populations by censusing the birds in three different forest areas, following up a study conducted in 1999. A key issue for tropical rainforests is whether it is possible to extract timber and still retain the biodiversity for which these forests are renowned. Some of the forests of South Trinidad have been managed for decades by two harvesting systems; others are conserved as wildlife sanctuaries. The first area surveyed was managed by the Open Range Method (ORM), which is the traditional method of exploitation. This involves identifying and removing any trees of marketable size and quality, with no thought for replacement or for the damage done to neighbouring trees during the felling process.

The second area surveyed was managed by the Periodic Block System (PBS), which uses a system of area control. An area of 150-300 ha is demarcated as a block and harvesting operations confined within this area for a two-year period. Trained silviculturalists go through the block and physically mark trees that are to be removed. Girth limits are not used and care is taken in selection of trees to be felled (Clubbe and Jhilmit, 1992). Extraction methods are designed so that as little damage as possible is done to the forest and a limited amount of timber is removed from each area. The third area studied was an area of virgin forest.

The second part of the study looked at the effects of disturbance on rainforest birds. At Simla Research Station in the Northern Range Mountains there are two quarries on neighbouring land. These quarries were idle for a number of years but since 1996 they have become active again causing a lot of disturbance to the surrounding area. In addition to the quarries the owners of the land adjoining Simla have been clearing small pockets of forest. A mist-netting study was conducted at Simla in 1995 and 1996 just before the quarries started operating again, as part of a long-term mark-recapture study (Johnston et al, 1997). The 2000 Glasgow University Expedition conducted a similar short study. A further study was conducted in 2001 and results from the four years compared to see if the increased levels of disturbance are having an effect on the numbers and species of birds in the area. Simla was also compared to another site in the Northern Range at Morne Bleu where no disturbance has taken place over the same period.

MATERIALS AND METHODS

The same methods were used in both parts of the study, with a combination of mist-netting, visual observation and recording of calls, using standard techniques as recommended by Bibby et al. (1992) – see following studies for details.

RESULTS

Effects of Exploitation

No significant differences were found between the results of 1999 and 2001 so the two data sets were combined and analysed together. The most heavily exploited area (OMS) gave the highest numbers of individual birds and species and the virgin area the lowest numbers. This is probably due to the canopy being significantly lower in the OMS area meaning that mist netting was sampling a higher proportion of the vertical area and it was easier to observe birds. Species were assigned into three guilds on habitat type utilised, Forest, Edge or Generalist. Using Shannon's index of diversity (H'), the PBS area had the highest diversity of forest species (H' = 2.64), followed by the Virgin area (H' = 2.39) and then OMS (H' = 2.16). This is important as it demonstrates that the PBS system of exploitation may be protecting the species that rely on forest. Diversity in the PBS area being higher than in the virgin area may be an effect of the greater sample size in that area. Similarly, species were divided into feeding guilds and the insectivores looked at in more detail. Many authors have found that insectivores are the most sensitive feeding guild in tropical forest and the first to be affected by disturbance of any sort. As before the PBS area showed the highest diversity of insectivores (H' = 1.78), followed by the virgin area (H' = 1.73) and then OMS (H' = 1.46). This provides more evidence that exploitation by the PBS method may be protecting the diversity of bird species relying on the forest.

Effects of disturbance

No significant differences were found between the results of 1995 and 1996 or 2000 and 2001 so the two data sets in each case were combined and analysed together. For all species at Simla diversity was greater in 2000/01 (H' = 3.02) than 1995/96 (H' = 2.87). Similarly, for Morne Bleu diversity was higher in 2000/01 (H' = 3.02) than in 1995/96 (H' = 2.94). Looking at forest specialist species, diversity was higher in 2000/01 (H' = 2.85) than 1995/96 (H' = 2.71) at Simla and slightly higher in 2000/01 (H' = 2.82) than 1995/96 (H' = 2.77) at Morne Bleu. For insectivores diversity was at higher in 1995/96 (H' = 2.37) than in 2000/01 (H' = 2.23) at Simla and also higher in 1995/96 (H' = 2.63) than 2000/01 (H' = 2.23) at Morne Bleu.

The above results give no clear indication of differences over the period 1995 to 2001. There is a suggestion that diversity of insectivores has gone down in both areas over the period but diversity of forest specialists has risen. There is certainly no suggestion of disturbance having had a major effect on the species of bird being found at Simla. It may be that other factors such as weather patterns are important, or that clear differences may not emerge until a longer period of time has elapsed since the onset of disturbance. It is intended to return and repeat the surveys in the future. The results from both the studies described above are being analysed in greater detail for submission to refereed journals.

REFERENCES

Bibby, C.J., Burgess, N. D. & Hill, A. (1992) Bird census techniques. London: Academic Press.

Clubbe, C.P. & Jhilmit, S. (1992) a case study of natural forest management in Trinidad. In: F.R. Millar and K.L. Adam (eds.), Wise management of tropical forests. Proceedings of the Oxford conference on tropical forests. Oxford Forestry Institute, Oxford.

Johnston, J.P., Peach, W.J., Gregory, R.D., and White, S.A. (1997). Survival rates of tropical and temperate passerines: A Trinidadian perspective. American Naturalist 150 (6):771-789

The effects of disturbance on rainforest birds: Arima Valley

Kirsty MacNaughton

ABSTRACT

Two locations in the Northern Mountain Range of Trinidad; Simla and Morne Bleu, were studied to ascertain if there were any differences in avifaunal diversity between the two sites as well as between the years 1995 and 2001. In particular, for Simla, comparisons were made between the two years to determine whether there had been any impact on avifaunal diversity due to the effects of human disturbance in this area.

Mist netting and sound recording was used to assess avifaunal diversity at both sites. Five mist nets were erected each day for two blocks of four days. Sound recordings were carried out during the dawn chorus from 0730 – 0800. Birds were identified and their feeding guild noted. The largest number of species was caught at Simla in 1995 followed by Morne Bleu in the same year suggesting an obvious decline in the number of species present between that year and 2001. There were no significant differences between feeding guild composition or habitat type at the two locations between the two years.

INTRODUCTION

Human activities in an ecosystem result in a restructuring of its communities through a variety of means. Focus has recently turned towards the cutting of moist tropical forests as a disturbance potentially responsible for major species loss (Myers, 1979, 1980; Lovejoy, 1980; Lovejoy and Brash, 1984; Raven, 1984). Several studies have assessed the ability of Neotropical birds to survive in conditions of habitat fragmentation, where only isolated patches of forest remain amidst cleared land (e.g. Lovejoy et al 1984, 1986, Willis 1979). Deforestation in the tropics often involves the conversion of landscapes with continuous forest to ones with remnant forest patches amongst non-forest vegetation. This manipulation of the forest environment has consequences for biodiversity at the landscape scale and forest-fragment level. Leck (1979) reported a loss of 25 species of bird from a highly isolated 87-ha forest fragment (Ecuador) in just 5 years. Studies over a longer period of time show even more of a marked impact on species numbers. A third of the species in a fragmented area of montane forest in the Colombian Andes has been lost over an 80-year period (Kattan et al. 1994).

This report aims to examine and compare the avifauna of two forest mosaics between the years 1995 and 2001 in North Trinidad. Many Neotropical rain forest birds adapt to heterogeneities of the environment by moving along habitat and microclimate gradients (Karr and Freenark 1983). If this fails to occur, and birds are excluded from a habitat by some intrinsic feature, such as unsuitability of microclimate or absence of suitable food resources then species that would be most affected by fragmentation would tend to be specialists of forest interiors. They are often unwilling to cross open areas (Diamond 1973, Willis 1974) and tend to be insectivorous species. By examining feeding guilds then, effects of disturbance may be assessed. Whereas many species may move away from a disturbed area, others will congregate within it (Johns 1986, 1989, Terborgh and Weske 1969; cf. Driscoll and Kikkawa 1989). The extent to which birds are restricted to particular habitats will indicate their ability to persist with increased disturbance (i.e. reduced access to intact forest).

Different types of vegetation present in a habitat can affect the bird species found within it. Some bird species are specialists and will only be present if particular plant species are also present. Vegetation, as well as structure, can also have an influence on particular species and determine whether they are present or absent. A bird’s foraging strategy can very often depend on vegetation density and vegetation will affect the birds found within a forest in terms of behavioural and morphological groups found therein (Pearson, 1975). For example, larger birds would tend to be inhibited in dense forests whereas smaller birds would be predicted to be present in large numbers, because of a greater number of insects being present.

This study proposes to examine the avifaunal diversity in Simla and Morne Bleu in Trinidad. The investigation aims to determine whether there has been any impact from disturbance in the Simla area concentrating on any differences in the six-year period between 1995 and 2001 and if there was, how it could be explained.

Simla and Morne Bleu were compared in 2001 as well as between 1995 and 2001. Techniques used for this study included mist netting and sound recording.

METHODOLOGY

Study sites

Two sites in the tropical forest of northern Trinidad were considered in this study:

Site 1 – Simla Research Station in the Arima Valley of the Northern Mountain Range foothills. This is within an area of seasonal deciduous rainforest located at around 250 metres above sea level comprising largely secondary growth of developing canopy trees towering over the cocoa and citrus trees which once were part of a plantation. The area is near two active quarry sites, which are about 200 and 800 metres away respectively. Both are worked from early morning to late at night i.e. approximately 0500 till 2200, and frequent blasting can be heard. Also, a little way to the north, some forested land has been cleared for agriculture.

Site 2 – Morne Bleu Ridge in the Northern Mountain Range, 6.5 km north of Simla. This is located in lower montane forest at around 750 metres above sea level and covers an area of approximately 7.5 ha. This higher elevation is characterised by thicker cloud cover and generally wetter conditions. There is little evidence of human disturbance in this location.

Mist Netting

At both selected locations 25 mist net sites were arranged in a grid formation. Vegetation was cleared from each site to produce rectangular zones of approximately 20 m by 2 m on the ground. Overhanging vegetation was also removed to avoid any tangling of the nets.

Each day, five sites were chosen randomly to avoid any bias, e.g. birds learning the locations of the nets. The nets were erected on poles and secured with guy ropes attached to any trees or vines nearby. They were raised early in the morning at around 0630 and the number of hours each individual net was up was recorded each day. Checking occurred every hour and any birds caught were placed in bird bags and taken to be identified, measured and weighed. This processing involved measuring bill length, back of head to tip of bill and tarsus length (all measured to +/-10 micrometers using callipers) as well as wing length in millimetres and the weight of each bird in grams. Each bird was also ringed as part of a long-term mark and recapture program. Ringing ensured that any birds captured more than once during the eight-day study period could be accounted for to avoid any biases.

Sound Recording

Sound recordings were carried out each morning at 0730 for a 30-minute period using a Marantz portable sound recorder and a direction sensitive microphone. A total of three hours of bird sound was obtained at each site. The sites chosen for sound recording were selected randomly to avoid any bias. Using a reference tape of Trinidad and Tobago bird sound, it was possible to identify the individual birds on the sound recordings.

Vegetation assessments

An indication of the basic forest structure at the two sites was determined by carrying out a number of different measurements. Five of the mist net sites were chosen randomly at each location and the following measurements were taken:

1) The circumference, at chest height, of 10 trees with a circumference > than 60 cm.

2) The approximate height of the same 10 trees using a clinometer.

3) The distance to the furthest of the 10 trees to give an estimate of tree density.

4) Percentage vegetation cover at ground, five metres, mid canopy and canopy levels.

Data Analysis

For bird data, abundance in terms of captures per hour was calculated for each study site. Various diversity indices were calculated, i.e. species richness, Simpson index and Shannon index together with the similarity coefficients Jaccard index, Sorenson index and the Sorenson quantitative (Thiollay, 1994; Magurran, 1988).

RESULTS

A Comparison Between Simla and Morne Bleu 2001: Bird Data.

Figure 1 shows the total number of species caught at the two sites and the overall total number of species caught between the sites. In Morne Bleu 39 different species were caught compared to 32 caught in Simla and an overall 53 species were caught in total. Of the 53 species caught, 16 were classed as canopy species and one was classified as an aerial species (Hayes and Samad, 1998) and so would not normally be caught in low mist net areas. The remaining 36 species were classified as being under-story species (Hayes and Samad, 1998) and would normally be caught in low-level mist net areas.

Figure 1: Total number of species caught at the two sites species for the year 2001.

From the graph, it seems that more species were caught in Morne Bleu than in Simla; however, it appears that more individuals were caught in Simla. At Morne Bleu 269 individuals were caught compared to 296 at Simla. As the amount of time mist netting may account for the greater number of individuals in one location number of captures per mist net hour using capture and re-capture data were also considered.

	Morne Bleu	Simla
Number of mist net hours	296	316
Number of captures	317	338
Captures per hour	1.07	0.93

Table 1: The numbers of mist net hours, captures and captures per hour at each study site.

Data showing rates of birds caught in the mist nets can determine species’ abundance for the two areas. From the results in Table 1, it can be said that species abundance was higher in Morne Bleu.

Figure 2 shows the cumulative number of species caught over the full eight days spent at Simla and Morne Bleu respectively. It can be seen that numbers of species caught increases with the number of days studied. The graphs show a greater increase in species numbers over the first few days with the appearance of new species beginning to decrease as the days go on. It is to be expected that after a certain amount of time the graphs would reach an asymptote after which no new species would be recorded. The rate at which new species were caught in the two study areas was similar however more species were caught in Morne Bleu.

Additionally to the mist netting data, 35 species were detected using sound recording at Morne Bleu and 32 at Simla. 25% of those recorded at Morne Bleu were exclusive to that site, compared to 18.75% of species at Simla. There is a statistically significant difference between species appearing on sound recording material for Simla and Morne Bleu 2001. Spearman’s Rank Correlation = -0.389, p =0.012, 65 d.f.

Diversity indices seek to characterise the diversity of a sample or community by a single number. Ecologists are interested in ecological diversity particularly as the measure of diversity is often used as an indicator of the well being of a particular ecological system.

Diversity indices used in this study were four of the most common, frequently and widely used. They are: Species Richness (S), Simpson Index (D), Shannon Index (H’) and Equability of Species (J’). Species richness (S) illustrates the total number of all species regardless of abundance. Simpson index (D) gives the probability of any two individuals drawn at random from an infinitely large community belonging to different species. The Shannon Index (H’) is a measure of the average degree of ‘uncertainty’ in predicting to what species an individual chosen at random from a collection of S species and N individuals will belong (Magurran 1988). The Equability of Species (J’) gives an idea of how equally abundant the species are. High evenness, when species are equal or virtually equal in abundance, is conventionally equated with high diversity. (Magurran, 1988). Table 2 shows results of these indices for Simla 2001 and Morne Bleu 2001.

Site	Diversity Indices
	Species Richness (S)	Simpson Index (D)	Shannon Index (H’)	Equability of Species (J’)
Simla 2001	33	7.73	2.84	0.82
Morne Bleu 2001	39	6.62	3.02	0.82

Table 2: Diversity Indices for Simla and Morne Bleu 2001.

It appears that species richness (S) and the Shannon index are higher in Morne Bleu whereas Simla shows a higher Simpson index (D). Both locations have a similar equability of species (J’). This suggests that there is a more diverse avifauna in Morne Bleu for this year.

Table 3 shows similarity coefficients between sites. The three selected are the Jaccard index (C_j) and two different forms of the Sorenson index, qualitative (C_s) and quantitative (C_N). These were chosen because they are recognised to be the most useful and widely used (Magurran, 1988). The closer the value of the indices to 1, the more similar the sites. The Jaccard and Sorenson indices indicate that the two habitats for this year were quite dissimilar; the Sorenson quantitative indicates much more similarity.

Habitat	Similarity indices
	Jaccard (Cj)	Sorenson (Cs)	Sorenson (Cn)
Simla vs. Morne Bleu	0.33	0.50	0.81

Table 3: Similarity coefficients for Simla 2001 and Morne Bleu 2001.

Table 4 & Figure 3 compare the number of species caught from mist netting and their feeding guild composition. The results show that for Simla, the greatest number of species was omnivorous. 42% of 33 species were omnivores compared to 23% of 39 species for Morne Bleu. Results from Morne Bleu show that the majority of species were insectivorous with 38% of 39 species being insectivorous compared to 30% of 33 for Simla.

Figure 3: Number of species in each feeding guild for Simla and Morne Bleu, 2001

A Chi-squared test was used to compare the composition and frequencies of feeding guilds between Simla and Morne Bleu 2001. The chi-squared results state that there is no significant difference between feeding guilds in the two habitats. (Chi-squared = 5.067, df = 4, P-value = 0.2814)

	Number of species
Feeding guild	Simla	Morne Bleu
Nectivore	5	7
Frugivore	2	7
Omnivore	14	9
Insectivore	10	15
Carnivore	2	1

Table 4: Feeding guilds of species caught in 2001.

There is however a tendency towards greater numbers of omnivores found at Simla and insectivores at Morne Bleu, omnivore numbers would be expected to be higher in more open disturbed habitat and so the patterns shown may indicate that disturbance may be having a preliminary effect on avifaunal community structures.

A Comparison Between Simla and Morne Bleu 2001: Vegetation.

This study found very little difference between the mean tree height (22.14m for Simla and 22.66m for Morne Bleu) and mean tree diameter in Simla and Morne Bleu (0.4648m for Simla and 0.4786 for Morne Bleu). There is also a difference in mean tree densities at each of the sites. It appears that Simla (168 trees/hectare) has a greater number of trees per hectare compared to Morne Bleu (114 trees/hectare).

Percentage vegetation cover values were arcsine transformed so that the data could be analysed. For all levels, ground to canopy, there is a trend towards greater percentage cover in Morne Bleu than in Simla. However statistically significant differences were detected only for vegetation cover at ground level: t = -2.62, p < 0.035 (Figure 4). There appears to be no significant difference for the other three levels of forest compared.

Figure 4: Vegetation cover at ground level in each study site with 95% c.i.

A Comparison of Simla and Morne Bleu Between 1995 and 2001.

Data used for this section consists mainly of mist net data, as there is no sound recording information available from the two locations for 1995. There was a greater number of species captured and identified in 1995 compared to 2001 in both of the study sites, however times spent at the sites were different between the two years and so capture rates need to be compared also (Table 5). The greatest capture rate occurred in Simla and Morne Bleu in 1995 and for the 2001 data capture rate was higher in Morne Bleu. This indicates that abundance of birds was higher in 1995 and that for the year 2001 abundance was higher in Morne Bleu.

		Study site
			Simla 1995		Simla 2001		Morne Bleu 1995		Morne Bleu 2001
Mist net hours	392			316		392		296
No of captures	641			338		643		317
Captures per hour	1.64			0.93		1.64		1.07

Table 5: No of mist net hours, captures and capture rate for Simla & Morne Bleu in 1995 and 2001.

Table 6 shows the feeding guild composition of birds caught at the different study sites for the two years. The majority of birds tend to be insectivores or omnivores. In Simla, the majority of birds caught from mist net data were omnivores (43% of species in 1995 and 42% of species in 2001). In Morne Bleu, the majority of birds caught from mist net data were insectivores (40% of species in 1995 and 38% of species in 2001). Results from chi-squared tests show that there is no statistically significant difference between feeding guilds in Simla and Morne Bleu between the years 1995 and 2001.

	Number of Species
	Simla 1995	Simla 2001	Morne Bleu 1995	Morne Bleu 2001
Nectivore	6	5	6	7
Frugivore	3	2	7	7
Omnivore	21	14	14	9
Insectivore	16	10	18	15
Carnivore	3	2	0	1

Table 6: Feeding guilds of species caught in the two different sites at two different years.

Rarefaction analysis is a technique developed by Sanders and is used to calculate the number of species expected from different communities if all samples are reduced to a standard size. It helps to overcome the incompatibility of unequal sample sizes from various communities. Figure 5 shows that at Simla there has been a marked difference in the number of expected species between the two years. At Morne Bleu on the other hand, total expected numbers of species between the two years appears to be much the same.

Figure 5: Rarefaction analysis: (A) Simla and (B) Morne Bleu in 1995 (open) and 2001 (filled)

Diversity indices were calculated as before: Results indicated that both species richness and Shannon index are lowest at Simla in 2001 (S = 33, H’ = 2.84) at all sites and across all years. This is in direct contrast to 1995, when the most diverse avifauna was found around Simla.

DISCUSSION

Many studies and projects have focused on the affects of various forms of disturbance on species diversities in tropical rainforests (Johns, 1991; Mason, 1996). Disruption of habitat as a result of human disturbance, in whatever form, will tend to alter the pattern of food availability and may constrict the amount of habitat space that can be occupied particularly by large arrays of similar species. Fewer species may occupy the area as a result. (Terborgh and Weske, 1969). Avian understorey communities undergo drastic changes in species composition following fragmentation. Studies by Willis (1979) in the state of Sao Paulo, Brazil, Leck (1979) in Ecuador, and Terborgh (1974) Willis (1974), Willis and Eisenmann (1979) and Karr (1982) on Barro Colorado Island in Panama have also reported the disappearance of various species of birds following fragmentation or insularisation of once continuous tropical forest into pockets of up to 1,500ha.

Birds may respond to a variety of environmental factors, such as changes in microhabitat, competition, predation and forest structure. As many of these factors are correlated, identifying the specific reasons for changes in species abundance is far from straight forward.

Species numbers and certain indices from mist net data are highest in 1995. A greater number of species were found in Morne Bleu with the rate of capture also higher for this area in 2001. Evidence from study results appears to point to vegetation playing little role in diversity differences. With the exception of estimated ground vegetation cover, there was no significant difference in respect of any other vegetation component. Variations in canopy cover would have an effect on the amount of light entering the forest interior, influencing the amount of understorey cover and therefore, the amount of microhabitats. It might be expected that, with a more open canopy, the forest would have a higher diversity. However, in this study, the opposite was true. Simla had a greater density but a lower diversity as compared with Morne Bleu.

Feeding guild composition appears to be as expected when looking at the results from capture data only. In this study, there were no significant differences in composition of feeding guilds between habitats although it was clear that the majority of birds caught in Simla were omnivores, with insectivores dominating in Morne Bleu. There appears to have been a decline in numbers of insectivores caught in Simla between 1995 and 2001 and perhaps, with a larger sample size, this decline would be more apparent. Understorey insectivores are among the species group particularly prone to habitat disturbance (Johns, 1987). Their relative abundance in disturbed habitats generally reflects both the loss of understorey vegetation and its associated insects and microclimate changes resulting from a decrease in canopy cover. A possible explanation for the high number of insectivores in Morne Bleu is the ant colonies that were often present at the mist net sites - an easy meal for insectivorous birds. Thiollay (1994) found that insectivores dominated in the understorey probably as a reflection of food availability. A greater number of insectivores present at Morne Bleu are expected because of a larger vegetation cover at ground level. Arthropod density is correlated with percentage ground cover (Nummelin, 1989) and consequently a greater number of insectivores might be anticipated with an increase in percentage ground cover and with significant differences between habitats, large variations in numbers of insectivores would be expected (Nummelin, 1989). In this study, this was not the case. In general, insectivores occupy more finely segregated niches and range less widely to other species. They are restricted to the understorey meaning specialists may not persist in heavily disturbed areas and consequently are less resilient than frugivores and folivores to the effects of disturbance (Johns 1985,1986). However, generalist species that can incorporate fruit into their diet are expected to be more successful (Johns, 1991).

Omnivores tend to have a more variable diet and because of this, they may be maintained in heavily disturbed areas. This evidence suggests reason for the high number of omnivores present in Simla even after only six years of disturbance in the area. A longer duration of study would show more convincing results.

Nectivores also usually persist well in disturbed forests where understorey plants respond to increased light levels resulting in an abundance of flowering plants (Johns, 1989). This is reflected in the study, having few differences between numbers of nectivorous species caught in the two habitats. Again, with a larger sample size, this is what would be expected. Hummingbirds are particularly successful in disturbed forests of Venezuela, probably as a result of increased floral resources.

Further evidence to support the concept of insectivorous species being most at risk in disturbed forest fragments is shown in the sound recording data. The data demonstrates that for those species with the greatest differences in ranks between the sites, more than half were insectivores (six out of nine species). For each insectivorous species, a larger abundance was found in Morne Bleu. Although there were no significant differences between species habitat types in the two areas, results from statistical analysis do show a convincing trend. It indicates that with a larger sample size and a longer study period a significant difference in habitat types amongst species caught would be expected for the two locations.

Results from rarefaction analysis in the data analysis section are persuasive. They demonstrate that for the disturbed Simla region there appears to be a smaller total expected number of species in 2001 compared to the number of species expected in 1995. Rarefaction analysis for Morne Bleu for the two years shows little difference in total expected number of species suggesting that disturbance in the Simla region may indeed be influencing species numbers in the area. Perhaps, over a much longer study period and with a larger study area, these results may become more even more convincing.

REFERENCES

Driscoll, P.V. and Kikkawa, J. (1989) Bird species diversity of lowland tropical rainforests of New Guinea and Northern Australia. Pp. 123-152 in Harmelin-Vivien, M. L. and Bourliere, F. (eds) Vertebrates in complex tropical systems Springer – Verlag. New York. 200 pp.

Hayes, F. E., Samad, I. (1998) Diversity, abundance and seasonality of birds in a Caribbean pine plantation and native broad-leaved forest in Trinidad, West Indies. Bird Conservation International. 8: 67-87.

Johns, A. D. (1986) Effects of selective logging on the ecological organisation of a peninsular Malaysian rainforest avifauna. Forktail 1: 89-105.

Johns, A. D. (1988) Effects of ‘selective’ timber extraction on rainforest structure and composition and some consequences for frugivores and folivores. Biotropica 20 (1): 31-37.

Johns, A. D. (1991) Responses of Amazonian rainforest birds to habitat modification. Journal of Tropical Ecology. 7: 417-437.

Karr, J. R. (1982) Avian extinctions on Barro Colorado Island. Panama: a reassessment. The American Naturalist 119: 220-239.

Karr, J. R. and Freemark, K. E. (1983) Habitat selection and environmental gradients: dynamics in the ‘stable’ tropics. Ecology 64: 1481-1494.

Kattan, G. H., Alvarez-Lopez, H. and Giraldo, M. (1994) Forest fragmentation and bird extinctions: San Antonio eighty years later. Conservation Biology. 8: 138-146.

Leck, C. F. (1979) Avian extinctions in an isolated tropical wet-forest preserve, Ecuador. Auk, 96:343-352.

Lovejoy, T.E. (1980) A projection of species extinction. In the Global 2000 Report to the President, ed. By G. O. Barney, 328-331. Washington, Dc, Council on Environmental Quality.

Lovejoy, T. E., Rankin, J. M., Bierregaard, R. O. Jr, Brown, K. S. Jr, Emmons, L. H. and van der Voort, M. E. (1984) Ecosystem decay in Amazon forest remnants. Pp 295-325 in Nitecki, M. H. (ed) Extinctions. University of Chicago Press, Chicago, Illinois. 354 pp.

Magurran, A. E. (1988) Ecological diversity and its measurement. Croom Helm.

Mason, D. (1996) Responses of Venezuelan understorey birds to selective logging, enrichment strips and vine cutting. Biotropica 28 (3) 296-309.

Myers, N. (1979) The sinking ark. Oxford, Pergamon Press.

Pearson, D. L. (1975) The relation of foliage complexity to ecological diversity of three Amazonian bird communities. Condor, 77, 453-466.

Terborgh, J. and Weske, J. S. (1969) Colonisation of secondary habitats by Peruvian birds. Ecology, 50: 765-782.

Thiollay, J. M. (1994) Structure, density and rarity in an Amazonian rainforest bird community. Journal of Tropical Ecology. 10: 449-481.

Willis, E. O. (1974) Population and local extinctions of birds on Barro Colorado Island, Panama. Ecological Monographs. 44: 153-169.

Willis, E. O. (1979) The composition of avian communities in reminiscent wood lots in southern Brazil. Papeis Avulsos Zoologica, Museu de São Paulo. 33: 1-25.

Hummingbird populations relative to host plant availability.

Neville Broadis

White-necked Jacobin (Florisuga mellivora) Photo: J. Smith

ABSTRACT

It is well known that hummingbirds feed on nectar secreted by flowers of many tropical plants. In some cases, a close relationship occurs whereby the hummingbird will only feed on one type of plant. The aim of this study was to determine whether abundance of certain plant species in a tropical rainforest caused the abundance of certain hummingbird species. Mist netting was carried out to obtain a sample of the population of birds at sites in primary and secondary growth forest. Differences in these two types of forest are manifest in the types of flora and fauna found in each. At Morne Bleu (primary forest), containing fewer H. bihai, fewer G. hirsuta were captured per hour suggesting that perhaps G. hirsuta feeds on this plant. Results must be treated with care as they are lacking in sample size. However, used with other data collected on hermit hummingbirds, results may prove useful.

INTRODUCTION

Many animals are involved in mutualistic relationships with their food source. One of the most striking examples of these types of mutual relationships is that of the hummingbirds and their nectar producing flowers. In the mainland tropics, many species of plant provide nectar for hummingbirds – which in turn provide the plants with a pollinator. Indeed hummingbirds and their food flowers have such a close relationship that the morphology of both has evolved to suit the other (Stiles, 1975). In some cases, the result has been that the plants have specialised morphologically and physiologically to the extent that only a single bird species can legitimately exploit its nectar resource (Linhart & Feinsinger, 1980). This mechanism may have evolved in response to the interspecific competition between several plant species in the community. In many breeding systems, not only does the plant require a pollinator but also it needs also to ensure that the visitor moves on to a conspecific plant in order to complete pollination. Therefore the tropical hummingbird species and their foraging patterns can affect the structure and even the communities of plant populations (Feinsinger et al, 1982). This study aimed to assess the impact of birds and plants on the structure of floral and faunal communities in two sites in Trinidad.

The north end of Trinidad is dominated by the Northern Range, a chain of mountains reaching up to 1,000 metres in places. It is a continuation of the coastal cordillera of Venezuela and is the location for this study. The vegetation of the Northern Range is highly varied, due to differences in altitude and historical land use. In places the Northern Range, and the primary rainforest found there, has remained untouched by man. Lianas, vines, epiphytes of numerous species, bromeliads and ferns are extensive throughout the Northern range (Marshall, 1939; Beard, 1946) but are predominantly more prolific in the primary forest areas. However, second growth forests are quickly becoming the most abundant types of forest as a result of human disturbance. Trees of mixed age, and the type of vegetation that can quickly take over following logging, farming and other disturbance processes, characterise this type of forest and in many areas disturbed habitat plants such as bamboo are found, along with the introduced shade tree Erythrina poeppigiana (Feinsinger & Swarm, 1982) and several Heliconia species.

Humming-bird pollinated plants are characterised by the possession of flowers of bright red colouration with a continuous large supply of available nectar (Wolf et al, 1976). Other characteristics of ornithophilous flowers are narrow tubular corollas, lack of odour, and thickened corolla basal tissues to protect against illegitimate nectar tapping (Baker, 1961; Snow & Snow, 1980; Stebbins 1989; Hurlbert et al, 1996) all of these characteristics can be seen to dominate in the flowers of many plants in the Northern Range with Heliconias (Musaceae) and bromeliads (Bromeliaceae) being the most prevalent of the ornithophilous groups found.

Most Heliconia species cannot tolerate deep shade (Linhart, 1973) and so whilst they are found (at least to some extent) in virgin forest, they are most abundant in second growth (Wolf et al, 1976). In primary forest they exist mainly as temporary inhabitants of light gaps created by fallen trees. One of the most numerous species of Heliconia in the study area was Heliconia bihai, a single clump of which is capable of meeting the daily nectar requirements of one hummingbird (Stiles, 1975).

The second important family, the Bromeliaceae, represents an extremely diverse group of vascular plants (Smith et al, 1986). Contained within the family are both terrestrial and epiphytic forms (Benzing, 1980). They can grow almost anywhere on a tree and at both study sites were seen at heights of up to 20 metres. Snow & Snow in 1972 carried out a survey of the plants used by hummingbirds. In Table 1, the bromeliad species that they found as important are listed (Snow & Snow, 1972). Other plants also capable of being utilised by hummingbirds, which are found in the Northern Range, are also shown.

Table 1: Ornithophilous plants

Bromeliaceae*	Non-bromeliad**
Aechmea dichlamydea	Calathea casupito (Marantaceae)
A. fendleri	Centropogon cornutus (Campanulaceae)
A. nudicaulis	Chrysothenus pulchella (Gesneriaceae)
Billbergia pyramidalis	Costus scaber (Zingiberaceae)
Gravis aquilega	Heliconia bihai L. (Musaceae)
Guzmanai monostachia	Heliconia hirsuta L.f. (Musaceae)
T. fasciculate	Monotagma spicatum (Marantaceae)
Vriesia procera	Psycotria poeppigiana (Rubiaceae)

* From Snow & Snow (1972) **National Herbarium of Trinidad and Tobago.

METHODOLOGY

Study Sites

The fieldwork for this study was carried out from June to August during the University of Glasgow Trinidad Expedition 2001. The expedition was based at Simla Research Station in the Arima Valley of the Northern Range (10⁰ 42' N, 61⁰ 18'W) and this provided the first site for the study. Located at the base of the northern range, Simla sits at 250m above sea level. The terrain is irregular; small cliffs and steep slopes create a canopy of mixed heights.

The second site is located approximately 6.5 kilometres north of Simla, at the head of the valley. It is situated on the Morne Bleu ridge and again covers a 7.5 ha area. It is the highest part of the Arima Valley and its elevation of 750m above sea level affords it a greatly increased rainfall. The steep slopes down from the ridge contain little or no cliffs. Subsequently, the canopy height is more stable with a mean height not significantly different to Simla. The ground cover is denser and contains fewer of the colourful plants that attract nectivorous birds.

Mist Netting

Using 5 nets of size 18m x 2.5m, with a mesh size of 36mm, netting was carried out separately in 4-5 day blocks at the two locations. Twenty-five net locations were available at both sites. Each of these was allocated an individual code. These codes were then used to randomly select which net locations would be operated on which day. After roughly clearing any protruding undergrowth from the net location, a net would be erected. Nets were checked hourly and any birds removed to bags. All birds caught were leg ringed with a numbered aluminium ring, checked for moults and brood patch, measured and released. In the event of a recapture, the ring number was noted and a relevant history number was assigned. The following measurements were taken; head length, from the back of the head to the base of the bill; bill length; wing length; maximum flattened cords; and weight.

Plant Data Collection

Using an 18m² quadrat, a survey of the flower types available for feeding hummingbirds and density at each net site could be estimated. These quadrats were placed at each net location at the two study sites. In each quadrat the number of flowers at ground level were counted and a sample collected for identification. Any plants that were out of reach (mainly bromeliads) were counted but not collected. The total area for each set of 25 quadrats combined to give 8.1 km². Flowers were taken to the National Herbarium at the University of the West Indies, St. Augustine, Trinidad to be identified by the curator, Doreen Oatham.

RESULTS

The Hummingbirds

A total of 7 hummingbird species were netted during the study (Table 2) all of which are resident throughout the year and are known to breed in the Northern Range (ffrench, 1991). Of the 7 species captured, 3 were hermit hummingbirds. All hermits are specialised trap-liners that often follow exact routes visiting undefended and isolated flowers. It has been observed that G. hirsuta may occasionally hold territories but the majority of published documentation suggests its main pattern of foraging is trap-lining (Cotton, 1998). The 4 other species captured in the study area were all typical hummingbirds. Unlike the hermit hummingbirds, these birds are known to frequently hold territories. The hummingbird A. tobaci was found only at Simla and the only capture of F. mellivora, a large non-hermit was at Morne Bleu.

Species	Number Caught at Simla	Number Caught at M. Bleu	Weight (g)	Wing (mm)	Culmen (mm)	Head (mm)
Rufous- Breasted Hermit Glaucis hirsuta	43	10	7.1 (0.007)	62.7 (0.035)	34.5 (0.016)	46.2 (0.014)
Green Hermit Phaethornis guy	17	24	6.2 (0.003)	63.5 (0.029)	42.6 (0.024)	55.7 (0.016)
Little Hermit Phaethornis longuemareus	1	1	3.0 (0.013)	45.5 (0.031)	28.0 (0.005)	38.4 (0.043)
White-chested Emerald Amazilia chionopectus	20	8	4.6 (0.004)	53.5 (0.028)	22.9 (0.022)	35.4 (0.011)
Copper-rumped Hummingbird Amazilia tobaci	12	0	4.5 (0.005)	52.3 (0.050)	21.4 (0.026)	33.8 (0.020)
Blue-chinned sapphire Chlorestes notatus	9	9	3.8 (0.005)	50.3 (0.037)	20.3 (0.023)	32.0 (0.017)
White-necked Jacobin Florisuga mellivora	0	1	8.5	73	23.8	40.7

Table 2: Biometrics of hummingbirds caught in study period: Mean (95% c.i.)

As a result of the difference in net days for the two study sites, results were standardised by calculating the number of birds caught per hour (Table 3). Mist netting was carried out for 12 days (381.5 hours) at Simla but only 8 days (296 hours) at Morne Bleu. Since it is reasonable to assume that frequency of capture reflects actual abundance then number of birds caught per hour can be compared for the two study locations. Certain species have been omitted from the following results, as they were not captured regularly or at both sites. Using statistical tests, the abundance of each of the species shown in Figure 1 can be assessed for differences.

	Mean number of birds caught per hour (95%c.i.)
Species	Simla	Morne Bleu
Glaucis hirsuta	0.11 (0.067)	0.04 (0.02)
Phaethornis guy	0.04 (0.017)	0.08 (0.04)
Amazilia chionopectus	0.07 (0.072)	0.03 (0.02)
Chlorestes notatus	0.02 (0.014)	0.03 (0.03)

Table 3: Capture rates of four species of Hummingbird at study sites.

Statistical tests were performed to compare numbers of birds caught per hour for all four species. The only species for which a significant difference was found between the two sites was G. hirsuta, which was caught at significantly higher rates at the Simla site (Mann-Whitney U: W = 153.00, p =0.041)

Figure 1: Hummingbird capture rates at Simla and Morne Bleu.

The Plants

Table 4 summarises the plant abundance found at both study sites. Bromeliads are a family that represents a wide variety of species in Trinidad. In the tables, there has been a distinction made however, between those that occur from the ground up to the height of the nets and those that occur higher than the nets (2.5 metres). This allows for the analysis of the feeding heights of those hummingbirds feeding on bromeliads. Heliconia numbers refer to numbers of bracts counted in each net site area. In some species of Heliconia, the flowers may only live for a day or less and it is common to find them wilting by the early afternoon, thus counting the bracts gives a more representative idea of the nectar resource available as the areas were sampled over a number of days.

	Simla			Morne Bleu
	Tree bromeliads	Ground bromeliads	Heliconias	Tree bromeliads	Ground bromeliads	Heliconias
Mean nos	11.74	0.83	41.39	9.68	2.84	3.52
Totals	270	19	952	242	71	88

Table 4: Summary plant abundance figures

There is a highly significant difference (p = 0.001) between the numbers of ground Bromeliads counted at the two study locations (t-value = 3.62) with more found at Morne Bleu. There is a highly significant (p = 0.0005) difference between the numbers of H. bihai counted at the two study locations (W = 699.5) with more found at Simla. Figure 2 shows the mean numbers of the plant types counted along with their confidence intervals (95%).

Figure 2: Plant abundance at study sites, Simla and Morne Bleu.

Photometer readings

Mean light meter readings for the two study locations were calculated. Readings are in lux, and have been standardised by taking a reading in direct sunlight and using this as a reference. There appeared to be no correlative relationship between occurrence of Heliconia at a net site and a necessity for bright light, however the mean light readings were greater at Simla, reflecting the fact that the forest is more ‘open’ in its canopy, probably due to its secondary forest type.

Hummingbirds and Flowers

Hermit hummingbirds are highly specialised in their dietary choices (Snow & Snow, 1963, 1964, 1972; Stiles, 1980; Brown & Bowers, 1985). The following graphs were compiled in an attempt to compare percentage abundance of certain species of hummingbird and plant. The graph in Figure 3 below shows the distribution of G. hirsuta (hummingbird) and H. bihai (plant) at the two study sites.

Figure 3: Distribution abundance of Glaucis hirsuta and Heliconia bihai at study sites

DISCUSSION

Plants

The plants available for use by hummingbirds tend to be available year-round at the study sites. However a peak in good flowering bloom occurs in the herbaceous plants and shrubs mostly in the late dry and early wet seasons. For the time period of this study, there was a peak flowering bloom in second growth and also in the understorey. Therefore, the numbers of flowers counted may represent the wider population well.

Hummingbird-plant interactions

Well-defined differences in foraging behaviour are known to exist between hermit hummingbirds and typical hummingbirds. The hermit hummingbirds were by far the most common type to be captured during this study (Table 2). Whilst it could be assumed that this is due to a larger abundance of hermits over other hummingbirds it may actually provide an indication of the niches where these birds feed. Snow & Snow conducted a study of the feeding niches of hummingbirds in the Arima Valley and found that the most common type of plants used by hermits were herbaceous plants up to three metres in height. Conversely, non-hermit hummingbirds tended to use these plants to a significantly lesser extent, preferring to feed at both small and large trees. As the net height was 2.5 metres it would be expected that we would capture many more foraging species of hermits than non-hermits. Indeed, this was the case and twice as many hermit hummingbirds were captured.

The hermit hummingbirds, with their long curved bills are highly specialised at feeding on plants whose corollas reflect this morphology (Stiles, 1975:Wolf et al, 1976). An important herbaceous plant utilised by the hermits is Heliconia bihai. Of the two species in this study, G. hirsuta is known to feed on the nectar of H. bihai more frequently than P. guy (Snow & Snow, 1972). The data in this study support this suggestion. More H. bihai were counted at Simla than Morne Bleu, and this difference was found to be highly significant. There was also a significant difference in the capture rate of G. hirsuta with more being captured per hour at Simla. On the basis of this, the data does support the suggestion that the main food source of G. hirsuta could be H. bihai and where this plant occurs in low numbers so too should G. hirsuta.

There are of course many alternative-feeding sources in a rainforest, insects form part of the diet of P. guy to a greater extent than in G. hirsuta. However it has been published (Feinsinger, 1976) that insects are never likely to be a deciding factor in hummingbird abundance. Hummingbirds have also been observed feeding on buds, and fruits of flowers in a Costa Rican rainforest (Stiles, 1985). The difference in the use of a habitat by two sexes of the same species could be one way in which these results are affected. For instance, if the male feeds in the canopy and the female in the understorey – we will only be able to catch approximately half the actual number of hummingbirds (assuming the sexes are equal in abundance). Whilst habitat separation is plausible, it is mostly confined to the non-hermits. Among hermits the sexes are quite indistinguishable in the field, but observations on marked, measured individuals do not indicate habitat separation of the sexes (Stiles, 1980).

Unlike the hermit hummingbirds, it is not possible with the data set here to draw any conclusions on which plant types are important to the non-hermits. In the four species of non-hermits captured at the two sites, size difference compared to the hermits was highly evident; the weight of a hermit may be twice that of a non-hermit (Table 2). The non-hermits also possess bills that are either straight or only slightly curved. It is these characteristics that give them the ability to feed on a wide variety of plants. They are also notorious for holding territories (Feinsinger, 1976). The needle-like bills usually utilised for peaceful feeding events can be equally useful as a weapon when a passing hermit is looking for a quick meal. This however may be the reason they have not become specialised to feeding on one or a few plant species like the hermits. If a bird stays in the same place; holding territories on the same tree for instance, and does not travel it will not be a useful pollinator (Linhart, 1973). A plant will not adapt (morphologically) to better suit this hummingbird at the risk of excluding others. It is known that non-hermit hummingbirds do not hold territories at H. bihai and H. hirsuta. These species have long flowers that are most effectively probed by long-billed hummingbirds (Feinsinger et al, 1982).

The result of these published accounts is that non-hermits are considered as generalists as they are capable of feeding at many flowers that produce moderate quantities of nectar (Linhart & Feinsinger, 1980). In several non-hermit species showing sexual dimorphism (e.g. C. notatus) there are quite definite sexual differences in habitat utilisation in the forest during the dry season. Males mostly occur in the canopy and females mostly in the understorey (Stiles, 1980). In the summer of 2001, there was an extremely dry wet season in Trinidad and this extension of the dry season may have caused this habitat difference to continue into the study period.

Even if habitat utilisation is different between sexes of the same species, the proportion of non-hermits should be the same at the two sites. The evidence in the results are as expected. There are no significant differences between either C. notatus or A. chionopectus at the two sites. Where these two non-hermits occur in the same area it is highly plausible that they will hold territories or utilise different flowers. The variety of plants from which these birds can extract nectar is great, only 50% of their foraging time is spent at red flowers (Snow & Snow, 1972). Not all non-hermits were found in equal abundances, indeed one species, A. tobaci was only captured at Simla. The only plant which had a significantly higher number of flowers counted was H. bihai. However A. tobaci is known never to feed on even dense clumps of Heliconia inflorescences (Feinsinger et al, 1982). The only other difference in plant community structure was that there were a wider variety of species at Simla. Whilst it is plausible to assume this could be a reason for only catching A. tobaci at Simla, the plants concerned did not occur in large abundance whereas the hummingbird did. Another explanation may be that A. tobaci is affected by altitude. Indeed, in past years (1999, 2000) mist netting at the study sites involved here, have captured none at Morne Bleu (White, personal communication). Lack (1973) found that A. tobaci can occur at both lowlands and highlands, however it was also noted by ffrench (1991) that A. tobaci is widespread at all levels up to 650 metres. Since Morne Bleu is at 750 metres elevation it would not be found there.

The question has to be asked: how do so many species of hummingbirds, utilising the same flowers co-exist? With the hermit hummingbirds, we have seen that co evolution with a particular plant species is one way of coping. Stiles (1975) suggested that if the number of potential pollinators is large in relation to the number of flowers then competition for pollinators might be less important than partitioning of pollinators. For non-hermits, the adaptation for coexistence is being able to feed on many plant species, thus allowing habitat separation and territoriality to occur.

Conclusions

We have shown that marked differences do occur in the abundance of some species of plant in two types of forest in Trinidad. Also, differences in the capture rates for some species of hummingbird. Suggestions can only be made to a very limited extent in this study as sample sizes were small and only two forests were studied. Further, hummingbirds in general are extremely sensitive to subtle changes in net energetic rewards, and quickly shift the best available sources of nectar (Wolf et al, 1972). Unless a hummingbird feeds on the same plant all the time it would be difficult to tell if a particular plant species is of great importance. In the majority of hummingbirds, strict mutualisms are the exception to the rule, most pollinators visit a wide variety of flowers on a non-exclusive basis. In some instances though, hummingbirds and plants do form close mutualistic symbiotic relationships and in these cases it is likely that population numbers of hummingbirds and flowers will be closely linked.

REFERENCES

Baker, H. G., (1961) The adaptation of flowering plants to nocturnal and crepuscular pollinators. Q. Rev. Biol. 36, 64-73

Beard, J.S. (1946) The natural vegetation of Trinidad. Oxford Forestry Mem. 20. Clarendon Press. 152pp

Benzing, D. H. (1980) The Biology of the Bromeliads. Mad River Press Inc., Eureka, California.

Brown, J. H., & Bowers, M. A. (1985). Community organisation in hummingbirds: relationships between morphology and ecology. The Auk 102, 251-269

Cotton, P. A. (1998). The hummingbird community of a lowland Amazonian rainforest. Ibis 140 512-521.

Feinsinger, P. (1976). The organization of a tropical guild of nectarivorous birds. Ecological Monographs 46, 257-291

Feinsinger, P., & Swarm, L. A. (1982) “Ecological release” seasonal variation in food supply and the hummingbird Amazilia tobaci on Trinidad and Tobago. Ecology 63, 1574-87

Feinsinger, P., Wolfe, J. A., & Swarm, L. A. (1982) Island ecology: reduced hummingbird diversity and the pollination biology of plants, Trinidad and Tobago, West Indies. Ecology 63(2), 494-506

ffrench, R. P. (1991) A guide to the birds of Trinidad and Tobago 2^nd Ed, Cornell Univ. Press

Hurlbert, A. H., Hosoi, S. A., Temeles, E. J., & Ewald, P.W. (1996) Mobility of Impatiens capensis flowers: effect on pollen deposition and hummingbird foraging. Oecologia 105, 243-246

Linhart, Y. B. (1973). Ecological and behavioural determinants of pollen dispersal in hummingbird-pollinated Heliconia. The American Naturalist. 107, 511-523

Linhart, Y. B., & Feinsinger, P. (1980) Plant-hummingbird interactions: effects of island size and degree of specialisation on pollination. Journal of Ecology 68, 745-760

Marshall, R. C. (1939) Silviculture of the trees in Trinidad and Tobago. London: Oxford Univ. Press 247pp

Smith, J. A. C., Griffiths, H. & Luttge, U. (1986). Comparative ecophysiology of CAM and C3 bromeliads. I. The ecology of the Bromeliaceae in Trinidad. Plant, Cell and Environment 9, 359-376

Snow, B. K., & Snow, D. W. (1972) Feeding niches of hummingbirds in a Trinidad valley. Journal of Animal Ecology 41, 471-485

Snow, D. W., & Snow, B. K. (1980) Relationships between hummingbirds and flowers in the Andes of Columbia. Bull. Br. Mus. (Nat. Hist.) Zool.38(2), 105-139

Stebbins, G. L., (1989) Adaptive shifts towards hummingbird pollination, In: Bock, J.H., Linhart, Y. B. (eds) The evolutionary ecology of plants. Westview press, pp39-60

Stiles, F. G., (1975). Ecology, flowering phenology, and hummingbird population of some Costa Rican Heliconia species. Ecology. 56 285-301

Stiles, F. G. (1980) The annual cycle in a tropical wet forest hummingbird community. Ibis 122, 322-343

Stiles, F. G. (1985) Seasonal patterns and coevolution in the hummingbird-flower community of a Costa Rican subtropical forest. Ornithological Monographs 36, 757-787

Temeles, E. J. (1996) A new dimension to hummingbird-flower relationships. Oecologia 105, 517-523

Wolf, L. L., Stiles, F. G., & Hainsworth, F. R. (1976). Ecological organisation of a tropical, highland hummingbird community. Journal of Animal Ecology. 45, 349-379

Avian species diversity in three types of Tropical rainforest

Carrie Finlayson

ABSTRACT

Three different areas of the Victoria-Mayaro region in the South of Trinidad were studied: untouched Virgin forest (see right), forest managed under the Periodic Blocking System (PBS) method, and that managed using Open Range Management (ORM)(see below). It was evident in the field that there were major differences between these forest areas, the ORM forest being ecologically in the worst condition, with large areas of clear cutting and open canopy. The PBS forest area was in better condition than the ORM forest with a more diverse range of tree heights and a denser canopy layer. In this region, care had been taken to preserve the forest and retain some of the more genetically valuable trees, and it was regenerating to the extent that it more closely resembled virgin forest. The only obvious evidence of logging was the timber extraction trails cutting 3m wide paths through the forest. Virgin forest was significantly higher in tree height and had a significantly denser canopy on statistical analysis. The results showed that there were no significant differences in the feeding preferences and habitat preferences of the bird species in each of the three forest areas. The preferred habitat was main forest areas in all three sites, with the ORM showing higher numbers of edge dwellers. The lack of a significant difference in capture rates would suggest a similar number of bird species in general in the three forest areas. In conclusion, it was found that bird species diversity did not vary between the three types of rainforest in the Victoria-Mayaro region of Southern Trinidad, although this was a short study and more research is needed in this area.

INTRODUCTION

A tropical rainforest is defined as a community of trees growing where the climate of the coldest month is greater than or equal to 18ºC, or where the difference between the warmest and coldest months is lass than 5ºC (Whitmore, 1998). Generally, these forests are cool, humid and dark near the floor, with interior microclimates differing at various levels. There are many different types of rainforest, and the species within them are dependant on the climate and soil in which they grow. Rainforests are areas of high species diversity, and the impact of man on this diversity is of prime concern for environmentalists the world over.

As in many rainforest areas, humans have changed a great deal of forest in Trinidad. The south of the island has several oil-producing districts and logging has significantly affected both these and other areas of the rainforest. Larger and more valuable timber continues to be removed, even from the reserves that make up 30% of the island, and evidence of this can be seen in the form of roads, trails and clearings, which open up the forest as a result. Forest managers have a difficult job, trying to struggle against the competition for land use by an ever-increasing population. Trinidad in particular faces removal of timber at a rate of 600 hectares per annum, from its total area of 165,000 ha of forest.

In the past there has been a tendency for over-exploitation of the forest using a system of Open Range Management – where girth limit is the only control of the trees being removed in the forest. The method is open to abuse in that the girth limit guidelines issued by forestry officers can be easily stretched by the presence of buttresses, knots and other distortions on the trunk (Clubbe and Jhilmit, 1992). There is very little consideration of the surrounding trees during felling and extraction of the timber trees and there are no limitations to the areas of the forest from which high value timber trees can be extracted. The most valuable trees are the first to be cleared, and specimens of poor quality timber are disregarded resulting in a decrease in diversity and decrease in economic viability in the area. This has resulted in the removal of key species in many areas, and these sites are now in poor ecological condition.

Recognising the problems of this method of extraction more recent management of rainforest areas in Trinidad have focused on allowing the removal of timber for profit whilst attempting to maintain the rainforests with as little disturbance as possible. This polycyclic selection system (Clubbe & Jhilmit, 1992) is known as the Periodic Blocking System. Skilled silviculturists are employed to select and physically mark trees within a designated block usually around 150-300 ha (Clubbe and Jhilmit, 1992) The silviculturalists deem trees suitable for removal and sale by taking into account factors such as tree maturity, usefulness to wildlife and protective value to soil stability. Extraction techniques are controlled, vines removed from trees and care taken to avoid the damage of surrounding trees and vegetation. Logging is confined to this ‘block’ for two years and after this time the area is left to recover for a minimum of 30 years. In this way the disadvantages of the open range method are avoided.

Whereas much work has been conducted in the North of the island, surveys of wildlife in the South of the island are rare; to some extent due to the more limited accessibility of this side of the island. The South of the island is also the one in which new ventures into logging are being concentrated and new forestry methods being put into practice. This study aims to compare three different types of forest in the Victoria-Mayaro region in the South if Trinidad, particularly with respect to diversity of bird species found within those areas. This data will be examined to determine whether the extraction method used for timber has an effect on the diversity of the forest in the exploited areas.

METHODOLOGY

Study sites

The sites studied were in three separate areas of the Victoria Mayaro Forest Reserve in the South of Trinidad. This reserve consists predominantly of Mora forest, dominated by tall evergreen Mora excelsa (White, 2001). Crappo (Carapa guianensis) helps form the canopy, while young Mora seedlings compose the ground level layers. Lianas and bromeliads are common features of the habitat as well. In its original unexploited state the Mayaro forest covered an area of 21,750 ha, with regularly spaced trees allowing few or no gaps. Three sites within the forest were chosen for study: Virgin Forest, which had not been previously exploited for timber extraction; Open Range Management Forest in which timber extraction was under limited control and Periodic Blocking System Forest where management had been closely controlled.

Four days were spent in each of the three areas. 5 mist nets were erected between 7am and 9am each day, and were left out for approximately 6 hours. Net sites were located in a grid, mapped over an area 600m x 400m. The sites were no less than 50m apart, and each day sites were picked at random to ensure good representation of all areas of the forest. The nets were checked every three-quarters of an hour and any birds caught were removed by hand and carried back to a central shelter. All birds captures were identified and a series of measurements (sex, weight, tarsus and bill length, head and wing length) were taken. Before release the birds were ringed to identify recaptures.

Vegetation assessments

An indication of the basic forest structure at the two sites was determined by carrying out a number of different measurements. Five of the mist net sites were chosen randomly at each location and the following measurements were taken:

5) The circumference, at chest height, of 10 trees with a circumference > than 60 cm.

6) The approximate height of the same 10 trees using a clinometer.

7) The distance to the furthest of the 10 trees to give an estimate of tree density.

8) Percentage vegetation cover at ground, five metres, mid canopy and canopy levels.

Presence of streams, roads gaps in the forest and trees were assessed according to architectural type A) Vertical branching above half height; B) Vertical branching above half height with scarring; C) Vertical branching below half height; D) Horizontal branching below half height, (Figure 1).

Figure 1: Tree Architectural types A – D.

RESULTS

Capture Rates

The capture rate for each forest is calculated by dividing the number of species caught by the number of hours that the mist nets were in use. This gives capture rates as follows for the study period:

Forest type	No birds caught	Mist-netting hours	Capture rate/hr
Virgin	109	103.5	1.05
PBS	146	111.50	1.31
ORM	171	105.75	1.62

Table 1: Capture rates for each forest area under study in Victoria-Mayaro forest

Daily capture rates and overall capture rates were compared using ANOVA analyses and found no significant differences between the sites (F = 1.44, P = 0.287, 2 d.f.).

Habitat preferences of birds

An assessment of the habitat preferences for each species caught in the mist net studies was carried out at each forest area (Figure 2). Habitat use is divided into three main categories: E) Forest edges of open areas F) Main forest areas (interiors) and G) Generalists. The proportion of species inhabiting these areas was calculated for each forest site and compared.

Figure 2: Habitat preferences of species of birds caught at study sites

Preference in all types of forest was for main forest areas, there is a visibly higher percentage of species inhabiting the forest edges in forest managed under the open range method. Chi-squared tests show no significant differences between habitat preferences of birds found within the three types of forest.

Feeding Guilds

There are 4 main feeding types into which the species caught during the study can be divided: (N) Nectivores; (O) Omnivores; (F) Frugivores and (I) Insectivores. As with habitat data, the proportions of species in each of these four groups were compared for each site.

Figure 3: Division into feeding guilds of birds caught in each forest type

Feeding guilds within Virgin Forest seem to be divided fairly evenly, with similar patterns in Periodic Blocking System managed forest. In Open Range Forest although generalist feeders and nectivores appear more numerous, statistically no significant differences exist. The reduction in number of the more ecologically sensitive insectivores may point to a forest in trouble but as yet differences are too small to draw any firm conclusions.

Vegetation analyses

Tree height distributions in the various types of forest show that in Virgin forest has distribution skewed to the left, though more tall trees are present in this type of forest than in any other. Skews to the left are exhibited in ORM forest with the lowest average tree height of all three types of forest. Distribution is normal in PBS. Average tree heights were highest in Virgin forest (31m), and a significant difference in the tree heights was detected through ANOVA analyses (F = 7.90, p = 0.001, 2d.f.)., though differences between the two exploited areas of forest are smaller.

Tree densities were calculated for each of the study areas within the Victoria-Mayaro Reserve. ANOVA analyses found no significant differences between the three types of forest. Average densities can be seen in Table 2:

Forest type	Average tree Density/ha
Virgin	109.07
PBS	116.74
ORM	108.01

Table 2: Tree densities in three forest types

Tree architecture was also compared: Type A (see earlier definitions) is typical of primary forest, C & D types are usually seen in older regenerating forest, and B is typical of newly disturbed forest. Chi-squared analyses suggest that a highly significant difference between the areas exists in terms of the architectural types of tree found within the forests.

Figure 4: Proportions of tree architecture in each forest area

Vegetation cover for ground, 1.5m and 5m and canopy level was estimated at each of the forest sites. ANOVA analyses found no significant differences in vegetation types between sites at ground, 1.5m or 5m levels but canopy cover in three forests were significantly different to one another (F = 15.14, p < 0.001, 2d.f.) as illustrated in Figure 5. Also evident in this graph is the tendency for Virgin Forest to have little ground cover, though this observation is as expected the difference is not significant.

Figure 5: Vegetation cover at different levels in three types of forest

DISCUSSION

Vegetation differences between Forest Types

Differences in tree heights were evident in the field during the study. The virgin forest is relatively untouched and as a result the trees are much older and mean tree height in these areas is greater than in the two exploited forest types. Distribution of tree heights is dominated by the presence of older, larger trees as the habitat has been undisturbed and allowed to develop over time. Most frequently recorded heights of trees ranged between 20-35 m, with the tallest being 80m in height. In comparison, PBS sites showed average heights between 17-23 m, with a more normal distribution over a smaller range of heights. This distribution reflects a range of heights throughout the forest, suggesting that the forest is regenerating, while having retained some of its older trees. Tallest trees here were around 55m, notably lower than that of the virgin forest. The sites studied in the ORM forest showed similar values for its most common tree heights, between 15-20m, but again the range was much smaller than that of the virgin forest with a large skew to the left; This is indicative of the lack of any larger trees of height, as would be expected in a forest where no effort had been made to conserve some ecological balance. Very few trees were greater than 45m in height in this area. Tree heights were significantly different between forest types, PBS was in a better condition in terms of height diversity and having a few tall, good-value trees when compared to ORM forest indicating that PBS management allows a certain amount of forest regeneration absent from ORM forest.

There was a highly significant difference in tree architecture found in each forest type. In virgin forest, the only tree architectural type was the A-type, branching above half-height. This type is typical of primary forest (Torquiebiau, 1986; Jones et.al 1995). This is as would be expected of an untouched area of forest which has not been utilised by man in any way, the natural falling of old trees causes gaps and damage to trees. PBS forest also had high levels of type A architecture, indicating that the area of forest is still in good condition i.e. comparable with Virgin forest. ORM forest on the other hand exhibited a number of types of tree, having greater numbers of B-type scarred trees, and of C & D types with vertical and horizontal branching below half height. C type trees are exhibited in areas where the felling of surrounding trees damages the upper branches and lower branches develop to grow into the canopy as growing tops have been irreparably damaged. Horizontal branching is typical of trees that are allowed space to grow and are the product of the opening up of the canopy such as is seen in ORM forest. Indeed levels of canopy cover were found to be significantly different between ORM forest and the other two types of forest. Virgin forest showed highest canopy cover and lowest ground cover as is typical of primary forest (Bergstedt & Milberg, 2001).

Bird Species Diversity and Composition

Capture rates in the three forest types showed no significant differences. Capture rate was highest in ORM forest, followed by PBS and lowest in Virgin forest. Reasons for this may be affected by capture techniques as low-level mist netting is known to have disadvantages in terms of sampling canopy species. Canopy level is significantly higher in areas of Virgin forest and so mist nets are not necessarily sampling representative species of that forest type. Webb et all (1999) found that ridge forest with low canopy and high disturbance, showed high species richness, a finding supported by Argentinean Chaco forest surveys by de Casenave et al (1998) who found bird species richness appeared greatest at forest edges when compared to forest interiors. As ORM forest consists of low canopy, open forest this may explain higher levels of species diversity within these areas. However statistically speaking the differences seen between site were not significant, echoing other studies which found that overall there was very little change in the number of species inhabiting damaged and regenerated forest with changes in species composition being the most relevant changes seen between forests as a result of disturbance.

The preference in all types if forest was for main forest areas, given sampling methods this was to be expected but differences in open area/forest edge species were more telling. More species caught within ORM forest belonged to this habitat preference group and to generalist species than in the other two forest types. Our results support previous work by Dranzoa (1998) on the avifauna of previously logged forest in Uganda, Dranzoa found that while the majority of species in damaged forest were generalist or forest dwellers, 84d% of the original numbers of main forest inhabitants had recolonised or persisted in the logged areas.

In all areas the strata preference was for the understorey layer. Understorey birds, such as the Ruddy Ground-Dove (ffrench, 2000) and those feeding in the lower forest layers are often the first to inhabit understorey layers. More birds would be expected to inhabit these layers in PBS and ORM forest therefore than in Virgin forest, whereas our results show a trend towards this levels of significance are low. It must be noted though that main sampling techniques being mist-netting may be less sensitive at showing differences in strata use by species, sound recording data taken supports this trend but analyses are problematic. A more accurate way of assessing changes in species composition by looking at life history strategies might be to look at feeding guilds of the species captured in the course of the study.

Virgin forest was found to have almost equal proportions of birds of each feeding type; omnivores, frugivores, insectivores and nectivores. There were notably smaller numbers of frugivores and insectivores in PBS and ORM forests, and higher numbers of nectivores. Generalist omnivores are present at all sites in similar numbers, reflecting the adaptability if these species. Thiollay (1999), studying impacts of logging and road building on an avian community in French Guiana found that omnivores and generalists were in an advantageous position in terms of maintaining numbers in recently disturbed areas as is reflected by the stability of numbers of omnivorous species captured in the different forest types examined in this study. Owiunji and Plumptre (1998) working in Uganda found increases in nectivores and omnivores in disturbed forest, the former observation being supported by observations within the study under consideration here. Our study also found decreased numbers of insectivores, known to be particularly affected by disturbance (Mason (1996), Thiollay (1997), and Johns (1996)) and vulnerable to logging pressures, though statistically changes were not significant. This trend may be further explored by classifying insectivores further according to their search tactics: bark-gleaning, leaf gleaning and sallying species are differentially affected by disturbance (Owiunji & Plumptre, 1998) and if the majority of insectivores found here were found to belong to the bark-gleaning species then differences would not be as pronounced between the three forest types – as is indeed the case (Figure 7).

Figure 7: Percentage of insectivores employing different foraging strategies

Conclusions

This study found that there was a significant difference in vegetation composition between the three forest types, as was evident in the field. Virgin forest had a much higher and denser canopy layer, while the Open Range Management forest had large areas of open canopy significantly lower in height. The ecological condition of the Periodic Blocking System was much better than that of the ORM forest, showing similarities in composition to the Virgin forest. There were no significant differences found in the composition of the bird community between the three forest sites, although preferences for habitat and diet guilds did emerge. A more prolonged and detailed study is needed to give more conclusive results.

REFERENCES:

Bergstedt, J. & Milberg, P. (2001) The impact of logging intensity on field layer vegetation in Swedish boreal forests. Forest Ecology and Management 154: 105-115.

Clubbe, C.P. & Jhilmit, S. (1992) A case study of natural forest management in Trinidad. In Wise management of tropical forests, F.R. Millar & K.L.Adams (eds) Proceedings of the Oxford conference of tropical forests. Oxford Forestry Institute, Oxford, U.K.

Dranzoa, C. (1998) The avifauna twenty-three after logging in Kibale National Park, Uganda. Biodiversity and Conservation 7:777-797

Ffrench, R. (2000) A guide to the birds of Trinidad and Tobago (Second Edition) A & C Black, London.

Johns, A.G. (1996) Bird population persistence in Sabah and logging concessions. Biological Conservation 75: 3-10.

Jones, M.J.; Linsley, M.D. & Marsden, S.J. (1995) Population sizes, status and habitat associations of the restricted-range bird species of Sumba, Indonesia. Bird Conservation International 5: 21-52

Mason, D. (1996) Responses of Venezuelan understorey birds to selective logging, enrichment strips and vine cutting. Biotropica 28: 296-309

Owiunji, I. & Plumptre, A.J. (1998) Bird communities in logged and unlogged compartments in Budango Forest, Uganda. Forest Ecology and Management 108: 115-126

Thiollay, J.M. (1997) Disturbanc, selective logging and bird diversity: a Neotropical forest study. Biodiversity and Conservation 6: 1155-1173

Thiollay, J.M. (2001) Responses of an avian community to forest degradation. Biodiversity and Conservation 8: 513-534.

Torquebiau, E.F. (1986) Mosaic patterns in dipterocarp rainforest in Indonesia and their implications for practical study. Journal of Tropical Ecology 2: 301-325

Webb, E.L.; Stansfield, B.J. & Jensen, M.L. (1999) Effects of topography on rainforest tree community structure and diversity in American Samoa, and the implications for frugivore and nectarivore populations. Journal of Biogeography 26: 887-889

White, S.A. (2001) A short mist-netting study in Guayaguayare and the Victoria-Mayaro Forest Reserve, Trinidad, W.I. (Short report for Petrotrin)

Whitmore, T.C. (1999) An introduction to Tropical rain forests. (2^nd edition) Oxford Uni Press, N.Y.

The role of sunning behaviour in temperature regulation in Black Vultures, Coragyps atratus

Jennifer Ward*

It is known that vultures have a body temperature that fluctuates between day and night. At night they allow their core temperature to drop, to save energy (Heath 1962, Larochelle et al. 1982, Bahat 1998). A recent study on Griffon Vultures (Gyps fulvus) has shown that these birds are able to save energy in the morning by using sunning behaviour (wing-spreading behaviour) to elevate their body temperature to normal daytime levels (Bahat 1995). I investigated whether New World Vultures use sunning behaviour in a similar manner.

The study was conducted at a site near Matura Beach, on the east coast of Trinidad, West Indies. Behavioural observations were made on Black Vultures (Coragyps atratus) in the early morning, together with environmental variables such as ambient temperature, wind speed and solar radiation levels. For comparison I also observed Turkey Vultures (Cathartes aura). I tested the hypothesis that birds would sun in the early mornings in response to solar radiation. I found, however, that birds were not utilising the first rays of sunlight to warm themselves in the mornings, probably because nighttime temperatures were not low enough to necessitate energy-saving physiological mechanisms.

Alternative functions for wing-spreading behaviour in diurnal birds of prey have also been proposed (review in Simmons 1986), so I then attempted to identify which of these could explain this behaviour in Black Vultures. Full wing-spreading behaviour appeared to function primarily in wing drying after prolonged soaking during rain showers. The length of time for which birds sunned after rain was correlated with the length of the rain shower. Brief bouts of wing spreading were also observed immediately after returning from flight; this probably assisted with restoring warped feathers after flight, or balance upon landing on branches. Further study is necessary. Turkey Vultures exhibited wing-spreading behaviour for longer periods, often when exposed to intense insolation; at such times Black Vultures were often seen crouching on branches, usually in the shade beneath foliage, in what appeared to be a prostrate position in response to solar heat.

Thus, it appears that Black Vultures in Trinidad do not use solar radiation to warm their bodies in the early mornings, unlike those observed in southern Florida by Kushlan (1973). Wing-spreading behaviour at this site seems to function primarily in wing drying, based on the current data set. It would be interesting to examine whether Black Vultures do use sunning behaviour in parts of their range where they experience colder overnight temperatures, and to determine the critical temperature at which this behaviour commences.

REFERENCES

Bahat, O. (1995). Physiological adaptations and foraging ecology of an obligatory carrion eater – the griffon vulture (Gyps fulvus). Ph.D. thesis, Tel-Aviv University.

Bahat, O., Choshniak, I. & Houston, D.C. (1998). Nocturnal variation in body temperature of griffon vultures. Condor 100: 168-171.

Heath, J.E. (1962). Temperature fluctuations in the turkey vulture. Condor 64: 234-235.

Kushlan, J.A. (1973). Spread-wing posturing in cathartid vultures. Auk 90: 889-890.

Larochelle, J., Delson, J. & Schmidt-Nielsen, K. (1982). Temperature regulation in the black vulture. Canadian Journal of Zoology 60: 491-494.

Simmons, K.E.L. (1986). The sunning behaviour of birds: A guide for ornithologists. Bristol Ornithological Club, Bristol

REPORTS OF THE TURTLE GROUP 2001

Turtle group with NatureSeeker’s Anderson Innes – Matura Beach: L to R: Rachel Thomas, Ally Humphries, Roxana McLennan, Oisin Sweeney (Photo: J.Ward)

Matura beach is situated on the East Coast of Trinidad and is one of the main nesting beaches for the endangered sea turtle, Dermochelys coriacea, also known as the ‘Leatherback’. The need for a leatherback conservation project was realised by the Matura community itself and the wildlife section of the Forestry Division due to the alarming number of egg-laying females being slaughtered by poachers during the 1970’s and 1980’s. Nature Seekers was formed by both the local community and Forest Division in 1990 to carry out such a project, doing nightly beach patrols and offering a tour guiding service to the public.

This years turtle group worked closely with the Nature Seeker group; helping with beach patrols at nighttime, and with ‘zoning’ the beach during the day. This zoning involved marking out areas to help define stretches of the beach so that data can be collected in a consistent manner as to where along the beach the turtles are preferentially choosing to nest. They also aided in a survey of the villagers to determine what local response to additional facilities to encourage higher levels of ecotourism within the village and carried out their own study into the levels of fly infestation within turtle nests along Matura Beach. For four

Insect infestation of leatherback turtle nests

Oisin Sweeney

Female Leatherback turtle (Photo J.Ward) and adult fly from infested nest (Image: O. Sweeney)

ABSTRACT

Leatherbacks nest on the beach between the months of March and late August / September. Nests are dug in the sand, approximately 60 – 120 cms in depth. Development time for the eggs is between 50 and 80 days. Nests were excavated after hatchlings had left for the open sea. The eggs found in these nests were sorted into three categories: hatched, unhatched and unfertilised. The unhatched eggs and any dead hatchlings in the nest or the sand column were checked for insect infestation. Infestation was typically found on the flippers and eye sockets of hatchlings.

Four species of flies were found to be infesting the nests: two species belonging to the family Phoridae (Puliciphora borinquenensis and Megaselia scalaris) and two to the family Sarcophagidae (Agoravinia rufiventris and 1 from the genus Tricharaea for which a species name is unknown). Both M. scalaris and A. rufiventris are widespread in their distribution and will feed on dead and decaying material.

Many previous studies carried out on various species of turtle disagree on the impact that infestation has upon their reproductive biology. In this case the evidence indicated that although, in absolute terms, infestation affected a large number of potentially viable hatchlings, the overall proportion of eggs that were infested was low, and hatching success differed very little between infested and uninfested nests. It must also be noted that not all of the unhatched eggs were infested, thus infestation alone cannot be held responsible for the failure of eggs to develop or for the death of turtles in the sand column.

INTRODUCTION

Leatherbacks (Dermochelys coriacea) are the largest of all the species of sea turtle, with carapace lengths measuring, on average, 160cm and a weight that can on occasion reach 0.5 tonnes. (National Research Council, 1990) The carapace differs from that of other species of turtles, as it is not rigid, nor formed by a series of scutes. The biology of these turtles is extremely interesting. They feed on jellyfish and other coelenterates, and as these creatures are not especially nutritious, the turtles must eat a great number per day to survive. Feeding generally takes place in the surface waters of the oceans, as this is where the ocean currents carry the leatherbacks’ prey. They are known to dive to great depths (~ 1km) – much greater than any other species of turtle. Interestingly the male leatherback is thought to swim continuously from the moment it enters the ocean as a hatchling until its death (perhaps 100 years later). The female will only stop swimming to come ashore to lay her eggs.

The eggs and hatchlings of the leatherback face many challenges before reaching adulthood. There are a host of predators that will feed on hatchlings: Vultures and ghost crabs on the beach (Whitmore and Dutton, 1985) sharks, fish and sea birds in the water (National Research Council 1990) and there have even been reports of plant roots invading nests and puncturing eggs (Hirth et al 1993). All of these affect hatchling survival, and of course there is the most serious predator of all to take into account – man. In many places e.g. Papua New Guinea, eggs are used as a valuable source of protein (Hirth et al 1993) and in Trinidad amongst other places the adults are occasionally used as food (Nature Seekers, pers. comm.) although this practice is becoming less frequent. Due to these varied and numerous threats, the chances of a hatchling surviving to reach adulthood are very small indeed, hence the large numbers of eggs produced by females.

It is these eggs – or more accurately the insects that attack them, that are the focus of this study. There have been several previous studies carried out on the presence of insect larvae on turtle eggs on a wide range of turtle species. (Acuna – Mesen and Hanson 1990, Broderick and Hancock (1997), McGowan et al (2001)) However there have been relatively few studies on insect infestation of leatherback nests. Those studies that have been carried out on leatherbacks, for example that of Andrade et al, (1992) and Vasquez (1994), have reported genera of sarcophagid flies infesting the turtle eggs. Members of the Sarcophagidae have also been reported infesting green (Chelonia mydas) and loggerhead (Caretta caretta) turtles (McGowan et al 2001).

There is considerable debate as to whether or not insect infestation actually has an effect on the reproductive success of turtles. To comment upon this it is necessary to look at the number of nests that are infested in a study. Lopes (1982) discovered that 90% of transplanted nests were infested and that this infestation decreased the success of hatching by 30%. Obviously in this case insect infestation did have a damaging effect on the turtle population, and if this is occurring on a global scale, it has severe implications for the future conservation of turtle species. McGowan et al found a much lower level of infestation in their study of loggerhead and green turtles in Cyprus (2001) – 13.4% and 17.4% in 1996 and ’97 respectively. In this case the impact of the insects on the reproductive success of the turtles is likely to be much less.

This study was intended to discover what species of insects infest the nests of leatherbacks nesting in Trinidad, West Indies. An attempt was also made to discover what made nests more susceptible to infestation and how and when the infestation was taking place, as well as attempting to show the severity of infestation (in terms of numbers of larvae present on the hatchlings, and number of individual eggs infested.)

MATERIALS AND METHODS

Study site

This study was carried out on Matura beach, east Trinidad between the dates of 3^rd July and 16^th August 2001. Matura is a 9km long beach divided by a river running into the sea into 2 major sections: Rincon and Orosco. Volunteers from Glasgow University Expedition 2001 divided the beach into 14 zones of 1400 feet, to aid the conservation volunteers working on the beach to collect accurate nesting data. The beach is fairly narrow, most parts being around 10-25 metres wide. Turtles nest all along the beach (with the exception of the narrowest sections) in roughly equal numbers. Work was carried out in conjunction with Nature Seekers, a voluntary organisation based in Matura who have been working with leatherbacks for 12 years. As leatherbacks are endangered and Matura beach closely monitored during the nesting period of the turtles, it was necessary to obtain a permit from the Forestry Division allowing the excavation of the nests.

Nest Location and Excavation

Hatching turtle nests were looked for on the beach in teams of two, five nights per week. They were located by following the tracks of newly emerged hatchlings back up the beach to their source. At the source it was typical to find a shallow depression that indicated the position of the nest. This depression was usually approximately 20 – 25 centimetres in diameter.

Upon discovery of a nest a decision was made whether to excavate or not. Due to the fragile status of leatherbacks in the wild, it was decided that on no account was the fieldwork to be in any way detrimental to the future survival of the species. Thus if there were only a few tracks leaving the nest, it was left alone and re-examined at a later date. If tracks were numerous, excavation went ahead. Excavation was carried out by hand and rubber gloves were worn to protect hands from decaying and infested eggs.

Nest depth was measured using a tape measure. This was taken to be the distance from the sand immediately below the lowest egg in the nest to the sand surface. Depth was measured to the nearest centimetre. The distance from the centre of the nest in a straight line back to the vegetation at the top of the beach was also measured, again with the use of a tape measure and again to an accuracy of 1cm.

After nests were excavated, the eggs were sorted into 3 separate categories: hatched, unhatched and unfertilised. Hatched eggs were immediately obvious, as they were opened and empty. All of the eggs that had not hatched before the nest was excavated were ‘lamped.’ This involved shining a torch through them to detect whether or not a yolk or embryo (both appear as dark patches in the egg) was present. If no dark patch was detected, the eggs were treated as unfertilised. Any eggs in which there was a dark patch present were said to be unhatched, and were opened to check for the presence of infestation. All unhatched eggs were checked for the presence of maggots. Particular areas that were checked were flippers and eye sockets of the hatchlings. If an egg was found to be infested, a sample of the infested hatchlings was taken, placed in a jam jar and covered with moist sand. A piece of muslin was placed over the mouth of the jar, and a lid with air holes placed on top of the muslin. The muslin was to ensure that there was no further infestation by flies after the samples had been taken. When the jars had been sealed they were then buried on the beach at approximately 4-6 inches depth and their positions discreetly marked so that they could be found again. This was to ensure that the developing larvae were in conditions very similar to those they encountered in the nest. The jars were left in this way for 7 days and following this period were excavated.

After a week, the jars were removed and maggots placed into a plastic vial (2-3 cm diameter) half filled with moist sand. These plastic vials were stored at room temperature in an upright position with adequate ventilation and were checked for the appearance of pupae on a daily basis.

Pupae were moved from the plastic vials into a smaller glass one. After the teneral flies emerged they were left for 24 hours in the vials to allow tanning of the cuticle (McGowan et al, 2001). The flies were then placed in the freezer to kill them and preserved in 75% alcohol along with the puparia for identification in the UK.

Sampling Adult Flies

Observations were also made on flies that were visible on the beach. Information from Nature Seekers stated that it was possible to see where a nest would hatch in the next day or two by looking for the presence of large swarms of flies on the sand immediately above the nest. To test whether or not these flies were the same as those that infested the nests a sample were captured, killed by freezing and preserved in 75% alcohol along with the puparia for identification in the UK.

Insect Identification

The flies were identified to family level according to protocol (Unwin, 1981). After this sorting, flies were sent to Thomas Pape of the Swedish Museum of Natural History and Henry Disney of Cambridge University for species level identification.

RESULTS

Infestation data

The results for the overall levels of infestation in the nests on Matura beach as well as data on the proportion of infested nests are shown in Table 1.

Total number of leatherback nests examined	77
Number of infested nests	66
Percentage of nests infested	85.7
Total number of eggs in infested nests	7062
No. of infested eggs in nests	280
Mean infested per nest (± SD)	4.12 (4.22)
Infested as proportion of total (%)	4

Table 1: Mean number of infested eggs per nest (+ / - SD), and as a proportion of total number of eggs in infested nests.

The majority of the eggs hatch, and unfertilised are the most numerous besides these. There is very little difference between the proportions of egg types between infested and uninfested nests. Table 2 gives mean values for many of the nest parameters measured.

	n	Mean (+/- SD)	Range
Nest depth	77	88.68 +/- 17.3	55 to 165
Number of eggs hatched	77	52.1 +/- 19.54	13 to 96
Number of eggs unhatched	77	15.16 +/- 14.33	0 to 88
Number of eggs unfertilised	77	40.1 +/- 17.22	6 to 88
Number of eggs infested	77	3.53 +/- 4.16	0 to 21
Distance to vegetation	77	7.68 +/- 6	1.9 to 31.8

Table 2: Means (+/- SD) and ranges of all of the data collected on Matura

Regression analyses were performed to assess the effect of these various measured parameters on levels of nest infestation. These included nest depth, number of eggs of various types and distances from vegetative shoreline. Results were as follows:

Infestation vs Nest depth: r = 0.17, p = 0.15, (N.S.) 75 d.f.

Infestation vs. distance to vegetation: r = 0.14, p = 0.69, (N.S.) 75 d.f.

Infestation vs. no. Hatched eggs in nest: r = 0.00, p = 0.95, (N.S.) 75 d.f.

Infestation vs. no Unfertilised eggs in nest: r = 0.03, p = 0.27, (N.S.) 75 d.f.

Infestation vs. no Unhatched eggs in nest: r = 0.23, p = 0.05, 75 d.f.

The only variable that shows a significant relationship with the number of infested eggs in a nest is the number of unhatched eggs in a nest. Figure 9 illustrates the relationship between these two variables.

Figure 9: Graph illustrating relationship between number of unhatched eggs (x) and number of infested eggs (y)

Approximately 4 eggs in every infested nest were infested. This is from a total of approximately 80 – 120 eggs. The figure in itself is not high. Nor is the number of infested eggs high as a proportion of the total number of eggs in infested nests (Table 4)

	Proportions of each egg	Proportions of eggs
	type in infested nests	in uninfested nests
Prop. Unhatched	0.13	0.15
Prop. Unfertilised	0.36	0.38
Prop. Hatched	0.47	0.47
Prop. Infested	0.04	0

Table 4: Proportion of each type of egg found in the infested and non-infested nests

Arc sin transformations on proportions of infested nests against other variables yielded similar results with the only significant correlation in this case occurring between levels of infestation and proportions of unfertilised eggs in the nest.

Fly Species Identified from Nests

Upon analysis of the flies using the methods outlined in the previous section, it was found that there were five different species of diptera infesting the leatherback nests:

1. Megaselia scalaris (Phoridae family)

2. Puliciphora borinquenensis (Phoridae family)

3. Argoravinia rufiventris (Sarcophagidae family)

4. Tricharaea spp. (Sarcophagidae family)

5. Unknown sp. (Muscidae family) – two individuals only.

The species overlapped in their temporal distribution, and on seven occasions the species were found in the same nest. Sarcophagid samples were only obtained from 34 of the total number of nests examined, (2 contained phorids) and so approximately 20% contained both species and 80% one or the other. The number of nests containing the different species is shown in Table 2 below. The vast majority of the flies were sarcophagids. The flies captured on the sand surface were analysed and found also to be sarcophagids.

Table 2: Table showing how many nests were infested with the different species of Diptera and Phoridae.

Data suggests there are some nests that are heavily infested and others that are not. The mean number of maggots per nest is just above 50, but a few nests with a very heavy infestation may influence this figure. Thus a median value has also been generated, as it will represent a more representative analysis of the numbers.

Species	Range	Mean no. of maggots after 7 days (± SD)	Median
Argoravinia	6 - 175	45.33 ± 35.34	40
Tricharaea	20 - 200	66.9 ± 64.58	37.5
Sarcophagidae	6 - 200	50.26 ± 43.53	50

Table 3: Table shows the range of maggot numbers, the median and the mean number of maggots in the infested turtle material. (+/- SD)

ANOVA analyses determined that no statistically significant differences existed in the measured nest parameters of nests infected by different species of fly.

DISCUSSION

Of the nests examined in the course of this study, very high proportions were infested (85.7 %). This contrasts sharply with most other studies on infestation (McGowan et al, 2001, Broderick and Hancock, 1997) where 23% of loggerhead and 9% of green nests were infested. However, these studies involved different species of turtle and so direct comparisons between these and the data found on leatherbacks are difficult to make, as there may well be interspecific differences.

Statistical analyses show that few nest parameters directly affect the likelihood of a nest being infested by insect larvae. The only parameter with which a positive correlation was found for infestation level was the number and proportion of unhatched eggs within the nest. Two explanations are possible for this; firstly, the more unhatched eggs within a nest, the more biological material available for consumption to flesh fly maggots such as the Sarcophagidae species. Secondly, infestation itself may have an effect of halting hatching in other eggs so that infestation itself was responsible for the increased number of unhatched turtles within an infected nest.

Species Infesting the Nests

The study on Matura beach found that the majority of infested nests were attacked by 2 species of fly from the family Sarcophagidae. Other studies have also reported finding a number of sarcophagid species infesting turtle nests. (Broderick and Hancock, 1997, Mc Gowan et al, 2001.) These investigations reported 3 and 6 sarcophagid species respectively. Accordingly it is not surprising to find species of sarcophagids present in the nests of leatherbacks, although in this case the genera are different to those previously found in this habitat. A. runfiventris is said to have a wide distribution, but due to a lack of published research on the species, it is impossible to draw any conclusions about whether or not the flies are exhibiting local specialisation on the turtle nests in Trinidad, or whether it is simply an opportunistic infestation.

Phoridae were also present in at least 2 of the nests excavated during the study. Broderick and Hancock, 1997, and McGowan et all (2001) reported finding one of these species in loggerhead (Caretta caretta) and green turtle (Chelonia mydas) nests in Cyprus. The species, found in all studies was the one found here, Megaselia scalaris. This particular phorid has a worldwide distribution and has been found in every available niche possible, from rotting flesh to blue gloss paint. (Disney 1994, Acuna-Mensen and Hanson, 1990). They have a remarkable ability to reproduce and feed on almost anything and so are likely to be opportunistically exploiting dead turtle hatchlings. The other phorid species found was a wingless one known as Puliciphora borinquenensis. This species of phorid has also been discovered in green and hawksbill nests (Bjorndal et al 1985, Fowler 1979) raising the question that these flies may be specialists in feeding on turtle eggs.

There is one other family that was found in the turtle nests, and that was the Muscidae. These flies belong in the same section as the Sarcophagidae (Calyperates) and include in their ranks the common house – fly. There was however only two individual flies in the whole of the sample, and so perhaps the infestation may have been opportunistic.

Why though do the other families of fly, (e.g. Calliphoridae) that are celebrated for their love of carrion, not have a presence amongst leatherback turtle eggs? One possible reason could be the differences in life history. The Calliphoridae, or blowflies, are known to be strictly diurnal (Smith, 1986) whereas, the Sarcophagidae definitely do fly at night, as we were able to trap some of the individuals that were swarming on the sand surface well after darkness had fallen. As most leatherback nests hatch at night – possibly an evolutionary trait to minimise predation – the implications are that these species that are flying at night will be the first on hand to infest the turtle nests when the hatchlings emerge, and thus outcompeting other families of flies.

Timing of Infestation

One major question that it is very important to ask about this study is how and when do the flies infest the nests? This question is very difficult to answer, as it is impossible to monitor nests closely throughout their incubation period (50 –80 days) but as the study progressed, views changed as to the time and method of infestation. To begin with, it was thought that the most plausible explanation for the infestation was that the flies attacked at the time of egg laying. However, as more and more nests were discovered that contained infested eggs with larvae at early stages of development, opinions began to change. The development period for a sarcophagid is between 11 and 17 days, (Busvine 1980, Parashar et al 1997) whereas that of a leatherback hatchling is between 50 and 80 days. Had the larvae been placed in the nest at its laying it would be very unlikely that any larvae at early stages would still be present in the nest. We therefore conclude that it is likely that infestation is occurring at the time of hatching. This is supported by observations on the stage of development of the hatchlings that were infested by larvae. They were fully pigmented which indicates a late stage of embryonic development (Whitmore and Dutton, 1985). If the infestation had occurred at the time of laying, the dead turtle would be at a much earlier stage of development, if indeed any remains resembling a turtle could be identified at all.

Method of Infestation by Dipterans

So far in the study an attempt has not been made to explain how the flies manage to get to the freshly hatched eggs, and what attracts them to do so. It is believed that the phorids burrow to reach the turtles and that they use olfactory cues to do so. (Acuna – Mensen and Hanson 1990, Disney 1994) Disney (1994) states that phorids were able to detect a dog corpse that was buried 50cm below the surface of the ground. This is considerably shallower than many of the nests excavated in Trinidad and if phorids are unable to detect olfactory cues at depths greater than this, it may explain the low incidence of phorid infestation found during this study.

After a female fly has located the nest, she must get to it to deposit her larvae. How does this happen? It was observed that the sand surface in places was highly populated with flies that appeared to be vying for a space on the surface. These flies were of the same species as those reared from the nests, and so it is possible that in this case the flies were attempting, in some way, to deposit larvae. None of the flies were observed to be burrowing below the surface. Having listened to Nature Seekers who suggest that this heavy fly presence is a common phenomenon above nests that are about to emerge, it is possible that larval deposition is occurring as the scent of the hatching eggs is emitted. Again, a closer study may reveal what exactly is occurring here.

Effects of Insect Infestation

In the introduction the point was made that there was debate about how much, if indeed any, impact insect infestation has upon the hatching success of turtles. After having studied this phenomenon and reviewed much literature, this author is of the opinion that insect infestation does not affect hatching success in a serious manner. It is believed that infestation occurs at the time of hatching and that the larvae are unable to enter the eggs until they hatch. Thus infestation does not prevent them from doing so. 85.7% of leatherback nests examined in the study period were infested with larvae. There was a mean of 15.16 (+/- 14.33) eggs unhatched per nest (Table 2) and the total number of unhatched eggs found in infested nests during the study was 979. However, the proportion of eggs infested was 4%. This is a low figure and so it would appear that although insect infestation definitely does affect the success of hatchlings getting to the water, the effect is minimal on the species as a whole. Indeed, the majority of eggs in nests hatched, even in the infested nests and the proportion infested was low. There is also very little difference in the proportions of the different egg types between infested and uninfested nests. Thus it would appear that infestation is not markedly affecting turtle reproduction.

REFERENCES

Acuna-Mensen, R.A. and Hanson, P.E. (1990) Phorid fly larvae as predators of turtle eggs. Herpetological Review 21, 13-14.

Bjorndal, K.A. Carr, A. Meylan, A.B. and Mortimer, J.A. (1985)Reproductive biology of the hawksbill, Eretmochelys imbricata, at Tortuguero, Costa Rica, with notes on the ecology of the species in the Caribbean.

Biological Conservation, 34, 353-368.

Broderick, A.C. and Hancock, G.E. (1997) Insect infestation of Mediterranean Marine Turtle Eggs. Herpetological Review, 28(4), 190-191.

Disney, R. H. L. (1994) Scuttle flies: The Phoridae, Chapman and Hall, London.

Kouki, J. and Hanski I. (1995) Population aggregation facilitates coexistence of many competing carrion fly species. Oikos, 72, no. 2, 223-227.

McGowan, A. Rowe, L. V. Broderick, A. C. Godley, B. J. (2001). Nest factors Predisposing Loggerhead Sea Turtle (Caretta caretta) Clutches to Infestation by Dipteran Larvae on Northern Cyprus. Copeia, 3, 808-812.

McGowan, A. Broderick, A. C. Deemings, J. Godley, B.J. and Hancock, E.G (2001) Dipteran infestation of loggerhead (Caretta caretta) and green (Chelonia mydas) sea turtle nests in northern Cyprus.” Journal of Natural History, 35, 573-581.

National Research Council. (1990) Decline of the sea turtles. Causes and prevention. National Academy Press, Washington D.C.

Parashar, B.D. Rao, Y.V.S and Rao, K.M. (1997) Effect of Environmental temperature on development, fecundity, survival and predation of the snail – predator Sarcophaga misera (Dipt., Sarcophagidae). Entomophaga, 42(3), 343-347.

Service, M. W. (2000) Medical Entomology for Students (2^nd Edition) Cambridge University Press, Cambridge, 2000

Smith, K.G.V. (1986) A Manual of Forensic Entomology. The Trustees of the British Museum (Natural History), London.

Unwin, D.M (1981) A Key to the Families of British Diptera Field Studies, 5(3)

Whitmore, C. P. and Dutton, P.H. (1985) Infertility, embryonic mortality and nest – site selection in leatherback and green sea turtles in Suriname. Biological Conservation, 34(3), 251-272.

Censusing Public Opinion on Tourism Development

Rachel Thomas

The group worked with Nature Seekers Inc, a non-governmental organisation responsible for the conservation of nesting turtles on Matura Beach, Northeast Trinidad. The company monitors nesting turtle numbers and behaviour and aim to produce data for further research. Nature Seekers Inc. is very prominent in the local community employing and educating locals. They also want to develop Matura and the surrounding area so it will be more attractive to tourists.

Matura has many attractions but very few facilities for tourists. Nature Seekers Inc. proposes sympathetic development of guesthouses, restaurants and shops. The surrounding area has many natural charms and the community is interesting with many local talents and skills. The only attraction currently developed and exploited is the nesting turtles and viewing tours are offered between April and August.

Nature Seekers Inc. appreciate that, in order to achieve their aim, they need the co-operation of the local residents. They asked us, as an impartial third party, to undertake a questionnaire to survey the feelings and ideas of the local community towards tourism development. We wrote a questionnaire inviting opinions, ideas and expectations, (the questionnaire is shown in Appendix V). 100 individuals were questioned and their answers recorded. The results were then summarised and reported to Nature Seekers Inc. They hope to use this data to gauge local opinion and appreciate the skills available locally. Nature Seekers also took this opportunity to recruit and we distributed employee application forms to interested individuals.

The group enjoyed the process of the questionnaire as it allowed them to get to know the local community with whom they had been living for over a month.

Everybody approached was forthcoming and happy to talk. We were invited into many homes and consequently, felt we had become a part of the community and befriended people. Initially, it was difficult to keep the interviewee focused but we gradually learnt a technique and were able to become more professional and confident. In this way, we were able to extract the relevant information in a reasonable period of time.

Most people interviewed welcomed tourism provided they were involved in major decisions and it progressed sympathetically. We asked for their perceived benefits and advantages of increased tourism in the area and received the following opinions;

¨ Increased employment and subsequent increased income

¨ Improved local and national economy

¨ Increased opportunities for the youth of the village

¨ The chance to meet people from other cultures and lifestyles

¨ The chance to improve the international opinion of Matura Village and surrounding area

¨ The chance to raise Trinidad and Tobago’s profile in the mind of the global population

¨ Improved facilities for locals e.g. roads, communication networks

¨ Improved environment e.g. cleaner, safer

¨ The chance to use skills e.g. craft making, cooking, local knowledge

¨ Increased activity in Matura as many considered it too quiet.

Contrary to what might have been imagined, most locals looked to the more developed islands e.g. Barbados and Jamaica, as examples to follow.

There were reservations about the possible spoiling of the landscape and loss of local control but generally, people trusted Nature Seekers and the Government to act in their best interests.

REPORTS OF THE TREE FROG GROUP 2001

The main body of work carried out by the tree frog in 20001 contributed to a long-term study of adhesion in several species of Trinidadian tree frog, the results of the full study are intended to be completed by September 2002. This fieldwork was essential to obtain data to help complete the understanding of how frogs of several species are affected by the increased demands on their adhesion system presented by the period of growth from new metamorph to adult frog. Additionally, the frog team spent time studying a population of the Trinidadian monkey frog, Phyllomedusa trinitatis, with respect to reproductive behaviour in particular. A number of interesting findings in this study may warrant further research by follow-up teams on future expeditions. Eggs and tadpoles were also sampled for a number of species for inclusion in a study by Glasgow University postgraduate student Mohsen Nokhbatolfogaghi into aspects of development in several species of Trinidadian frog. As a side project, complementary to the collection of bromeliads for the sampling of tadpoles and adult frogs of the marsupial tree frog, Flectonotus fitzgeraldi, one member of the frog team collected the fauna found dwelling in the leaves and central well of the bromeliad to attempt to discover whether there was an effect of height and tree type on species diversity.

The tree frog team would like to thank in particular; Mary Alkins-Koo and Karen Duhn at UWI for helping to gather equipment for the adhesion studies, Ronnie Hernandez for top herpetological and entomological advice. Jo would additionally like to thank Andrew Barnes and Darren Croft, Beth Arrowsmith, Kirsten Skinner, Liz White and Jon Bielby of Leeds University for all their help before the expedition arrived.

Mating behaviour in the frog Phyllomedusa trinitatis

Rebecca Johnson

Identifying frogs – Chin patterns and nuptial pads in Phyllomedusa trinitatis (Photos: J.Smith)

ABSTRACT

A population of Phyllomedusa trinitatis in Trinidad’s Arima Valley, was observed for a six-week period during June and July. The mating system was found to be that of a prolonged or complex breeding population, where the female (who is generally significantly larger than the male) selects a male mate. A two-metre zone around standing water was significantly the most popular perch site, but at both high rainfall and humidity, males were found to perch significantly further from water. Large males tended to perch closer to the ground than small males, but weight also correlated positively with humidity. Males perched significantly further from one another on nights with high moon illumination, and calling males perched further apart than silent males. There is evidence that Phyllomedusa trinitatis exhibits perch fidelity, though more study is required to assess this. Males were observed in combat, and exhibiting ‘leg-waving’ behaviour that is thought to be a display to indicate fighting ability. There is some evidence that alternative mating strategies exist in this species, but further study is needed to provide more evidence for this. Colour change was noted in the species, and there is evidence that this is linked to environmental conditions and may play a role in hydro- or thermoregulation.

INTRODUCTION

The hylids are a family of anurans that show a mostly arboreal way of life. Their feet have adhesive pads and their digits are elongated with and extra cartilaginous element (Allaby, 1999). The genus Phyllomedusa represents a distinctive phyletic line within the hylids, differing by having a vertical pupil, a distinctive composition of amines & polypeptides in the skin and by laying their eggs arboreally (Duellman, 1968; Cei, 1963; Kenny, 1966). Their fingers and toes have greatly reduced or absent webbing, with small terminal discs and an opposable first finger and toe, which gives them a distinctive loris-like walking, gait. They are found in lowland to moderate elevations in eastern South America including Trinidad, where the arboreal species Phyllomedusa trinitatis can be found inhabiting secondary forest and savannah island wide.

The overall aim of this project was to investigate the breeding behaviour of the tree frog species Phyllomedusa trinitatis, in particular we aimed to establish the type of mating system in place, as described by Wells (1977a), by studying the breeding behaviour of members of a population, and comparing it to what is known from other hylids, and, where possible, other phyllomedusine species. This was done by mapping locations of individuals around a breeding site, and recording interactions between them. In particular, we focussed on behaviour of interest, such as perch-site selection, advertisement calling, interactions within and between the sexes, aggressive encounters, and the occurrence of alternative mating strategies. An incidence of multiple-male mating was observed in this species on a previous expedition to Trinidad and we hoped to establish how common an occurrence this strategy was within this species. This behaviour, where a non-calling male adopts an alternative mating strategy to avoid wasting the energy required to call for females by intervening in the mating behaviour of an already amplexed pair, has been reported in other Phyllomedusines (Pyburn, 1970; D’orgeix & Turner,1995) but never specifically in this species. This study hoped to be able to confirm an observation made in the field on a previous expedition to Trinidad that seemed to suggest that this species may too exhibit such behaviour. However, this proved problematic as Trinidad was besieged by an atypically dry ‘wet’ season this year and the densities of frogs observed at sites that had previously had high density conditions were far lower than in other years. As the expression of this behaviour is dependent on the high levels of competition involved in breeding at high-density sites then it came as little surprise that there were no observed incidences of sneak mating behaviour in these frogs in this field season. However a number of other factors can give clues as to the population dynamics of an anuran population and these other factors were also examined. In particular analysis of physical biometrics of the population in relation to factors such as behaviour and environmental conditions, including typical characteristics of anurans, such as sexual dimorphism and colour change.

MATERIALS AND METHODS

Study Site

This project was carried out in the Northern Range mountains in Trinidad, at the Simla Research Station (owned by the Asa Wright Nature Centre) near the town of Arima. The site used for all the work in this project was a man-made pond, measuring 7m x 7m, in the secondary rainforest surrounding Simla. This site was chosen because of its accessibility, and because it was known from previous years to be inhabited by a substantial population of tree frogs during the wet season, belonging to a number of different species including P. trinitatis, H. minuta and H. crepitans.

The pond was built of concrete and was raised off the ground by approximately 1 metre around the perimeter. It was comprised of four smaller ponds, which were individually designated as North-East (NE), North-West (NW), South-East (SE), and South-West (SW). The SE and SW pools contained no substantial volume of water during the period of study but had vines and small leafy plants growing in them. The NE pool contained a small volume of water, the depth of which being dependent on recent rainfall, and the NW pool was always full of water, and aquatic plants. There were a number of branches overhanging the NE and NW ponds, some of which were fairly low hanging, which made it a possible P. trinitatis nest site (Kenny, 1966).

Since Phyllomedusa trinitatis are not always found calling directly over the water (Kenny, 1966), the main area of study was extended to include bushes and trees within a 14 x 14m area around the ponds. The edges of this area were then marked out using string, and the study site further subdivided using electricians tape to mark 1m intervals along the boundary string. Each resulting square metre was then assigned a grid reference, composed of a letter and number, dependant on its position.

Observations

Observations were made at this site over the course of a six-week period during July and August 2001. Four observers were also involved in the collection of data, work which was mainly carried out between the hours of 11pm and 2am (the parameters for this being set by examining published accounts of mating behaviour in Phyllomedusa trinitatis to establish the most active times (Kenny, 1966; Murphy, 1997) The study area was searched thoroughly for half an hour with handheld and head torches, to locate any frogs in the study site. Pairs of observers then recorded behaviour in these frogs over the course of the night.

At the end of the field observation period, the frogs were collected and a quick field-sketch made of the chin pattern to aid recognition in the lab. The frogs were then transported to the research station in a cloth enclosure for further work. Snout-vent length (SVL) was measured using a pair of callipers accurate to 0.1mm. Weight was recorded by placing the frog in a small cotton bag suspended from a small Pesola balance (50g maximum weight). Skin colour was defined using Royal Horticultural Society colour charts.

Each frog was then checked against previous identification charts to establish its identity. If a frog was found to be a new capture, an identification chart was completed with a more detailed sketch of the frog (dorsal and ventral views, plus both side views), noting down any distinguishing features to aid recognition on recapture. Each frog was given a number and sex was determined by the presence or absence of nuptial pads and by weight.

Environmental Data Collection

Temperature and humidity was recorded on a nightly basis using a digital thermometer (to 0.1 degrees C) and humidity gauge. Information on rainfall in the Arima Valley was obtained from weather recording equipment at the Asa Wright Nature Centre (approximately 5km further up the valley), and information on lunar cycles from the USA Navy Astronomical Applications Department.

RESULTS*

Capture results and analysis

Figure 1: Numbers of frogs observed on nights of study.

Figure 1 shows the number of frogs observed over the course of the study at the study site from night to night. Total observed differs from total captures as frogs were occasionally too high in the dense foliage to be captures. Different numbers were present at the pond from night to night suggesting that an outside influence affects the population’s activities.

Data was analysed with respect to environmental factors to determine what these factors were. Temperature and rainfall had little effect on capture numbers but a weak correlation was exhibited seen between capture rate and humidity (Figure 2).

Figure 2: Frogs captured on nights of varying humidity

Capture rate of new individuals showed a steady decline across the time of study. This suggests that the population present at the pond is reasonably fixed, since fewer new captures occurred with time – the second small peak (Figure 3) occurring towards the end of the study may suggest that new individuals were beginning to visit the area.

Figure 3: Numbers of new and recaptured frogs visiting the site on study nights

Physical data results

Variations were found both between and within individuals in terms of weight and dorsal colouration, with both fluctuating over the season. Reasons for changes in mean weight of frogs were examined with respect to a number of environmental factors. Weight was found to have a weak negative correlation with temperature and a stronger positive correlation with humidity (Figure 4). There was also a tendency for mean weight to be significantly lower on nights with high moon illumination.

Figure 4: Mean weights of frogs vs. percentage humidity

Colouration of frogs also varies nightly, there is some evidence that darker frogs are more prevalent on warmer nights but the most convincing relationship between environmental factors and frog colouration is evident when examining the levels of moonlight illumination and the different percentages of frogs of different shades (Figure 5) The greater the proportion of the moon illuminated, the higher the number of light frogs found and conversely the lower the levels of moon illumination the higher the proportion of dark-shaded frogs seen.

Figure 5: Percentages of light, medium and dark shades of frog found according to lunar cycle

Numbers of frogs observed in each square within the grid are shown in Figure 6 (left) with light blue squares denoting the pond area. Numbers of frogs found within areas at certain distances from the water were analysed.

	Over	0-1m	1-2m	2-3m
Frogs	7	31	29	7

Table 1: No of frogs at distances from pond.

Chi-squared analyses found that frogs were found in significantly greater numbers within 0-2m of the pond (p < 0.01).

Perching preferences

	N	M	L	K	J	I	H	G	F	E	D	C	B	A
1						1	1
2	1				3	1		1		1	1		1
3	1			1	2	2	5				1	4	2
4			1				2				2	1	1
5			1							1		3		1
6			1	3				1		1		2
7			1	4
8						1								1
9				2			1	1				1		2
10					1	1		3	2	1	1
11									1
12
13
14

Figure 6: Number of frogs sampled in grid squares.

Although some evidence exists of individual perch fidelity there is variation in where frogs are found to perch on different nights in terms of heights and distances from water and nearest neighbours. These were examined with respect to nightly temperatures, rainfall and humidity and the lunar cycle. Humidity and rainfall appear to play a key role in the distance from the water that a frog chooses to perch on a nightly basis. Figure 7 shows the relationship between these factors.

Figure 7: Distance from water as affected by percentage humidity

Pearson’s correlations were performed between arcsine-transformed data and distances from water and statistical significance was found between horizontal distance from the water and percentage humidity (r = 0.62, p < 0.01, 30d.f.) Perch height and distance from water was also examined in terms of the physical parameters measured for each frog. A weak negative correlation appears to exist between height and weight of males such that heavier frogs tend to prefer lower perches but this relationship is a tenuous one

Calling behaviour

Numbers of frogs calling varied on a nightly basis, with calling frogs being more likely to be mobile than non-calling frogs. Strategies varied within individuals on a nightly basis also. No demonstrable relationship between calling incidence and the measured environmental variables were found but mean distances from nearest neighbours were significantly higher amongst calling frogs and a tendency was shown by calling frogs to be closer to the water both vertically and horizontally (Figure 8).

Figure 8: Differences in distance variables between calling and non-calling males

Mean weights of calling males and non-calling males were examined and although a trend exists such that calling frogs were heavier than non-calling ones this is not a significant one (Figure 9).

Figure 9: Mean weights of calling and non-calling males

DISCUSSION

Mating System

The observations made in this study suggest that Phyllomedusa trinitatis’ mating system is that of a ‘prolonged’ or ‘complex’ breeder, rather than an ‘explosive’ system as suggested by Wells (1977b) and Roberts (1994). Kenny’s (1966) description of courtship, mating and nest building in this species is also consistent with that of a prolonged breeding population. The unseasonably dry weather and lack of females observed in this study meant that based on this study mating system was difficult to determine, and it is problematic to assess whether the frogs are mating and ovipositing in their home range (a characteristic of a prolonged breeding species), since no work has ever been carried out to assess the home range of this species. The best evidence that this breeding site is in the phyllomedusines’ home range is that the rate of new captures declined as the study progressed. This implies that a reasonably fixed population of frogs live in the area in question throughout the season, and come to the breeding site to attempt to attract a mate on certain nights.

Perching Location

The observations on perch site position in this study, mirror that of Kenny (1966) and Murphy (1997), with males perching on branches and leaves of bushes and trees. Less than 7% of occupied perch sites were further than 2 metres from the edge of the pool, even though numerous suitable perches were available among the dense vegetation. This is strong evidence that Phyllomedusa trinitatis show a preference in perch site location, in relation to distance from the body of water in question. This has been observed in other hylids (e.g. Fellers, 1979; Kime et al., 2000) where certain perch sites were found to be particularly advantageous calling sites, that interfere less with the projection of the call than others.

Perches occupied from night to night did vary, and this variation has been found to correlate with the environmental conditions on a given night. The higher the humidity on a given night, the further the mean distance from the pond for all the males observed that night – this is a significant correlation. Also, on nights after heavy rainfall, males were observed perching significantly further from the water and higher above the ground. When seeking an explanation for this, it is important to remember that anurans are highly dependent on moisture (Tyler, 1994). In fact, the mean weight of the males present on each night at the study site was found to positively correlate (significantly) with humidity, suggesting that water loss plays a large role in weight. Perhaps on nights with highly elevated humidity, or after heavy rainfall, males can venture further from their nearest water source, since they can go longer between rehydrations than they would be able to on dry nights.

The variation in perch location may depend not on external factors, but on variations within the frogs themselves. No significant correlation is seen between SVL or weight, and perching location, though there does seem to be a weak relationship between mean weight and mean height above ground. Larger frogs tended to perch closer to the ground than smaller frogs. Perhaps if a frog is closer to the ground he can intercept females entering the site via the ground, and large frogs are dominant so they can select these optimal sites. The small number of females observed in this study were not observed as they entered the site, so it is unclear whether they do enter breeding sites via the ground.

Spatial Separation, Predation and the Evolution of Vocalisation

In many anurans, there is a degree of spatial separation between calling males (Wells, 1977b), but in this population, calling males were often observed perching very close to one another, with no agonistic consequences. The only environmental factor, which correlates significantly with an individual’s distance from its nearest neighbour, is moon illumination. On nights when a higher proportion of the moon was illuminated, frogs were found perching further from each other than on nights with lower moon illumination. Tropical forest birds are known to prey upon frogs (Poulin et al., 2001), and nocturnally, bats have been observed swooping on Phyllomedusa trinitatis frogs at this study site (Smith, J., pers. comm.). One night during the study, a Fer-de-lance snake was observed striking an individual at the site, so snakes are obviously a threat too. Perhaps on bright nights, predation pressure is higher, so frogs perch further from one another to avoid detection. Predators often locate frogs by their vocalisations, so two frogs, both calling close together will give double the number of calls, and therefore double the vocal cue used by a predator to detect them. Tuttle & Ryan (1981) found that the rate at which bats capture frogs is significantly higher when the frogs are calling, and is higher still when the call is rapid. Assuming that two calling frogs sitting close together would not reduce their rate of calling, this would increase their chance of detection by a bat predator would be greater than if they were calling alone. Males in this population were found to perch significantly further from their nearest neighbour when calling, than when they were silent.

Perch Fidelity, Territoriality and Aggression

Certain individuals in this phyllomedusine population show a degree of perching fidelity. A recent study by Luddecke (1996) on Hyla labialis, found that almost 100% of frogs recaptured were in the same perch site as they had previously been captured in. Stewart and Rand (1991) found a similar pattern of perch fidelity, with recaptured males consistently being caught within five metres of their original perch site.

Agonistic encounters were observed, but they were not necessarily between close-perching individuals. Fights took the form of those described by Murphy (1997), with the males pushing and grappling while standing on (or hanging by) their hind legs. Males were not observed to attack those in amplexus, as is noted by Pyburn (1970) for two other Phyllomedusa species.

Game theory predicts that at least some of the behaviour patterns displayed during aggressive encounters are used to assess asymmetries in variables that indicate fighting ability (Hoffman & Schildberger, 2001), and it seems that this may be the function of the ‘leg-waving’ behaviour this species exhibits. This occurred between males, and often before fights, suggesting that the males may have been ‘sizing each other up’, to see if there would be an obvious victor from a fight. Out of the four frogs that exhibited leg-waving, only one did not begin to fight immediately afterward, suggesting that the males observed fighting were very closely matched physically, so fights ensued. This seems likely, since the one male who did not fight after leg waving, had the third largest SVL out of all the males observed, suggesting that he successfully indicated that he had a greater fighting ability, and the other male retreated.

Colouration

Kenny (1966) stated that colour change in Phyllomedusa trinitatis is dependent on mating condition. This could not be investigated in this study due to the low number of matings observed. Variability in colour was observed throughout the population, and within individuals themselves. Kenny states that two colour phases are seen in this species (vivid green and dull olive-brown), but the use of colour charts in this study revealed a far more complex pattern of colour change. Colour varied from frog to frog and within a single individual, and certain colours were more abundant than others, with ‘medium’ shades being seen most frequently.

Colour is linked with moon illumination, such that light shaded frogs were more commonly seen on nights with high moon illumination, and dark frogs were seen on nights with low moon illumination, though the significance of this could not be assessed. Amphibians can shift in colour to match their background (Bagnara & Hadley, 1969) and in this case is presumably due to the need for camouflage, since leaves would appear a lighter shade on lighter nights, and a darker shade on darker nights. There also seems to be a relationship between humidity and colour, in that a greater proportion of medium-shaded frogs were observed at mid-range humidities. This could mean that colour is related to hydroregulation or thermoregulation, as suggested by Hoppe (1979), though more study would be required to assess whether this is the case.

Alternative Mating Strategies

Jennions et al. (1992) state that alternative mating strategies are a consequence of male-male competition for a limited number of females, and since the sex ratio in this population was so highly skewed, with only 7.4% of captures being female, the population fits this requirement.

They also state that stationary/calling behaviour is the dominant reproductive strategy in anurans. Stationary/calling behaviour was exhibited by males in this population, along with the alternative strategies of being mobile/calling, mobile/silent, and stationary/silent. No satellite males were observed directly, but evidence is available from other variables and observations of the population. Calling frogs were on average heavier than silent males, and the proportion of nights a male spent calling was found to be positively correlated with SVL, suggesting that larger males may be dominant to smaller males, though neither of these relationships is significant. There was a weak negative correlation between SVL and the proportion of nights a male spent being mobile, supporting this theory, but the correlation is not significant. Calling and mobility were observed in around 60% of all captures, suggesting that this is the dominant strategy. Nearly every male captured exhibited each of these behaviours at least once, though few adopted one strategy permanently. ‘Switching’ between these strategies was also observed in a number of males, a common feature of populations with alternative mating strategies in place (Perril et al., 1982).

REFERENCES

Allaby, M. (1999) Oxford Dictionary of Zoology (2^nd Edition). Oxford Uni Press, London

Bagnara, J.T. & Hadley, M.E. (1969) The control of bright coloured pigment cells of amphibian and fishes. American Zoologist 9; 465-478

Cei, J.M. (1963)Some precipitin tests and preliminary remarks on the systematic relationships of the four South American families of frogs. Bulletin of the Serological Museum 30: 4-6

D’Orgeix, C.A. & Turner, B.J. (1995) Multiple paternity in the red-eyed tree frog, Agalychnis callidryas. Molecular Ecology 4(4); 505-508

Duellman, W.E. (1968) The genus of Phyllomedusine frogs. University of Kansas Publications Museum of Natural History 18 (1): 1-10

Fellers, G.M. (1979) Aggression, territoriality and mating bahaviour in the North American tree frogs. Animal Behaviour 27: 107-119

Hoffman, H.A. & Schildberger, K. (2001) Assessment of strength and willingness to fight during aggressive encounters in crickets. Animal Behaviour 62 (2): 337-348.

Hoppe, D.M. (1979) The influence of colour on behavioural thermoregulation and hydroregulation. In (Burtt, E.H. ed.)The Behavioural Significance of Colour. Garland Press; New York.

Jennions, M.D.; Backwell, P.R.Y. & Passmore, N.I. (1992) Breeding behaviour of the African frog, Chiromantis xerampelina: a multiple spawning and polyandry. Animal Behaviour 44; 1091-1100

Kenny, J.S. (1966) Nest building in Phyllomedusa trinitatis. Caribbean Journal of Science 6: 15-22

Kime, N.M.; Turner, W.R. & Ryan, M.J. (2000) The transmission of advertisement calls in Central American frogs. Ecology 11(1): 71-83

Luddecke, H. (1996) Site fidelity and homing ability in Hyla labialis. Alytes 13 (4): 167-178.

Murphy, J.C. (1997) Amphibian and Reptiles of Trinidad and Tobago. Florida: Krieger Publishing.

Perrill, S.A. ; Gerhardt, H.C. & Daniel, R. (1982) mating strategy shifts in male green tree frogs (Hyla cinerea): an experimental study. Animal Behaviour 30; 43-48

Poulin, B.; Lefebvre, G.; Ibanez, R., Jaramillo, C.; Hernandez, C. & Rand, A.S. (2001) Avian predation upon lizards and frogs in a neotropical forest understorey. J. Tropical Ecology 17(1): 21-40.

Pyburn, W.F. (1970) Breeding behaviour of the leaf frogs Phyllomedusa callidryas and Phyllomedusa dacnicolor in Mexico. Copeia 1970 (2): 209-218

Roberts, W.E. (1994) Explosive breeding aggregations and parachuting in a neotropical frog Agalychnis saltator. Journal of Herpetology 19(3): 432-436

Stewart, M.M. & Rand, A.S. (1991) Vocalisations and the defence of retreat sites by male and female Eleutherodactylus coqiu. Copeia 1991: 1013-1024

Tuttle, M.D. & Ryan M.J. (1981) Bat predation and the evolution of frog vocalisations in the neotropics. Science 214: 677-678

Tyler, M.J. (1994) Climatic change and its implications for the amphibian fauna. Trans. Roy. Soc. S. Austral. 118(1-2): 53-57

Wells, K.D. (1977a) The social behaviour of anuran amphibians. Animal Behaviour 25: 666-693.

Wells, K.D. (1977b) Territoriality and male mating success in the green frog. Ecology 58: 750.

Adhesion and Behaviour in Trinidadian Tree frogs

Joanna Smith

Many species of frog, over a wide range of taxa, possess specialised and often expanded digital pads, which allow them to adhere to smooth and often vertical surfaces. This adaptation is particularly prevalent amongst arboreal frogs in the families Hylidae, Microhylidae, Centrolenidae, Rhacophoridae and Hyperolidae (Emerson and Diehl, 1980). This observed prevalence amongst arboreal frogs has meant that the possession of specialised toe pads has become an identifying characteristic in giving such frogs the common name ‘tree frog’.

A number of hypotheses have been put forward to explain the physical mechanism by which tree frogs adhere to a surface. These include interlocking, friction, dry adhesion, wet adhesion and suction. Recent evidence suggests that wet adhesion is the mechanism employed (Emerson & Diehl, 1980; Green, 1981; Hanna & Barnes, 1991). Here, a fluid-filled joint resists cavitation as a result of two separate forces. The first of these is a surface tension and the second is the viscosity of the mucosal layer between pad and substrate, Recent evidence (Barnes, 1999) suggests that the most important of these two forces is the surface tension, dependant on the surface area in contact.

If frogs do indeed adhere by wet adhesion, then this system of adhesion may be expected to have difficulty with the large mass increase involved when a frog develops from new metamorph to adult, dependent as it is on increasing surface area. The adhesive ability of several species of frog is currently under investigation in relation to whether the frogs are indeed able to keep pace with changes in their weight as far as their adhesive abilities are concerned, and if so how are they able to do this? This study is investigating adhesive ability in relation to growth rates in terms of increasing weight and toe pad area in several species of Hylid frog. This is a long-term study and results will be available in full detail in early 2003.

MATERIALS AND METHODS

Experimental animals

On two previous visits to Trinidad adults and juveniles from 10 species of frog of different sizes were collected and their adhesive abilities determined. In the expedition to Trinidad 2001, I was able to complete adult data sets for two of these previously studied species, Hyla geographica and Hyla boans, regarding their adhesive abilities. Due to the tree climbing skills of Jimmy McVeigh and Dan Thornham we were able to collect three times the number of adult Hyla boans, Trinidad’s largest tree frog, than has been possible in previous years. As this study of adhesion in tree frogs examines the ways by which the frogs are able to scale their adhesive abilities to match increase in mass, the collection of adults and juveniles belonging to this particularly large species is essential for the success of a comprehensive comparative study of adhesion. Data sets for two new species were also obtained. We were able to collect a number of adults belonging to Trinidad’s smallest two species of frog, Hyla miniscula and Flectonotus fitzgeraldi.

Studies were also undertaken to determine whether differences exist between the sexes in adults of the species under study. Females were examined in both gravid and non-gravid states. Initial findings suggest there are some differences in adhesive abilities in the sexes, such that females have larger angles of last contact than males in both their gravid and non-gravid states though when gravid their adhesive abilities were significantly and detrimentally affected. This is an interesting finding as there are few examples of sex differences affecting physical performance in herpetology. These results are not discussed here due to constraints on size but will be further examined for inclusion in the long-term study being conducted by myself.

Frogs were taken back to the lab where they were weighed and measured using callipers to obtain a snout-vent length (SVL). Markings were noted on ID sheets. At least one individual in each species, and where possible a male and female of each species, had front and back feet photographed through acetate marked with a grid of one millimetre square using a Polaroid MicrosSLR camera attached to a binocular microscopes. The images were then analysed using Cherry Digitiser software to determine toe pad area on return to Glasgow.

Measurement of sticking ability

The sticking ability of frogs was determined using the protocol of Hanna & Barnes (1991). Each frog was placed on a Perspex platform attached to a kymograph spindle and smoothly rotated from 0º (horizontal) through 90º(vertical) to 180º (upside down). The angle at which the frogs fell from the platform was then recorded from a pointer attached to the platform, which showed the angle on a 360º protractor attached to the base of the spindle. Each frog was rotated ten times on the platform. The adhesive force can then be calculated simple geometry thus:

Adhesive force (mN) = cos (180-fall angle) x body mass x force of gravity

Frogs are placed in an orientation such that they faced up the rotation platform as it turned and records only taken if they maintained this orientation throughout, this precaution is due to findings that sticking ability is impaired in frogs forced into a downwards facing orientation (Hanna & Barnes, 1991). Most of the larger species of frog will automatically orient themselves into a head upwards position and so this was only problematic for smaller, jumpier species such as Flectonotus fitzgeraldi.

Frogs were prevented from jumping by the light cupping of hands around the frog as the platform rotated, this prevented the frog from seeing any objects onto which it might jump to escape and calmed the frog down to some extent. Any incidences where there was any indication of the frog jumping from the platform were not recorded as a valid result.

Many frogs also utilise additional areas of skin on the stomach and upper thighs when adhering naturally. Any incidences of utilisation of extra skin were noted and fall angles discounted if any skin other than toe pads was in contact.

If the frog exhibited signs of stress, the bringing of the lower eyelid over the eye being a main indicator of this, along with the cessation of active attempts to maintain a hold on the surface, then the experiment was stopped. The frog was then misted with water and transferred to a darkened tank to calm down for at least half an hour before the experiment was resumed.

RESULTS*

Flectonotus fitzgeraldi

Flectonotus fitzgeraldi are a small species of frog with adults caught in Trinidad this year ranging between 18 and 22 mm SVL. They spend the majority of their lives in bromeliads found particularly in the wet Montane Forest and were found by us in numbers at a site on the Morne Bleu Ridge. They are unique amongst Trinidadian frogs in being ‘marsupial’, with the female carrying fertilised eggs in a specialised skin pouch on their back until they are well developed and the late stage tadpoles released into the wells of bromeliad where they develop into tiny froglets measuring an average 7.93mm SVL (± 0.26, p < 0.05, n = 10).

Adhesive abilities are able to keep pace with the increase in weight exhibited in this species: Figure 1A shows that the increase in adhesive force is as SVL^2.51 as opposed to the increase in weight which increases as SVL^2.53 (Figure 1B), with these two slopes not significantly different to one another (t = 0.125, non-sig, 22df). This increase in weight is lower than expected from isometric growth (Difference from 3: t = 3.98, p < 0.05, 22df). The increase in toepad area (Figure 1C) is as would be expected at a rate of SVL^2.06 (Difference from slope of 2: t = 0.35, non-sig, 4df).

Flectonotus fitzgeraldi are very active little frogs and dorsoventrally flattened, a feature particularly pronounced when observing them on the rotation platform. They are extremely competent at adhering to the rotation platform, with the tiny froglets able to rotate through 360 degrees without having to alter their position on the board (Average angle: 179.33º ± 1.40, p = 0.05). Adults were active on the board and walked around readily, often maintaining a hold to 180 degrees (Average angle: 174.63 ± 8.04, p = 0.05).

Figure 1A: Log-log plot: Adhesive force vs SVL in Flectonotus fitzgeraldi

r = 0.99, y = 2.51x –2.51, t = 22.27, p < 0.001, 24df.

Figure 1B: Log-log plot: Weight vs SVL in Flectonotus fitzgeraldi

r = 0.99, y = 2.53x – 3.52, t = 21.46, p < 0.001, 24df.

Figure 1C: Log-log plot: Toe pad area vs. SVL in Flectonotus fitzgeraldi

r = 0.99, y = 2.07x – 1.92, t = 11.91, p < 0.01, 14df.

Hyla geographica

Hyla geographica is a medium sized frog, of a comparable size to Hyla crepitans. Adults average at 57.39mm SVL (± 2.38, p = 0.05, n = 14). They are docile frogs and easily stressed, found calling from riverside grass land.

Hyla geographica are large sized froglets at metamorphosis, at around 20-21mm SVL and 0.6g. Juvenile frogs of this species are very difficult to maintain in lab conditions, and so data relating to juvenile frogs are for the first three months of development, the largest size attained by the longest surviving froglet being 25mm SVL and 1.31g. All further attempts to maintain the frogs in Glasgow for longer periods than this have been unsuccessful.

Adhesive abilities are not only able to keep pace with the increase in weight exhibited in this species but in fact increase at a greater rate than expected: Figure 2A shows that the increase in adhesive force is as SVL^2.90 as opposed to the increase in weight which increases as SVL^2.48 (Figure 2B), with the first of these slopes being significantly greater than the second (t = 2.73, p < 0.01, 66df). The increase in weight is lower than expected from isometric growth (Difference from 3: t = 6.93, p < 0.05, 66df). Although the increase in toe pad area shows a trend towards a greater than expected rate of increase as SVL^2.15 (Figure 2C) this is not a statistically significant trend (Difference from slope of 2: t = 0.74, non-sig, 4df).

As adults Hyla geographica have a low angle of initial slip (68.2°±4.08, p = 0.05, n=10). They are less jumpy frogs than many species, becoming fairly subdued when handled. They press themselves close to the board and are able to maintain this highly flattened pose until angles beyond 90°, preferring to stick using their belly skin until this point. The frogs have a lesser tendency to adjust their centre of gravity through spread of the back legs as is seen in other species. Average angle of last contact in adults is relatively low (127.55 ± 6.48, p = 0.05, n=21), with higher angles seen in the juvenile ageclass (149.52 ± 7.66, p = 0.05, n = 21)

Figure 3A: Log-log plot: Adhesive force vs SVL in Hyla geographica

r = 0.97, y = 2.90x –3.30, t = 22.27, p < 0.001, 66df.

Figure 3B: Log-log plot: Weight vs SVL in Hyla geographica.

r = 0.99, y = 2.48x – 3.45, t = 33.24, p < 0.001, 66df.

Figure 3C: Log-log plot: Toe pad area vs. SVL in Hyla geographica.

r = 0.99, y = 2.15x – 1.78, t = 11.35, p < 0.01, 14df.

Hyla boans

Hyla boans is Trinidad’s largest species of frog, with adult SVL measurements in 2001 averaging at 95.50 ± 1.66, p = 0.05. Like Hyla geographica these frogs breed in the dry season but males do continue to call all year round, possibly to mark territorial boundaries. These frogs are more easily captured than females who are not known to call and consequently males dominate in the adult sample and values for average size of frogs caught are therefore lower than expected (110-115mm SVL, Murphy, 1997). Newly metamorphosed froglets are in size so that by the time the frogs have matured they will have undergone a hundredfold increase in mass.

This is a considerable weight increase for the adhesion system to cope with, and this species isn’t in fact able to maintain adhesive abilities with weight increase (t = 25.0, p < 0.05, 29df), with adhesive forces increasing as SVL^2.34 (Figure 3A) as opposed to increase in weight as SVL^2.84 (Figure 3B). This is in spite of the fact that as with other species, Hyla boans shows a lower than expected mass to body length ratio than expected (Difference of weight slope to slope of 3: 6.15, p <0.05, 30df). Toe pad area increase is difficult to assess as there is insufficient data available to examine increase directly, generating data by doubling the available points (Figure 3C) gives an increase in toe pad area as SVL^1.93, an increase of around the expected rate (no stats available).

Adults slip at average angles of 72.67°±27.54 (n=40) at which point they start to shuffle on the rotation platform drawing their feet beneath them. As the rotation progresses the frogs spread their front toes which are heavily webbed and walk up the platform, the average angle of last contact for this species is the second lowest of the ten studied at 111.63°±27.54 (n=40).

Figure 4A: Log-log plot: Adhesive force vs SVL in Hyla boans

r = 0.98, y = 2.34x, t = 16.74, p < 0.001, 28df.

Figure 4B: Log-log plot: Weight vs SVL in Hyla boans

r = 1.00, y = 2.84x, t = 110.61, p < 0.001, 30df.

Figure 4C*: Log-log plot: Toe pad area vs SVL in Hyla boans

r = 1.00, y = 1.93x – 1.43, t = 85.71, p < 0.001, 4df.

* from generated data – created by doubling available data.
Summarised results

It is important to note that the juveniles studied here have not reached adulthood and that presuming the fit of the lines in these cases, where gaps in the data exist are unsatisfactory and results treated with care. However these results are similar to those obtained for other similarly sized species for which greater data sets are available and so cannot be disregarded. It is also important to note that it seems likely that at some stages of a young frogs development toe pad area does increase disproportionately such that overall results for all age classes may not adequately illustrate this trend and its importance at specific stages of a frog’s development.

Species in order of increasing size as adults	Average fall angle (adult)	Log-log slope of adhesive force vs SVL	Log-log slope of weight vs SVL	Log-log slope of toe pad area vs. SVL
Flectonotus fitzgeraldi	174.633	2.51	2.53	2.06
Hyla geographica	127.55	2.90	2.48	2.15
Hyla boans	106.37	2.34	2.84	1.93

Table 1: Summary statistics for three species of hylid frog.

v In all three species, the rate of increase of adhesive force was greater than would be expected by increase in surface area of toe pad alone. In the two smaller species the rate of increase in adhesive force is able to keep pace with the increase in weight involved as the frog grows.

v In all three species, the increase in mass is lower than expected from isomeric predictions. The difference from expected rate of increase is lowest in Hyla boans, the only species in which the adhesive force was unable to match increase in mass.

v In all three species the increase in toepad area is not statistically significant from expected. If toe pads are not proportionally bigger in larger frogs then perhaps some alteration is occurring in the structure of the pad to allow the greater than expected increase in adhesive abilities.

Work is still under way on preserved toe pads for species described here but preliminary observations certainly suggest that changes in the structure of the toe pad are occurring between metamorphosis and adulthood that may affect toe pad efficiency. Toe pad cell size is visibly different between juvenile and adult Flectonotus fitzgeraldi and statistical data obtained from SEM images (see right for an example of Flectonotid toe pad) backs up these observations. Average cell size in juveniles is 120.67μm² ± 37.31 (p = 0.05, n = 7), compared to 86.57 μm² ± 10.26 (p = 0.0, n = 9) in Adult 1, with the cells in adults being significantly smaller (t = 2.39, p< 0.05, 14df). This is a pattern seen in previously examined species and the confirmation of this pattern within a newly examined species such Flectonotus fitzgeraldi is an exciting one. There is also evidence that toe pads are becoming more complex in terms of the depth of circumferal grooves at the margins of the adhesive pad and that the elevation of the pad itself is greater in the adults of this species as compared to the juveniles. SEM work is continuing to examine the significance of these observed differences in toe pad structures on observed adhesive abilities both within and between species. As has been alluded to previously, the data collected in 2001 in Trinidad is being included in a larger and more extensive study of adhesion in Trinidadian tree frogs, intended for completion late in 2002. Data obtained for the new species described here although difficult to draw firm conclusions from alone show similar trends to other data sets obtained and extending the species range studied to include Trinidad’s two smallest and its largest species is extremely important and I am grateful for the opportunity to have done so.

REFERENCES

Barnes, W.J.P. (1999) Tree frogs and tire technology. Tire technology International (March) 42-46.

Emerson, S.B. & Diehl, D (1980) Toe pad morphology and mechanisms of sticking in frogs. Animal Behaviour 45, 893-899.

Green, D.M. (1981) Adhesion and the toe pads of tree frogs. Copeia 1981, 790-796.

Hanna, G. & Barnes, W.J.P. (1991) Adhesion and detachment of the toe pads of tree frogs. Journal of Experimental Biology 155, 103-125.

Murphy, J.C. (1997) Amphibian and Reptiles of Trinidad and Tobago. Florida: Krieger Publishing.

Fauna in Bromeliad wells

Dan Thornham

Dan collecting bromeliads at Simla

(Photo J.Ward)

climbing trees, chosen according to the high density of bromeliads on their branches. The collection height was recorded for each climb, as was the tree/collection site. The bromeliads were taken apart leaf by leaf, either on site or in the laboratory at Simla. For a proportion of these bromeliads all the visible fauna were collected and logged though as each bromeliad took approximately three quarters of an hour to dismantle in this way, so only a small number could be analysed each day in addition to the other work being carried out by the group. All except the tadpoles and hoverfly larvae were preserved immediately in alcohol. The hoverfly larvae were incubated in small vials and transported back to Britain, along with the preserved samples for identification. Identification is still underway on the 400+ samples collected in this study. The bromeliads had

Bromeliad centipede and babies (Photo: J.Smith)

APPENDICES

I: Accommodation

Simla Research Station

Views around Simla (Photos: C. McCulloch)

The expedition this year stayed at three different locations – using Simla Research Station in the Arima Valley as a main base. Simla Research Station is situated in secondary forest in the Arima Valley. The Research Station has had a long history of occupation by large and riotous parties of people, having been a governor’s retreat and US field hospital in previous incarnations. In the 1950s the building was purchased by the New York Zoological Society as a research base for expeditions under the leadership of the eminent and much-travelled naturalist, William Beebe. William Beebe’s love of the area was such is that he was recorded to say that the view of the Arima Valley from Simla ‘surpasses anything I have before seen’, a view to which by the end of the expedition many of our members would have ascribed. The Research Station is now owned and run by the nearby Asa Wright Nature Centre.

The grounds around Simla are still rich in wildlife, in spite of the encroaching threat of the nearby quarries. In the course of our stay we recorded over 40 species of birds and 10 species of frog, including a population of over 40 Phyllomedusa trinitatis at the lower ponds.

Although slightly more expensive than accommodation we have stayed in previously having our own water supply, (especially this year when the rest of the island was under severe water restrictions), and the extreme proximity to suitable sites for research for both the bird and frog groups, literally on our doorstep, meant that for us the benefits of working at Simla far outweighed the extra costs. We also had the added benefit of being able to work with Simla’s caretaker, Ronnie Hernandez, whose knowledge and enthusiasm about the Arima Valley and its wildlife proved to be invaluable.

II: Personnel

Back row: Jen Ward, Jimmy McVeigh, Dr. Stewart White, Dan Thornham, Carrie Finlayson, Gemma Doran, Roxy McLennan, Oisin Sweeney. Front row: Carrie McCulloch, Rachel Thomas, Ally Humphries, Nev Broadis, Kirsty MacNaughton, Jo Smith, Beccy Johnson.

Name	Status	Group	Expedition role
Nev Broadis	3^rd year UG (Zoology)	Birds	Driver/ project leader
Gemma Doran	2^nd year UG (Biology)	Birds	Machete wielder
Carrie Finlayson	3^rd year UG (Zoology)	Birds	Project leader
Ally Humphries	3^rd year UG (Zoology)	Turtles	Driver/ project leader
Rebecca Johnson	3^rd year UG (Zoology)	Tree frogs	Project leader
Carrie-Ann McCulloch	2^nd year UG (Biology)	Tree frogs	Enthusiasm supply
Roxana McLennan	2^nd year UG (Biology)	Turtles	Community outreach!
Kirsty MacNaughton	3^rd year UG (Zoology)	Birds	Project leader
James McVeigh	3^rd year UG (Zoology)	Tree frogs	Tree climber
Joanna Smith	3^rd year PG (Zoology)	Tree frogs	Expedition leader
Oisin Sweeney	3^rd year UG (Zoology)	Turtles	Driver/ project leader
Rachel Thomas	2^nd year UG (Medicine)	Turtles	First aider
Daniel Thornham	4^th year G (Zoology)	Tree frogs	Driver/ first aider
Jennifer Ward	3^rd year PG (Zoology)	Birds	Ringer/ first aider
Stewart White	Associate lecturer	Birds	Staff leader/ driver

III: Finances

INCOME

SOURCE £

Previous balance 652.58

Personal contributions (15 people + 3 flights) 10399.30

Trust Funds

Albert Reckitt Trust 750.00 British Ecological Society 1250.00

Carnegie Trust 2000.00

Dennis Curry Charitable Trust 2000.00

Duke of Edinburgh 500.00

Glasgow Natural History Society (Blodwen Lloyd Binns Bequest) 600.00

Glasgow University Court 1600.00

Local businesses 110.00

Royal Geographic Society (minus registration fee of £10) 740.00

Tescos 250.00

Walter Guinness Charitable Trust 500.00

Fundraising Events

Snack bar 775.00

Pub golf 71.85

Blackfriar’s ska night 235.00

Uisge Beatha Pub quiz night 80.40

Reds Rum Dubh (minus deposit (£50)) 170.00

Rum Dubh 3 101.30

Sale of t-shirts 360.00

Raffle (minus initial ticket cost of £40) 1077.00

In Trinidad

NGS kids camp* 172.00

TOTAL 24394.43

EXPENDITURE

SOURCE £

Insurance 1375.00

Flights 8381.10

T-shirt printing 513.00

Accommodation (Simla) 4936.11

Accommodation (Nature Seekers) 2569.44

Food 1788.81

Car Hire 1986.11

Car repairs 118.64

Petrol 635.59

Equipment 38.98

First Aid Courses (and kits) 80.00

Departure Taxes 169.50

TOTAL 22461.68

BALANCE (for use in future expeditions) 1658.50

IV: NGS Kids Camp

Group of children from Camp Olympia 2001 – The NGS Summer camp at Couva (Photo: J.Smith)

In 2001, a personal contact established a relationship between Glasgow University Exploration Society Expeditions and the National Gas Company of Trinidad and Tobago who run a summer camp for children of their employees over a fortnight in July. Following this highly successful initial collaboration we were approached again to take part. In recognition of the camps theme of a Trinidadian biodiversity alphabet we became involved in field trips to inform the camp of around 50 children aged 5-11, about the reasons for us coming all the way to Caribbean from Britain to study the animals of Trinidad. The children were, in the majority, from the South of the island and for many these were their first trips to Nature Reserves on the North side of the island.

The first of these was the beautiful man-made reservoir of Hollis Dam, in the North of the island. Members of the expedition Dan Thornham, Carrie McCulloch and Joanna Smith joined Roger Downie, Naomi Barron and Sara Jackson to show children the bugs and various creepy-crawlies on a short nature hike around the reservoir. This gave the children a unique opportunity to see the shy spectacled caiman that are visible in the shallows and basking on the muddy banks of the reservoir and to learn about where the water in their taps comes from!

The second set of field trips involved the children coming to see where we worked and what we were doing there. Children were split into two groups; with one half coming to Simla and the other going to Asa Wright to look at hummingbirds at feeders on the observation verandah, the groups changed over at lunch time so the children got to see both sites. At Simla, Stewart White and his birders showed the children birds being caught in nets, giving them the opportunity to see birds up close. The frog group showed the children a selection of the tree frogs found in Trinidad, explaining life histories and allowing them the chance to handle some of the larger species; an activity which caused some excitement and consternation as many of the children believed folklore about frogs being poisonous, and were very sceptical about us telling them it was okay to touch them! In this way, both kids and Expedition members got a lot out of the collaboration as the kids were incredibly enthusiastic and interested and asked a lot of searching questions.

We were given the opportunity to attend the end of camp performance at Couva where the children, in their environmentally assigned groups of flora and fauna, gave a sung and spoken recital of the Biodiversity alphabet, written by a local poet. It was a fantastic show, we enjoyed it immensely and were very impressed at the children’s performances, which were incredibly adept. The show ended with the camp leaders singing self-composed calypsos, which had kids and the audience alike in tears of laughter, as did the humorous sketch ‘The Statue’, performed by the camp counsellors.

V: Field trips

Nariva Swamp

The Nariva swamp is the largest freshwater swamp in Trinidad. It has been run as a Nature Reserve by the Governmental Wildlife Division for several decades. Unique vegetation conditions, particularly with respect to the trees such as the Moriche Palm make the reserve particularly important for the islands three species of macaw parrot, including the recently reintroduced blue-and-gold macaw. Under the expert guidance of David Boodoo and ‘Shortman’ students were taken into the heart of the swamp, walking across a land bridge into ‘Bush-Bush Island’, important as a refuge for several troops of red howler monkeys and the weeping capuchin (right). We were also fortunate to see a number of more elusive creatures such as the tree porcupine, which is only found in any significant numbers at this site in Trinidad. In his position as head of the Manatee Conservation Trust*, David took us into a less accessible area of the swamp where many of the students were thrilled to see the West Indian Manatee in its natural surroundings from a treetop platform in the centre of the swamp. The trip was fascinating and the students certainly welcomed the opportunity to see animals that weren’t birds, frogs or turtles for a whole day!

Caroni Swamp

We spent another day at a swamp whilst in Trinidad, albeit a very different kind of swamp. Caroni Swamp, a brackish mangrove swamp on the Eastern Coast, is one of the world’s most famous breeding sites of the spectacular Scarlet Ibis, Trinidad’s national bird. With Wilson Nanan**, whose family has exclusive access to a number of waterways around the swamp and who has been working at Caroni since the nineteen-seventies, we went to the centre of the swamp where we watched the Scarlet Ibis come in to roost. Students benefited from the knowledgeable commentary about the ecology of mangrove swamps and we also saw a number of creatures exclusive to the mangrove swamps, such as Tree Boas and Mangrove crabs, notable for their forward locomotion as they climb through the aerial root systems of their host mangroves.

Steel pan Scots!

We were lucky to be able to go to a number of Steel Pan Festivals in the time we were in Trinidad, visiting the Sforzata Pan Yards in Tunapuna, the Emancipation Day Festival at Laventille and the Arima Centenary Celebrations. Recognising our enthusiasm, a friend whom we met in our local rum shop brought his pan and a number of friends to teach us some rudimentaries, in return we added ‘Scotland the Brave’ to their extensive repertoire. Many of these nights were also a great excuse to eat huge amounts of roti and to partake of possibly unhealthy amounts of Fernandes Black, Trinidad’s finest rum! Thanks to everyone ‘Down the Valley’ for all their nights of hospitality.

VI: Ecotourism Community Survey 2001, Matura

You have been selected as part of a survey on the Matura Beach Ecotourism Project. This study is being conducted by students from The University of Glasgow, Scotland, to find out how the Matura Community feels about tourism development.

Thank you for your cooperation in advance.

Respondent Details

Age: under 20__ 21-30__ 31-40__ 41-50__ 51-60__ over 60__

Sex: male__ female__

Place of residence: In Matura___ Near Matura (within 2 miles)____ Over 2miles from Matura____

Occupation:

Place of occupation: In Matura___ Near Matura (within 2 miles)____ Over 2miles from Matura____

Annual Income bracket: under $10,000____ $10,000- 20,000_____ $21,000-30,000____ $31,000-40,000_____ over $41,000 _____

Number of persons dependent on your income: ____

Questionnaire

1. How do you feel about Matura becoming a tourist attraction? __________________________

___________________________________________________________________________

2. What fears or reservations, if any, do you have about Matura becoming a tourist attraction? ___________________________________________________________________________

___________________________________________________________________________

3. What benefits do you feel tourism can bring to Matura? ______________________________________________________________________________________________________________________________________________________

4. What new facilities do you think tourists would want or need? ___________________________________________________________________________

___________________________________________________________________________

5. What attractions do you feel Matura and the surrounding area have to offer tourists?

___________________________________________________________________________

___________________________________________________________________________

6. Would you be interested in helping to make Matura more attractive to tourists?

Yes ________ no_________

7. If yes, how? ___________________________________________________________________________

8. Are you interested in providing any of the following? Please tick below;

Local food to tourists? ____________________________________________________________

Accommodation to tourists? _______________________________________________________

Transportation for tourists? ________________________________________________________

Making/ selling local crafts?______ what? ___________________________________________

Tour guiding?___________________________________________________________________

Working in a wildlife park?________________________________________________________

Other__________________________________________________________________________

9. What resources do you have to make this happen?___________________________________ ___________________________________________________________________________

10. What skills do you have to offer the community?____________________________________

___________________________________________________________________________

11. Do you have any fund raising ideas that may benefit the community?____________________

___________________________________________________________________________

12. Are you aware of the turtle ecotourism project at Matura run by Nature Seekers?

Yes ________ no_________

13. If yes, what do you think of this project? ___________________________________________________________________________

14. How prominent are Nature Seekers in the local community?___________________________

15. Have you or your family benefited as a result of the establishment of the turtle ecotourism project? Yes _______ how? ____________________________________________________

no _______

16. Have you or your family been disadvantaged as a result of the establishment of the turtle ecotourism project? Yes _________ how? _________________________________________

no _________

17. Do you feel proud of the achievement of the sea turtle project at Matura?

Yes ________ no__________

Why?______________________________________________________________________

18. Are you or any family member interested in becoming involved with Nature Seekers? ___________________________________________________________________________

Cheers for all your Help!

---

* N.B. Work carried out under a postgraduate grant from Glasgow University – this is a short report only: full report available by contacting J. Ward, c/o Graham Kerr Building, Glasgow University, Glasgow G12 8QQ.

* Sample only of some of the statistical analyses carried out on collected data.

* Only a summary of results will be discussed here: further information is available from J.Smith, Graham Kerr Building, Glasgow University, Glasgow G12 8QQ, Scotland, U.K. e-mail: frojo2525@hotmail.com

* See Appendix 4 (Page 107) for details.

* Contact through Wildlife and Forestry at St.Joseph ( 868-668-3133

* Nanan’s Tours 868 645 1305