Introduction 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 Along the Labi Road in the Tiny Sultanate of Brunei Lies a Series of Sandy 13 14 Stream Beds

Cambridge University Press 978-0-521-12000-5 - Food Webs and Container Habitats: The Natural History and Ecology of Phytotelmata R. L. Kitching Excerpt More information

1 1 1 2 2 3 Introduction 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 Along the Labi Road in the tiny Sultanate of Brunei lies a series of sandy 13 14 stream beds. Each is raised above the surrounding forest, presumably by eons 14 15 of silt deposition during occasional periods of inundation and stream flow. 15 16 But now, in May, the water flow is a mere trickle, winding around patches 16 17 of tar seepage; here, not the product of human error and environmental insen- 17 18 sitivity, but a natural phenomenon reflecting the oil-bearing strata that under- 18 19 lie this region of northern Borneo. But this is the perhumid tropics and wet- 19 20 ness is the order of the day – any day. Treacherous quicksands lie centimetres 20 21 below the scorched white sand and the peat-swamp forest on each side of the 21 22 stream bed is permanently inundated with tea-coloured peaty water standing 22 23 thigh-deep around the tangle of buttress roots, fallen trees, scrambling vines 23 24 and creepers. 24 25 I first came to this nutritionally poor but biologically rich ecosystem in 25 26 1989. During the day the biological riches, at least of the more obvious kinds, 26 27 are largely to be inferred rather than experienced directly. The clean smooth 27 28 sand is criss-crossed with tracks of mammal and bird, reptile and insect: here 28 29 the measured marks of a monitor lizard scavenging for carrion, eggs and 29 30 nestlings; there, the dainty steps of forest rats. A honking flight of bushy- 30 31 crested hornbills sweeping overhead and the distant exuberance of a gibbon 31 32 troop call me back to nature immediate rather than nature previous, and 32 33 direct my attention to the smaller scale concerns of the ecologist and the 33 34 mundane necessities of field data collection. 34 35 For Odoardo Beccari, the egocentric Italian naturalist who visited this 35 36 region from 1866 to 1868, this was the land of the ‘maias’ – the orang utans: 36 37 he describes graphically the shooting, dissecting and killing of any number 37 38 of the apes, from infants to magnificent old males (Beccari 1904). Alfred 38 39 Russel Wallace, visiting Borneo in the 1850s in pursuit of beetles, butterflies 39 40 and birds, noted the faunal contrasts between the Bornean animals and those 40

© in this web service Cambridge University Press www.cambridge.org Cambridge University Press 978-0-521-12000-5 - Food Webs and Container Habitats: The Natural History and Ecology of Phytotelmata R. L. Kitching Excerpt More information

2 Introduction 1 of the islands to the east, in particular that of the Celebes (Wallace 1869). 1 2 The mass of natural historical information accrued by Wallace gelled in a 2 3 burst of inspiration that came to him, immured by rain, and light-headed with 3 4 malaria, in a Ternate bungalow. The end result of this concatenation of ideas 4 5 and observations was Darwin and Wallace’s joint presentation to the Lin- 5 6 naean Society of London on July the 1st 1858: On the Tendency of Varieties 6 7 to Depart Indefinitely from the Original Type. The theory of evolution by nat- 7 8 ural selection had finally seen the light of day and biology would never be 8 9 the same again! 9 10 For these and many other less tangible reasons, Borneo, for me, was yet 10 11 another of those ecological destinations that had seemed impossibly distant 11 12 during my English incubation as a biologist. But by 1989 Borneo had become 12 13 in my mind the land of pitcher plants, and the opportunity to follow up this 13 14 impression with my graduate student Charles Clarke, an indomitable pitcher- 14 15 plant enthusiast and field worker, had been just too much to resist. 15 16 We walked gingerly through the quicksands of the Labi stream beds: cas- 16 17 cades of the delicate pitchers of Nepenthes gracilis hanging from their scram- 17 18 bling stems over exposed shrubs, or forming perfect reflections in the mir- 18 19 ror-like surfaces of the pitch pools. In the most exposed areas the larger and 19 20 more robust red and green pitchers of N. mirabilis marked out the parent 20 21 plants, free standing in the stream beds themselves. Scrambling among the 21 22 branches of the larger trees at the edge of the stream beds the fleshy green 22 23 leaves of N. rafflesiana terminated in dramatic trumpet-shaped upper pitch- 23 24 ers hanging like red and yellow decorations on an equatorial Christmas tree. 24 25 In the peat swamp itself, we waded through the opaque water. Here the squat 25 26 pitchers of N. ampullaria hung in tight wreaths around the stems of their host 26 27 plants. Last, but by no means least, we stood engrossed by the extraordinary 27 28 giant pitchers of Nepenthes bicalcarata, which diverted even Beccari from his 28 29 pursuit of the inedible, by its extraordinary combination of features as both 29 30 an insectivorous pitcher plant, and a mutualistic ant plant. 30 31 31 32 I shall return to the details of our pitcher-plant studies in due course: let them 32 33 now stand as an introduction to all of the so-called phytotelmata – plant-con- 33 34 tainer habitats. Pitchers, of course, act as traps for insects and other arthro- 34 35 pods which drown in the fluid contained in the pitchers. They are digested 35 36 by plant exudates, and the nitrogen-rich nutrients so produced are absorbed 36 37 by the plant – enabling them, inter alia, to live in the nutrient-low substrates 37 38 of the kerangas forest. But these small perched water bodies offer themselves 38 39 as habitats for other species of animals which exploit the isolation and pro- 39 40 tection offered by the pitchers and cream off some of the nutrient base for 40

Phytotelmata – defined and described 3 1 themselves. A complete food web exists within each of the pitchers and the 1 2 species of organisms which make up the pitcher fauna are largely restricted 2 3 to this very special environment, breeding nowhere else. 3 4 4 5 5 6 Phytotelmata – defined and described 6 7 Pitchers are by no means the only aquatic habitats that occur as plant-based 7 8 containers and, in 1928, the German biologist Ludwig Varga coined the term 8 9 phytotelmata for the whole class of such ecological situations. Varga carried 9 10 out extensive work on the flora and fauna associated with the water-filled leaf 10 11 axils of the European teasel Dipsacus silvestris and, in order to put his work 11 12 into literary context, he drew attention to earlier work on comparable habi- 12 13 tats. Varga’s work confirmed and established a vigorous interest in the biol- 13 14 ogy of phytotelmata by European, particularly German, ecologists which con- 14 15 tinued from the very early days of organised limnology and ecology through 15 16 the 1950s to the present day. 16 17 Varga’s interest in container habitats appears to have been stimulated by 17 18 the work of Müller (1879 et seq.), Picado (1912, 1913) and van Oye (1921, 18 19 1923). These earlier writers focused on the ‘tanks’ formed by the overlap- 19 20 ping leaf-bases of bromeliads which retain water and provide a habitat for a 20 21 wide range of freshwater and terrestrial organisms. Varga, writing in 1928, 21 22 was also aware of the existence of an aquatic fauna within the pitchers of the 22 23 Old World genus Nepenthes and the American Sarracenia and he quotes the 23 24 works of Sarasin & Sarasin (1905), Jensen (1910), Günther (1913) and van 24 25 Oye (1921) as his sources. His designation ‘phytotelma’ also included water- 25 26 filled tree holes in branch axils and stumps of the European beech tree Fagus 26 27 sylvatica as well as the water bodies collected in a variety of leaf axils such 27 28 as those he himself studied in the teasel. Russian work by Alpatoff (1922) 28 29 on the water bodies collected within the inflated sheathing petioles around 29 30 the inflorescences of Angelica sylvestris had also come to his notice. Varga 30 31 was not the first author to attempt to draw general attention to this class of 31 32 habitats but the earlier designations of Müller (1879) (‘Miniatür-Gewässern’), 32 33 Brehm (1925) (‘Hängende Aquarien’) and Alpatoff (1922) (‘Mikrogewäss- 33 34 ern’) were all superseded by Varga’s ‘phytotelma’. 34 35 Of course the mere bestowal of a name on an object does not indicate its 35 36 first discovery and Frank and Lounibos (1983) have followed long tradition 36 37 by tracing the earliest record of insects originating from water-filled plant 37 38 containers to a classical Chinese source. Ch’en Ts’ang-ch’i, writing during 38 39 the T’ang dynasty sometime between 618 and 905 ad, is quoted by them as 39 40 follows: 40

4 Introduction

1 Beyond the Great Wall there is a wen mu t’sai (mosquito-producing plant), in the 1 2 leaves there are living insects which change into mosquitoes. 2 3 (Pen T’sao Shih-yi) 3 4 4 5 Nevertheless Varga (1928) brought this class of habitats into the ken of sci- 5 6 entists and, in particular, to the attention of Albrecht Thienemann. Thiene- 6 7 mann is deservedly regarded as one of the great organising forces in ecology 7 8 and limnology in Europe from the early part of the twentieth century until 8 9 the 1950s. As director of the Max-Planck Institüt in Plön he was influential 9 10 in directing a large part of German aquatic science throughout much of this 10 11 period. Thienemann organised a major German research effort in the East 11 12 Indies, the Deutsche Limnologische Sunda-expedition, the Proceedings of 12 13 which appeared as supplements to the Archiv für Hydrobiologie. It was here 13 14 that Thienemann drew attention to the fauna of Nepenthes pitchers in 1932. 14 15 This work was followed by an equally influential general treatment in 1934, 15 16 Die Tierwelt der tropischen Planzengewässer, which was encyclopaedic in 16 17 nature and was the basis for the extended treatment he gave the subject in 17 18 his massive work, Chironomus, published in 1954. 18 19 Phytotelmata occur in one form or another in a very wide range of ecosys- 19 20 tems from subarctic bogs (Sarracenia pitchers) to anthropogenic road verges 20 21 (Dipsacus axils, Angelica axils, etc.) to deciduous woodlands (tree holes). 21 22 Container habitats occur on some of the most inhospitable and inaccessible 22 23 of locations on the face of the Earth. Heliamphora pitchers are commonly 23 24 found only on the remote Venezuelan outcrops known as tepuyos (see George 24 25 1989). The tiny Australian native pitcher plant Cephalotus follicularis shares 25 26 with other genera of pitchers its predilection for nutrient-poor swamplands. 26 27 But it is in tropical rainforests that phytotelmata display their full range and 27 28 ubiquity. To return, by way of example, to the forests and creek beds of the 28 29 Labi region of Brunei, I found, in addition to six species of Nepenthes pitcher 29 30 plants, water-filled tree holes, water-filled bamboo internodes, a wide range 30 31 of water bodies in the leaf axils of fleshy plants and the bract axils of inflo- 31 32 rescences, together with pools of water in fallen palm leaves, horizontal logs 32 33 and animal-damaged woody fruits such as coconuts. This range, varying 33 34 slightly with location, can be found in tropical rainforests around the world. 34 35 35 36 36 37 Phytotelmata within modern ecology 37 38 There is little doubt that the early interest of naturalists in these habitats was 38 39 fuelled by the fascination with the unusual that drives most natural histori- 39 40 ans. The range of plants which host phytotelmata, the range of animals which 40

Phytotelmata within modern ecology 5 1 occur within them, and the sometimes curious and always interesting eco- 1 2 logical interactions and other processes which occur within them form a legit- 2 3 imate base for their continued study. 3 4 But with the extension of community ecology from the descriptive to the 4 5 predictive stage that has occurred in recent years, phytotelm studies have 5 6 come to play a much more central role (Pimm 1982, Young 1997). The mul- 6 7 titude of ideas on how natural communities have developed through evolu- 7 8 tionary time, how they change through ecological time, and how and why 8 9 they vary geographically call upon theoretical concepts relating to competi- 9 10 tion, predation and dispersal, as well as more holistic ideas which address 10 11 structural limitations inherent (or otherwise) in the complex objects we know 11 12 as food webs. It turns out that the communities which occur in phytotelmata 12 13 are pre-eminent and predisposed to provide field tests of many of these ideas. 13 14 The reasons for this are fourfold: 14 15 15 • Phytotelm communities are aquatic but are located within terrestrial or 16 16 semi-terrestrial ecosystems such as forests, woodlands, or swamps. 17 17 Accordingly the aquatic organisms which occur within them encounter 18 18 distinct edges which impose upon the community a discreteness not read- 19 19 ily found in other more complex ecosystems. 20 20 • The communities of metazoan animals which occur in phytotelmata are 21 21 relatively simple with the numbers of species ranging from one or two 22 22 up to twenty or thirty. This relative simplicity allows the feeding links 23 23 and hence the food web within them to be defined readily. This is in 24 24 marked contrast to larger and more diffuse habitats such as lake beds or 25 25 forest canopies in which species richness may be measured in hundreds 26 26 of species and the number of potential feeding links is much larger. 27 27 • Most phytotelmata occur within larger ecosystems as a series of units 28 28 scattered spatially within the forest, woodland, road verge or swamp. 29 29 They are, to all intents and purposes, replicated across the ecological 30 30 landscape. Of course, this is not a feature unique to phytotelmata and is 31 31 comparable to the discrete habitat units provided by dung pats, macro- 32 32 fungus fruiting bodies, fallen logs and so forth. 33 33 • The replicated, faunistically simple and (usually) accessible units pro- 34 34 vided by phytotelmata together with their structural simplicity allows 35 35 them to be manipulated and/or imitated in the field so that experiments 36 36 can be designed and carried out on manageable scales of space and time. 37 37 38 Drawing upon these features and the rich base of natural historical informa- 38 39 tion available to us, a number of workers including myself have been using 39 40 phytotelm communities to test ideas within community ecology. 40

6 Introduction 1 This book, then, is devoted, first, to describing the natural history and basic 1 2 biology of phytotelm communities and the excitement this engenders. Build- 2 3 ing on this I go on to describe recent and on-going work which uses these 3 4 communities to test a range of ideas within community ecology. Most of the 4 5 investigations have generated at least as many questions as they have 5 6 answered. But this is not unusual. 6 7 7 8 8 9 The information base 9 10 Since the early 1960s my students and I have maintained an interest in phy- 10 11 totelm ecology. This received a major boost in the early eighties when Howard 11 12 Frank, Phil Lounibos and Durland Fish called a Symposium which was held 12 13 during the 16th International Congress of Entomology held in Kyoto in 1980 13 14 (Frank & Lounibos 1983). During this time I renewed contact with Stuart Pimm 14 15 who added a critical theoretical dimension to my thinking and field work. This 15 16 has resulted in a number of collaborative efforts and has underpinned the field 16 17 research of two of my most recent students, Bertram Jenkins and Charles Clarke. 17 18 The observations that are used repeatedly to illustrate points within this 18 19 book result from our field work in various parts of the world. In general the 19 20 results have been published in the primary literature and details of study sites 20 21 and circumstances are to be found in these works. These field investigations 21 22 were carried out on different occasions during a period of nearly thirty years. 22 23 Naturally my experience with and appreciation of phytotelmata and what I 23 24 now see as the key questions to be addressed grew during this period. I sum- 24 25 marise here the circumstances surrounding crucial phases in this development 25 26 so that particular results can be placed in context. 26 27 27 28 28 29 Tree holes in Wytham Woods 29 30 During the period 1966 to 1969 I studied a set of tree holes located within 30 31 beech trees in deciduous woodland at Wytham Woods near Oxford in the 31 32 United Kingdom. This work formed the basis of a doctoral dissertation, an 32 33 overview of the results of which is provided by Kitching (1971). Details of 33 34 population fluctuations in two of the commonest species of Diptera from these 34 35 sites are presented in Kitching (1972a,b). 35 36 36 37 37 38 Tree holes in Lamington National Park 38 39 In 1978 I began examination of the community found within water-filled tree 39 40 holes in the subtropical rainforest of Lamington National Park in south-east 40

The information base 7 1 Queensland. This has remained a research site from that time but was stud- 1 2 ied intensively from 1978 to about 1984. Student collaborators at the time 2 3 included Catherine Callaghan, Rosemary Lott, Bertram Jenkins and Stewart 3 4 Jackson. Stuart Pimm also participated in elements of this work. The basic 4 5 food web involved is described by Kitching & Callaghan (1982) and com- 5 6 parative analyses presented in Kitching (1983), Kitching & Pimm (1985) and 6 7 Kitching & Beaver (1990). The process of community reassembly following 7 8 complete disruption is described by Jenkins & Kitching (1990). Basic work 8 9 on spatial and temporal variability in food-web structure was also based on 9 10 Lamington data and is described in Kitching (1987a) and Kitching & Beaver 10 11 (1990). Results of our earliest manipulative experiments, carried out here, are 11 12 in Pimm & Kitching (1987). 12 13 13 14 14 15 Phytotelmata in Sulawesi 15 16 In January and February 1985 I participated in the Royal Entomological Soci- 16 17 ety’s centenary expedition (‘Project Wallace’) to northern Sulawesi, Indone- 17 18 sia. Field work on tree holes and bamboo containers was carried out in the 18 19 tropical rainforest of the Dumoga-Bone National Park and work on the fauna 19 20 of Nepenthes maxima at nearby Lake Mooat. Basic descriptions of the food 20 21 webs encountered are in Kitching (1987b), a specific account of a tree-hole 21 22 dragonfly that was encountered is Kitching (1986a), and Kitching & Schofield 22 23 (1986) provide a semi-popular account of the pitcher-plant work. 23 24 24 25 25 26 Tree holes in New England 26 27 The northern Tablelands of New South Wales provide ready access to a vari- 27 28 ety of rainforest ecosystems and during the period 1986 to 1991 we studied 28 29 water-filled tree holes in subtropical and, so-called, warm temperate rainfor- 29 30 est in Dorrigo National Park, and cool temperate Nothofagus-dominated rain- 30 31 forest at New England and Werrikimbe National Parks. The basic results are 31 32 incorporated in the comparative analyses presented later in this work. Exper- 32 33 imental manipulations of energy supply patterns to analogues of these tree 33 34 holes were made by Bertram Jenkins and are described in his thesis (Jenkins 34 35 1991) and in Jenkins et al. (1992). 35 36 36 37 37 38 Phytotelmata in the northern New Guinea 38 39 In January and February 1988 I received a research fellowship to work at the 39 40 Christensen Institute just north of Madang in New Guinea. I was able to iden- 40

8 Introduction 1 tify food webs from water-filled tree holes, bamboo containers and the bract 1 2 axils of the zingiber, Curcuma australasica. These are described in Kitching 2 3 (1990). Further results from this work appear for the first time in this volume. 3 4 4 5 5 6 Tree holes in northern Florida 6 7 In mid-1988, under the aegis of the US/Australia Cooperative Science 7 8 Scheme, I spent an extended period working in northern Florida. Studies of 8 9 the food web occurring in tree holes on the Tall Timbers Reserve of the Uni- 9 10 versity of Florida, Tallahassee, intended to build on the earlier work of Brad- 10 11 shaw and his colleagues, were only partly successful due to the extended 11 12 drought being experienced at the time! Some of these results are presented 12 13 here. 13 14 14 15 15 16 Phytotelmata in the Daintree 16 17 In January and February of both 1989 and 1990 the University of New Eng- 17 18 land mounted expeditions to the Daintree region of far north Queensland 18 19 based at Cape Tribulation. During this period Jenkins and I carried out the 19 20 most extensive study of tree-hole communities that I am aware of, sampling 20 21 and analysing the contents of over eighty sites. In addition we examined the 21 22 water bodies in the leaf axils of an undescribed species of Freycinetia (Pan- 22 23 danaceae) and the bract axils of the inflorescences of the so-called 23 24 ‘backscratcher ginger’, Tapeinocheilus ananassae. A few observations of ani- 24 25 mals occurring in fallen rat-opened coconuts were made, along with obser- 25 26 vations on the fauna of water-filled leaf bases of fallen palm fronds. Some 26 27 of these results are presented in this volume for the first time. 27 28 28 29 29 30 Nepenthes pitcher plants in northern Borneo 30 31 As indicated in the opening paragraphs of this chapter, in May and June 1989, 31 32 I assisted Charles Clarke in his studies of Nepenthes pitcher plants in north- 32 33 ern Borneo. The results of these studies are presented in Clarke & Kitching 33 34 (1993). A full account of the work comprises Clarke’s doctoral dissertation 34 35 (1992). 35 36 36 37 37 38 Tree holes in tropical Queensland 38 39 In the wet season of 1993, in preparation for writing this book, we under- 39 40 took studies of the faunas of water-filled tree holes at a number of rainforest 40

Structure and content 9 1 sites in Queensland selected to fill gaps in our knowledge, hitherto restricted 1 2 to sites in the subtropical extreme south-east, and the lowland tropical North. 2 3 Additional studies were carried out in higher altitude forests at Eungella 3 4 National Park, on the Atherton Tablelands, and at Paluma; more southerly 4 5 tropical lowland forests at Bellenden Ker National Park and in the subtropi- 5 6 cal forests of the Conondale Ranges north of Brisbane. This work was done 6 7 with the assistance of Beverley Kitching, Charles Clarke, Peter Juniper and 7 8 Heather Mitchell and the results appear for the first time in the present work. 8 9 9 10 10 11 Other studies 11 12 Of course the extensive work of other writers also forms a vital part of the 12 13 information on which the present analyses are based. In particular the work 13 14 of the following researchers form an indispensable part of the foundation for 14 15 this work: Beaver on Nepenthes pitcher plants in West Malaysia; Bradshaw, 15 16 Frank, Lounibos and many others on phytotelm mosquitoes in North, Cen- 16 17 tral and South America; Seifert on the axil waters of Heliconia species in 17 18 Central America; Kurihara, Mogi and other Japanese colleagues on phytotelm 18 19 mosquitoes in Japan and south-east Asia; and, Machado-Alison and his stu- 19 20 dents on Venezuelan phytotelms. 20 21 21 22 22 23 Structure and content 23 24 This work is designed to be encountered both as a whole and in parts. Accord- 24 25 ingly, for the student of phytotelmata, I hope the traditional beginning-to-end 25 26 read will be educational, profitable and thought-provoking. But, in addition, 26 27 each chapter presents an essay within areas of natural history and ecology 27 28 and those with more specialised interests will find value in reading them in 28 29 at least partial isolation from the remainder of the work. 29 30 The basic structure and intended logic of the book are summarised in Fig- 30 31 ure 1.1. I have grouped the chapters into five parts after this Introduction. I 31 32 have prefaced each of these parts with a word picture of some of the field 32 33 experiences involved, deliberately to try to recreate some of the excitement 33 34 of the naturalist, to offset in part the objectivity of the ecologist. 34 35 Part I contains three substantial ‘background’ chapters describing the 35 36 plants, animals and physico-chemical environment of phytotelmata in gen- 36 37 eral. These provide information on the underlying biology, natural history 37 38 and environmental science which are the vital context of later discussions on 38 39 the ecology of phytotelm communities. An extended annexe at the end of the 39 40 book is a family-by-family treatment of the phytotelm fauna. 40

10 Introduction 1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20 21 21 22 22 23 23 24 24 25 25 26 26 27 27 28 28 29 29 30 30 31 31 32 Fig. 1.1. The structure and logic of the book. 32 33 33 34 34 35 Part II of the work contains two important chapters. The ﬁrst introduces 35 36 the idea of the food web by presenting examples from each phytotelm type. 36 37 The chapter deﬁnes key formalisms concerning the food-web statistics which 37 38 will be used without further explanation in later chapters. The second chap- 38 39 ter (Chapter 6) in this section reviews segments of the theory of food webs 39 40 from the recent literature. In addition the chapter explores the biological mech- 40