Biological studies on two European termite species: establishment risk in the UK
Laetitia Virginie Laine B.Sc. M.Sc. D.I.C.
A thesis submitted for the degree of Doctor of Philosophy of the University of London
November 2002
Department of Biological Sciences, Imperial College, Silwood Park, Ascot, SL5 7PY, Berkshire Abstract
The discovery of an accidental introduction of termites into Devon in 1994 generated great interest as termites were previously thought to be unable to establish in the UK due to unfavourable climatic conditions. Information about the species present in Devon, Reticulitermes grassei, was found to be lacking and the present study was undertaken to determine the importance of various abiotic and biotic factors in establishment of this species.
The factors included in the study were the minimum termite number for establishment, the consumption of wood and its effect on survival and temperature and soil type. A review of the literature was also conducted, detailing the problems with the taxonomy of this termite genus, their present distribution pattern and the life cycle of Reticulitermes species.
Two populations of both R. grassei and R. santonensis were studied. The effect of the minimum termite number was found to be significant in both laboratory and field conditions. However, survival decreased in the laboratory and increased in the field with increased number of termites. Consumption experiments were performed using blocks of Scots pine, beech and oak. In most cases termite populations were found to consume and survive best on oak. Consumption was also tested on live seedlings but these results were inconclusive. Survival was observed to increase with increased temperature. Soil type experiments provided very mixed results for termite survival and these did not seem to be linked to soil moisture content. The trigger for alate production was not determined but fluctuating temperature seemed to have a positive effect on fecundity.
The two most important factors for termite establishment appear to be the actual termite number and temperature. The final chapter looks at climate change and the various factors involved in termite establishment in order to put the results of this study into context.
2 Acknowledgements
There are many people that have helped me during this PhD project. I would firstly like to thank my supervisor Denis Wright for all his help and advice, The Forestry Commission who funded this project and Robert Verkerk who initiated the project. Simon Leather, Hugh Evans and Geoff LePatourel gave me much appreciated guidance during this research project. Thanks to Janet Phipps, for always knowing Denis' whereabouts (or not!). I am also indebted to Mick Crawley for his help (and patience) as I waded through the various problems I encountered with the interminable 'S TAT S'.
I would like to thank Dave Yeoman (Rothamstead) for providing the soil samples for Chapter 7 and Ann Kelly and Richard Hartley (University of Plymouth) for doing the particle analysis of the soils. I would like to thank the various members of the CTBA who allowed me to use the laboratory in Bordeaux and the field site on L'ile d' Oldron. More specifically, Ivan Paulmier, especially for helping me to lug half a tonne of soil to Bordeaux and back, and Sandra Girardi for keeping an eye on my various experiments. I would like to thank Mr. Chaigneau for allowing me the use of the 'Office National des Forets' facilities and Paul Beasley, for producing various pieces of equipment, especially the field equipment (Chapter 4). Thanks also to Andy Wilby for proof reading part of the thesis. I am also grateful to Florent Vieau for the interesting discussions about termite life cycles and Brian Forschler for my week in Athens (Georgia) and for initiating me to 'house inspections'.
I would like to thank all the people that made Silwood a little sunnier (even on a wet and windy winter's morning). Ek, Anna, Dan and Lou (the inseparable pair of pranksters!)... as well as Paul B., Kate, Rob B., Aunty Anne, Pauline and Christine.
I would like to thank my family for all their support and love. Finally I would like to thank Fabien, my husband to be, without whom I would probably never had got through it. I thank him for all his support, encouragement, patience (the halo is in the post) and love.
3 And last but not least... thank you to all those termites that gave their lives for my research. This is for you...
The Termite by Ogden Nash
Some primal termite knocked on wood And tasted it, and found it good! And that is why your Cousin May Fell through the parlor floor today.
4 Contents
ABSTRACT 2
ACKNOWLEDGEMENTS 3
CONTENTS 5
LIST OF FIGURES 8
LIST OF TABLES 13
CHAPTER 1 14
INTRODUCTION 1.1 Pest Status 14 1.2 Control tactics 15 1.2.1 Preventative 15 1.2.2 Remedial treatment 16 1.2.3 Conclusion 17 1.3 Project background 17 1.4 Aims and Objectives 18 CHAPTER 2 19 LITERATURE REVIEW 2.1 Introduction 19 2.2 Species studied 19 2.3 Taxonomy 20 2.4 Distribution 23 2.4.1 Genus 23 2.4.2 Reticulitermes grassei 25 2.4.3 Reticulitermes santonensis 26 2.4.4 Reticulitermes flavipes 26 2.5 Movement of termite species in France 27 2.6 Differences between the species studied 29 2.7 The life cycle of Reticulitermes spp. (Isoptera, Rhinotermitidae): what do we know? 32 2.7.1 Abstract 32 2.7.2 Introduction 33 2.7.3 Life history 34 2.7.4 Comparisons of life cycles 48 2.7.5 Conclusion 56 CHAPTER 3 58
GENERAL MATERIALS AND METHODS 3.1 Termite species and strains 58 3.2 Insect culture 58 3.3 Experiments using vermiculite 58 3.1 Wood block quality 59 3.4 Data analysis 59
5 CHAPTER 4 60 MINIMUM NUMBER FOR SUCCESSFUL TERMITE ESTABLISHMENT 4.1 Background 60 4.2 Materials and Methods 61 4.2.1 Survival of termites on Scots pine 61 4.2.2 Survival of termites in the field 62 4.2.3 Survival of termites in different substrate types 65 4.2.4 Survival of termites in different substrate volumes 66 4.3 Results 66 4.3.1 Survival of termites on Scots pine 66 4.3.2 Survival of termites in the field 69 4.3.3 Survival of termites in different substrate types 70 4.3.4 Survival of termites in different substrate volumes 71 4.4 Discussion 72 4.4.1 Survival of termites on Scots pine 72 4.4.2 Survival of termites in the field 74 4.4.3 Survival of termites in different substrate types 74 4.4.4 Survival of termites in different substrate volumes 75 4.5 Conclusion 75 CHAPTER 5 77
STUDIES ON TERMITE CONSUMPTION 5.1 Background 77 5.1.1 Consumption 78 5.2 Materials and Methods 80 5.2.1 Consumption on different wood species - no choice 80 5.2.2 Consumption on different wood species - choice 81 5.2.3 Consumption and survival on Scots pine over time 81 5.2.4 Survival on Corsican pine seedlings 81 5.2.5 Survival on Scots pine seedlings and in combination with wood blocks 82 5.2.6 Survival on Scots pine and cherry seedlings 83 5.3 Results 83 5.3.1 Consumption on different wood species - no choice 83 5.3.2 Consumption on different wood species - choice 85 5.3.3 Consumption and survival on Scots pine over time 87 5.3.4 Survival on Corsican pine seedling 89 5.3.5 Survival on Scots pine seedlings and in combination with wood blocks 90 5.3.6 Survival on Scots pine and cherry seedlings 92 5.4 Discussion 93 5.4.1 Consumption on different wood species - no choice 93 5.4.2 Consumption on different wood species - choice 94 5.4.3 Consumption and survival on Scots pine over time 94 5.4.4 Survival on Corsican pine seedlings 95 5.4.5 Survival on Scots pine seedlings and in combination with wood blocks 95 5.4.6 Survival on Scots pine and cherry seedlings 95 5.4.7 Overview 96 5.5 Conclusion 97 CHAPTER 6 99
REPRODUCTIVE STRATEGIES 6.1 Introduction 99 6.2 Materials and Methods 100 6.2.1 Observation Apparatus 100 6.2.2 Preliminary observations on R. grassei UK and R. santonensis BRE 101
6 6.2.3 Observational study on two populations of R. grassei and R. santonensis.101 6.2.4 Effect of cold treatment on reproductive strategy 102 6.3 Results 103 6.3.1 Preliminary observations on R. grassei UK and R. santonensis BRE 103 6.3.2 Observational study on two populations of R. grassei and R. santonensis104 6.3.3 Effect of cold treatment on reproductive strategy 109 6.4 Discussion 116 6.4.1 Preliminary observations on R. grassei UK and R. santonensis BRE 116 6.4.2 Observational study on two populations of R. grassei and R. santonensis 116 6.4.3 Effect of cold treatment on reproductive strategy 117 6.5 Conclusion 118 CHAPTER 7 120 TEMPERATURE AND SOIL TYPE AND THEIR EFFECT ON TERMITE SURVIVAL 7.1 Background 120 7.2 Materials and Methods 122 7.2.1 Survival of termites on different soil types 122 7.2.2 Survival of termites at different temperatures 124 7.3 Results 125 7.3.1 Survival of termites on different soil types 125 7.3.2 Survival of termites at different temperatures 126 7.4 Discussion 127 7.4.1 Survival of termites on different soil types 127 7.4.2 Survival of termites at different temperatures 128 7.5 Conclusion 129 CHAPTER 8 131
SUMMARY AND GENERAL DISCUSSION 8.1 Summary 131 8.2 Factors affecting termite establishment 132 8.3 Climate and termite distribution 134 8.4 Climate change 137 8.5 Conclusion 138
GLOSSARY 139
REFERENCES 140
APPENDIX 159
7 List of Figures
CHAPTER 2
LITERATURE REVIEW
FIGURE 2.1 WORLD-WIDE DISTRIBUTION OF RETICULITERMES SPECIES (AFTER PEARCE (1997)). 24
FIGURE 2.2 DISTRIBUTION OF RETICULITERMES SPECIES IN EUROPE (CONSIDERED TO BE NATURAL). CROSSHATCHED AREAS SHOW SYMPATRIC ZONES BETWEEN R. GRASSEI AND R. SANTONENSIS IN THE NORTH AND R. GRASSEI AND R. BANYULENSIS IN THE IBERIAN PENINSULA (AFTER CLEMENT (2001)). 24
FIGURE 2.3 URBAN DISTRIBUTION OF RETICULITERMES SPECIES IN EUROPE (AFTER (CLEMENT, 2001)). 25
FIGURE 2.4 DISTRIBUTION OF RETICULITERMES FLAVIPES (SHADED AREA) IN THE U.S.A. (AFTER NUTTING (1990)) 27
FIGURE 2.5 LIFE CYCLE OF RETICULITERMES SPP. ACCORDING TO F. VIEAU (PERSONAL COMMUNICATION), AMENDED FROM BUCHLI (1958). 36
FIGURE 2.6 LIFE CYCLE OF RETICULITERMES SPP. BASED ON SNYDER (1915) (L - LARVA, LB - LARVA WITH LARGE HEAD, EL- EMERGENCY RESERVE REPRODUCTIVE, N - NYMPH, WL - WORKER LARVA, NL - NYMPH WITH LONG WING BUDS, NS - NYMPH WITH SHORT WING BUDS, W -WORKER, SL - EMERGENCY LARVAL SUBSTITUTE, SN - EMERGENCY NYMPHAL SUBSTITUTE, BN - BRACHYPTEROUS NEOTENIC, PS - PRE- SOLDIER, S - SOLDIER). 49
FIGURE 2.7 LIFE CYCLE OF RETICULITERMES SPP. BASED ON HARE (1934). 50
FIGURE 2.8 LIFE CYCLE OF RETICULITERMES SPP. BASED ON THE DESCRIPTION BY FEYTAUD (1946), TAKEN FROM GRASSI AND SANDIAS (1893) (E - ERGATOID, NN - NYMPH WITH NO WING BUDS) 51
FIGURE 2.9 LIFE CYCLE OF RETICULITERMES SPP. BASED ON THE DESCRIPTION BY GRASSI (1893) AND LATER BY GRASSE (1949) (LS - LARVA WITH SMALL HEAD, NE - NEOTENIC LARVA). 52
FIGURE 2.10 ONTOGENIC POTENTIAL OF RETICULITERMES SPP. BASED ON THE DESCRIPTION BY BUCHLI (1958) (P - PSEUDERGATE). 53
FIGURE 2.11 LIFE CYCLE OF RETICULITERMES SPP. BASED ON THE DESCRIPTION BY VIEAU (1991, 1994A) PERSONAL COMMUNICATION) 54
FIGURE 2.12 LIFE CYCLE OF RETICULITERMES SPP. BASED ON THE DESCRIPTION BY NOIROT (1985). 55
8 CHAPTER 4 MINIMUM NUMBER FOR SUCCESSFUL TERMITE ESTABLISHMENT
FIGURE 4.1 DETAIL OF STEPS TO CONSTRUCT FIELD EQUIPMENT (SEE TEXT FOR DESCRIPTION). (A) - 50 L BUCKET (B) - BUCKET INVERTED, CIRCLE REMOVED FROM LID (34.6 CM), BASE REMOVED (CUT MADE AT 3.7 CM FROM BASE) AND CIRCLE (30.6 CM) CUT FROM BASE (C) BASE INVERTED AND PLACED BACK IN BUCKET, NYLON MESH PLACED ON TOP OF BUCKET AND LID PUSHED OVER MESH 63
FIGURE 4.2 DETAIL OF TOP OF CONTAINER SHOWING CROSS SECTION THROUGH SIDE OF CONTAINER AND 'LIP' TO PROVIDE A PHYSICAL BARRIER TO PREVENT TERMITE ESCAPE. 63
FIGURE 4.3 LAYOUT OF PART OF FIELD SITE ON L'ILE D'OLERON (SEE APPENDIX, FIGURE A.1 FOR COMPARISON ON TERMINATION OF EXPERIMENT) 64
FIGURE 4.4 EXPERIMENTAL DESIGN LAYOUT OF FIELD TRIAL USING DIFFERENT INITIAL TERMITE NUMBERS OF RETICULITERMES SANTONENSIS FRANCE (R.s. F) AND RETICULITERMES GRASSEI FRANCE (R.G. F). 65
FIGURE 4.5 PROPORTION OF RETICULITERMES SANTONENSIS BRE (R.s. BRE) (0) RETICULITERMES SANTONENSIS FRANCE (R.s. F) (o), RETICULITERMES GRASSEI UK (R.G. UK) (.) AND RETICULITERMES GRASSEI FRANCE (R.G. F) (.) WORKERS / PSEUDERGATES SURVIVING AFTER 12 WEEKS AS A FUNCTION OF INITIAL TERMITE NUMBER. FITTED VALUES WERE BACK TRANSFORMED FROM THE LINEAR MODEL; Y= 0.207 - 0.0018x (R.s. BRE), Y = 0.207 - 0.0018x (R.s. F), Y = 1.132 -.0.0018x (R.G. UK), Y = 0.433 - 0.0018 (R.G. F). WHERE S. E. = 0.04, 8E-5 (R.s. BRE); 0.04, 8E-5 (R.s. F); 0.043, 8E-5 (R.G. UK) AND 0.039, 8E-5 (R.G. F) RESPECTIVELY. 67
FIGURE 4.6 PROPORTION OF RETICULITERMES GRASSEI UK (o) AND RETICULITERMES SANTONENSIS BRE (A) WORKERS / PSEUDERGATES SURVIVING AFTER 12 WEEKS AGAINST THE INITIAL NUMBER OF TERMITES AT T = 0. FITTED VALUES WERE BACK TRANSFORMED FROM THE LINEAR MODEL; Y= -1.161 + 0.0647x. WHERE S.E. = 0.22 (INTERCEPT), 0.011 (SLOPE). 69
FIGURE 4.7 MEAN PROPORTION SURVIVAL (± S. E.) OF RETICULITERMES GRASSEI FRANCE (R.G. F) AND RETICULITERMES SANTONENSIS FRANCE (R.s. F) WORKERS / PSEUDERGATES IN THE FIELD AFTER 2 YEARS AS A FUNCTION OF INITIAL WORKER NUMBER; N = 4. 70
FIGURE 4.8 MEAN PROPORTION SURVIVAL (± S. E.) OF RETICUITERMES SANTONENSIS BRE WORKERS / PSEUDERGATES ON TWO DIFFERENT SUBSTRATES, SAND AND VERMICULITE, FROM DIFFERENT INITIAL NUMBERS OF 125 (STRIPED) AND 250 (DOTTED) INDIVIDUALS AFTER A PERIOD OF 12 WEEKS; N = 6 71
FIGURE 4.9 MEAN PROPORTION SURVIVAL (± S. E.) OF RETICULITERMES SANTONENSIS BRE (R. s . BRE), RETICULITERMES SANTONENSIS FRANCE (Rs. F), RETICULITERMES GRASSEI UK (R. G. UK) AND RETICULITERMES GRASSEI FRANCE (R. G. F) WORKERS / PSEUDERGATES AFTER 12 WEEKS AT LOW (WHITE) AND HIGH (GREY) SUBSTRATE VOLUMES AGAINST INITIAL TERMITE NUMBERS OF 125 AND 250 INDIVIDUALS AT T=0. 72
9 FIGURE 4.10 FLOW CHART SHOWING WHICH FACTORS HAVE AN EFFECT ON TERMITE SURVIVAL AND HOW THEY INTERACT (TAKEN FROM LENZ (1980)). 75
CHAPTER 5
STUDIES ON TERMITE CONSUMPTION
FIGURE 5.1 MEAN PROPORTION SURVIVAL (± S. E.) OF 100 RETICULITERMES GRASSEI UK (R. G. UK), RETICULITERMES GRASSEI FRANCE (R. G. F), RETICULITERMES SANTONENSIS BRE (R.S. BRE) AND RETICULITERMES SANTONENSIS FRANCE (R.S. F) WORKERS / PSEUDERGATES ON BEECH (DOTTED), OAK (GREY) AND PINE (STRIPED) AFTER 8 WEEKS; N = 10 84
FIGURE 5.2 MEAN CONSUMPTION (G) (± S. E.) BY 100 RETICULITERMES GRASSEI UK (R. G. UK), RETICULITERMES GRASSEI FRANCE (R. G. F), RETICULITERMES SANTONENSIS BRE (R. S. BRE) AND RETICULITERMES SANTONENSIS FRANCE (R. S. F) WORKERS / PSEUDERGATES ON BEECH (DOTTED), OAK (GREY) AND PINE (STRIPED) AFTER 8 WEEKS; N = 10 85
FIGURE 5.3 MEAN PROPORTION (± S. E.) OF 125 RETICULITERMES GRASSEI UK (GREY) AND RETICULITERMES SANTONENSIS BRE (WHITE) WORKERS / PSEUDERGATES SURVIVING ON COMBINATIONS OF BEECH, OAK AND PINE AFTER 8 WEEKS; N = 10. 86
FIGURE 5.4 MEAN PROPORTION OF TOTAL WOOD WEIGHT CONSUMED BY 125 RETICULITERMES GRASSEI UK (R.G.) AND RETICULITERMES SANTONENSIS BRE (R.S.) WORKERS / PSEUDERGATES DEPENDING ON COMBINATIONS OF WOOD TYPES AFTER 8 WEEKS. PROPORTION RELATING TO EACH WOOD TYPE SHOWN; BEECH (DOTTED), OAK (GREY) AND PINE (STRIPED); N = 10. 87
FIGURE 5.5 PROPORTION OF RETICULITERMES SANTONENSIS BRE (❑ & DOTTED LINE) RETICULITERMES GRASSEI UK (A & CONTINUOUS LINE) WORKERS / PSEUDERGATES SURVIVING OVER 84 DAYS. FITTED VALUES WERE BACK TRANSFORMED FROM THE LINEAR MODEL; Y = 1.674 - 0.022X (R.S. BRE) AND Y = 2.572 - 0.022x (R.G. UK). WHERE S. E. = 0.05 (INTERCEPT), 9.8E-4 (SLOPE) (R.s. BRE) AND 0.06 (INTERCEPT), 9.8E-4 (SLOPE) (R.G.UK); N = 3. 88
FIGURE 5.6 RELATIVE CONSUMPTION RATE (LOG (FINAL WEIGHT / INITIAL WEIGHT)) OF
RETICULITERMES SANTONENSIS BRE (❑) AND RETICULITERMES GRASSEI UK (A) OVER 84 DAYS. REGRESSION EQUATION: Y = -0.012 -0.0008X. WHERE S. E. = 0.004 (INTERCEPT) AND 0.0001 (SLOPE); N = 3 89
FIGURE 5.7 MEAN PROPORTION SURVIVAL (± S. E.) OF 125 RETICULITERMES SANTONENSIS BRE WORKERS / PSEUDERGATES ON LIVE CORSICAN PINE SEEDLINGS AND SCOTS PINE BLOCKS, AFTER 8 WEEKS; N = 15. 90
FIGURE 5.8 MEAN PROPORTION SURVIVAL (± S. E.) OF RETICULITERMES GRASSEI UK (GREY) AND RETICULITERMES SANTONENSIS BRE (WHITE) WORKERS / PSEUDERGATES ON SCOTS PINE BLOCKS ALONE AND IN COMBINATION WITH LIVE SCOTS PINE SEEDLINGS AFTER 8 WEEKS; N = 6. 91
FIGURE 5.9 WEIGHT LOSS (± S. E.) BY WOOD BLOCKS WITH AND WITHOUT THE PRESENCE OF LIVE SCOTS PINE SEEDLINGS IN THE ABSENCE (DOTTED) OR PRESENCE OF 125 RETICULITERMES GRASSEI UK (GREY) AND RETICULITERMES SANTONENSIS BRE (WHITE) WORKERS / PSEUDERGATES AFTER 8 WEEKS; N = 6. 92
10 FIGURE 5.10 MEAN PROPORTION SURVIVAL (± S. E.) OF 125 RETICULITERMES GRASSEI UK (GREY) AND RETICULITERMESSANTONENSIS BRE (WHITE) WORKERS / PSEUDERGATES ON LIVE CHERRY AND SCOTS PINE SEEDLINGS AND SCOTS PINE WOOD BLOCKS AFTER 8 WEEKS; N = 7. 93
CHAPTER 6
REPRODUCTIVE STRATEGIES
FIGURE 6.1 GLASS PLATE ASSEMBLY USED FOR REPRODUCTIVE STRATEGY EXPERIMENTS. 101
FIGURE 6.2 NUMBER OF EGGS (WHITE), FIRST INSTAR LARVAE (GREY) AND SECOND INSTAR LARVAE (BLACK) OBSERVED OVER 567 DAYS FROM AN INITIAL POPULATION OF 125 WORKERS / PSEUDERGATES OF RETICULITERMES GRASSEI UK. 104
FIGURE 6.3 RANK (4 = HIGHEST) IN TERMS OF THE LEVEL OF ACTIVITY OBSERVED IN RETICULITERMESSANTONENSIS BRE (R.S. BRE) (o), RETICULITERMES SANTONENSIS FRANCE (R.s. F) (❑), RETICULITERMES GRASSEI UK (R.G. UK) (•) AND RETICULITERMES GRASSEI FRANCE (R.G. F) (•) COLONIES IN GLASS PLATES OVER 469 DAYS. 106
FIGURE 6.4 NUMBER OF SQUARES ON WHICH CLUMPING OF RETICULITERMES SANTONENSIS BRE (R.s. BRE) (o), RETICULITERMES SANTONENSIS FRANCE (R.s. F) (❑), RETICULITERMES GRASSEI UK (R.G. UK) (.) AND RETICULITERMES GRASSEI FRANCE (R.G. F) (■) INDIVIDUALS WAS OBSERVED IN GLASS PLATES OVER 469 DAYS. 107
FIGURE 6.5 NUMBER OF EGGS (WHITE), FIRST INSTAR LARVAE (GREY) AND SECOND INSTAR LARVAE (BLACK) OF RETICULITERMES SANTONENSIS BRE PRODUCED OVER 469 DAYS FROM AN INITIAL NUMBER OF 250 WORKERS / PSEUDERGATES, 2 NYMPHS AND 2 SOLDIERS. 108
FIGURE 6.6 NUMBER OF EGGS (WHITE) AND FIRST INSTAR LARVAE (GREY) OF RETICULITERMES GRASSEI UK PRODUCED OVER 469 DAYS FROM AN INITIAL NUMBER OF 250 WORKERS / PSEUDERGATES AND 2 NYMPHS 108
FIGURE 6.7 SCORE (5 = HIGHEST) IN TERMS OF RETICULITERMES GRASSEI UK (+) AND RETICULITERMESSANTONENSIS BRE (o) SURVIVAL IN GLASS PLATES OVER 350 DAYS. 109
FIGURE 6.8 SCORE (3 = HIGHEST) IN TERMS OF LEVEL OF ACTIVITY OF RETICULITERMES GRASSEI UK (+) AND RETICULITERMES SANTONENSIS BRE (0) COLONY IN GLASS PLATES OVER 350 DAYS. 110
FIGURE 6.9 CUMULATIVE CONSUMPTION OF SCOTS PINE SQUARES BY RETICULITERMES GRASSEI UK (+) AND RETICULITERMES SANTONENSIS BRE (0) IN GLASS PLATES OVER 350 DAYS. 111
FIGURE 6.10 MEAN NUMBER OF RETICULITERMES GRASSEI UK EGGS (WHITE) AND LARVAE (GREY) OBSERVED OVER 350 DAYS AT A CONSTANT TEMPERATURE IN GLASS PLATES FROM AN INITIAL POPULATION OF 500 WORKERS / PSEUDERGATES AND 4 BRACHYPTEROUS NEOTENICS; N = 3. 113
11 FIGURE 6.11 MEAN NUMBER RETICULITERMES GRASSEI UK EGGS (WHITE) AND LARVAE (GREY) OBSERVED OVER 350 DAYS AT FLUCTUATING TEMPERATURES IN GLASS PLATES FROM AN INITIAL POPULATION OF 500 WORKERS / PSEUDERGATES AND 4 BRACHYPTEROUS NEOTENICS; N = 3. 114
FIGURE 6.12 MEAN NUMBER RETICULITERMES SANTONENSIS BRE EGGS (WHITE) AND LARVAE (GREY) OBSERVED OVER 350 DAYS AT A CONSTANT TEMPERATURE IN GLASS PLATES 114
FIGURE 6.13 MEAN NUMBER RETICULITERMES SANTONENSIS BRE EGGS (WHITE) AND LARVAE (GREY) OBSERVED OVER 350 DAYS AT FLUCTUATING TEMPERATURES IN GLASS PLATES FROM AN INITIAL POPULATION OF 500 WORKERS / PSEUDERGATES AND 4 BRACHYPTEROUS NEOTENICS; N= 3. 115
CHAPTER 7
TEMPERATURE AND SOIL TYPE AND THEIR EFFECT ON TERMITE SURVIVAL
FIGURE 7.1 MEAN PROPORTION SURVIVAL (± S. E.) OF 125 RETICULITERMES GRASSEI UK (GREY) AND RETICULITERMES SANTONENSIS BRE (WHITE) WORKERS / PSEUDERGATES ON DIFFERENT SOIL TYPES AT 25 ± 2 °C AFTER 12 WEEKS; N = 6 125
FIGURE 7.2 MEAN PROPORTION SURVIVAL (± S. E.) OF 125 RETICULITERMES GRASSEI UK (GREY) AND RETICULITERMES SANTONENSIS BRE (WHITE) WORKERS / PSEUDERGATES AFTER 3 MONTHS AT FOUR DIFFERENT TEMPERATURES; N=8 126
FIGURE 7.3 MEAN PROPORTION SURVIVAL (± S. E.) OF 125 RETICULITERMES GRASSEI UK (GREY) AND RETICULITERMES SANTONENSIS BRE (WHITE) WORKERS / PSEUDERGATES AFTER 6 MONTHS AT FOUR DIFFERENT TEMPERATURES; N = 8. 127
CHAPTER 8
GENERAL DISCUSSION AND CONCLUSIONS
FIGURE 8.1 CONCEPTUAL MODEL OF FACTORS INFLUENCING TERMITES ESTABLISHMENT IN THE UK. 133
FIGURE 8.2 MEAN ANNUAL AIR TEMPERATURE ISOTHERMS FOR WESTERN EUROPE (FROM DATA COLLECTED BETWEEN 1950 AND 2001). IMAGE PROVIDED BY THE NOAA- CIRES CLIMATE DIAGNOSTICS CENTER, BOULDER COLORADO (WWW.CDC.NOAA.GOV) 135
FIGURE 8.3 MEAN WINTER AIR TEMPERATURE ISOTHERMS FROM DATA BETWEEN 1950 AND 2001. IMAGE PROVIDED BY THE NOAA-CIRES CLIMATE DIAGNOSTICS CENTER, BOULDER COLORADO (WWW.CDC.NOAA.GOV) 136
FIGURE 8.4 NUMBER OF COMMUNES IN EACH DEPARTMENT OF FRANCE REPORTED AS BEING INFESTED WITH TERMITES. IMAGE PROVIDED BY THE CENTRE TECHNIQUE DU BOIS ET DE L'AMEUBLEMENT (CTBA) (WWW.TERMITES.COM.FR). 137
12 List of Tables
CHAPTER 2
LITERATURE REVIEW
TABLE 2.1 COMPARISONS BETWEEN RETICULITERMES SANTONENSIS AND RETICULITERMES LUCIFUGUS/ RETICULITERMES GRASSEI. 30
CHAPTER 4
MINIMUM NUMBER FOR SUCCESSFUL TERMITE ESTABLISHMENT
TABLE 4.1 SURVIVAL OF RETICULITERMES SANTONENSIS BRE (R.S. BRE), RETICULITERMES GRASSEI U.K. (R.G. UK), RETICULITERMES SANTONENSIS FRANCE (R.s. F) AND RETICULITERMES GRASSEI FRANCE (R.G. F) WORKERS / PSEUDERGATES AFTER 12 WEEKS AS A FUNCTION OF INITIAL TERMITE NUMBER. 66
TABLE 4.2 SURVIVAL OF RETICULITERMES SANTONENSIS BRE AND RETICULITERMES GRASSEI UK WORKERS / PSEUDERGATES AFTER 12 WEEKS AS A FUNCTION OF SURVIVAL. 68
CHAPTER 5
STUDIES ON TERMITE CONSUMPTION
TABLE 5.1 HARDNESS OF WOOD SPECIES IN NEWTONS (LAVERS, 1983) 80
CHAPTER 6
REPRODUCTIVE STRATEGIES
TABLE 6.1 LEVEL OF ACTIVITY SHOWN BY DIFFERENT STRAINS OF RETICULITERMES SANTONENSIS AND RETICULITERMES GRASSEI AFTER 0, 224 AND 452 DAYS IN THE GLASS PLATE APPARATUS. 105
TABLE 6.2 RANGE, MEDIAN AND GRAND MEAN OF SCORES FOR SURVIVAL, ACTIVITY AND CONSUMPTION FROM AN INITIAL NUMBER OF 500 WORKERS / PSEUDERGATES AND 4 NEOTENICS OF R. SANTONENSIS BRE (R.S. BRE) AND R. GRASSEI UK (R.G. UK) AT CONSTANT AND FLUCTUATING TEMPERATURES IN GLASS PLATES AFTER 350 DAYS. 112
CHAPTER 7
TEMPERATURE AND SOIL TYPE AND THEIR EFFECT ON TERMITE SURVIVAL
TABLE 7.1 NAME, CLASSIFICATION AND SOURCE OF SUBSTRATES USED TO ASSESS TERMITE SURVIVAL IN DIFFERENT SOIL DESCRIPTIONS. 123
TABLE 7.2 SUBSTRATE MOISTURE CONTENTS AND SUBSTRATE WEIGHTS USED TO ASSESS TERMITE SURVIVAL IN DIFFERENT SOIL TYPES. 124
13 Chapter 1
Introduction
This chapter provides a brief overview of the problem posed by termite species of the genus Reticulitermes. Past and current modes of control are summarised, leading onto the reasons behind the initiation and aims of the current PhD project.
1.1 Pest Status Termites, in a natural situation, are useful recyclers of organic matter, it is only when they start feeding on cellulose that is important to man that problems occur and thus that they become pests. The main damage caused is by termites feeding on cellulose material within buildings but damage of other substrates can also be caused by termites when trying to reach potential food sources.
Subterranean termites, of which Reticulitermes are a genus, are extremely important pest species, costing 11 billion dollars per year (if building repair cost is included) in control and damage in the USA alone (Haverty et al., 1991; Forschler & Jenkins, 1999; Culliney & Grace, 2000; Su, 2002). Reticulitermes species live in or in contact with the soil, although this is not thought to be a necessity. Their nests tend not to be defined and are very diffuse, tunnels are formed over impenetrable substances (Snyder, 1926; Grasse, 1949; Weesner, 1965). Reticulitermes colonies are said to be made up of a few hundreds of thousands of individuals (Grasse, 1949; Nutting, 1969), although this seems quite a low estimation as other members of the Rhinotermitidae have been reported to have several million individuals per colony (Dangerfield & Mosugelo, 1997). The colony structures can vary from simple families with alate-derived parents to interconnecting nests with several supplementary reproductives (Thorne et al., 1999). There are six species of Reticulitermes that are commonly found in Europe, R. santonensis, R. lucifugus, R. grassei, R. balkanensis and R. banyulensis and one of which remains unnamed (currently termed R. sp. nov.) (Clement et al., 2001).
The main problem involved in the control of Reticulitermes species is their complex life history, which will be discussed in Chapter 2. They are able to reproduce via
14 Introduction primary reproductives or alate forms as well as secondary reproductives. The latter are of great importance as they enable any group of termites to potentially form a new colony and the initial group does not necessarily need to have reproductives (Castle, 1934; Buchli, 1951). It is, therefore, not sufficient to kill the queen to eliminate the colony (Feytaud, 1911). A further problem is that termite alates are often mistaken for ants and the presence of termites is often confirmed only once the infestation has reached serious proportions (Perez, 1895; Serment & Tourteaux, 1991).
1.2 Control tactics A thorough knowledge of the biology of the pest species is a prerequisite for successful control or eradication. Recent reviews by Grace and Yates (1999) and Su (2002) cover the various aspects involved in subterranean termite control. The various methods are briefly listed below.
1.2.1 Preventative Preventative treatment may be achieved by simple sanitation (such as ensuring that termites neither have a water source nor food source) and via good architectural design of buildings (Feytaud, 1946; Jacquiot, 1955, 1956; Fougerousse & Perlade, 1975; Krishna, 1989; Verkerk, 1990; Serment & Tourteaux, 1991; Forschler, 1998b; Grace & Yates, 1999). A recent example of an anti-termite design in construction are anti-termite cups for floor posts used in Japan (Takahashi & Yoshimura, 2002). The type of material used in construction is also important, for example, the hardwood of some tree species is less susceptible to termite attack (Grace & Yates, 1999). Wood treatment, physical barriers in the form of insecticide impregnated plastics, stainless- steel mesh, crushed rock (e.g. granite, basalt, quartz) and sand (e.g. coral and silica), and termiticide soil treatments are other forms of preventative treatment (Verkerk, 1990; Grace et al., 1997; Myles, 1997; Forschler, 1998b; Grace & Yates, 1999; Culliney & Grace, 2000; Su, 2002). Caution must be taken with physical barrier around points which traverse them, for example around pipes entering the building, as termites have been known to follow edges, cracks in soils as well as gradients, for example moisture (Forschler, 1994b; Forschler & Lewis, 1997; Grace et al., 1997). The level of control provided by insecticide soil treatments may be dependent on the
15 Introduction soil types as certain soils make the pesticide less available (Forschler & Townsend, 1996). The way in which particulate physical barriers work is that the size of the particles are such that they are too big for termites to pick up and yet the spaces between them are too small for termites to tunnel through (Grace & Yates, 1999). This means that particle sizes depend on the termite species, thus control is difficult when many species are present (Myles, 1997; Grace & Yates, 1999).
1.2.2 Remedial treatment In the past DDT or arsenic based substances have been used to control termites (Feytaud, 1946; Jacquiot, 1955, 1956; Fougerousse & Perlade, 1975; Krishna, 1989; Verkerk, 1990). Today the substances used can be naturally-occurring insecticides, synthetic chemicals or biological control agents (Forschler, 1998b; Su, 2002; Takahashi & Yoshimura, 2002). These treatments either repel the termites or kill them (Forschler, 1998b).
1.2.2.1 Sanitation Sanitation also falls under remedial treatment as any unsound and heavily infected timber should be removed and all timber treated (Jacquiot, 1955, 1956; Fougerousse & Perlade, 1975; Serment & Tourteaux, 1991; Culliney & Grace, 2000).
1.2.2.2 Soil & Masonry treatment Soil treatment is the conventional method of treating subterranean termite infestations (Forschler, 1998a, b; Culliney & Grace, 2000; Takahashi & Yoshimura, 2002). A trench is usually dug and the insecticide is applied in the trench, alternatively the insecticide is injected directly into the soil (Forschler, 1998b). However, this is not environmentally friendly and movement away from these methods is to be anticipated in the future (Su, 2002). Masonry may also be treated by injection into ground floor level and by treating basement and cross walls (Forschler, 1998b).
Baiting - this is an environmentally friendly method, as it has no residual effect. However, this can also be considered to be a flaw as it means that control only occurs when the baits are present. The aim is to provide a low acting toxicant that is non-repellent in a palatable food source (Forschler, 1998a, b). The toxicant must be
16 Introduction slow acting and in a high enough concentration to be passed through the colony (Suarez & Thorne, 2000). Baits are usually composed of cellulose matrices that are impregnated with a chitin synthesis inhibitor, such as hexaflumuron, and diflubenzuron or insecticides with other modes of action, for example, sulfluramid which disrupts energy metabolism (Vieau, 1991; Ferrari & Marini, 1999; Culliney & Grace, 2000; Su, 2002). The use of baits requires a thorough understanding of the termite life histories and colony structures (Pawson & Gold, 1996; Forschler, 1998a).
Biological control — this may be possible commercially but requires more development (Fougerousse & Perlade, 1975; Le Bayon et al., 1999). Many forms of biocontrol have been tested including predators and parasitoids but pathogens are probably the only practical option. However, to date, field trials with the pathogens have been rather limited (Culliney & Grace, 2000). The greatest potential has been shown in the use of fungal pathogens (Culliney & Grace, 2000) and these seem to be promising in combination with baiting (Takahashi & Yoshimura, 2002).
1.2.3 Conclusion In each method of control it is useful to know the pest life history to correctly use these various methods. Other information could also be useful, for example, when designing houses, knowing which timbers may be more resistant to termite attack. Knowing how fast colonies of particular species develop is particularly useful when considering biocontrol options, as these are often slow acting products.
1.3 Project background
The discovery of a colony of subterranean termites (Reticulitermes grassei) in Devon in 1994 (Verkerk, 1998; Verkerk & Bravery, 2000) sparked great interest. The initial concern was the risk of future introductions of these termites into the UK.
It is thought to have been accidentally introduced from the Mediterranean, and to have spread from an initial introduction point on one of the properties (Verkerk, 1998). In 1998, when a survey was performed, the colony was found to be spread over two properties, to a total area of approximately 2400 m2 (Verkerk & Bravery, 2000). It was decided that eradication was a viable solution as this was the only colony known in the UK. A project was therefore established, which was to run over
17 Introduction
10 years and was to include control and monitoring. This control programme used baits (Sentricon® and SentriTech®) both of which use the chitin synthesis inhibitor hexaflumuron, but it was found that there were problems in that high recruitment but low consumption was achieved (Verkerk & Bravery, 2000). The problem was found to be due to the bait matrix (this species preferred a pine matrix to a paper matrix) and bait repellence. The lack of available information on this species was made apparent during the implementation of this control programme and led to the development of this Ph.D. project.
1.4 Aims and Objectives The aim of the current work is to provide data that can be used to help assess the risk posed by termites being able to establish and multiply in the UK. This risk can be judged by determining the minimum number of individuals and other biotic and abiotic conditions required for establishment of a termite colony. The triggers required for reproductive formation and the developmental strategies of the species under study are also of interest.
The two species that were chosen for the study are R. grassei, the species colonising Devon, and R. santonensis, the Reticulitermes species with the most northerly distribution in Europe and which therefore would appear poses the greatest potential risk of establishment to the UK.
The objectives are to:
• determine the minimum number of each species required for establishment under optimum conditions (25°C, 90% RH) (Chapter 4); • establish whether laboratory data can be related to field conditions with regard to minimum number (Chapter 4); • determine food preferences using different types of cellulose sources (Chapter 5); • determine the precise reproductive strategies of the two species (Chapter 6); • determine soil type preferences (Chapter 7); • determine the effect of temperature on each of these species (Chapter 7); • identify areas in the UK that may have suitable conditions to enable termite colonies to establish themselves.
18 Chapter 2
Literature Review
2.1 Introduction Termites belong to the order Isoptera and in the past have been called 'white ants' even though they are not related to ants and in fact their closest ancestors are cockroaches (Krishna, 1989). There are more than 2700 known species of termites world-wide of which approximately 70 or 80 are considered pests (Culliney & Grace, 2000); many are in fact also very useful in the role of nutrient -recyclers (Thorne, 1998; Yamashita & Takeda, 1998).
The genus Reticulitermes seems to have been in existence since the Oligocene epoch (24 to 37 million years ago) (Grasse, 1949; Plateaux & Clement, 1984). Species from this genus are also often called 'subterranean' termites as they commonly have an association with the soil, although this association is not obligatory (Weesner, 1965; Forschler, 1998b). They used to be considered to be part of the lower termites but are now thought to be an intermediate between the higher and lower termites and are, therefore, also termed 'middle' termites (Snyder, 1920; Harris & Sands, 1965; Weesner, 1965; Noirot & Pasteels, 1987; Takematsu, 1992).
Reticulitermes species are from temperate regions and their colonies undergo seasonal changes (Grasse, 1954). This seasonality may be witnessed by a change in feeding habits, activity and movement deeper into the soil during colder periods (Grasse, 1954; Forschler, 1998a). The genus occurs in Europe, Asia and North America (Palearctic and Nearctic regions) but has also, in recent decades, been accidentally introduced into South America (Weesner 1965, Aber and Fontes 1993, Myles, personal communication).
2.2 Species studied
The two species that are being considered in the present work are Reticulitermes santonensis and Reticulitermes grassei. Reticulitermes grassei, as mentioned in Chapter 1, is the termite that was discovered to have established a colony in Devon,
19 Literature Review
UK. The lack of information available in the literature concerning this species became evident when a control programme was put into place.
Reticulitermes santonensis was included in this study as it is found at a greater proximity to the UK than any other termite species, including R. grassei. The details of the distribution of each species will be discussed in Section 2.4. The literature on R. flavipes will also be briefly discussed as it has recently been found that this species may be synonymous with R. santonensis (Jenkins et al., 2001).
2.3 Taxonomy Termite taxonomy, like taxonomy in general, has undergone many changes. Initially, in the 18th century, termites were put into the genus 'Terms' by Linneus (Feytaud, 1946). This causes particular difficulty when referring to the old literature (Thorne, 1998; Forschler & Jenkins, 1999). The original name of R. lucifugus was therefore Termes lucifugus which was given by Rossi in 1792 (Gassies, 1855; Lespes, 1856; Perez, 1907; Bagnere et al., 1990). The genus was later changed to Leucotermes Silvestri (Feytaud, 1920). Holmgren eventually created the genus Reticulitermes in 1913, which groups the main termites found in the temperate regions of the Northern hemisphere (Feytaud, 1951).
The problems in the taxonomy of termites also extends to the species level. The need for revision of the genus Reticulitermes has been mentioned by various authors yet is still a problem today (Haverty et al., 1991; Haverty et al., 1999a). It has been suggested that, for North American species, there have been no improvements since the 1920's revision by Snyder and Banks (Haverty et al., 1996).
Initially, there was only thought to be one Reticulitermes species in France, R. lucifugus (Perez, 1895; Perez, 1907; Feytaud, 1911). This species was first described by Rossi in 1794 (Grasse, 1954). The word `lucifugus' means 'light shunning' in Latin (Feytaud, 1911; Snyder, 1926). De Quatrefages (1853) was probably one of the first to mention the possibly of there being two termite species. Feytaud (1924) later concluded that there were indeed two Reticulitermes species, the second being R. santonensis, an introduced species that was given this name because of its presence in the formerly called region of Saintonge (Vieau, 1991). The idea that the termites present in Saintonge were an introduced species had been mentioned by authors during the 19th century (de Quatrefages, 1853; Lespes, 1856). Clement et al. (2001)
20 Literature Review explained that the hypothesis that R. flavipes present in Europe (where in fact it was first identified) was accidentally imported from the U.S.A. can never be verified as the initial description was incomplete. Reticuliterms flavipes has also had taxonomic changes and was previously named R. claripennis Banks in the U.S.A. (Weesner, 1965). Reticulitermes santonensis, however, has been variously thought to be synonymous with Calotermes flavicollis, a hybrid of R. flavipes and R. lucifugus, a subspecies of R. lucifugus, a subspecies of R. flavipes, a systematically distinct species from both R. lucifugus and R. flavipes and most recently, likely to be synonymous to R. flavipes (Feytaud, 1912, 1925, 1946, 1951; Grasse, 1954; Buchli, 1961; Hickin, 1961; Hrdy, 1961; Becker, 1970; Clement, 1976; Clement, 1977a; Clement, 1977b; Clement, 1978a; Clement, 1978b, 1979a, b; Bagnere et al., 1990; Vieau, 1991, 1996; Forschler & Jenkins, 1999; Jenkins et al., 2001). Reticulitermes santonensis has also been suggested to be composed of several semi-species (Clement et al., 1986).
Reticulitermes lucifugus species has since undergone many changes as it was split into subspecies (Clement, 1976; Clement, 1977a, 1978a; Clement, 1982b, 1984) one of which, R. lucifugus grassei, has now been elevated to species status as R. grassei (Clement et al., 2001). The presence of several subspecies within R. lucifugus had previously been suggested to exist due to behavioural differences observed within this species but these had never been described (Feytaud, 1946). Reticulitermes lucifugus has in addition been found to be genetically close to R. speratus which is found in Japan (Forschler & Jenkins, 1999) and Reticulitermes lucifugus banyulensis has been identified as a separate species, R. banyulensis (Clement, 1979a).
The problem of being able to identify termite species and the intraspecific variation seen has been known of for many years (Feytaud, 1925, 1946; Grasse, 1954). In the past, there were few taxonomic characters and these are now thought to be unreliable (Grasse, 1954). The situation has not really changed as researchers are still experiencing problems with termite taxonomy (Clement et al., 1986; Forschler, 1998a). New species are still being identified in the order Isoptera. At present, however, the majority are being found in China, although this may be because less taxonomic work is being performed in the rest of the world (Eggleton, 1999).
There are various methods that have been used for the identification of species and phenotypes. These include: aggression (also termed agonism), comparisons of
21 Literature Review symbionts present in the gut, colour (of cuticle and wings), wing size, ocelli-eye distance, ocelli size and frons-postclypeus suture shape (Feytaud, 1924; Weesner, 1965; Clement, 1978a; Clement, 1981, 1982a; Plateaux & Clement, 1984; Getty et al., 1999; Getty et al., 2000; Clement et al., 2001).
Reticulitermes species have been shown to hybridize in the laboratory. The absence of this phenomenon in the field has been explained by the presence of aggression as well as the difference in swarming times between the species (Grasse, 1954; Becker, 1970; Clement, 1976; Clement, 1977a; Clement, 1977b, 1978b, 1979a, b, 1981). Aggression has been said to prevent copulation but not colony fusion (Bagnere et al., 1990). The level of aggression is measured as agonism (Ag), where Ag = 2.5 x (M + m/2); M = mean no. of termites dead, m = mean no. of termites with amputated legs or antennae (Clement, 1982b). Agonism is considered to be present when Ag > 25.
Colour and aggression have been rejected as taxonomic criteria as they are very unreliable for species differentiation (Hrdy, 1961; Forschler, 1994a; Polizzi & Forschler, 1998; Forschler & Jenkins, 1999; Clement et al., 2001). Aggression is also thought to decrease with time in the laboratory (Polizzi & Forschler, 1998). Distinctive characteristics have not as yet been identified in each caste and this has been briefly covered by Vieau (1993) who worked with mandibular measurements in soldiers. The main problem with termite keys is that they are very old and it is often difficult to find alates for identification. Sands (1998) produced keys for identification using workers but these do not allow the user to go further than genus level.
Present day approaches to taxonomic classification include the use of chemotaxonomic characters such as long-distance attraction and contact sex- pheromone tests, the analysis of the defence secretions released by soldiers (mostly terpenes) and cuticular hydrocarbons (Clement, 1985; Haverty et al., 1991; Haverty et al., 1996; Forschler & Jenkins, 1999; Getty et al., 2000; Clement et al., 2001). Another method is by the analysis of DNA. To date, sequencing has only been done with 16s ribosomal RNA and NADH dehydrogenase 1 gene fragments (Forschler & Jenkins, 1999; Clement et al., 2001) although, the cytochrome oxidase II (COII) gene has been used to aid in establishing phylogenetic relationships between families and subfamilies (Forschler & Jenkins, 1999). Jenkins et al. (2000) used both mtDNA and chemical phenotypes in their study and found that these correlated, although
22 Literature Review they did caution that slight differences in cuticular hydrocarbon mixes should not be over-interpreted. These may also not be as accurate as thought as Vauchot et al. (1996) found that there was a transfer of cuticular hydrocarbons between species after only two hours of cohabitation.
2.4 Distribution
2.4.1 Genus
The general distribution of Reticulitermes species is shown in Figure 2.1. Although this is not very accurate, it gives a general idea of their world-wide distribution. This genus is found in North America, Asia, North Africa and southern Europe (Plateaux & Clement, 1984). Light (1934) gave the 10 °C isotherm as the most northern limit for Reticulitermes species. This genus is thought to be the only one present where severe winters occur (Kofoid, 1934). Harris (1962) proposed that R. lucifugus would not be able to establish in the UK due to the short summer, which, he suggested, would not be sufficient for reproduction. Many previous descriptions in the literature of the distribution of the genus Reticulitermes have been too general to be able to produce reliable maps (Feytaud, 1912; Grasse, 1949, 1954; Ionescu, 1959; Weesner, 1965; Fougerousse & Perlade, 1975). More detailed maps, depicting the distribution of the species being studied, will be provided later in this section.
The taxonomy of the genus Reticulitermes is still in the process of being defined and many changes have been made, even recently (Clement et aL, 2001). Thus R. lucifugus is often referred to as, in older literature, no differentiation was made as they were considered to be one and the same species (see section 2.3). The most recent known distribution of the genus in Europe is shown in Figure 2.2.
23 Literature Review
Figure 2.1 World-wide distribution of Reticulitermes species (after Pearce (1997)).
U.K.
R. santonensts .
PORT
/ R. banyulensls
R.I. carat, R. luciti:rgu'
grdssei R. balkan nsi
Figure 2.2 Distribution of Reticulitermes species in Europe (considered to be natural). Crosshatched areas show sympatric zones between R. grassei and R. santonensis in the north and R. grassei and R. banyulensis in the Iberian Peninsula (after Clement et al. (2001)).
24 Literature Review
Reticulitermes are also found in areas which are not considered to be part of their 'natural' distribution range and these are depicted in Figure 2.3. Although, it must be noted that the position of the R. grassei colony in the UK is incorrect and should be further west, in Saunton, Devon.
A R. grassed Hamburg R. banyidensis * R. lucifugus * R. I. COISICUS GERMANY • R. sp * R. sapronensIs • R. balkanensis
Ar: 4A A ATA A LAA ,_6A`Ata- -A ArA AA, A . jia,A•ArArA,A, A AWAVAS: A•Alki4eA•A "A" a- WA*AWA.Navir at Ali
500 kms thetIS
Figure 2.3 Urban distribution of Reticulitermes species in Europe (after (Clement et al., 2001)).
Reticulitermes species are also present in North Africa. However, the particular species present are unknown (Plateaux & Clement, 1984).
2.4.2 Reticulitermes grassei
The main distribution of R. grassei is in South-Western France, South and North- Western Spain and the whole of Portugal. (Figure 2.2). This species has only recently been defined as it was previously considered to be a subspecies of R. lucifugus (Clement et al., 2001). Thus, the information on its distribution is not specific and in older literature is grouped under R. lucifugus. Reticulitermes lucifugus was said by Grasse (1954) to be the most widely distributed species in France. This is now true for R. grassei (Clement et al., 2001). There have also been recent introductions of R. lucifugus into Uruguay (Aber & Fontes, 1993), however,
25 Literature Review the subspecies was not specified and thus further investigation is required. Reticulitermes lucifugus is also said to be found in Africa and Asia minor (Gay, 1967; Sands, 1998).
2.4.3 Reticulitermes santon e nsis
The distribution of R. santonensis is mainly in the area defined in Figure 2.2. This species is also found further north in urban areas of France such as Paris and Rouen (Normandy) (Figure 2.3) (de Feytaud, 1955; Jacquiot, 1955; Cals-Uscati & de Frescheville, 1963; Fougerousse & Perlade, 1975; Clement, 1976; Serment & Tourteaux, 1991; Vieau, 1991; Clement et al., 2001). Some authors consider it to be an urban termite and it has been suggested that the presence of R. santonensis colonies in natural environments is simply because these areas used to be inhabited (de Quatrefages, 1853; Feytaud, 1925; Vieau, 1993; 1994a; 1999). A species of Reticulitermes has also been found in Chile. Due to the very complex taxonomy of this genus, the species was originally identified as being R. hesperus, though this is now thought to be R. santonensis and further investigation is required (Clement et al., 2001; Ripa et al., 2002).
2.4.4 Reticulitermes flavipes
In Europe, R. flavipes is present in Hamburg, Germany (Harris, 1970), although this has recently now said to be R. santonensis (Clement et al., 2001). The general distribution of this species is from South-Eastern Canada to North-Eastern Mexico (Weesner, 1965; Gay, 1967; Grace, 1996) (Figure 2.4). Weesner (1965) believed this species to have a northerly limit, in North America at the —6.7 °C annual minimum air temperature boundary. However Grace (1989) listed Winnipeg as the furthest northern point these termites are found, which has a mean minimum temperature of - 3.1 °C. The first observation of termites in Ontario, Canada is said to be in 1938 (Gay, 1967; Grace, 1989). Reticulitermes flavipes is said to me the most common Reticulitermes in the Eastern United States (Jenkins et al., 1998).
26 Literature Review
Figure 2.4 Distribution of Reticulitermes flavipes (shaded area) in the U.S.A. (after Nutting (1990)).
2.5 Movement of termite species in France
The first reference to termites is said to be in 1442 when the King of France is reported to have fallen through the floor of a hotel at La Rochelle (46° N) (Fougerousse & Perlade, 1975). Although this was simply a suggestion of the cause of the collapse by de Feytaud (1959). At the end of the 18th century the presence of termites into the South-West of France was officially recognised (de Quatrefages, 1853; Feytaud, 1911; Fougerousse & Perlade, 1975; Serment & Tourteaux, 1991; Vieau, 1991). Many regions have been suggested as being the origin of these termites, namely: India, South America, West Indies, Middle East and the Dominican Republic (de Quatrefages, 1853; Gassies, 1855; Feytaud, 1911; Feytaud, 1924; Grasse, 1954; de Feytaud, 1959). Gassies (1855) believed the species at La Rochelle and in Bordeaux were the same and rejected the idea of the accidental introduction of termites. He suggested that indigenous termites had been accidentally brought into town with firewood. Jacquoit (1955) also believed R. santonensis to be an indigenous species.
27 Literature Review
During the 1830s to 1850s there were many accounts of damage by termites in South-Western France (Feytaud, 1911; Thompson, 1917; Feytaud, 1924; Vieau, 1991). The presence of termites in Bordeaux was observed in 1853 and shortly after confirmed by other authors (Gassies, 1855; Perez, 1895; Feytaud, 1911, 1946). Termite swarms were first described in South-Western France in 1895 (Perez, 1895; Perez, 1907). The movement of each species is difficult to follow as there has been much confusion regarding the taxonomy (section 2.3). At the turn of the 20th century R. lucifugus (which at the time was not split into various subspecies and species) was thought to only be found in Maritime pine forests, 100 km north and south of Bordeaux (Feytaud, 1912).
One of the first observations of the presence of termites in Paris was that of Lesne (1923) in 1922 (Fougerousse & Perlade, 1975). Several well established infestations, which were identified as being R. lucifugus and R. santonensis, but are now thought to only have been R. santonensis, were discovered in the 1940s and 1960s in Paris (de Feytaud, 1955; Jacquiot, 1955; Cals-Uscati & de Frescheville, 1963; Fougerousse & Perlade, 1975).
Reticulitermes santonensis has been gradually spreading northwards since the first accounts of damage. In the 1950s, termites thought to have been R. santonensis were said to have spread up to the town of Nantes (47° 15' N) in North-West France (Grasse, 1954; Jacquiot, 1955; Vieau, 1991). More recently, only R. santonensis was thought to be present in Nantes whereas Bordeaux is thought to have both species (Vieau, 1993). During the 1950s the limit of distribution between the two species was the Gironde (Jacquiot, 1955), a river that lies just to the north of Bordeaux.
In the 1970s, R. santonensis was found to be 50 km into the interior from the Charente Maritime coast (Central Western France) and was also found as far north as Rouen (49° 30' N), Normandy (Fougerousse & Perlade, 1975; Clement, 1977a). Reticulitermes lucifugus was found south of the Gironde river (between 46° and 47° N) (Clement, 1978a).
In recent decades there has been a marked extension of termite distribution towards the North-East (Serment & Tourteaux, 1991). Reticulitermes lucifugus has moved much further north than its distribution at the turn of last century (Feytaud, 1912;
28 Literature Review
Vieau, 1999). Reticulitermes santonensis is now also found in and around Bordeaux (Vieau, 1999).
Termite dispersal is slow in temperate regions of the world. Some authors, such as Grasse (1954), did not believe man to have played a significant role in termite spread. However, the spread of Reticulitermes species is thought to be mainly due to artificial distribution and particularly through the movement of freight via railways (Fougerousse & Perlade, 1975; Serment & Tourteaux, 1991; Vieau, 1991, 1993). Man's influence on the dispersal of termites was thought to be particularly true for R. santonensis which is considered to be a mainly urban termite. The presence of central heating in urban areas has meant that establishment of this species has been more successful and may explain the presence of R. santonensis in areas situated much further north than its normal distribution, e.g. Paris (Kofoid, 1934; Cals-Uscati & de Frescheville, 1963; Esenther, 1969; Plateaux & Clement, 1984). The wood that the termites feed on may also act as an insulation against the cold (Grace, 1989). Termite establishment has been seen in Bordeaux where the introduction of central heating had a great impact on the increase in termite populations (Vieau, 1991, 1993). Heating could well have aided the establishment of the accidentally introduced R. flavipes in Hamburg, which, as mentioned above, is now though to be the same species as R. santonensis (de Feytaud, 1955; Jacquiot, 1956; Plateaux & Clement, 1984; Clement et al., 2001; Jenkins et al., 2001). Reticulitermes lucifugus is thought to have increased in occurrence due to the increase in the availability of pine (de Feytaud, 1955). Global warming may also mean that termites that are accidentally introduced may be able to sustain themselves in areas previously thought to be too cold (Jacquiot, 1956).
2.6 Differences between the species studied
The main differences reported between the species studied are listed in Table 2.1. However, very few of these are substantiated by laboratory experiments.
29 Literature Review
Table 2.1 Comparisons between Reticulitermes santonensis and Reticulitermes lucifugus/ Reticulitermes grassei.
Topic R. santonensis R. lucifugus and R. Reference grassei Activity More virulent/fecund, Less virulent/fecund Feytaud, 1946; more robust more feeble Grasse, 1954; Jacquiot, 1956; Becker, 1970; Harris, 1970
No difference in voracity No difference in Feytaud, 1946 voracity
Higher activity Lower activity Vauchot et al., 1996
Greater gallery building Less gallery building Becker, 1970 activity activity
More easily detected Less easily detected Fougerousse & Perlade, 1975
Food Higher consumption Lower consumption Becker, 1970 except at high humidity where becomes equal
Attacks live plants more Attacks live plants less Grasse, 1954; Harris, often often 1970
Environment Better able to spread to Less able to spread to Jacquiot, 1956 cold region cold regions
Optimum temperature for Optimum temperature Becker, 1970 consumption is 29 °C. for consumption is 26.5
More tolerant to low Less tolerant to low Buchli, 1951; Grasse, moisture levels moisture levels 1954; Becker, 1970; Harris, 1970; Fougerousse & Perlade, 1975; Clement, 1977a; Clement, 1979b
Lower positive Higher positive Buchli, 1961 geotropism geotropism
Distribution Present further north Present further south Plateaux & Clement, 1984
30 Literature Review
Topic R. santonensis R. lucifugus and R. Reference grassei Distribution Urban termite Rural termite de Quatrefages, cont. 1853; Feytaud, 1925; Vieau, 1993, 1994a, 1999
Reproduction Reproduces more via Reproduces more via Jacquiot, 1956; neotenics (also found in alates Fougerousse & R. flavipes) Perlade, 1975; Clement, 1977a; Vieau, 1993, 1994a, 1999
Initial theories thought Initial theories thought Clement, 1977a
that nymphs that nymphs
overwintered with short overwintered with long
wing buds wing buds
Not known Different reproductive Feytaud, 1946; strategies in different Grasse, 1949; areas e.g. Italy and Feytaud, 1951; France Harris, 1958
Greater fecundity Less fecund Harris, 1970
Biological Some gut fauna present Specific gut fauna Becker, 1970; Harris, difference which are not common to 1970 R. lucifugus
Spirotrichonympha Spirotrichonympha Clement, 1979a kofoidi present, kofoidi absent,
Spirotrichonympha Spirotrichonympha flagellata absent flagellata present
Colony Always anagonistic Aggression varies Clement, 1982b; interaction seasonally. Clement, 1986 Aggression in winter
Swarming in March Swarm in May and June de Quatrefages, in morning 1853; Feytaud, 1912, 1946
Swarm in mid April Swarm in April to May Feytaud, 1912; Plateaux & Clement, 1984;Clement, 1977a
Swarm in April and May Weesner, 1965; (same is true for R. Vieau, 1994a flavipes)
Swarming May and June (considered both spp. to Perez, 1895 be the same)
Swarming between Harris, 1958 March and May
31 Literature Review
Reticulitermes santonensis is more active that R. lucifugus and R. grassei in terms of voracity, consumption, gallery building etc. An interesting point to note is that R. santonensis is thought to be present in urban environments and this may explain its presence further north compared to R. grassei, although the results by Becker (1970) seem to contradict this as he found that consumption of R. santonensis was optimum at higher temperatures than R. lucifugus. It may not be that R. santonensis is better able to spread to cold regions as mentioned by Jacquiot (1956) but that it is able to take advantage of the urban environment to proliferate.
Neoteny is considered to be a characteristic of an introduced species that needs to develop rapidly so as to find a niche (Vieau, 1991). Although this is particularly true with regard to R. santonensis, it is also seen in R. lucifugus, which in Italy, has no primary reproductives in the colonies, whereas in France these are observed (Feytaud, 1946).
De Quatrefage (1853) was the first to put forward the idea that R. santonensis may have come from a warmer climate as he observed that swarming occurred earlier in R. santonensis compared to R. lucifugus and R. grassei. He considered this to be unusual as R. santonensis had a more northerly distribution in France. The information in Table 2.1 shows that there is some confusion as to when the two species swarm.
2.7 The life cycle of Retic ulitermes spp. (Isoptera, Rhinotermitidae): what do we know?*
2.7.1 Abstract
The subterranean termites in the genus Reticulitermes (Isoptera, Rhinotermitidae) have a complex and plastic life cycle, which has been the subject of a number of publications over the past century. Given the inherent difficulties in studying such cryptic, eusocial organisms it is perhaps not surprising that the literature on their biology has failed to reach a consensus. In this section, an overview of the literature is given leading into a discussion of the various theories on the life cycle. A substantial proportion of the review focuses on the French literature, which
* Submitted to Bulletin of Entomological Research (with authorship as follows: Lail* L. V. and Wright, D. J.) and presented here with minor modifications.
32 Literature Review constitutes the majority of the primary sources and which can be difficult to access. There are many discrepancies in the literature in terms of the number of instars. The definition of workers is a further problem. There are outstanding questions of whether they should be termed pseudergates or whether there should be an additional terminology to differentiate between pseudergates and true workers, as seen in the higher termites. It remains very difficult to compare publications to date as there is little conformity and this is further aggravated by the absence of drawings of the relevant instars. Further work on Reticulitermes biology is required as is the need for researchers to agree on a standard terminology for this genus. This section aims to compare and consolidate the different life cycles proposed to date, a glossary is provided at the back of this thesis to aid the reader through this difficult topic.
2.7.2 Introduction
The life cycle of Reticulitermes spp. (Isoptera, Rhinotermitidae) has been a subject of much debate (Grassi & Sandias, 1893; Snyder, 1915; Hare, 1934; Feytaud, 1946; Grasse, 1949; Grasse et al., 1950; Buchli, 1958; Harris & Sands, 1965; Noirot & Pasteels, 1988; Takematsu, 1992; Vieau, 1994a, 1999). This is primarily due to the difficulties that arise when termites are studied; they are subterranean and therefore have very cryptic habits. They also have an extremely plastic biology.
Life cycle studies on Reticulitermes spp. have not been done for several decades, possibly because it is assumed that sufficient knowledge has already been gained. However, looking at the primary literature, most of which is in French, it becomes apparent that the life cycle is not well understood (Lespes, 1856; Feytaud, 1946; Grasse, 1949; Grasse et al., 1950; Buchli, 1958; Esenther, 1969; Noirot, 1985; Noirot & Pasteels, 1988). There have been recent reviews that have dealt with Reticulitermes (Thorne, 1998; Thorne et al., 1999), but these do not go into detail about problems involved in the various life cycle theories. Central to the debate is the work of Buchli (1958) which has been the base of most reviews and discussions thereafter (Noirot & Pasteels, 1987; Thorne, 1996, 1998).
Our aim is to compare and consolidate the different life cycles proposed to date. We initially explain the current understanding and compare previous theories to this. This section is especially relevant in the light of the recent finding that Reticulitermes santonensis (Feytaud) is likely to be synonymous with Reticulitermes
33 Literature Review flavipes (Kollar) as this makes the availability of the (European) literature on R. santonensis more widely applicable (Jenkins et al., 2001).
2.7.3 Life history Termite colonies are composed of individuals of different castes; these castes have a similar function in all termites (Noirot, 1985). These castes are the alates (primary reproductives), neotenics (replacement or secondary reproductives), nymphs (developing individuals in the sexual line), workers (individuals in the neutral line), pseudergates, larvae and soldiers. In Reticulitermes spp. all castes are made up of individuals from both sexes (Perez, 1907; Snyder, 1926).
European termites were thought to have the same life cycles as those from the tropics (Thompson, 1917). However, not enough is known about each species to support this view and Thorne (1998) states that species within the genus Reticulitermes should not all be considered to have the same life history. Yet, to this day, general comments are often made about the genus as a whole and this makes the review of this topic extremely difficult as in some circumstances broad statements have had to be made.
The most recent, detailed investigation of the life cycle of R. santonensis and R. lucifugus (Rossi) was performed by Buchli (1958), although some of his methodology was problematic and whether all of the developmental stages he observed occur, has been questioned (Noirot, 1985; Noirot & Pasteels, 1987; Vieau, 1994b). He performed very few replicates and he marked some termites by amputation of a leg, possibly altering their behaviour, for example, cannibalism of such an individual by other members of the colony is more likely to have occurred. Buchli (1958) also concentrated on the study of R. santonensis and is thought to have confused the two species, an important drawback since R. santonensis and R. lucifugus may have different developmental pathways and reproductive strategies. Both species have the capacity to form secondary reproductives, but, dispersion in R. santonensis is mainly via secondary reproductives and R. lucifugus via alates (Vieau, 1999).
Many of the post-embryonic forms described by Buchli (1958) are not thought, by others, to exist in nature (Noirot, 1985; Noirot & Pasteels, 1987; Vieau, 1994a) and the title of Buchli's thesis, 'The ontogenic potential...' does not suggest that these
34 Literature Review were all naturally occurring individuals. However, it would seem unlikely that so many forms not seen in nature would occur under laboratory conditions, which are far more stable than the external environment. The caste organization described by Buchli (1958) was extremely complex and was later simplified by Noirot (1985). The explanation of neoteny (reproduction via supplementary reproductives), one of the most important developmental paths, was omitted by Noirot (1985) (Vieau, 1996).
The study of post-embryonic development in termites is extremely complex due to the difficulty in counting the number of moults that individuals undergo, the presence of polymorphism, and because individuals may undergo deviations in their development. Their social behaviour is also very complex and they have a lengthy developmental time (Weesner, 1965). That any instar can eventually develop into a reproductive, either a primary or secondary reproductive, means that a viable colony could form from any group of individuals (Noirot, 1990). The existence of parthenogenesis in Reticulitermes spp. has had mixed support and has only been seen in a few species (Grassi & Sandias, 1893; Buchli, 1950b; Weesner, 1956; Nutting, 1969; Howard & Haverty, 1981; Pawson & Gold, 1996; Matsuura & Nishida, 2001).
2.7.3.1 Terminology The terminology of termite stages is a source of much confusion (Forschler & Jenkins, 1999) due to the use of the same terms to describe different stages. This is particularly evident when comparing the French and American literature, especially when looking at the older literature (Thompson, 1917; Grasse, 1949; Weesner, 1965; Vieau, 1994a; Thome, 1996, 1998).
35 Literature Review
SOLDIER
N6 WHITE SOLDIER
inn N5 W5
N4 W4
N3 4 L2 W3
Li
Figure 2.5 Life cycle of Reticulitermes spp. according to F. Vieau (personal communication), amended from Buchli (1958).
36 Literature Review
The terminology used in this article is that of Buchli (1958) with the numbering based on the work by Vieau (1994a; 2001) (Figure 2.5). The preferred terminology is shown in bold. The term larva is used for the first two stages after hatching (L1 & L2) (Buchli, 1958). After this stage, two lines are formed, the worker line (where the individuals have no wing buds) (W3 to W7) and the sexual line or nymphal line (where the individuals have wing buds). Workers can develop either into soldiers, via an intermediary stage called 'white soldiers', or alternatively into secondary reproductives called apterous neotenics, third form reproductives or ergatoids, which have no wing buds (Weesner, 1965; Krishna, 1989; Thorne, 1996, 1998). The nymphal line is made up of four initial nymphal stages (N3 to N6). After the N6 stage there is a split in the line where individuals can either develop into nymphs with long wing buds (LWBN) or nymphs with short wing buds also termed a pre- neotenic brachypterous stage (PBNEO) (Vieau, 1994a; 2001). The LWBN will go on to develop into an alate, also termed imago, (termed primary or first form reproductive once they have lost their wings), and those nymphs with short wing buds will develop into brachypterous neotenics (BNEO), also termed second form reproductives (secondary reproductives with wing buds) (Weesner, 1965; Krishna, 1989; Thorne, 1996, 1998; Vieau, 2001). The three adult forms were first defined by Thompson (1917). Although the first brachypterous neotenics were originally described by Lespes (1856), whose terminology was taken up by Grassi and Sandias (1893), the term 'secondary reproductive' refers to any reproductive apart from the primary reproductive (i.e. alates that have lost their wings) that founded the colony. In contrast, the term 'supplementary reproductive' is given to secondary reproductives that develop whilst the primary pair are still alive (Thorne, 1996).
2.7.3.2 Instar differentiation Termite instars may be separated by the width of the cephalic capsule (Hare, 1934; Buchli, 1958; Clement, 1979a). Measurements of other structures, such as antennae and wing pad development, have also been used (Hare, 1934; Buchli, 1958). Nymphal stages are said to be determined by measuring the antennal length and head width (Buchli, 1958). Individuals preparing to moult almost completely empty their stomachs, stop eating and become milky in appearance (Grasse, 1949; Buchli, 1958). The moult is either eaten by the individual itself or by larvae or workers that
37 Literature Review surround the individual (Grasse, 1949); the unfit and injured individuals are also eaten by colony members (Grasse, 1949).
2.7.3.3 Life cycle: egg to L2 In Reticulitermes spp. eggs are laid approximately 15 days after mating and the colonies' eggs are generally found in clumps of 500 to 1000 (Feytaud, 1946; Grasse, 1949; Vieau, 1991, 1996). Lespes (1856) reported that in R. lucifugus eggs are found in clumps of about 100, but, the age of the colonies in these studies was not stated and the number of eggs laid is dependent upon the colony size (see below). The eggs are often moved around and are continuously stacked and restacked. They are also polished and turned in the workers' mouthparts (Weesner, 1965). In young colonies, the eggs are initially attended by the reproductives and later by the workers (Grasse, 1949). Egg laying rates that have been quoted for R. santonensis include several thousand eggs per individual per year and 4.8 eggs indiv."1 111 (Grasse, 1949; Vieau, 1996). These rates of egg laying probably relate to mature colonies as Reticulitermes spp. have a very slow initial egg laying rate where young reproductives tend to lay five to six eggs in the first year and 25 to 30 eggs in the second (Feytaud, 1912; Snyder, 1926; Weesner, 1965). Thorne (1998) reported on various studies which showed that, in their first year, young Reticulitermes colonies reach no more than 30 individuals yet another study (Beard, 1974) showed the number to be 70 in R. flavipes colonies. Egg laying varies, a period without egg laying occurs each year in termites from temperate regions (Grasse, 1949; Weesner, 1965; Nutting, 1969).
There is considerable variation in the time taken for egg development, varying from 15 days to 55 days, at temperatures between 17 and 25 °C (Grassi & Sandias, 1893; Feytaud, 1946; Buchli, 1950a; Weesner, 1956; Beard, 1974; Vieau, 1991). The period quoted by Buchli (1958) is 20 to 30 days at 25 °C. Young termites emerge and go through two stages or instars called larvae or white immatures (Buchli, 1958; Harris & Sands, 1965; Thorne, 1996; Vieau, 1996). Harris & Sands (1965) claimed that the larvae were the first 2 to 3 stages. The first instar is approximately 1 to 2 mm long and can take between 7 to 17 days and the second 10 to 18 days to develop (Lespes, 1856; Weesner, 1956; Buchli, 1958; Vieau, 1991). Though their size is also said to vary depending on the environment (Buchli, 1950a). At the second moult differentiation into the sexual and neutral lines occurs (Hare, 1934; Noirot, 1985; Noirot, 1990; Thorne, 1998). The worker line is said to diverge irreversibly from the
38 Literature Review nymphal line, except for the ergatoids (see section 2.7.3.4) (Thorne, 1998). The two lines can be differentiated at the third stage by the presence (sexual) or absence (worker) of wing pads (Feytaud, 1912). Some authors believed that caste determination occurs at the egg stage (see section 2.7.3.10) (Feytaud, 1912; Thompson, 1917; Snyder, 1925, 1926; Miller, 1969).
2.7.3.4 Life cycle: workers Workers are approximately 4 to 6 mm in length, wingless, have antennae with 14 to 17 segments, no compound eyes and a pronotum that is trapezoid in shape (Lespes, 1856; Feytaud, 1912; Snyder, 1926; Feytaud, 1946; Serment & Tourteaux, 1991; Vieau, 1991; Thorne, 1996). The definition of an isopteran worker according to Noirot and Pasteels (1987) is: ' ...morpholo gically specialised individuals whose development has diverged early and irreversibly from the imaginal line. Workers constitute a sterile, morphologically distinct, eusocial caste.' This definition proves difficult when we consider that workers can become reproductives. Perhaps a more appropriate term may be used, such as pseudo-worker. This will allow the differentiation between an individual that can develop into a soldier or an ergatoid. Though, conversely it has been said that workers in Reticulitermes spp. are close to true workers (Miller, 1969).
Workers are the most numerous caste in the colony and forage to feed the colony, as well as tend to the nursery and groom nestmates (Feytaud, 1911, 1912; Snyder, 1926; Noirot & Pasteels, 1987; Krishna, 1989; Vieau, 1996; Thorne, 1998). Their activity tends to increase in the early spring and decrease over the winter (Haverty et al., 1999b). They also play a role in defending the colony and this aggressive behaviour has been well studied (Grasse, 1949; Buchli, 1961; Clement, 1978a, 1986; Thorne & Haverty, 1991; Polizzi & Forschler, 1998; Haverty et al., 1999c; Getty et al., 2000). They have mouthparts that are adapted for chewing and possess flagellate protists in their hindgut which digest the wood, the products of which are used by the termites (Feytaud, 1946; Krishna, 1989). Workers are the first caste to develop when a new colony is founded by alates, their developmental time lasts approximately one year and they can live for several years. Buchli (1958) suggested that workers can live 9 to 10 years in a natural situation. However, little information is available on their longevity (Feytaud, 1912; Snyder, 1926; Nutting, 1969; Noirot, 1990).
39 Literature Review
Buchli (1958) (Figure 2.10) stated that there were nine worker instars, including the first two larval instars, after which the workers continued to moult but did not change in size (i.e. go through stationary moults). Harris and Sands (1965) and Plateaux and Clement (1984) both believed that there were a varying number of instars (five to seven from egg) in the Rhinotermitidae before the workers' heads became fully sclerotized, juvenile cuticles are unsclerotized. These differences may stem from the fact that the young worker instars are difficult to differentiate from the larvae as mentioned by Grasse et al. (1950). Termite workers are not necessarily at their final instar and are, therefore, able to retain the capacity to become either a soldier or a supplementary reproductive with no wing buds (ergatoid) (Noirot, 1985; 1988). Workers still have prothoracic glands even when they reach the final instars (Thorne, 1996). Interestingly this goes against the definition of a worker, mentioned above, where they are said to diverge irreversibly from the sexual line. This is true if one considers the sexual line to be individuals with wing buds but not so if we consider them to be individuals that are able to reproduce.
Workers start to feed at the third instar in a small colony and at the forth instar in a large colony (Buchli, 1958). The workers can develop into soldiers at and after the forth moult (the most common point being at the sixth moult) and the seventh moult is the point at which development into ergatoids starts to occur (Buchli, 1951, 1958). Thompson (1917) was the first to suggest that ergatoids may be fertile workers. Ergatoids go through two moults to develop from a worker and are approximately 7 to 9 mm in length (Noirot, 1990; Thorne, 1998). The differentiation of worker to ergatoid can take ten months, although in general it is thought to be faster in larger groups (Grasse et al., 1950; Thorne et al., 1999).
2.7.3.5 Life cycle: soldiers Soldiers are varied in length, not only between species, for example, R. flavipes soldiers are 6 to 7 mm in length and R. virginicus soldiers are 4.5 to 5 mm in length, but also within species (Snyder, 1915; Buchli, 1958). Soldiers have a thorax and abdomen that resemble that of workers. They are blind and have biting mandibles for defence (Feytaud, 1912; Harris & Sands, 1965; Krishna, 1989; Serment & Tourteaux, 1991). Soldiers are thought to have a defence role, although in some cases they are rather passive and they sometimes go out with the workers on food
40 Literature Review scouting expeditions (Grasse, 1949; Thorne, 1998). They are completely dependent on the workers for food (Snyder, 1926; Harris & Sands, 1965).
Soldiers are present in higher proportions in young natural colonies. However, they are always outnumbered by the workers (Feytaud, 1920; Snyder, 1926; Feytaud, 1946). They usually make up 1 to 3 % of a Reticulitermes spp. colony population though values of less than 1 % and between 8 and 10 % have also been mentioned (Hrdy, 1961; Haverty, 1977; Haverty & Howard, 1981; Grace, 1996; Forschler & Jenkins, 1999). Weesner (1956) stated that the number of soldiers present in a colony was influenced by environmental or genetic factors in the colony from which the initial reproductives originated. However, Noirot (1990) thought that the proportion of soldiers present was dependent on the ecological niche that the termites occupy, that is to say that more soldiers are present in termite colonies that are more prone to predation due to the niche that they occupy. The formation of soldiers may also be artificially induced through the use of juvenile hormones and large numbers of soldiers are thought to be detrimental to the proper functioning of the colony (Hrdy & Krecek, 1972; Haverty & Howard, 1981).
There is no equivalent to the soldier caste in other social insects and it is the first to disappear when the colony is under threat (Feytaud, 1946; Noirot & Pasteels, 1988). Reticulitermes soldiers were previously thought to be either all males or all neutrals. However, as with all other castes, they are comprised of both sexes and have rudimentary reproductive organs (Lespes, 1856; Feytaud, 1912; Grasse, 1949). Their lack of fertility may be due to the presence of prothoracic glands (Miller, 1969). As mentioned above, soldiers develop from workers and not from a special developmental line. This is true for all Heterotermitinae (Grasse et al., 1950; Noirot, 1985; Vieau, 1996). Soldiers are formed in two steps from workers. After the first moult they form a white soldier (also called presoldier, callow soldier, pseudosoldier, soldier nymph or soldier larva), which is unpigmented and unsclerotised with a soldier-like morphology. After the second moult, a fully sclerotised soldier is formed, which will no longer moult (Grasse, 1949; Harris & Sands, 1965; Weesner, 1965; Noirot, 1985; Noirot & Pasteels, 1987, 1988; Vieau, 1991; Thorne, 1996, 1998). It was thought by some authors that soldiers developed via four to six instars (i.e. they could develop from the L2 to the W4) (Weesner, 1956; Plateaux & Clement, 1984). Noirot (1985) thought that soldiers could develop from any
41 Literature Review immature stage from the second stage upwards. Other authors suggested that soldiers could develop from nymphs (Grasse, 1949; Buchli, 1958). The white soldier stage is said to last 11 to 20 days (Weesner, 1965). This development is irreversible, and once the soldiers are formed, they are unable to regress to another caste (Thorne, 1996). They are thought to live for up to several years, but there is little information available (Noirot, 1990).
2.7.3.6 Life cycle: nymphs Nymphs are individuals with wing pads that develop into alates or secondary reproductives with wing pads (Harris & Sands, 1965). They have, to some degree, a similar role as the workers and aid in the daily running of the colony (Noirot & Pasteels, 1987). Nymphs are said to start forming 8 months after the nest is started (Grasse, 1949). Buchli (1958), however, stated that the nymphal line only developed in a R. santonensis colony when a minimum of 200 workers were present (after 18 months), and in a R. lucifugus colony when a minimum of 1000 workers were present (after four years). Weesner (1965) simply mentioned that a large colony is required for alate formation and Grasse (1949) that nymphs formed when the alates were no longer present.
Antennal segments may be used to differentiate between instars, for example, nymphs (N6) have 17 antennal segments compared with 18 in secondary reproductives and alates (Feytaud, 1912; Buchli, 1958; Ionescu, 1959). It is sometimes a problem using antennal segments to separate instars as they can often become damaged (Buchli, 1958).
The most recently postulated life cycle has seven nymphal stages, including the first two larval stages that are common to both lines (Vieau, 1994a) (Figure 2.12). After the first two larval stages (L1 & L2) the nymphs go through four stages (N3 to N6) before either forming nymphs with long wing buds (LWBN), which will develop into alates (primary reproductives), or nymphs with short wing buds, which will moult into brachypterous neotenics (Vieau, 1994a). In R. santonensis, the N3 are distinguishable from other nymphal instars, whereas the N4 to N6 cannot be separated (Vieau, 1999, 2001). In R. lucifugus, the PBNEO is distinguishable earlier and is more easily separated from the other instars (Vieau, 1999; 2001). Noirot (1988) stated that the forth instar nymphs were able to feed if they were not fed by
42 Literature Review workers but that this ability only lasted until the sixth instar. Whereas Buchli (1958) thought that only the fifth and sixth instar nymphs were able to feed themselves. The amount of work carried out by a forth instar nymph in a natural situation is believed to be negligible (Noirot & Pasteels, 1988).
The PBNEO was said (Buchli, 1958) to be a stage 7 nymph and the LWBN to be a stage 8 nymph, which then becomes an alate. This idea, which was also supported by Noirot (1985), was disputed by Vieau (1994a) who studied the evolution of the proportions of PBNEOs, LWBNs, BNEO and alates in colonies, compared the individuals' histology and looked at levels of Juvenile Hormone and ecdysteroids. This study led Vieau to conclude that PBNEOs did not originate from the LWBNs but that both these stages originated from the stage 6 nymph (Figure 2.5). The likely misinterpretation made by Buchli (1958) also seems to be repeated by Thorne (1998), whose explanation implies that the brachypterous neotenics are formed from pre-alate nymphs. Thus, it would suggest that she thought, as did Buchli (1958), that PBNEO develop into LWBN. Grasse (1949) also caused confusion because he believed that there were three forms of neotenics, some with short wing buds, some with no wing buds and, finally, some with long wing buds.
Vieau's (1994a) opinion is not a recent one. It was first mentioned by Fritz Muller in 1887 (Perez, 1895; Feytaud, 1912; Grasse, 1949) and reiterated by Grasse (1949) who described two forms of N7 nymphs (his forth instar), which were the same size but had wing buds of different lengths. Feytaud (1912) also thought this but stated the stage as being the fifth instar.
2.7.3.7 Life cycle: neotenics Neotenics, supplementary reproductives that develop from nymphs or workers whose internal development is accelerated with only a few changes in external anatomy, are distinguished from nymphs and workers by their darker pigmentation, slight sclerotisation, longer abdomen and the presence of eyes and ocelli (Weesner, 1965; Plateaux & Clement, 1984; Krishna, 1989; Serment & Tourteaux, 1991; Thorne, 1996, 1998). They have no wing development and retain many juvenile characteristics. Brachypterous neotenics in R. lucifugus are 12 mm in length (Thorne, 1998). There are conflicting theories as to whether neotenics are able to feed
43 Literature Review themselves initially and then lose this capacity or whether they are entirely unable to feed themselves (Snyder, 1925; Grasse & Chauvin, 1946).
Secondary reproductives are a common occurrence in R. santonensis and R. lucifugus (Snyder, 1925; Feytaud, 1946; Vieau, 1991, 1994b). However, Buchli (1958) was of the opinion that this caste only developed when there was a disturbance. Ergatoids seem to be rarely found in nature (Feytaud, 1946; Thorne, 1998). Brachypterous neotenics are not produced in the first few years after colony foundation (Weesner, 1965). The development of neotenics is said to take place from between 6 weeks to 4 months after a group of termites have separated from the main colony (Grasse, 1949; Thorne, 1998), though it is believed by some that in Reticulitermes spp., neotenics are formed in the presence of primary reproductives and several neotenics may be present in any one colony (Noirot, 1990; Serment & Tourteaux, 1991). Weesner (1956) did not believe that neotenics developed in the presence of primary reproductives. The process of forming new colonies by the formation of neotenics within a group of individuals that has been separated from the colony is commonly termed 'budding' (Perez, 1907; Feytaud, 1946; Harris, 1958; Harris & Sands, 1965; Plateaux & Clement, 1984; Serment & Tourteaux, 1991). The extent to which budding occurs in nature is uncertain, although this is commonly observed in the field in France (Vieau, 1999). The first person to clearly explain budding was Snyder (Snyder, 1912, 1920; Thorne, 1998; Thorne et al., 1999). The groups of termites that bud off a main colony and form a reproductive unit are termed 'satellite' colony units (Thorne et al., 1999).
Egg laying starts approximately 4 to 8 weeks after the formation of neotenics has occurred (Feytaud, 1946). Female neotenics lay eggs at a lower rate than true queens. However, they are usually present in large numbers and therefore produce much larger communities, even though their egg laying capacity is lower (Harris, 1958; Harris & Sands, 1965; Noirot, 1990; Thorne, 1996, 1998; Thorne et al., 1999). Grasse (1949) thought this to be particularly true of the genus Reticulitermes where, he said, large numbers of neotenics are found. Conversely, Feytaud (1946), Pawson and Gold (1996) and Thompson (1917) reported that neotenics had a higher fecundity than primary queens, although Thompson (1917) went on to mention that this is only at the start of colony foundation and that the primary reproductives later overtake neotenics in terms of fecundity. The male to female sex ratio can vary from
44 Literature Review
1:1 to 1:15, though Feytaud (1946) does not specify the species, it is likely to be R. lucifugus which is the one that he studied.
2.7.3.8 Life cycle: alates Alates, the primary or first form reproductives (Thorne, 1996), develop approximately two years after colony foundation (Feytaud, 1920). They are produced at particular times of the year, and there is usually only one batch formed per year although there can be more if in an urban environment due to more favourable environmental conditions, for example, higher temperatures through central heating (Feytaud, 1912; Weesner, 1956; Krishna, 1989; Noirot, 1990). A period of warmth and increased humidity is said to be required for alate development, however, not much information is available with regard to actual alate dispersion (Perez, 1907; Feytaud, 1912; Grasse, 1949; Weesner, 1956; Harris, 1958; Harris & Sands, 1965; Nutting, 1969). There is much controversy in the literature with regard to when R. santonensis and R. lucifugus swarm. In general this is between March and June (de Quatrefages, 1853; Lespes, 1856; Perez, 1895; Feytaud, 1912, 1946; Buchli, 1956; Weesner, 1965; Vieau, 1994b). An increase in the proportion of alates, compared to the proportion of neotenics, is seen in R. santonensis as colony size increases (Vieau, 1994a). After swarming, the alates shed their wings along a suture, although this may occur without a flight, and then shortly after pair and mate, each pair forming a new colony (Grasse, 1949, Krishna, 1989; Thorne, 1996). Nutting (1969) has reviewed flight and mating behaviour in great detail. The primary reproductives are initially able to feed. However, this ability is lost at the first development of workers, who take over the duty of feeding the primary reproductives (Perez, 1895; Krishna, 1989). Supplementary reproductives may be formed from alates that have their wings removed and remain in the nest. In this case, the more correct term is adultoid reproductive (Feytaud, 1912, 1946; Grasse, 1949; Harris & Sands, 1965; Thorne et al., 1996).
Alates are imagoes that are fully sclerotised, dark brown to black in colour, winged, 8 to 10 mm in length (including wings) and have compound eyes (Feytaud, 1912; Banks & Snyder, 1920; Krishna, 1989; Thorne, 1996, 1998). The male and female can be identified using the following characteristics (Lespes, 1856; Feytaud, 1912, 1946; Grasse, 1949; Weesner, 1956, 1965; Hickin, 1969):
45 Literature Review
• the female has an enlarged seventh stemite which hides the 8th and 9th sternites.
• the male has a seventh sternite that is similar to the previous stemites and also has a pair of styli on the nineth sternite.
The dealated female primary reproductive will later become physogastric (have a distended abdomen), but will not reach the proportions found in the Termitidae (Grasse, 1949; Thorne, 1996). These queens stay relatively mobile (Lespes, 1856; Feytaud, 1912; Snyder, 1926; Feytaud, 1946). Primary reproductives are thought to live for up to seven to ten years (Thome et al., 1999) and even up to 18 years in a natural environment (Feytaud, 1946) and have been seen to survive for 25 years in artificial colonies (Snyder, 1926).
Physogastric queens are said to be rare, and in R. santonensis they have not been found. However, if one considers that in an urban situation, which is mainly where R. santonensis is found, spread tends to occur via budding and thus physogastric reproductives are less likely to be found. The more commonly found reproductive form in all cases is the neotenic (Feytaud, 1912, 1951; Vieau, 1991, 1993, 1994b, 1996; Thorne, 1998; Vieau, 2001). A mix of physogastric and neotenic reproductives can be found in a colony, although Snyder (1926) thought that only one or the other were found in a given colony. As mentioned in the section on neotenics, Buchli (1958) thought that neotenics only occurred after a disturbance and were, therefore, not a common occurrence. Interestingly, over a decade earlier, Bathellier (1941) suggested that neotenics be considered a common caste. It is thought, even though swarming occurs in R. santonensis, that colony establishment by alates is never successful (Vieau, 1996) and there is controversy as to whether swarming occurs in R. flavipes colonies (Vieau, 1993; Thorne et al., 1999). The lack of swarming and thus of physogastric queens in R. santonensis has been said to be a sign that this species is in the process of establishing itself and is therefore an introduced species (Vieau, 1993; 1999). This phenomenon may be comparable to that seen in tramp species of ants where the ability for nuptial flights has been lost (Passera, 1994). Interestingly, Weesner (1965) mentioned having witnessed all three forms of reproductive types in R. flavipes in the U.S.A., which is the same case for R. lucifugus in France where both alates and neotenics are formed. However, very few data are available concerning the reproductives present in nature (Vieau, 1999, 2000). Swarming of R. hageni and R. virginicus in the U.S.A. has also been well
46 Literature Review documented (Weesner, 1970). Buchli (1958) did not think swarming to be successful in R. lucifugus. This is probably because he confused R. lucifugus with R. santonensis, which was a common mistake at that time (Vieau, 1999). Reticulitermes lucifugus in Italy, where R. santonensis does not occur, was, however, not thought to have successful swarming (Grassi & Sandias, 1893; Feytaud, 1912; Snyder, 1926).
2.7.3.9 Life cycle: pseudergates Reticulitermes species have an intermediary stage in the evolution of life cycles in termites (Noirot & Pasteels, 1988). The reason for this is that the worker line can be separated from the sexual line at the third instar. However, nymphs in the sexual line can develop into pseudergates (Harris & Sands, 1965; Noirot & Pasteels, 1988). Pseudergates are defined as individuals that have diverged from the sexual line at a late instar by undergoing a regressive moult (where some nymphal characteristics are lost) or a stationary moult (where growth occurs without changes in structures) and have stabilised to a state where they function as workers (Grasse et al., 1950; Noirot, 1985; Noirot & Pasteels, 1987, 1988; Vieau, 1991; Thorne, 1998). Pseudergates are no different from nymphs apart from the fact that they have greatly reduced or no wing buds (Noirot, 1985; Noirot & Pasteels, 1987; Krishna, 1989). They also never attain true worker morphology (Noirot, 1985). Pseudergates conserve the ability to develop into neotenics, alates or soldiers and are usually present in conjunction with workers (Grasse et al., 1950; Buchli, 1958; Thorne, 1998). Some articles have interchanged the term 'pseudergates' with the term 'worker' (e.g. Grace (1996), Pawson & Gold (1996) and Weesner (1956)). This may be due to the difficulty in differentiating between workers and pseudergates. The identification of workers cannot, therefore, be limited to the absence of wings buds and the presence of a pigmented gut (Noirot & Pasteels, 1987). Pseudergates, which are able to be distinguished from workers morphologically (Buchli, 1958) (i.e. they are an intermediatary stage between nymphs and workers and can be distinguished from workers by the fact that their pronotum is narrower than their mesonotum) and have therefore not undergone a complete regression, are said to be present in very low number in R. lucifugus colonies (4 to 5%) and are not found in R. santonensis colonies (Vieau, personal communication).
47 Literature Review
2.7.3.10 Life cycle: caste determination Grassi and Sandias (1893) thought that the developmental pathway taken depended upon environmental factors whereas Snyder (1925) believed that determination was embryonic. The exact mechanisms involved in caste development are still not known. However, some early workers believed that it was purely via feeding or due to extrinsic factors (Perez, 1907; Thompson, 1917; Hare, 1934; Grasse, 1949; Esenther, 1969). Grasse (1949) thought that there were two mechanisms involved in caste differentiation; (1) sociohormones whose formation inhibited the presence of certain castes and (2) sensoral stimuli that have been seen to have an effect on ovarian activity in worker Polistes. The latter theory is supported by insufficient arguments and requires further research according to Noirot (1990). Thompson (1917) believed that caste determination occurred in the embryo.
In general, this divergence in termites is thought to occur mainly during post- embryonic development. The expression of genes is controlled by hormones, which are in turn under the influence of complex environmental factors within the colony (Buchli, 1956; Luscher, 1960; Weesner, 1965; Noirot, 1990). One theory is a combination of inhibitory and stimulatory actions by, or mainly by, pheromones on the development of certain castes (Noirot, 1990).
2.7.4 Comparisons of life cycles
There has been a variety of possible life cycles put forward for Reticulitermes spp.. The reproductive life cycle of R. santonensis has often been confused with that of R. grassei and remains poorly understood (Vieau, 1994a). The first diagram depicting the life cycle of a Reticulitermes species that is found in the literature is that by Snyder (1915). This interpretation was based on various works, but clearly the same ideas later represented by Feytaud (1946) were based on the work of Grassi and Sandias (1893). The inclusion of 'large-headed' larvae is a distinguishing factor.
48 Literature Review
Hatches T , Egg
Figure 2.6 Life cycle of Reticulitermes spp. based on Snyder (1915) (L — larva, LB — larva with large head, EL- emergency reserve reproductive, N — nymph, WL — worker larva, NL — nymph with long wing buds, NS — nymph with short wing buds, W —worker, SL — emergency larval substitute, SN —emergency nymphal substitute, BN — brachypterous neotenic, PS — pre-soldier, S — soldier).
49 Literature Review
IMAGO T N ()
(PS) A
N
Hatches 4
Egg
Figure 2.7 Life cycle of Reticulitermes spp. based on Hare (1934).
The complex diagram shown in Figure 2.6 is sharply contrasted by the simplicity of Figure 2.7. Hare (1934) (Figure 2.7) thought that there were three or four worker instars before both the 'adult worker' and the pre-soldier. The adult reproductive is
50 Literature Review said to go through five nymphal instars prior to development into an alate. In all cases there are two larval instars before the split into the two lines and this is also the case in Coptotermes spp. (Roisin & Lenz, 1999). Interestingly Hare (1934) does not mention the presence of neotenics and does not describe their development path.
Hatches 11/4 Egg
Figure 2.8 Life cycle of Reticulitermes spp. based on the description by Feytaud (1946), taken from Grassi and Sandias (1893) (E - ergatoid, NN — nymph with no wing buds).
In contrast, Grassi and Sandias (1893), Snyder (1915) (Figure 2.6) and later Feytaud (1946) (Figure 2.8) thought that there was only one undifferentiated larval instar and that the second larval instars could be differentiated by the size of their head.
51 Literature Review
Hatches
{ Egg
Figure 2.9 Life cycle of Reticulitermes spp. based on the description by Grassi (1893) and later by Grasse (1949) (LS — larva with small head, NE — neotenic larva).
This idea was reiterated by Grasse (1949) (Figure 2.9). The larvae with small heads were thought to develop down the sexual line and those with large heads down the 'neutral' line. The split between the neotenic line and the imago line occurred at the second moult and there were two types of neotenics forming from nymphs with no wing buds, which were probably workers and nymphs with short wing buds. The alate was said to develop from nymphs with long wing buds. The only real difference with the present theory on development (Vieau, 1991, 1994a) (Figure 2.5)
52 Literature Review is that the ergatoid forms from the nymphal line in Feytaud's (1946) explanation (Figure 2.8), whereas they are presently thought to form from workers.
IMAGO /1/V
N L OR >. OR NS
OR N
Hatches
Egg
Figure 2.10 Ontogenic potential of Reticulitermes spp. based on the description by Buchli (1958) (P — pseudergate).
53 Literature Review
There are also discrepancies with the number of moults that are said to occur but, as stated above, it is very difficult to separate out each moult, even today, so these anomalies are not surprising. Another difference of Feytaud's explanation (Figure 2.8) compared with that of Hare (1934) (Figure 2.7) and Buchli (1958) (Figure 2.10), is that the soldiers were shown to develop from second instar larvae (called large headed larvae) and a split occurred at this point into a soldier line. This is also seen in Snyder's (1915) interpretation (Figure 2.6).
Hatches 4
Egg
Figure 2.11 Life cycle of Reticulitermes spp. based on the description by Vieau (1991, 1994a, personal communication)
54 Literature Review
Feytaud (1946) and Grasse (1949) (Figure 2.8 and Figure 2.9) confused the sexual and neutral line, and also showed a cross-over from what they considered to be the sexual line (individuals developing from the small headed larvae) and the worker line (individuals developing from the big headed larvae). The present theory (Figure 2.11) shows the ergatoids or third form reproductives developing from the worker line, whereas in Figure 2.9 all three types of reproductives develop from a nymph or small-headed larva. Grasse (1949) also assumed that soldiers came from the sexual line. However, the current theory (Figure 2.11) is that soldiers originate from the worker line (Noirot, 1985; Vieau, 1991). Although Buchli (1958) (Figure 2.10) also suggested the possibility of soldiers developing from the sexual line, Bathellier (1941) thought that the small headed larvae depicted in Figure 2.9 were nymphs, which, according to the plates, seems correct and would be in agreement with the present-day theory (Figure 2.12). However, the discrepancy with regard to the ergatoid still holds true.
IWO
Figure 2.12 Life cycle of Reticulitermes spp. based on the description by Noirot (1985).
55 Literature Review
Buchli's (1958) interpretation (Figure 2.10) was somewhat complex, his work described the ontogenetic 'potential' of R. santonensis and R. lucifugus and this included various developmental pathways which were not thought to be found in a natural situation (Noirot, 1985; Noirot & Pasteels, 1987; Vieau, 1994b). His was the first life cycle where the presence of pseudergates was shown. Brachypterous neotenics were shown to develop from either pseudergates or from later nymphal instars.
Noirot (1985) (Figure 2.12) later simplified Buchli's life cycle, but incorporated neither the pseudergates nor the neotenics, which were the least understood castes in terms of their developmental path. The life cycle represented in Figure 2.12 was virtually identical to that of Reticulitermes speratus by Takematsu (1992), however, the pre-soldiers were seen to develop from one stage later (i.e. W4 and W5 instead of W3 and W4).
The current theory (Figure 2.11), as explained in the life cycle section, supports the split at the sixth instar of the sexual line. The LWBN develop into alates and the PBNEO develop into brachypterous neotenics (i.e. brachypterous neotenics). The ergatoids develop from the sixth or seventh instar of the worker line and the soldiers develop from the sixth instar in the worker line. Development of pseudergates is thought to follow the path described by Buchli (1958). Figures 2.5 and 2.11 show the same theory. The difference is that Figure 2.5 provides detailed drawings of the different stages and is slightly simplified for clarity and Figure 2.11 provides a diagrammatic representation for easy comparison with the other proposed theories.
2.7.5 Conclusion
Overall, there have been many changes with regard to the interpretation of the life cycle of this genus (Figure 2.6 to Figure 2.12). Today, the life cycle that is often referenced is that of Buchli (1958) although as mentioned earlier in this review, Buchli's methodology showed some flaws. However, it was Snyder (1915) who probably provided the first diagrammatic representation, and Lespes (1856) who provided the first insight into the life cycle of this genus. The most recent update to the life cycle is that of Vieau (1994a). The life cycle depicted in Figure 2.5 and Figure 2.11 is the presently accepted life cycle in the French scientific community
56 Literature Review for R. lucifugus and R. santonensis as these are the species on which the studies were completed.
Other investigators have concentrated on the life cycle but did not further explain the number of stages involved (Lespes, 1856; Noirot, 1985; Serment & Tourteaux, 1991; Thorne et al., 1999). This causes difficulty when trying to compare different ideas, as there is considerable inconsistency, particularly concerning the number of moults taken to develop into workers and soldiers. The nymphal line is, however, better understood.
There is much that needs to be done in the area of Reticulitermes biology and with the use of modern techniques, especially biochemical and genetic, further clarification may be accomplished. In particular, the inconsistency with sexual reproductives developing from the supposedly 'neutral' worker line needs to be clarified. Perhaps the term 'neutral' line should no longer be considered and the term pseudo-worker line used instead. There are also known differences between Reticulitermes spp., yet they are still assumed to have the same life cycles. The literature is made even more difficult to interpret as statements are often made about the genus rather than specifying the species. If anything, the discrepancies seen in the literature reiterate the plasticity of Reticulitermes spp., as it is probably the basis for all this confusion.
57 Chapter 3
General materials and methods
3.1 Termite species and strains
The Reticulitermes santonensis BRE population was obtained from the Building Research Establishment (BRE), Watford, UK and established at Silwood Park on the 1St October 1998. This strain was originally collected from Saintes, France in the late 1960s. The Reticulitermes santonensis France population was collected by I. Paulmier of the Centre Technique du Bois et de l'Ameublement (CTBA) from St. Trojan les Bain, L'ile D'Oleron, France in 2000 to 2002. The Reticulitermes grassei UK population was collected by R. H. J. Verkerk from Saunton, Devon on the 6th and 7th August 1998. The Reticulitermes grassei France population was collected by I. Paulmier of the CTBA from the Foret de Lacanau in 2000 to 2002.
3.2 Insect culture
Termite colonies were kept in large stacking boxes (80 x 60 x 42.5 cm) (Mailbox Mouldings International Ltd, Stalybridge, UK), which had been filled to a third of their volume with a 50:50 mix of John Inns No. 2 and sharp sand (William Sinclair Horticulture Ltd, Lincoln, UK). Plastic dishes (10 cm dia. x 4 cm) were part filled with distilled water and placed on the surface of the soil to maintain an optimum Relative Humidity (RH) (30 to 50 %). Slats of untreated, oven dried, Scots pine (Pinus sylvestris) (30 x 7.5 x 1 cm) obtained from BRE provided a food source. The cultures were kept in a controlled environmental room at 25 ± 2 °C and 90 ± 5 % RH. The boxes were covered with lids to keep the colonies in darkness.
3.3 Experiments using vermiculite
Vermiculite (3-5 mm, Dupre Vermiculite, Hertford, UK) was washed with tap water in a sieve (1.18 mm) to remove any fine particles. Distilled water was used for the final rinse. The vermiculite was left to air-dry for 10 to 12 hours at ambient temperature. Approximately 50 g of vermiculite was placed into a 500 ml glass jar,
58 General materials and methods equivalent to a depth of approximately 2 cm. The jars were covered with aluminium foil, which had 1 mm holes punched in it to allow aeration. Specific details of methods are provided in each experimental chapter.
3.1 Wood block quality The wood block and square wafer quality conformed to the European standard EN118: 1990. Thus blocks were from sound sapwood, with straight grains and without knots. The blocks also had a low resin content, which was specific to the Pinus species.
3.4 Data analysis The data analysis was performed either by doing regressions, analyses of variance (ANOVAs) or analyses of covaraince (ANCOVAs) depending on the type of explanatory or response variables (using S Plus 2000 Professional Release 1 (1998- 1999, MathSoft, Inc.)). The method of handling proportion data and count data was slightly different compared to the standard analyses.
The proportion survival data were analysed using generalised linear models (GLM) with binomial errors and the link used was a logit in order to account for boundedness of the fitted values.
The count data (e.g. Chapter 6) were analysed using GLMs, however, the error structure used was Poisson and the link used was a log, which meant that the fitted values could only be positive.
Further details of the methods used above may be found in Crawley (2002).
59 Chapter 4
Minimum number for successful termite establishment
4.1 Background The number of termites is known to have an effect on survivorship of individuals under experimental conditions (Forschler, 1996). However, very little literature is available, with regard to R. grassei and R. santonensis. Some authors have mentioned that only 50 individuals of Reticulitermes species are required for colony establishment (Jacquiot, 1955; Serment & Tourteaux, 1991). While Hrdy (1961) believed that more than fifty individuals of R. lucifugus were required to start a colony and that even with 250 individuals, survival was very low. Sen-Sarma (1994) who reviewed laboratory tests on a number Reticulitermes species stated 200 - 500 workers to be the optimum number. In contrast, Becker (1969b) mentioned having initiated colonies with less than ten larvae and workers. The arena and substrate used in Hrdy's (1961) study were not mentioned although these factors may be of great importance for termite survival (Lenz & Williams, 1980).
Haverty (1979) found that Reticulitermes species took longer than Coptotermes formosanus to reach a caste ratio equilibrium, possibly because of the former's slow rate of differentiation. This may in turn mean that the rate of establishment under ideal conditions may be slower for Reticulitermes species relative to other termite genera. The number of individuals present in natural colonies has also been thought to dictate the number of individuals that should be used in a laboratory colony (Sen- Sarma, 1994). This may be an important fact to consider during experimental design to enable easier comparison between laboratory tests and the field. In Reticulitermes species natural colonies are said to comprise thousands of individuals (Nutting, 1969).
In the present study, experiments have been conducted primarily using workers although in some cases pseudergates may have been included due to the difficulty in distinguishing between these stages.
60 Minimum number for successful termite establishment
The aim of this study was to assess the minimum number of individuals for survival of a group of R. santonensis or R. grassei workers / pseudergates under favourable conditions in the UK. This is particularly important when posing the question of how many termites would need to be introduced accidentally for a viable colony to form. The specific objectives of the experiments were to determine the relative importance of termite number, substrate type and substrate volume in the laboratory experiments presented here, identify differences between the populations studied, estimate the minimum number of termites required for establishment and finally to attempt to draw comparisons between laboratory experiments and the field.
4.2 Materials and Metho ds
The laboratory experiments were conducted in a controlled environment room at 25 ± 2 °C and 90 % RH and were run for 12 weeks. Distilled water was added as required (to keep the RH between 50 to 70 %) and tunnels that were formed were broken to prevent escapees. The individuals used were workers or pseudergates of a uniform stage (approximately W5 - see Chapter 2). Their morphology (Buchli, 1958) and their stomach contents were used to judge whether or not they were functional workers.
4.2.1 Survival of termites o n Scots pine
Five, 25, 125, 250 or 500 individuals of R. santonensis (BRE or France) or R. grassei (UK or France) were placed into 500 ml glass jars containing vermiculite (Section 3.3). Five 3 cm squares of Scots pine, conditioned in a controlled environment room (see above, 4.2) for 3 days, with a thickness of 0.2 cm were used as a food source. The termites were handled with a moist paintbrush.
The experiment was repeated to look at the effect at low densities, using 5, 10, 15, 20 and 25 individuals of R. santonensis (BRE or France) or R. grassei (UK or France). Five Scots pine squares were added to each of the jars.
In all cases each treatment was repeated six times. Survival was determined by the number of individuals present at the end of the experiment as all dead individuals were cannibalised.
Regression analysis and ANCOVA were used for data analysis.
61 Minimum number for successful termite establishment
4.2.2 Survival of termites in the field A field experiment was established at Saint Trojan les Bains, L'ile D'Oleron, France (45° 50' 34" N, 1° 12' 42" W) on 6th April 2000 in conjunction with the CTBA. This experiment was run for two years.
4.2.2.1 Field Site The field site used was in a maritime forest, which was composed largely of two tree species: Pinus pinaster and Quercus ilex. The forest had been badly affected by the storms of December 1999. The fact that there were many fallen trees in the forest provided ideal conditions for termites in the area due to the large food resources that were available. The area used was 5 x 5 m and was on a south-east facing 30° slope.
4.2.2.2 Containers Sixteen 50 1 plastic buckets with lids (Fenton Packaging Ltd, Hemel Hempstead, UK) were used as containers. The base of each bucket had a 30.6 cm diameter circle removed and discarded (Figure 4.2 (b)). The circle was then chiselled to give a sharp edge (Figure 4.2). The bottom of each bucket was cut at a distance of 3.7 cm from the base (Figure 4.1 (b)). This was then inverted and placed back within the sides of the bucket and secured with the aid of 8 rivets to provide a physical barrier to prevent termite escape (Figure 4.1 (c) and Figure 4.2).
A 34.6 cm diameter circle was cut out of the lid. A square (57 cm length sides) of 490/500 micron polyamide mesh (Heath Filtration Ltd, Stoke-on-Trent, UK) was placed on the top of the container before replacing the lid. Nylon mesh has previously been used in experiments looking at leaf litter decomposition (Crossley & Hoglund, 1962; Yamashita & Takeda, 1998). The container was then inverted so that the end with the lid was used as the base. Details of the method of removal of the containers from the field site are provided in the Appendix, Figures A.2 and A.3.
62 Minimum number for successful termite establishment
Figure 4.1 Detail of steps to construct field equipment (see text for description). (a) - 50 1 bucket (b) - bucket inverted, circle removed from lid (34.6 cm), base removed (cut made at 3.7 cm from base) and circle (30.6 cm) cut from base (c) base inverted and placed back in bucket, nylon mesh placed on top of bucket and lid pushed over mesh.
Figure 4.2 Detail of top of container showing cross section through side of container and 'lip' to provide a physical barrier to prevent termite escape.
63 Minimum number for successful termite establishment
4.2.2.3 Field Site Layout Fifty centimetre diameter holes were dug to a depth of 28 cm in a 4 x 4 layout with one metre spacing (Figure 4.3). The soil was removed and placed into heavy-duty plastic bags and frozen at —20 °C for 27 h to ensure any endogenous termites were killed. The containers were placed in the ground (lid side down) and filled with 30 1 of the thawed soil. Fifteen 1 x 7 x 20 cm slats of untreated, oven dried Scots pine were placed side by side on the 20 cm edge on the soil surface. The termites were placed on top of the slats. A 60 cm square of galvanised garden wire netting with a mesh size of 25 mm (Wickes Building Supplies, Harrow, UK) was placed over the top of the container and secured with U-shaped brackets (aluminium extrusion) with a hole drilled through one side to insert a no. 8 self tapper screw. The wire netting was to prevent any large debris from trees from falling into the container and thus forming a means of escape for the termites.
Figure 4.3 Layout of part of field site on L'ile D'Oleron (see Appendix, Figure A.1 for comparison on termination of experiment).
4.2.2.4 Termites The termites used were R. santonensis France, collected from the experimental site and R. grassei France, collected from the Forest of Lacanau and Carcans. The
64 Minimum number for successful termite establishment termites were counted using a suction pump. The number of termites used for the experiment was 250 workers / pseudergates with 2 nymphs and 2 soldiers or 1000 workers / pseudergates with 8 nymphs and 8 soldiers per container, respectively. Each treatment was replicated four times. The design used was a stratified complete randomised block (Figure 4.4), taking the slope into account.
ANCOVA was used for the data analysis.
R.s. F 250 R.g. F 1000 R.g. F 250 R.s. F 1000
R.s. F 1000 R.g. F 1000 R.s. F 250 R.g. F 250
CD R.s. F 250 R.s. F 1000 R.g. F 1000 R.g. F 250
R.g. F 1000 R.s. F 250 R.s. F 1000 R.g. F 250
Figure 4.4 Experimental design layout of field trial using different initial termite numbers of Reticulitermes santonensis France (R.s. F) and Reticulitermes grassei France (R.g. F).
4.2.3 Survival of termites in different substrate types
One hundred and fifty or 250 individuals of R. santonensis (BRE & F) and R. grassei (UK & F), respectively, were placed in 500 ml jars containing either vermiculite (Section 3.3) or 150 g of Supamix silver sand (Pioneer Supamix, Nuneaton, UK) moistened with 30 ml of distilled water. The different weights of the media were to ensure the same volume, i.e. 2 cm depth of substrate in the jars. Five 3 cm Scots pine squares, 0.2 cm thick, were added as a food source to each jar. Each treatment was repeated six times. The termites were handled using a paintbrush.
ANCOVA was used for data analysis.
65 Minimum number for successful termite establishment
4.2.4 Survival of termites in different substrate volumes
One hundred and fifty or 250 individuals were placed in 500 ml jars containing the standard 50g volume of vermiculite (Section 3.3) or 100 g. Five 3 cm Scots pine squares, 0.2 cm thick were added to each jar. Each treatment was repeated six times. The termites were handled using a paintbrush.
ANCOVA was used for the data analysis.
4.3 Results
4.3.1 Survival of termites o n Scots pine
An initial termite number of 125 gave the greatest proportion survival except for the R. grassei France population (Table 4.1). In each case the standard error of the mean number surviving increased with increased initial termite number, which is what would be expected as the variation is dependent on the number of termites present.
Table 4.1 Survival of Reticulitermes santonensis BRE (R.s. BRE), Reticulitermes grassei U.K. (R.g. UK), Reticulitermes santonensis France (R.s. F) and Reticulitermes grassei France (R.g. F) workers / pseudergates after 12 weeks as a function of initial termite number.
Initial Mean number surviving ± S.E. (mean proportion surviving )1 termite R.s. BRE R.g. UK R.s. F. R.g. F. number
5 0 1.2 ± 0.8 (0.23) 0.3 ± 0.7 (0.23) 1.2 ± 0.3 (0.07)
25 9.3 ± 3 (0.37) 15.3 ± 1.5 (0.61) 8.3 ± 3 (0.57) 14.2 ± 3.7 (0.33)
125 67 ± 10 (0.53) 100 ± 3 (0.81) 66 ± 6 (0.63) 79 ± 21 (0.53)
250 124 ± 19 (0.49) 198 ± 5 (0.79) 131 ± 10 (0.53) 134 ± 42 (0.34)
500 152 ± 60 (0.30) 240 ± 75 (0.48) 196 ± 11 (0.39) 271 ± 88 (0.54)
n = 6
A logistic regression (which was found to be the minimum adequate model) showed
that initial termite number (F1, 118 = 7.28, p<0.01) had a significant effect on the survival (Figure 4.5). The interaction between species and initial termite number was
66
Minimum number for successful termite establishment
not significant (F 1, 141 = 0.63, p = 0.43). However, the interaction between species
and origin was significant (F1, 115 = 10.38, p<0.01).
• • • ■ co. Q s a ...Q. . . s ....s . • .... a
l 0 • ...... iva CO ...... rv 0 — —4--- • 0 ...... 1'4" R.g. UK su n • -----.0 ,.. R.s.F io t ❑ or 0 ✓ — ■ • ❑ 0 R.g.F Prop R.s. BRE 0
0
O
O • ■ ■ ■ ■
0 100 200 300 400 500 Initial termite number
Figure 4.5 Proportion of Reticulitermes santonensis BRE (R.s. BRE) (0) Reticulitermes santonensis France (R.s. F) (o), Reticulitermes grassei UK (R.g. UK) (•) and Reticulitermes grassei France (R.g. F) (■) workers / pseudergates surviving after 12 weeks as a function of initial termite number. Fitted values were back transformed from the linear model; y= 0.207 - 0.0018x (R.s. BRE), y = 0.207 - 0.0018x (R.s. F), y = 1.132 - 0.0018x (R.g. UK), y = 0.433 - 0.0018 (R.g. F). Where S. E. = 0.04, 8e-5 (R.s. BRE); 0.04, 8e-5 (R.s. F); 0.043, 8e-5 (R.g. UK) and 0.039, 8e-5 (R.g. F) respectively.
The results from the experiment with five to 25 individuals of both R. santonensis BRE and R. grassei UK are shown in Table 4.2 and Figure 4.6. An attempt was made to repeat the experiment using the French populations, however, this failed. The reason for this failure is unknown.
67
Minimum number for successful termite establishment
The proportion survival was greatest at an initial termite number of 20 for R. santonensis BRE and 25 for R. grassei UK. The lowest proportion survival in both cases was five individuals. Interestingly, the standard error did not always increase with initial termite number.
Table 4.2 Survival of Reticulitermes santonensis BRE and Reticulitermes grassei UK workers / pseudergates after 12 weeks as a function of survival.
Initial Mean number surviving ± S.E. termite (mean proportion surviving)1 number R. santonensis BRE R. grassei UK
5 1.2 ± 0.8 (0.24) 1.2 ± 0.8 (0.24)
10 4 ± 1.7 (0.40) 2.8 ±1.6 (0.28)
15 8.3 ± 1.7 (0.55) 7.3 ± 0.6 (0.49)
20 13.3 ± 0.8 (0.67) 8.5 ± 2.3 (0.43)
25 14 ± 2.9 (0.56) 15.3 ± 1.5 (0.61)
= 6
The results as presented in Figure 4.6 show an increase in proportion survival with increased initial termite number.
The analysis showed that initial termite number had a significant effect on termite
survival (F1, 58 = 9,611, p < 0.01), but that species had no significant effect (F1, 52 = 1.465, p = 0.231).
68 Minimum number for successful termite establishment
l CO _
iva A v sur n io
t ..... r 0 o ...... ❑ Prop ......
...... ❑
("! O 0
O O 0
5 10 15 20 25 Initial termite number
Figure 4.6 Proportion of Reticulitermes grassei UK (o) and Reticulitermes santonensis BRE (A) workers / pseudergates surviving after 12 weeks against the initial number of termites at t = 0. Fitted values were back transformed from the linear model; y= -1.161 + 0.0647x. Where S.E. = 0.22 (intercept), 0.011 (slope).
4.3.2 Survival of termites in the field
Preliminary observations were taken one year after establishment of the trial. Termites were seen in all containers that had received an initial 1000 workers. Three out of four R. grassei France replicates with 250 workers were also active but only one out of four of the R. santonensis France replicates with 250 workers showed activity.
Both species (F1, 4 = 9.11, p < 0.01) and initial termite number (F1, 13 = 8.088, p < 0.05) had a significant effect on survival after two years (Figure 4.7). Reticulitermes grassei France showed a greater mean proportion survival (0.14) at 1000 initial number compared to R. santonensis France (0.03).
69
Minimum number for successful termite establishment
0.25 R.g. F
0.2 -a
c en 0.15 c O
c c R.s. F w 0.05 R.g. F I R.s. F 0 1 I 250 1000 Initial worker number
Figure 4.7 Mean proportion survival (± S. E.) of Reticulitermes grassei France (R.g. F) and Reticulitermes santonensis France (R.s. F) workers / pseudergates in the field after 2 years as a function of initial worker number; n = 4.
4.3.3 Survival of termites in different substrate types
The experiment with R. santonensis BRE was the only one to be considered further (Figure 4.8); experiments with the other populations showed too high mortality.
There was no significant interaction between density and substrate type (Fi, 20 = 1.81,
p = 0.19) and neither initial termite number (F1, 22 = 0.032, p = 0.86) nor substrate type (F1,21= 0.45, p = 0.51) had a significant effect on survival.
70 Minimum number for successful termite establishment
0.6
0.5 l iva v 0.4 - sur ion t
or 0.3 - rop p
n 0.2 - Mea
0.1
Sand Vermiculite Substrate type
Figure 4.8 Mean proportion survival (± S. E.) of Reticulitermes santonensis BRE workers / pseudergates on two different substrates, sand and vermiculite, from different initial numbers of 125 (striped) and 250 (dotted) individuals after a period of 12 weeks; n = 6.
4.3.4 Survival of termites in different substrate volumes
Neither initial termite number nor substrate volume had a significant effect on termite survival (Figure 4.9; F1, 90 = 0.44, p = 0.51; F1, 91= 0.51, p = 0.48). The interaction between origin (i.e. France, BRE or UK) and species was, however, significant (F1, 92 = 8.99, p < 0.01).
71 Minimum number for successful termite establishment
1 0.9 - (7; 0.8 - 0.7 - c 0.6 - o I I c.a 0.5 2 0-0.4 • 0.3 0.2 0.1
0 125 250 125 250 125 250 125 25U R.s. BRE R.s. F R.g. UK R.g. F Initial termite number
Figure 4.9 Mean proportion survival (± S. E.) of Reticulitermes santonensis BRE (R.s. BRE), Reticulitermes santonensis France (R.s. F), Reticulitermes grassei UK (R.g. UK) and Reticulitermes grassei France (R.g. F) workers / pseudergates after 12 weeks at low (white) and high (grey) substrate volumes against initial termite numbers of 125 and 250 individuals at t=0.
4.4 Discussion
4.4.1 Survival of termites o n Scots pine The results for the first experiment, presented in Table 4.1 seem to contradict the logistic regression shown in Figure 4.5 as the data in the table suggest an initial increase in survival with increased initial termite number followed by a levelling off or decrease in survival. The simplest adequate logistic model was linear and thus a quadratic model (as was suggested by the data presented in Table 4.1) did not best describe the data. This may have been due to the high number of zero values, which may have been influential. The analysis, therefore, showed a downward trend in survival from five to 500 termites per container for both species and origin. However, in the second experiment with lower termite numbers (5 - 25) there was a positive trend. Grasse &
72 Minimum number for successful termite establishment
Chauvin (1946) also found that survival increased with increased number at low numbers (1 - 10). This positive dependence on initial termite number, at low initial number only, is called an Allee effect (Crawley, 1998) and is particularly interesting when compared to a study by Lenz and Williams (1980), which showed low termite numbers of Coptotermes and Nasutitermes species to give low survival.
The total mortality observed in the first experiment for R. santonensis BRE at the lowest initial termite number (5) may be explained by the fact that the termite numbers were too few to combat colonisation by fungi or other pathogens (Becker, 1969b) or to maintain optimum humidity (Gay et al., 1955). Alternatively, five individuals may be too few to form a functional colony. While proportion survival was higher in the second experiment, it may be that these small 'colonies' may be less vigorous and therefore less able to establish (Lenz & Williams, 1980). Termites are social insects and require the presence of several other individuals to be able to thrive (Grasse & Chauvin, 1946; Gay et al., 1955).
The decrease in survival at the highest initial termite numbers in the first experiment may be explained by over-crowding and hence high levels of cannibalism due to limited food resources or simply by a lack of resources such as moisture. Crowding in small containers is said to offset the advantage of a larger group (Light &lg, 1945). The number of termites used in the experiments was a compromise between available resources and what was judged to be a reasonable colony size. Density effects may be specific to the experimental methods used, e.g. container size and matrix used (Lenz & Williams, 1980). Thus, the same must be true in a natural situation and experiments performed to date have not used volumes large enough to make a suitable comparison to the field situation.
The significance of the interaction term in the analysis of the first experiment was that survival was affected differently by the two species depending on their origin. Reticulitermes grassei UK showed a greater level of survival compared to R. santonensis BRE yet in the French strains this was inverted, R. santonensis France showing the greater survival. The latter is in agreement with previous observations that R. santonensis is the more robust species (Feytaud, 1946; Grasse, 1954; Jacquiot, 1956; Becker, 1969a) although these were purely field observations and had no experimental proof This could, however, suggest that the population of R. grassei that established in Devon was a more robust population and thus was more
73 Minimum number for successful termite establishment easily able to establish or alternatively, has developed a greater robustness during its development in the UK. Perhaps both factors could play a role in its increased ability to survive.
4.4.2 Survival of termites in the field
The mean proportion survival in each case was significantly higher in R. grassei France compared with R. santonensis France. There was also a significant effect of initial worker number, yet, compared to the results from the laboratory experiment, this effect was opposite as increased initial worker number gave an increase in survival. This may suggest that laboratory experiments may not be ideal in judging termite survival. The number of termites required to establish a colony may be higher due to the increased stresses imposed, for example, by climatic factors.
4.4.3 Survival of termites in different substrate types
There was no difference between the two substrates. Thus, both can be considered to be equally suitable. It was, however, decided that vermiculite be used in the experiments as it retains high humidity for a longer period of time, which is beneficial for termites (Howick, 1975; Haverty, 1979; Lenz & Williams, 1980; Lenz et al., 1982). Sand has also been said to collapse easily and thus pose a risk of causing harm to termites (Becker, 1969a). The only problem with using vermiculite is that it is difficult and thus time consuming to separate the termites from this substrate (Haverty, 1979; Esenther, 1980). Fontainebleau sand has been and still is used by researchers in France (Grasse & Chauvin, 1946). Other substrates that have been used are sawdust and a mixture of sawdust and agar (Light & Weesner, 1947; Weesner, 1956; Smith et al., 1969), these are not ideal as they provide a source of nutrition and cannot be used for consumption experiments (Howick, 1975). Thus, soil and humus, which have been used in the past (Becker, 1969a), are not recommended as substrates. The probable reason for their use is that researchers may have thought Reticulitermes species to require soil to survive, whereas this is not the case (Weesner, 1965; Forschler, 1998b).
74 Minimum number for successful termite establishment
4.4.4 Survival of termites in different substrate volumes
Termite survival was affected in different ways in the two species depending on their origin. The results also showed that neither substrate volume nor initial termite number had a significant effect on survival. Therefore, using a higher substrate volume in the experiments was not justified. These results contradict Lenz and Williams (1980) who showed that termite survival decreased with an increase in the amount of space provided. Conversely Esenther (1980) found that living space density did not affect termite performance. However, Esenther (1980) did not use a substrate in his experiments and thus calculated surface area from the area of the base of the container and the area where termites had formed 'mud' with their faecal material. The expected result for this experiment was that volume would have an effect due to the greater difficulty of termites to control their environment. Indeed, Butterworth et al. (1958) observed a species of Cryptotermes reduce the size of its environment by forming walls of frass.
4.5 Conclusion
The experiments presented here had several aims. Firstly to investigate the minimum termite number for colony survival and establishment. Secondly to verify whether factors such as substrate type which had previously been mentioned as being important (Forschler, 1996) and volume had an effect on survival. In terms of the latter aim, both were found to have no effect, which is in contrast to a previous study that showed that volume had an effect on survival (Lenz & Williams, 1980).
Figure 4.10 shows how container size and volume fill interact to have an effect on moisture, which in turn has an effect on survival. In the laboratory experiments presented here the container size was kept constant and the volume fill was shown to have no effect (Section 4.3.4). This difference may be due to the species as Lenz and Williams (1980) used Coptotermes and Nasutitermes species.
75 Minimum number for successful termite establishment
Figure 4.10 Flow chart showing which factors have an effect on termite survival and how they interact (taken from Lenz (1980)).
Laboratory trials are important as they allow experiments to be performed under controlled conditions and allow for easy comparison. However, the results from the present experiments showed an opposite effect of initial termite number on survival between experiments in the field compared to laboratory experiments. This may be an important observation as survival experiments in the past have only been conducted in a laboratory environment without cross-reference to the field situation.
In the present field experiment the initial number of termites was known, whereas in many experiments, namely ones using mark-release-recapture, the total number of termites present is unknown. Mark-release-recapture has also been considered to be rather inaccurate especially regarding the validity of the assumptions (Esenther, 1980; Thorne et al., 1996). In addition, Reticulitermes species are considered to be particularly sensitive to disturbance (Paulmier et aL, 1997), a problem that may be more prevalent in mark-release-recapture studies than in the field work presented here. The design of the field work equipment of the present study may be interesting for use on larger scale survival experiments. The materials used have proved to be able to withstand external environmental conditions for two years without degrading.
76 Chapter 5
Studies on termite consumption
5.1 Background
The Rhinotermitidae feed on various types of material containing cellulose, including wood, paper and textiles, fruit, seeds, dried lichen, roots and stems of live plants, such as vegetables (Chaine, 1910; Feytaud, 1911; Lesne, 1923; Feytaud, 1946; Grasse, 1949; de Feytaud, 1955; de Feytaud, 1959; Harris & Sands, 1965; Thorne, 1998). Wood is the preferred source and limited availability of this source makes it more likely for termites to feed on other cellulose sources. They may even destroy other materials (e.g. cement or metal) during their search for food (Serment & Tourteaux, 1991; Forschler, 1998b). Reticulitermes santonensis and R. lucifugus tend to prefer wood that has been partially decomposed by fungi and is moist and soft (Feytaud, 1946; Grasse, 1949; de Feytaud, 1955; Harris & Sands, 1965; Becker, 1969a, 1970; Serment & Tourteaux, 1991; Forschler, 1998b). This may be because some fungi are thought to produce attractants and even extend termite survival (Lund, 1962).
Termites usually prefer softwood compared with hardwood species (Becker, 1969a), in fact, some hardwoods are naturally termite resistant (Carter & Dell, 1981; Serment & Tourteaux, 1991; Forschler, 1998b). This is true for many tropical species of wood (Serment & Tourteaux, 1991; Grace & Yates, 1999). This resistance only applies to the heartwood as the sapwood of nearly all tree species is susceptible to attack by termites (Kennedy et al., 1994; Grace & Yates, 1999). Various chemicals within the wood have been found to be important in affecting termite attraction or repulsion (Grace, 1997). The repulsion of termites to certain pine species may be due to the high flavinoid content (Mannesmann, 1973; Kennedy et aL, 1994). The presence of silica makes wood very difficult to work with but wood species with this characteristic would be ideal for use in construction due to their hardness and therefore lack of susceptibility to termite attack (Jacquiot, 1956).
77 Studies on termite consumption
Some of the wood types that Reticulitermes species tend to prefer are birch, poplar, beech, apple, pear and certain tropical hardwoods. All European forest species, such as oak, tamarind and elder, elm, and other trees such as magnolias, acacias, poplar and ash can be attacked by R. santonensis and R. lucifugus (Lespes, 1856; Chaine, 1910; Becker, 1969a; Serment & Tourteaux, 1991). Pine is the most commonly infested under natural conditions (Clement, 1981). In France, R. santonensis has been seen to attack fruit trees but in a forest environment the more attractive species seem to be evergreen oak (Quercus ilex) and Maritime pine (Pinus pinaster) (Vieau, 1991). The latter is thought to be particularly attractive to Reticulitermes species and it is also the natural habitat of R. lucifugus (Perez, 1907). Oak is thought to be less attractive to Reticulitermes species compared to Pinus species (Ionescu, 1959; Cals- Uscati & de Frescheville, 1963). Reticulitermes lucifugus has also been seen to feed on vines, carrots as well as live plane trees (Feytaud, 1946; de Feytaud, 1955; Ionescu, 1959; Cals-Uscati & de Frescheville, 1963; Plateaux & Clement, 1984). In fact, Grassi and Sandias (1893) stated that R. lucifugus fed mainly on plants and rarely on wood. Conversely, de Feytaud (1959) said that it was R. santonensis that fed on live plants. Reticulitermes flavipes is also reported to feed on live trees (Grace, 1997).
5.1.1 Consumption
There are no standard laboratory tests to determine the consumption rate of termites because of the large number of variables that are involved (Becker, 1969b; Edwards & Mill, 1986). The suggested minimum time period for consumption experiments is 4 weeks to cancel out the effect of aperiodic feeding (Howick, 1975) but the presently accepted test period is 8 weeks, which is the interval used in the present study (Lenz & Williams, 1980; Lenz et al., 1982).
Consumption in laboratory experiments is affected by behaviour, survivorship or termite biomass, and aperiodic feeding (Becker, 1969b; Haverty & Nutting, 1974; Howick, 1975; Su & La Fage, 1984; Forschler, 1996; Thorne, 1998). These are in turn influenced by factors such as moisture, temperature, volume, density, substrate as well as the colony vigour and geographic variation in termite size (Becker, 1969b; Haverty & Nutting, 1974; Howick, 1975, 1978; Lenz et al., 1982; Forschler, 1996; Thorne, 1998). It is also difficult to identify which individuals out of a group are
78 Studies on termite consumption feeding and this means that an accurate measure of consumption per individual is virtually impossible (Forschler, 1996). Su and La Fage (1984) performed experiments taking survivorship into account to thus provide better estimates of consumption per individual. A further factor to consider is that of wood block size used in experiments as workers tend to feed at higher rates on larger blocks of wood (Thorne, 1998). Esenther (1980) found that it was feeding space density and not survival that affected termite performance. Choice experiments compared to no choice experiments may also give varying consumption rates (Lenz et al., 1982).
In nature the type and condition of the wood, the availability of other food sources, temperature and the physical health, species and number of termites all play a role in determining consumption by termites (Forschler, 1998b). It is important not to extrapolate from feeding rates that are calculated in the laboratory as these will not take all of the above factors into account (Lenz et al., 1982; Thorne, 1998).
Termites can survive for several weeks or months without a cellulose source as they commonly cannibalise each other (Feytaud, 1946; Forschler, 1998b). Smythe and Carter (1970) quoted a period of two to three weeks of survival without food for R. flavipes. Termites also consume the remains of moults as well as enemy cadavers (e.g. termites from other colonies) (Feytaud, 1946). A further factor that may affect consumption is the colony structure as, for example, the presence of soldiers has been seen to increase the consumption by colonies (Su & La Fage, 1987).
In the present study, the consumption on various wood species was assessed. Wood preference may be an important factor when considering the risk of introduction. The availability of a particular food source could mean the difference between thriving and surviving, and could be the make or break point to determine whether a colony is successful at establishing itself. In this experiment, consumption and survival rates of R. santonensis and R. grassei on oak (Quercus robur), beech (Fagus sylvatica) and Scots pine (Pinus sylvestris) were recorded. These were chosen as they are some of the most common species of tree in the UK. Since no-choice experiments may be inaccurate in assessing feeding preference since consumption on a particular wood species can depend on the other wood species also present (Gay et al., 1955; Oi et al., 1996), choice experiments were also performed.
79 Studies on termite consumption
Table 5.1 Hardness of wood species in newtons (Lavers, 1983).
Wood Species Hardness at 12 % humidity (N)
Pinus sylvestris 2980
Fagus sylvatica 4670
Quercus robur 5470
A second group of experiments assessed the survival of R. santonensis individuals on live seedlings of Corsican pine (Pinus nigra), Scots pine and cherry (Prunus avium) compared with wood blocks of Scots pine. The consumption rate of R. santonensis and R. grassei on live seedlings was also recorded. Pinus nigra is common in the UK and is present in sandy soils, soils which are thought to be preferred by Reticulitermes species (Kofoid, 1934). Cherry was chosen as it is one of the most common garden plants. Whether or not these species feed on live trees is important information in predicting where these species are likely to pose a threat.
5.2 Materials and Methods
In each experiment, the wooden blocks of each tree species were placed in a drying oven at 60 °C for at least 48 h. (Waller, 1988). They were then left to cool, weighed and placed for at least 7 days in a controlled environmental room at 25 ± 2 °C and 90 ± 5% RH, to allow the blocks to reach a moisture equilibrium before use in an experiment. The wood blocks used conformed to the European Standard EN 118 : 1990 in terms of wood quality.
5.2.1 Consumption on different wood species - no choice
Twenty blocks (5 x 2.5 x 1.5 cm) of each species (oak, beech and Scots pine) were used as food sources for the experiments. Sixty (2 termites species x 3 tree species x 10 replicates) 500 ml glass jars containing vermiculite were prepared (Section 3.3). A single wood block was placed in each jar as well as either 100 R. santonensis BRE
80 Studies on termite consumption or R. grassei UK workers / pseudergates. The experiment was run for 8 weeks at 25 ± 2 °C and 90 ± 5 % RH. Termite survival and dry weight consumption were measured at the end of the experiment.
ANOVA was used for data analysis.
5.2.2 Consumption on different wood species - choice
Pairs of Scots pine, beech and oak blocks (5 x 2.5 x 1.5 cm) were used as food sources. Ten replicates of each pair were prepared, thus using a total of 40 blocks of each wood species. Sixty (2 termite species x 3 wood combinations x 10 replicates) 500 ml glass jars containing vermiculite were prepared (Section 3.3) and the blocks placed on top of the vermiculite. The termites, 125 R. santonensis BRE or R. grassei UK workers / pseudergates were then added to the jars. The experiment was run for 8 weeks at 25 ± 2 °C and 90 ± 5 % RH. Termite survival and dry weight consumption were measured at the end of the experiment.
ANOVA was used for data analysis.
5.2.3 Consumption and survival on Scots pine over time
A total of 78 (2 termite species x 13 sample days x 3 replicates) jars were set up with vermiculite, a single Scots pine block (5 x 2.5 x 1.5 cm) and 125 R. santonensis BRE or R. grassei UK workers / pseudergates. Thus three replicates were made per day per species.
Three days after the start of the experiment three jars were chosen at random and removed from the experiment, termite survival was noted and the wood blocks were dried to calculate consumption. This was repeated after a further 4 days and, thereafter, at seven-day intervals. The experiment was run for 12 weeks at 25 ± 2 °C and 90 ± 5 % RH.
Regression was used for data analysis.
5.2.4 Survival on Corsican pine seedlings
Corsican pine seedlings, approximately 2 years-old, had the soil from their roots removed. The seedlings were then placed in a bucket of distilled water to wash off any excess soil. The foliage of the seedlings was cut to the same height,
81 Studies on termite consumption approximately 25 cm from ground level. The roots were trimmed to facilitate assessment of root consumption.
Images of the trimmed roots were taken using a Sony Digital Still Camera (Mavica, MVC-FD7). Vermiculite (Dupre vermiculite, 3-5 mm) was used to fill garden pots (h= 7.4 cm, dia. = 74 mm). Once the seedlings had been placed in the pots, 30 ml of a solution of fertilizer (Phostrogen House Plant Food, NPK Fertiliser solution 6:6:6) in the proportions of 35 ml to 1 1 of distilled water was added to each pot to saturate the vermiculite. The plants were left to establish for two to three days at 15 ± 2 °C and 90 ± 5 % RH.
An acetate cone was placed around the pot to prevent the plants from rubbing against each other and thus preventing the movement of termites. Fifteen replicates of each were performed. The pots were placed into 500 ml glass jars which had their bases covered with 30 ml of Supamix silversand (Pioneer Supamix Ltd, Nuneaton, UK) to absorb any excess water. The same procedure was used for the control, where, instead of the live seedlings, 5 x 2.5 x 1.5 cm blocks of Scots pine were used. To each pot, 125 R. santonensis BRE workers / pseudergates were added. The experiment was run for 8 weeks at 15 ± 2 °C and 90 ± 5 % RH. Ten ml of the fertilizer solution was added to each pot on a weekly basis. On termination of the experiment, the roots of each plant were re-photographed and the number of surviving termites was noted.
Consumption on the live plants was measured using P C Image (Ver 2.2.05, Foster Findlay Associates Ltd.). The root area was selected and the number of pixels equivalent to the roots was noted. The consumption was calculated by taking the difference of the number of pixels at the start and end of the experiment.
ANOVA was used for data analysis.
5.2.5 Survival on Scots pine seedlings and in combination with wood blocks
Scots pine seedlings, approximately 1.5 years-old were prepared as in Section 5.2.4.
After fertilizer was added to each pot to saturate the vermiculite, the plants were placed in a closed propagator and left to establish for 7 to 11 days at 15 ± 2 °C and 90 ± 5 % RH. The pots were then prepared using the method in Section 5.2.4.
82 Studies on termite consumption
The different treatments were Scots pine seedlings with and without termites, Scots pine blocks (5 x 2.5 x 1.5 cm) with and without termites, Scots pine seedlings combined with blocks with termites and finally termites with no food source. Six replicates were performed for each treatment.
To each treatment with termites, 125 R. santonensis BRE or R. grassei UK workers / pseudergates were added. The experiment was run for 8 weeks at 15 ± 2 °C and 90 ± 5 % RH. The roots of each plant were re-photographed and the number of surviving termites noted.
ANOVA was used for data analysis.
5.2.6 Survival on Scots pine and cherry seedlings
The same method as described in Section 5.2.5 was used in this experiment. However, in this case both Scots pine and cherry (Prunus avium) seedlings, approximately 1.5 years-old, were used. The treatments included Scots pine seedlings with and without termites, Cherry seedlings with and without termites and Scots pine blocks with and without termites. These treatments were repeated for both species and all replicates were performed seven times.
ANOVA was used for data analysis.
5.3 Results
5.3.1 Consumption on different wood species - no choice
Survival of R. santonensis BRE and R. grassei UK on different wood types is shown in Figure 5.1. The three-way interaction term between wood type, species and origin (i.e. French, BRE or UK) was found to be significant (F 2, 108 = 3.59, p < 0.03), indicating that survival of the different species was dependent on the origin of the species and that in turn the effect of species was dependent on the wood type. The significant interaction term means that all the terms themselves were significant, thus wood, species and origin all had a significant effect on survival.
83 Studies on termite consumption
0.8
0.7
To 0.6
2 0.2 -
0.1
R.g. UK R.g. F R.s. BRE R.s. F Species
Figure 5.1 Mean proportion survival (± S. E.) of 100 Reticulitermes grassei UK (R.g. UK), Reticulitermes grassei France (R.g. F), Reticulitermes santonensis BRE (R.s. BRE) and Reticulitermes santonensis France (R.s. F) workers / pseudergates on beech (dotted), oak (grey) and pine (striped) after 8 weeks; n = 10.
The analysis for consumption (Figure 5.2) was conducted by using relative consumption rate as the response variable (i.e. log (final weight / initial weight)). The three-way interaction term between wood type, species and origin was significant (F 2, 103 = 7.15, p < 0.01) as in the analysis of survival. Thus species had a significant effect on consumption observed depending on the origin. However, the effect that the species had on consumption observed was dependent on the wood type. As with survival, all terms (wood type, species and origin) were thus significant. The number of termites alive also had a significant effect on the consumption as the two-way interaction term between alive and origin was significant (F1, 103 = 6.13, p < 0.05).
84 Studies on termite consumption
-El 0.7
0.6
Ei 0.5 0 0.4
0.3
R.g. UK R.g. F R.s. BRE R.s. F Species
Figure 5.2 Mean consumption (g) (± S. E.) by 100 Reticulitermes grassei UK (R.g. UK), Reticulitermes grassei France (R.g. F), Reticulitermes santonensis BRE (R.s. BRE) and Reticulitermes santonensis France (R.s. F) workers / pseudergates on beech (dotted), oak (grey) and pine (striped) after 8 weeks; n = 10.
5.3.2 Consumption on different wood species - choice
The interaction term between wood combination and species was significant (F2,34 = 3.47, p < 0.05). Thus, both the combination and species had a significant effect on survival (Figure 5.3), though the effect that they had was different. Reticulitermes santonensis BRE showed a reduction in survival compared to R. grassei UK across the different wood species. However, wood combinations did not produce a set effect, there was no great difference in R. grassei UK survival, whereas the differences were noticeable in R. santonensis BRE between the wood species.
85 Studies on termite consumption
0.9
0.8 To 0.7
co 0.6 0 0E 0.5 00. cti 0.4 - co w 0.3 2 0.2
0.1
0
Oak & Beech Oak & Pine Beech & Pine Wood
Figure 5.3 Mean proportion (± S. E.) of 125 Reticulitermes grassei UK (grey) and Reticulitermes santonensis BRE (white) workers / pseudergates surviving on combinations of beech, oak and pine after 8 weeks; n = 10.
The analysis for consumption was conducted using relative consumption rate as the response variable (i.e. log (final weight / initial weight)). A highly significant
interaction term was found between wood type and species (F 2, 112 = 12.43, p < 0.001). Thus, both wood type and species had a significant effect on consumption (Figure 5.4). However, the patterns observed on consumption were different. For example, in the oak and pine combination, R. santonensis BRE caused a reduction in consumption compared to R. grassei UK, but the proportion of pine consumed as a measure of total wood weight was higher in R. santonensis BRE compared to R. grassei UK. The wood combination also had a highly significant effect on consumption (F 2,112 = 9.07, p < 0.001).
86
Studies on termite consumption
0.06
0.05
0)
-a 0.04 0 0 76 0 0.03
45 c 0 fo 0.02 a
0.01
0 R.g. R.s. R. g . R.s. R.g. R.s. Oak & Beech Oak & Pine Beech & Pine
Figure 5.4 Mean proportion of total wood weight consumed by 125 Reticulitermes grassei UK (R.g.) and Reticulitermes santonensis BRE (R.s.) workers / pseudergates depending on combinations of wood types after 8 weeks. Proportion relating to each wood type shown; beech (dotted), oak (grey) and pine (striped); n = 10.
5.3.3 Consumption and survival on Scots pine over time
Species (F1, 76 = 10.74, p < 0.01) and the number of days after the start of the experiment (F 1, 75 = 18.45, p < 0.001) were found to have a significant effect on the proportion of termites surviving (Figure 5.5).
87 Figure 5.5Proportionof surviving over84days.Fittedvalueswerebacktransformedfromthelinear significant effectonrelativeconsumptionrate(F 1, E. =0.05(intercept),9.8e-4(slope)(R.s.BRE)and0.06 Reticulitermes (R.g. UK);n=3. experiment (Fi, model; y=1.674-0.022x(R.s.BRE)and2.572(R.g.UK).WhereS. consumption rate. The relativeconsumptionrate(log(finalblockweight/initialweight))with time wasalsoplotted(Figure5.6).Onlythenumber ofdaysfromthestarthada species (F1, Proportion survival 74 = 3.97, p=0.05)northenumberoftermitesalive on removalfromthe grassei UK(A&continuousline)workers/pseudergates 75 = 20 1.05, p=0.31)hadasignificanteffecton the relative
Reticulitermes santonensis 40 i
Day 76 = 107.15, p<0.001).Neither 60 1
Studies ontermiteconsumption BRE
( ❑ & dotted line) 80 1 88 Studies on termite consumption
0 _
A IN 0 O
A
te O _ O ra A ion t CD O 0 A A 9 A A ❑
consump c0 o _
ive O t
la 0
Re O O
9 A
0 20 40 60 80 Day
Figure 5.6 Relative consumption rate (log (final weight / initial weight)) of
Reticulitermes santonensis BRE (o) and Reticulitermes grassei UK (A) over 84 days. Regression equation: y = -0.012 -0.0008x. Where S. E. = 0.004 (intercept) and 0.0001 (slope); n = 3.
5.3.4 Survival on Corsican pine seedling
The survival of R. santonensis BRE (Figure 5.7) on live seedlings was significantly lower than on wood blocks (F1,28 = 28.24, p < 0.001).
89 Studies on termite consumption
0.45
0.4 I
0.35
0.3
0.25
0 0.2 0 0 0.15 a_ 0.1
0.05
0
Seedling Block Food source
Figure 5.7 Mean proportion survival (± S. E.) of 125 Reticulitermes santonensis BRE workers / pseudergates on live Corsican pine seedlings and Scots pine blocks, after 8 weeks; n = 15.
5.3.5 Survival on Scots pine seedlings and in combination with wood blocks
The effect of food type on termite survival is shown in Figure 5.8. The values for survival on live plants were omitted as in all cases the survival was zero. Neither food type (F1, 21 = 1.35, p = 0.26) nor species (F1, 22 = 3.0, p = 0.1) had a significant effect on survival.
90
Studies on termite consumption
0.9
0.8
0.7
I 0.6
= 0.5 co 0 fo 0.4 0.
0.2
0.1
0
Block Block & Seedling Food source
Figure 5.8 Mean proportion survival (± S. E.) of Reticulitermes grassei UK (grey) and Reticulitermes santonensis BRE (white) workers / pseudergates on Scots pine blocks alone and in combination with live Scots pine seedlings after 8 weeks; n = 6.
Consumption of live plants alone showed no significant difference between controls
and treatments (F1, 33 = 2.84, p = 0.1), i.e. with and without termites, and thus were not further analysed (see 5.4.5 for explanation). The food type (i.e. treatment with blocks or in combination with live seedling) was found to have a significant effect on the block weight change (F1,31 = 21.44, p < 0.001) (Figure 5.9). Model simplification combining the two termite species showed a significantly different effect of the
species on the block weight change (F 1, 32 = 17.32, p < 0.001). Thus R. santonensis BRE had a significantly greater consumption than R. grassei UK. The number of termites alive had a significant effect on consumption (F1, 22 = 50.89, p < 0.001).
91 Studies on termite consumption
0.3
0.25
0.2
) (g
loss 0.15 - ht ig
k we 0.1 - Bloc 0.05
N/A
:77:771.7 Block Block and seedling
-0.05 Food Source
Figure 5.9 Weight loss (± S. E.) by wood blocks with and without the presence of live Scots pine seedlings in the absence (dotted) or presence of 125 Reticulitermes grassei UK (grey) and Reticulitermes santonensis BRE (white) workers / pseudergates after 8 weeks; n = 6.
5.3.6 Survival on Scots pine and cherry seedlings
The analysis of termite survival (Figure 5.10) showed that species did not have a significant effect on survival (F 1, 38 = 0.30, p = 0.59). Food source had a significant effect on survival (F 1, 40 = 25.65, p < 0.001) although termite survival was not significantly different on the two species of seedlings (F 1, 39 = 0.38, p = 0.54).
92 Studies on termite consumption
0.7
0.6
0.4 - co c 0 -Eo 0.3 0. 2 a_ 0.2 -
0.1 -
Cherry Pine Wood block Food source
Figure 5.10 Mean proportion survival (± S. E.) of 125 Reticulitermes grassei UK (grey) and Reticulitermes santonensis BRE (white) workers / pseudergates on live cherry and Scots pine seedlings and Scots pine wood blocks after 8 weeks; n = 7.
Analysis of weight change of wood blocks (F1,25 = 3.68, p = 0.07) and seedlings (F1,
53 = 0.66, p = 0.42) revealed that there was no difference if the termites were present or not. Thus, these data were not further analysed.
5.4 Discussion
5.4.1 Consumption on different wood species - no choice
For all termites tested except R. santonensis BRE, oak was found to have the highest survival and this was also true for all the termites in terms of consumption. The observation that survival was highest on oak contradicted results by Waller (1988) who found that red oak decreased survival. Reticulitermes santonensis France showed the lowest survival and yet showed the highest consumption on both beech and oak. There was a further anomaly that is difficult to explain. In each case, R.
93 Studies on termite consumption grassei UK and R. santonensis BRE showed a higher survival than their equivalent French populations yet in terms of consumption the inverse was true. As one of the problems with this experiment was that survival was only determined at the end, it is unknown at which point in time mortality occurred. Perhaps mortality occurred slightly later in the French population, thereby explaining the higher presence of individuals on completion of the experiment?
5.4.2 Consumption on different wood species - choice
Reticulitermes grassei UK showed a higher proportion survival compared to R. santonensis BRE and the same was true for the wood weight consumed for the oak and beech and the oak and pine combinations. The relative consumption level was:
• Reticulitermes grassei UK - oak > pine > beech
• Reticulitermes santonensis BRE - oak E pine > beech
In both species, pine had a higher observed consumption over beech. However, it is important to note that beech is of a higher density than pine and thus greater volumes must be consumed to equate to the same weight.
5.4.3 Consumption and survival on Scots pine over time
The proportion survival between the two species was significantly different even though the consumption rate was not. The much lower regression line (dotted) for R. santonensis BRE was caused by the zero values which were highly influential due to the low replication. The reduction in relative consumption rate over time appeared to be due to the reduction in survival. Interestingly the zero survivals seen did not necessarily mean that consumption was low and this was seen in the non- significance of the number of termites alive on termination of the experiment. An explanation for high consumption yet corresponding high mortality, which was given by Ripa et al. (2002), is that although the termites prefer this wood type there may be toxins present which cause mortality.
The present results contrast with those of Su and La Fage (1984) who fitted a negative exponential survivorship curve, however, their experiment only lasted eight weeks compared to the 12 weeks of the experiment presented here. The survival in the present experiment between 0 and 56 days could probably have a similar curve
94 Studies on termite consumption
fitted. However, this would assume that mortality virtually stopped thereafter, whereas this was not the case.
5.4.4 Survival on Corsican pine seedlings
This initial experiment showed that although there was a significant difference between the food sources, survival was still seen on seedlings. Thus, R. santonensis BRE workers feed on live plants. However, there were problems with keeping the plants alive in such artificial environments and, to help the plants to establish, the methods were improved for the experiments that followed by allowing the plants to acclimatise in the propagator prior to being exposed to termites.
5.4.5 Survival on Scots pine seedlings and in combination with wood blocks
Survival was not significantly influenced by either the food source or the species. This and the fact that there was zero survival on the seedlings alone would suggest that the termites were not feeding on the seedlings.
The weight loss seen on the block was significantly higher in the presence of R. santonensis BRE compared to R. grassei UK. The food source was also found to have a significant effect on weight change and in both cases it was higher on the wood block on its own. Two possible explanations may be that the seedling was having a detrimental effect on the termites or that consumption was occurring on the seedlings but that the method of measurement was not sensitive enough to detect such small quantities of consumption. Perhaps both explanations play a role, as indeed, survival was lower on the combination of block and seedling than on the block on its own. The number of termites alive also had a significant effect on the level of consumption, in comparison to the experiment of consumption and survival over time (Section 5.4.3) where the number of termites alive had no effect on consumption.
14.6 Survival on Scots pine and cherry seedlings
Survival was low on all the food sources, even after repeating the experiment the same was true and insufficient time was available to repeat the experiment for a third time. Survival was, however, highest on wood blocks and was the same on the two
95 Studies on termite consumption seedling species. Thus, while feeding on live plants occurs, termites do better on dead wood. Survival by R. santonensis BRE was only seen on the cherry seedlings and by R. grassei UK on the Scots pine seedlings. Reticulitermes santonensis BRE thus did not survive on the Scots pine seedlings in this case but survived on Corsican pine seedlings (Section 5.4.1). Further study would be required to provide a valid explanation.
5.4.7 Overview
In general, the number of termites per container affected consumption, which is something which would be expected (Smythe & Carter, 1969). Yet in the experiment of consumption over time (Section 5.3.3) the number of termites alive were not found to have a significant effect on consumption. This may be due to aperiodic feeding or that certain factors during these experiments suppressed consumption by termites. Total mortality was also observed in some replicates which has been mentioned in other studies (Mannesmann, 1973).
Consumption on Scots pine blocks seemed to be consistently higher by R. santonensis BRE compared to R. grassei UK across all the experiments. This was not true for the French populations in the no-choice experiment, which tended to show a higher consumption but lower survival than the R. grassei UK and R. santonensis BRE populations. Therefore, R. grassei UK may be more robust than R. grassei France.
It has been previously stated that there are differences in consumption between choice and no-choice experiments (Oi et al., 1996). The repetition of the no-choice experiments with two blocks to remove the effect of increased block size was not possible due to logistical constraints and thus any comparison between these two experiments must be approached with caution. The survival seen in R. grassei UK was consistently higher than R. santonensis BRE in the no-choice experiment which was not the case in the choice experiment.
Beech was generally the wood the least consumed except for consumption by R. santonensis France in the no-choice experiment (Figure 5.2). Oak was the most consumed wood except for consumption by R. santonensis BRE in the combination of oak with pine (Figure 5.4). This once again contradicts the results by Waller (1988) who found consumption to be highest on pine. It is difficult to consider
96 Studies on termite consumption consumption in terms of weight as this does not measure the structural effect termite damage has on the wood blocks. For example, the same weight of wood consumed produced very different visual damage on oak compared to pine, a much softer wood. This is an aspect that may warrant further investigation.
The preference for pine by R. santonensis BRE was evident in the choice experiment. This preference for pine is in agreement with the findings of Behr et al. (1972) who showed that consumption decreased with wood hardness. However, the no-choice experiment did not show this difference. Behr et al. (1972) also found there to be no difference between the no-choice with two blocks and the choice experiments which contradicts Lenz et al. (1982). A conclusion cannot be reached as the no-choice experiment only used one wood block and, as block size has been said to have an effect on consumption (Thorne, 1998), any differences seen could be attributed to this reason.
Higher survival by R. grassei UK compared to R. santonensis BRE was observed in the no-choice, the choice experiment and the consumption over time. This supports the conclusion that R. grassei UK had a higher survival. A contradiction was seen in the seedling experiments, where R. santonensis BRE had the higher survival, although the latter experiments had many problems and the results obtained may not be reliable. The lower survival seen in the seedling experiments may be because these were subjected to a lower temperature.
The data from the experiments with seedlings provided very mixed results. This would suggest that survival did occur on seedlings but is only at very low levels. The possibility of cannibalism as a means of continuing the population in this case was not a factor that played a role as zero survival was seen when termites were placed without a food source. Gay et al. (1955) found some survival with starved colonies of Nasutitermes exitiosus. Perhaps higher termite numbers are required for this to be true. The method for measuring consumption of seedlings was not sensitive enough and this meant that no conclusions could be made about consumption for these experiments.
5.5 Conclusion
Consumption studies still remain difficult due to the various factors (see Section 5.1.1) that have an influence (Carter et al., 1972; Mannesmann, 1973; Carter & Dell,
97 Studies on termite consumption
1981). This in turn makes it virtually impossible to make direct comparisons between studies by different authors and is further complicated by slight differences in methodology.
Survival on live plants provided very mixed results. The survival seen on the initial experiment with Corsican pine may have been due to feeding on dead parts of the seedlings as it was very difficult to keep these plants alive under such artificial conditions.
The aim of this study was to observe the ability of the termite species studied here to survive on various wood species. Though these termite species do not necessarily thrive on all the wood species investigated here, they will survive. Oak and pine are both used in construction throughout the UK and the high structural damage caused to pine could lead to structural problems in a building should a termite infestation be present. Larger scale studies are required to be able to judge consumption rates by termites and laboratory trials have a high variance and cannot necessarily be extrapolated to the field to provide values for consumption.
98 Chapter 6
Reproductive strategies
6.1 Introduction
The triggers for alate formation have been, and still remain, a matter of conjecture. Snyder (1920) even went as far as suggesting that swarming may be due to the 'spirit of the colony'.
Climatic factors have been said to play an important role in alate formation and the need for a period of warm weather and soil dryness are two factors mentioned in the literature (Feytaud, 1912; Buchli, 1956; Clement, 1986). A threshold for the minimum number of individuals required before nymph production is thought to exist (Buchli, 1958; Weesner, 1965). An explanation, from another perspective, for the need of dispersion is the greater availability of food for the future young within the colony (Buchli, 1956). Nutrition has been thought to be an influential factor on the number of nymphs that are produced (when there is high food quality) or regress to pseudergates (when food quality is low) (Perez, 1907; Hare, 1934; Buchli, 1958; Luscher, 1960; Harris & Sands, 1965; Lenz, 1976). Inhibition, possibly via hormonal regulation, has also been mentioned as a form of control of nymph and thus alate formation (Castle, 1934; Luscher, 1960; Harris & Sands, 1965; Miller, 1969). Although, a combination of external cues and genetics have been said to be triggers (Hare, 1934).
The point at which individuals are susceptible to triggers has been discussed and the most surprising theory is possibly that castes are predetermined in the egg (Thompson, 1917; Snyder, 1925, 1926). The most recent view is that gene expression is controlled by hormones and neurohormones which are in turn influenced by external cues as well as cues within the colony (Noirot, 1990).
Neotenics were thought by Buchli (1958) to only be formed when disturbances occur, yet earlier authors considered them to be commonly found in a colony and even went as far as suggesting that they were the true adult caste (Snyder, 1925). Reticulitermes santonensis and R. lucifugus are thought to have different
99 Reproductive strategies
reproductive strategies. Reticulitermes santonensis is thought to reproduce mainly by secondary reproductives, whereas R. lucifugus is thought to reproduce by forming alates (Jacquiot, 1956; Fougerousse & Perlade, 1975; Clement, 1977a; Vieau, 1993, 1994a, 1999, 2000). This is possibly due to R. santonensis potentially being an introduced species (see Chapter 2). However, all references to this phenomenon have been based on field observations. Grace (1996) talked about a difference in strategy, where more northerly distributed R. flavipes did not form alates, so the observations on R. santonensis may just be that it is not receiving the required environmental triggers to produce alates. This requirement of certain cues for alate production has been mentioned in Coptotermes (Lenz & Barrett, 1982). It could be argued that R. santonensis primary reproductives may simply be difficult to find and not easily encountered, which was the reason given by Feytaud (1912) for R. lucifugus before the primary reproductives were encountered.
This chapter includes mainly observational studies that have been performed together with a study on whether subjecting termites to a period of cold will influence their reproductive strategy. These species have not been observed to produce alates in a constant environment such as in a CT room (L. Laine, personal observation in the present study). However, when a R. santonensis colony was exposed to 15°C for a period of approximately 3 months, alates were found (T. Wilkinson, personal communication). Termites may, therefore, require a trigger, such as a period of low temperature, to change from only using secondary reproductives to using primary reproductives (alates) for reproduction and dispersal.
6.2 Materials and Methods
6.2.1 Observation Apparatus
The apparatus used were based on the design by Luscher (1949). A glass sheet, 0.3 x 30 x 40 cm was used as the base plate (Figure 6.1, no.1). On top of this was placed glass spacer strips (Figure 6.1, no. 2) of 1 cm width and 0.3 cm thickness. Two glass strips, 30 cm in length, were also placed along each 30 cm edge of the base plate. A second pair of 38 cm glass strips that had been cut into three even pieces were placed along the 40 cm edge of the base plate. The small gap produced by the strips being cut (Figure 6.1, no.3) provided for aeration and allowed for water to be inserted
100 Reproductive strategies
between the plates. A top plate (Figure 6.1, no. 4) was made from a 40 x 30 cm glass sheet cut into eight 30 x 5 cm strips and placed over the top of both the base sheet and glass strips. This was to be able to remove certain sections for cleaning without having to disturb the whole colony. The whole apparatus was held together using twenty 1.9 x 1 cm fold-over clips. The apparatus was kept upright along one of the 40 cm edges in a stand (see Figure A.4 of the Appendix).
40 cm 4
Figure 6.1 Glass plate assembly used for reproductive strategy experiments.
6.2.2 Preliminary observations on R. grassei UK and R. santonensis BRE
Vermiculite was used as the medium and covered approximately 2/3 of the internal surface area between the glass sheets (Section 3.3). Twelve 3 cm Scots pine squares of 0.2 cm thickness were used as a food source and were placed in contact with but above the level of the vermiculite (see Figure A.5 of the Appendix). In this initial experiment, one replicate was performed for each species, using 125 workers. The plates were maintained at 25 ± 2 °C at 90 ± 5 % RH for 567 days. Observations and egg and larval counts were made on a weekly basis.
6.2.3 Observational study on two populations of R. grassei and R. santonensis.
Plates were set up with vermiculite covering the total internal surface area of the glass sheets and 14 pine squares were arranged evenly within the vermiculite. Two
101 Reproductive strategies
hundred and fifty workers were used for each species, but the number of nymphs and soldiers used varied due to shortages in these caste members within the laboratory colonies. Differences in caste proportion has been observed previously and is also true in the field depending on the seasons (Light & Illg, 1945; Howard & Haverty, 1981). The number of soldiers and nymphs used for each species and strain is shown below:
R. grassei UK: 2 nymphs
R. grassei France: 2 nymphs, 1 soldier
R. santonensis BRE: 2 nymphs, 2 soldiers
R. santonensis France: 2 soldiers
Each treatment was replicated once and the plates were incubated as in Section 6.2.2. The weekly observations made were: activity in each plate (ranked 1 to 4; 1 = least active plate, 4 = most active plate), the number of squares out of 14 where clumping was seen (i.e. 5 or more individuals), and the number of eggs, first instar and second instar larvae present in the plate.
6.2.4 Effect of cold treatment on reproductive strategy
The same method was used as in section 6.2.3. Fourteen pine squares were provided as a food source. Five-hundred R. grassei Devon or R. santonensis BRE workers / pseudergates and 4 brachypterous nymphs were added to the glass plates. Three plates of each species were maintained at 25 ± 2 °C and a further three plates per species were initially kept at 25 ± 2 °C for 16 weeks (until day 112) and were then put at 15 ± 2 °C and later returned to 25 ± 2 °C after 16 weeks (at day 224). They were left at 25 ± 2 °C for a further 18 weeks (until day 350).
Three series of measurements were taken from these plates:
Survival —judged on a scale of 1 to 5 (1 = 0 to 20 %, 2 = 20 to 40 %, 3 = 40 to 60 %, 4 = 60 to 80%, 5 = 80 to 100 %).
102 Reproductive strategies
Activity —judged on a scale of 1 to 3 (1 = low termite movement within plate, 2 = moderate termite movement, 3 = high termite movement).
Consumption — measured by scoring consumption on each block out of five. These were then added together to get a reading for total consumption per plate.
The number of termites surviving on termination of the experiment was counted. This was broken down into the number of eggs, L 1 , L2, W3, workers, soldiers, ergatoids and brachypterous neotenics present.
ANOVA and ANCOVA were used for data analysis.
6.3 Results
6.3.1 Preliminary observations on R. grassei UK and R. santonensis BRE
The difference in behaviour between the two species was marked after nine days. Reticulitermes santonensis BRE was much more active and started forming tunnels over the food source. Reticulitermes grassei UK did not form any tunnels over the food source, preferring to remain within the vermiculite. Reticulitermes santonensis BRE thoroughly explored the food source whereas R. grassei UK consumed the areas that were closest to the vermiculite. All the R. santonensis BRE had died after 121 days in the apparatus, whereas R. grassei UK survived for up to 567 days. Egg formation was first observed in the R. grassei UK colony 173 days after experimental set up (Figure 6.2).
103 Reproductive strategies
90
80
70
60
50
0 0 40
1 30
20 - - - -
10 LI rut 11 np, e p10 1.0 01P eg' te. Day
Figure 6.2 Number of eggs (white), first instar larvae (grey) and second instar larvae (black) observed over 567 days from an initial population of 125 workers / pseudergates of Reticulitermes grassei UK.
6.3.2 Observational study o n two populations of R. grassei and R. santonensis
In this experiment, differences in activity were observed after only 6 weeks. One hundred percent mortality was seen in R. grassei France after 45 days. There seems to have been a shift in activity over time (Table 6.1).
104 Reproductive strategies
Table 6.1 Level of activity shown by different strains of Reticulitermes santonensis and Reticulitermes grassei after 0, 224 and 452 days in the glass plate apparatus.
Species Level of activity (4 = highest)
Start of Mid-way reading Last reading experiment (+224 days) (+ 452 days)
R. santonensis France 4 4 3 R. santonensis BRE 3 3 4
R. grassei France 1
R. grassei UK 2 2 4
* 100 % mortality
During the experiment, R. grassei UK only showed a higher level of activity compared to R. santonensis BRE on three occasions. Activity levels at different stages during the experiment are shown in Table 6.1. After 38 days, evidence of previous activity was seen on all squares except for R. grassei France.
In Figure 6.3 the highest activity was seen in the two R. santonensis populations. Reticulitermes grassei UK only attained highest activity at the end of the experiment and R. grassei France consistently showed the lowest activity before complete mortality.
105 Reproductive strategies
00000000000000000000000000000 00013 0001704 0000000000 ■a 000 OS
5 3- N• 00•00000•0000000000•000•00•00 0000 •••••• OOOOO ••1111•0 O• ❑• O ❑ ❑ 0
C 2- .00 OOOOOOOOOOOOOOO mom 000000 0•• • • • C
• • •
50 100 150 200 250 300 350 400 450 500 Day
Figure 6.3 Rank (4 = highest) in terms of the level of activity observed in Reticulitermes santonensis BRE (R.s. BRE) (o), Reticulitermes santonensis
France (R.s. F) (❑), Reticulitermes grassei UK (R.g. UK) (.) and Reticulitermes grassei France (R.g. F) (0) colonies in glass plates over 469 days.
The inverse of what was seen in activity was seen in the clumping behaviour. Reticulitermes grassei UK showed the greatest number of squares on which clumping was observed (Figure 6.4). This is particularly true in the second half of the experiment. Both R. santonensis populations did not exceed five squares on which clumping was observed.
106
Reproductive strategies
100 -
80 -
0
cab \ Act, ,bb ,‘rb b23 <,jo co gp op 41 ()Cs N^0, h9.. op C;) ,bcp ,43% op NNNNNN 'le rl‘ "C 1, `1, ntA, ' le 'b c'T Day
Figure 6.5 Number of eggs (white), first instar larvae (grey) and second instar larvae (black) of Reticulitermes santonensis BRE produced over 469 days from an initial number of 250 workers / pseudergates, 2 nymphs and 2 soldiers.
100
80
4E' 60 0
40
20
0 •\9' ^gL Pc') 99 fib` ugh b00 cb Co cbQ. ts<')% Day
Figure 6.6 Number of eggs (white) and first instar larvae (grey) of Reticulitermes grassei UK produced over 469 days from an initial number of 250 workers / pseudergates and 2 nymphs.
108 Reproductive strategies
6.3.3 Effect of cold treatment on reproductive strategy
The analysis performed using survival as the response variable showed a significant
interaction term between temperature regime and time (F1, 532 = 27.17, p < 0.05) and
significant terms for both species (F1, 536 = 78.57, p < 0.01) and consumption (F1, 535
= 32.98, p < 0.001). Activity did not have a significant effect on survival (F 1, 530 = 24.05, p = 0.11). There was a decrease in survival over time (Figure 6.7) although, in general, the survival of R. grassei UK appeared lower than that of R. santonensis BRE.
LO - E4000
-o€DoeeED(f)e00000 00000 l a iv urv f s o - + + + +WEBEDGEBEDED GE131,33EDED99(0 ,9199 eEDEBEDO EDEDW 90$ EDCBCDEBED ,DED El) ED terms k in
Ran - +++++++++++ gEDEBEBEREDED 0 0
100 200 300 Day
Figure 6.7 Score (5 = highest) in terms of Reticulitermes grassei UK (+) and Reticulitermes santonensis BRE (o) survival in glass plates over 350 days.
109 Reproductive strategies
ri G00100000000000 0000000000
ity iv
t N ac ite
f term O N ED ED + + + + + + + + + + + + + + + + + ++9 00000 $e9 em) EDED 00ED09 ED o terms in k n Ra
O _ +u 000000 0$000++++++ ++ ++++
0 100 200 300 Day
Figure 6.8 Score (3 = highest) in terms of level of activity of Reticulitermes grassei UK (+) and Reticulitermes santonensis BRE (o) colony in glass plates over 350 days.
The results of the analysis with activity as the response variable showed a significant three way interaction term between survival, consumption and temperature regime (Fi, 524 = 9.92, p < 0.01) and between consumption, temperature regime and time (F1,
524 = 48.10, p < 0.01). Species was also found to be significant (F1, 524 = 52.71, p < 0.05). The terms were therefore all significant. As with survival, activity decreased with time (Figure 6.8), this could be considered to be self-intuitive, as activity is a function of survival. Although, with activity the difference between the two species is clearer, where R. grassei UK has lower activity than R. santonensis BRE.
110 Reproductive strategies
▪ t 0
°TT-€D ;- g ++00 00 oo + -1r)) 8000 o?oet 0 0
+00 f ++oe + + ion ++ J t +8+° 8 mp + + +8P+%) ;?eo °°+; OP°
consu o@it° ° $ee61.c+ ++ 0
ive co +++ t /66' la
Cumu 0 (NI
+e 0 0
0 100 200 300 Day
Figure 6.9 Cumulative consumption of Scots pine squares by Reticulitermes grassei UK (+) and Reticulitermes santonensis BRE (o) in glass plates over 350 days.
The analysis was performed with consumption as the response variable and a four- way interaction term between species, survival, temperature regime and time was highly significant (F1, 509 = 17.32, p < 0.001) as was the four-way interaction term between survival, activity, temperature regime and time (Fi, sos = 16.60, p < 0.001). Thus species had a significant effect on consumption even though this difference is not noticeable in Figure 6.9. The range of the cumulative consumption increased with time.
111 Reproductive strategies
Table 6.2 Range, median and grand mean of scores for survival, activity and consumption from an initial number of 500 workers / pseudergates and 4 neotenics of R. santonensis BRE (R.s. BRE) and R. grassei UK (R.g. UK) at constant and fluctuating temperatures in glass plates after 350 days.
Species R.s. BRE R.g. UK R.s. BRE R.g. UK
Temperature Constant Constant Fluctuating Fluctuating (25°C) (25°C) (25—•15—•25°C) (25.15.25°C)
Survival
Range 1 - 5 1 - 5 3 - 5 3 - 4
Median 3 3 3 3
Grand Mean 3.2 2.6 3.3 3.1
Activity
Range 1 - 3 1 - 3 1 - 3 1 - 3
Median 3 2 2 2
Grand Mean 2.5 1.7 2.0 1.7
Consumption
Range 6-58 0-61 6-49 5-53
Median 36 37 33 31
Grand mean 37.0 36.1 32.0 30.8
Reticulitermes grassei UK had the lower mean survival (Table 6.2), which is in agreement with Figure 6.7. The ranges for activity were no different between all species and origins. Mean activity was lowest in R. grassei UK. Reticulitermes grassei UK showed a larger range of consumption but lower mean consumption compared to R. santonensis BRE. Overall, mean consumption was greater in the constant temperature regime.
112 Reproductive strategies
In one of the R. grassei UK plates, which had been subjected to cold treatment, a long wing bud nymph was observed to be moulting into an alate. However, this moult was never completed due to cannibalism of the individual. Long wing bud nymphs were only produced in two of the 12 plates and in each case in R. grassei UK plates. One of which had a constant temperature regime and the other fluctuating temperatures.
180
160 -
140
120
r
be 100 num 80 Mean
60
40
20 -
r
7 42 77 119 154 196 232 266 315 Day
Figure 6.10 Mean number of Reticulitermes grassei UK eggs (white) and larvae (grey) observed over 350 days at a constant temperature in glass plates from an initial population of 500 workers / pseudergates and 4 brachypterous neotenics; n = 3.
113 Reproductive strategies
180
160
140 -
120
100 -
c
(13 80
60
40
20
0
7 42 77 119 154 196 232 266 3 5 Day
Figure 6.11 Mean number Reticulitermes grassei UK eggs (white) and larvae (grey) observed over 350 days at fluctuating temperatures in glass plates from an initial population of 500 workers / pseudergates and 4 brachypterous neotenics; n = 3.
180
160
140 -
120
60 -
40 -
20 -
42 77 119 154 196 232 266 31 Day
Figure 6.12 Mean number Reticulitermes santonensis BRE eggs (white) and larvae (grey) observed over 350 days at a constant temperature in glass plates from an initial population of 500 workers / pseudergates and 4 brachypterous neotenics; n = 3.
114 Reproductive strategies
180
160
140
120
ber 100 num 80 - Mean
60 -
40
20 -
n JUL 7 42 77 119 154 196 232 266 315 Day
Figure 6.13 Mean number Reticulitermes santonensis BRE eggs (white) and larvae (grey) observed over 350 days at fluctuating temperatures in glass plates from an initial population of 500 workers / pseudergates and 4 brachypterous neotenics; n= 3.
Analyses were performed to assess which factors had an effect on egg, larval, brachypterous neotenic and ergatoid development. The data for the egg and larval development for the two species at different temperature regimes are presented in Figures 6.10 to 6.13. Brachypterous neotenics took three weeks to be observed in R. grassei UK and 15 weeks in R. santonensis BRE, whereas ergatoids were not observed before 11 and 15 weeks respectively.
Analysis of egg production showed a significant interaction between temperature regime and species (F 1, 175 = 9.30, p < 0.01). The analyses on the larval production data gave a significant three way interaction term between temperature, species and day (F 1, 532 = 4.42, p < 0.05).
The three factors (temperature regime, time and termites species) had a significant effect on brachypterous neotenic and ergatoid production. The brachypterous neotenic production only showed significance with the main terms, temperature
115 Reproductive strategies
regime (F 1,538 = 1.68, p < 0.001), species (F 1, 537 0, p < 0.001) and time (F 1, 536 = 0, p < 0.001). Ergatoid production, however, showed a significant three-way
interaction term between temperature regime, species and time (F 1, 532 = 6.04, p < 0.05).
Analysis of the effect of temperature and species on final count from the glass plates only revealed two significant effects. These were the effect of the temperature
regime on the number of third instar workers (W3) (F 2, 8 = 9.06, p < 0.01) and the
effect of temperature regime on the number of ergatoids (F 2, 8 = 4.411, p < 0.05).
6.4 Discussion
6.4.1 Preliminary observations on R. grassei UK and R. santonensis BRE
Reticulitermes santonensis BRE showed a much higher activity compared to Reticulitermes grassei UK. This behaviour may explain observations that R. santonensis is a more robust species. In fact, this may just be that due to the higher activity R. santonensis is much more visible through its tunnelling.
Figure 6.2 demonstrated a cyclic production of eggs and, nearing the end of the experiment, a near complete reduction in egg production.
6.4.2 Observational study on two populations of R. grassei and R. santonensis
The level of activity at the start of the experiment was greatest in the R. santonensis populations. There was then a gradual shift at the end of the experiment, R. grassei UK showed similar activity to R. santonensis BRE. Reticulitermes santonensis BRE populations tended to be the most active.
First and second instar larvae were seen in the R. santonensis BRE plate after 164 days, this meant that earlier egg production must have been missed. In the R. grassei UK plate, egg production only started after 178 days and first instar larvae were not observed before 192 days after the start of the experiment. So therefore there was approximately a month's difference in first larval occurrence between the two
116 Reproductive strategies
species. Thus R. grassei UK, under the same conditions as R. santonensis BRE, showed a slower reproductive rate.
6.4.3 Effect of cold treatment on reproductive strategy
Survival, activity and consumption were found to have a significant effect on each other, which makes the data difficult to interpret. Looking at termite survival against time (Figure 6.7), R. grassei UK showed a lower survival compared to R. santonensis BRE and this remained true even when R. santonensis BRE showed a reduction in survival. The reduction in survival over time may be due to the low initial colony numbers, as was explained in Chapter 4. The same can be said for termite activity, although the increase in larval numbers with time meant that the termites spent a large proportion of time caring for the larvae and perhaps the reduction in activity was in part due to this (Figure 6.8). The observation of mean consumption being higher at higher temperatures might be expected as the termites would have a higher metabolic rate and thus be able to assimilate more food. The cumulative consumption observed in Figure 6.9 did not seem to be different between the two species even though statistical analysis showed that the difference was significant.
There was a much more cyclic production of eggs when the plates were subjected to fluctuating temperatures (Figure 6.11 and Figure 6.13) compared to constant temperatures (Figure 6.10 and Figure 6.12). This was particularly apparent in the number of larvae produced. This phenomenon, of decreased egg production during cold periods, has been observed in the past (Feytaud, 1912). In fact, fluctuating temperatures have also been said to be good for growth, reproduction and temperature resistance (Lenz et al., 1982). A surprising result was the higher fecundity and number of larvae produced by R. grassei UK compared to R. santonensis BRE, particularly under the fluctuating temperature regime (Figure 6.11 and Figure 6.13). Under constant conditions, R. grassei UK larval production was more rapid than R. santonensis yet in the previous experiment (6.3.2) the inverse was true.
The observation of a long wing bud nymph being cannibalised by workers may be explained by the report that workers tend to kill nymphs and neotenics if too numerous in the colony (Buchli, 1958; Esenther, 1977). In contrast, Lenz and Runko
117 Reproductive strategies
(1993) stated that the presence of neotenics was thought to be proof of colonies passing the critical period for survival. Although, they did mention biting behaviour by colony member on individuals with wing buds. This, they said, inhibited progressive moults. Cannibalism of these potential alates may be a form of energy recycling. Esenther (1977) showed that the addition of dead termites to colonies improved the diet quality. Thus, an explanation may be that when alates form and the colony cannot afford to lose these individuals via swarming, they are cannibalised.
The time intervals taken for the development of ergatoids and brachypterous neotenics were in agreement with Thorne (1998) who mentioned six to eight week or three to four months. In contrast, Esenther (1977) mentioned a period of only four weeks. Although interestingly in each case development was faster in R. grassei UK. This could have been due to a higher proportion of neotenics present in these colonies and thus the greater likelihood of nymphs being mistaken for workers. Apterous reproductives have been said to have a higher reproductive potential than brachypterous reproductives (Light & Illg, 1945). In the present experiment, more apterous neotenics were observed in the R. santonensis BRE plates than the R. grassei UK plates. Yet, when comparing this to the number of eggs produced, there was clearly a higher production by R. grassei UK. The fact that ergatoids may not necessarily be more fecund cannot be concluded here but is an interesting possibility.
An anomaly was observed between the two different sets of larval data, the observation data over time and the final count data. The observation data showed a significant effect of both species and temperature regime whereas the final count data showed neither. However looking at Figures 6.10 to 6.13, there was a drop in the larvae near the end of the experiment and this could explain the lack of significant differences. In both the experiments, however, the number of ergatoids was significantly affected by the temperature regime.
6.5 Conclusion
Termite colonies kept in glass plates have been said not to increase in number (Hickin, 1961). Perhaps this statement is not far from the truth as, in the experiment on the effect on cold treatment on reproductive strategy, only two of the twelve plates showed a higher final number compared to initial termite number. This did not mean that no reproduction occurred. However, although this method of
118 Reproductive strategies experimentation may not be ideal, it is a convenient way in which to do observational studies.
The conclusions that can be drawn from these experiments and observations are that R. santonensis BRE shows greater tunnelling behaviour and this may mean that the presence of this species is more obvious. However, this does not mean that R. santonensis is necessarily more robust or fecund. In contrast, the higher clumping behaviour observed in R. grassei UK, makes this species less 'visible'. Reticulitermes santonensis BRE seemed to be more adept at surviving and consuming more. Reticulitermes grassei UK appeared to have a higher fecundity. The marked difference observed between the fecundity of the two species, with and without a constant temperature regime may support the theory that R. santonensis is an introduced species as the artificial seasonality that was imposed did not have a beneficial effect on this species.
There are many more studies that could be done with these species, for example, doing similar studies to those of Light and Illg (1945) who used different combinations of workers, nymphs and soldiers. A more extreme drop in temperature may also be required for alate formation. The triggers required are most likely not straightforward and are probably a complex combination of caste make-up, termite density, food source, temperature and humidity.
119 Chapter 7
Temperature and soil type and their effect on termite survival
7.1 Background
Temperature greatly affects colony life in terms of survivorship, behaviour, activity, colony makeup as well as moulting ability (Snyder, 1920; Buchli, 1958; Weesner, 1965; Forschler, 1996, 1998b). Moulting of Reticulitermes species has been reported to stop below 15 °C (Buchli, 1958). Thus termites require high ambient temperatures (Serment & Tourteaux, 1991).
Termites in general are found in areas where the mean annual temperature does not drop below 10 °C (Williams, 1934; Fougerousse & Perlade, 1975) and R. lucifugus is thought to occur in areas where the mean annual temperature is 12 °C (Clement, 1977a). Conversely, Reticulitermes santonensis is thought to be particularly temperature dependent compared to other species in the genus (Becker, 1970) even though it has a more northerly distribution (see Chapter 2). The sensitivity of this genus to temperature thus means that with global warming its physical distribution may change. Temperature increase has been observed in the soil, Baker and Baker (2002) measured a rise of as much as 2 °C in the mean soil temperature in Minnesota, U.S.A. at depths between 3 and 13 m over 37 years and have said this to be true in other regions of the world (e.g. Irkutsk, Russia).
Reticulitermes species are present in the wood above ground in the summer and move to below the frost line in the winter (Banks & Snyder, 1920; Snyder, 1920; Buchli, 1958, 1961; Esenther, 1961). Esenther (1961) showed that individuals of R. flavipes were able to survive at very low temperatures. They have been observed in wood that is covered with frost as well as imbedded in frozen ground (Esenther, 1961).
120 Temperature and soil type
Perhaps the reason for their ability to survive these conditions is that the subterranean environment is very stable and thus soil temperature is said to remain nearly constant (Jacquiot, 1956; Harris & Sands, 1965). Indeed, at increased depth, temperature fluctuations decrease and the mean temperature at greater than 30 cm depth increases (Gupta et al., 1982; Davidoff et al., 1986; Walsh et al., 1991; Baker & Baker, 2002). Fluctuations in soil temperature are said to be highest in the top 10 cm of soil (Walsh et al., 1991). Temperatures that sustain termite establishment can also be provided by artificial sources such as central heating (Serment & Tourteaux, 1991).
The maintenance of humidity is said to be vital for survival, behaviour and activity of termites (Perez, 1907; Forschler, 1996, 1998b). Nest humidity of between 96 to 100 % is required for survival (Harris & Sands, 1965; Forschler, 1998b). Wood humidity of 15 % is sufficient as the interstitial air spaces are said to have 100% RH (Forschler, 1998b). Rhinotermitids are less tolerant of desiccation compared to Kalotermitidae (Feytaud, 1912; Becker, 1969a) and termites in the genus Reticulitermes in particular are said to require a constant supply of moisture and have a high level of humidity in their tunnels (Hendee, 1937; Weesner, 1965). Becker (1969a) commented that termites were better able to cope with low humidity if drinking water was provided and if in large groups. While this may be true, R. santonensis is said to have lower moisture requirements compared to R. grassei (Clement, 1977a) and R. flavipes was found to be more tolerant to drying than R. tibialis (Strickland, 1950). It is important that excessive moisture should be avoided in the laboratory (Feytaud, 1912; Light & Weesner, 1947; Osmun, 1957), even though termites have been observed to be able to survive periods of submersion (Forschler & Henderson, 1995).
Little information as to soil preferences of the genus Reticulitermes is available in the literature. Lee and Wood (1971) provided a thorough review of studies on termites and soils, however, the work that was covered was almost entirely on tropical genera. In terms of soil type, Reticulitermes species have been said to prefer sandy soils to clay soils (Kofoid, 1934). In contrast, Williams (1934) mentioned the inability of R. tibialis and R. hesperus to establish in sand. Aber and Beltrami (2002) described the distribution of R. lucifugus in Uruguay and mentioned the higher presence of this species in light to medium soil. This they attributed to the fact that
121 Temperature and soil type the termites were able to disperse more easily. The ease of dispersion in light soil may also mean that the termites are more likely to come across greater food sources and thus propagate faster.
The present experiments aimed to provide an initial insight into the effect of soil type and temperature on survival of R. santonensis BRE and R. grassei UK.
7.2 Materials and Methods
The experiments were conducted in controlled temperature rooms at 25 ± 2 °C and 90 % R.H. unless specified otherwise. The jars, in both experiments, were placed into plastic lidded containers, filled to a depth of approximately 3 cm with water. This was to ensure that the air in the boxes was saturated and thus to aim to keep the moisture of the substrates constant. The termites were handled using a paintbrush.
7.2.1 Survival of termites on different soil types
Samples (100g) of each substrate (Table 7.1; see Section 3.3 for vermiculite preparation) were placed into an oven at 60 °C for a minimum of 3 days to dry. The moisture content of each soil was thus verified (Table 7.2). A set weight (Table 7.2) of each substrate equating to a 2 cm depth of fill in the 500 ml jars was placed into separate zip-lock bags and put in a freezer at -20 °C for 48 hours to kill as many pathogens (e.g. nematodes) that may have been present as possible. The option of heat-sterilisation was not feasible as, due to the high clay content of some of the soils, the soils would have had to be homogenised which would have changed the soil structure.
122 Temperature and soil type
Table 7.1 Name, classification and source of substrates used to assess termite survival in different soil descriptions.
Name Description % Sand' % Silt' % Clay' Source
Great Knott III Batcome / Carstens with 20 65 15 Rothamsted, sandier inclusions.2 UK
Stackyard Cottenham, a brown sand 57 33 10 Woburn Farm, on Lower Husborne Crawley, Greensand.2 Bedford, UK 3
Warren Field Alluvial gley soil on 34 42 24 Woburn Farm, recent alluvium.2 Husborne Crawley,
Bedford, UK 3
Sand 82 14 4 Supamix silver sand (Pioneer Supamix, Nuneaton, UK)
Vermiculite Hydrated laminar Dupre magnesium-aluminium- Vermiculite, ironsilicate Hertford, UK
Soil analyses performed by A. Kelly and R. Hartley (University of Plymouth) (see Section A.1 and Figure A. 7 of the Appendix for further details).
2 Soil series of England and Wales
3 See Figure A.6 in Appendix for map.
123 Temperature and soil type
Table 7.2 Substrate moisture contents and substrate weights used to assess termite survival in different soil types.
Substrate name Moisture content (%) Weight used calculated from 100g
Stackyard 14 225 g
Great Knott III 19 235 g
Warren field 28 220 g
Sand 2 125 g
Vermiculite 68 50 g
One hundred and twenty-five workers / pseudergates of R. santonensis BRE or R. grassei UK were each placed in 500 ml jars containing one of the substrates described in Table 7.1. Five 3 cm Scots pine squares, 0.2 cm thick, were used as a food source in each jar. Each treatment was replicated six times. The jars had a square of cling film and then foil placed over the top which was secured with an elastic band. Five 1 mm holes were then pierced through the two layers to provide ventilation and the jars placed into the afore-mentioned large plastic boxes (Section 7.2 ). The experiment was run for 12 weeks.
Survival was determined by the number of termites present at the end of the experiment. ANOVA was used for the data analysis.
7.2.2 Survival of termites at different temperatures
One hundred and twenty-five workers / pseudergates of R. santonensis BRE or R. grassei UK were each placed in 500 ml jars containing vermiculite (Section 3.3). Five 3 cm Scots pine squares, 0.2 mm thickness, were used as a food source. The termites were handled with a moist paintbrush. The jars were then placed into the large plastic containers.
124 Temperature and soil type
Eight replicates were prepared for each temperature (10, 15, 20 and 25 ° C). The experiment was run for three months. The same experiment was replicated, but this time was run for six months. Survival was determined by the number of termites present at the end of the experiment.
ANCOVA was used for the data analysis.
7.3 Results
7.3.1 Survival of termites o n different soil types
Soil and species had a significant interaction term (F 4, 80 =- 14.28, p < 0.001). Both these factors therefore had a significant effect on survival. The experiment with R. grassei UK was repeated due to the low survival of the species. However, this did not provide higher survival and the results from both experiments were used in the analysis. The moisture contents of the soils were measured at the end of the experiment and these were found to be approximately the same as at the start (± 1 %).
0.8
0.7
0.6
ra 0.5 -
5 0.4
0.2
0.1
0 Knott Sand Stackyard Vermiculite Warren Soil types
Figure 7.1 Mean proportion survival (± S. E.) of 125 Retieulitermes grassei UK (grey) and Reticulitermes santonensis BRE (white) workers / pseudergates on different soil types at 25 ± 2 °C after 12 weeks; n = 6.
125 Temperature and soil type
7.3.2 Survival of termites at different temperatures
The interaction term between time and species was found to be significant (F1,187 = 14.24, p < 0.001). Thus, both species and time had a significant effect on survival, although these effects differed. Temperature, though not present in an interaction, also had a significant effect on survival (F 1, 190 = 12.58, p < 0.001).
0.4
0.35
0.3 -
7°> 0.25
0.2 -
0.1 -
0.05 -
10 15 20 25 Temperature
Figure 7.2 Mean proportion survival (± S. K) of 125 Reticulitermes grassei UK (grey) and Reticulitermes santonensis BRE (white) workers / pseudergates after 3 months at four different temperatures; n=8.
Reticulitermes grassei UK showed a greater change in survival from three (Figure 7.2) to six (Figure 7.3) months compared to R. santonensis BRE. Third stage workers were observed at 15 °C and above in the R. santonensis BRE replicates for the six-month experiment.
126 Temperature and soil type
0.4
0.35
0.3
Tv> 0.25
a. 0.15
I 0.1
0.05
0 I I
10 15 20 25 Temperature
Figure 7.3 Mean proportion survival (± S. E.) of 125 Reticulitermes grassei UK (grey) and Reticulitermes santonensis BRE (white) workers / pseudergates after 6 months at four different temperatures; n = 8.
7.4 Discussion
7.4.1 Survival of termites o n different soil types
The survival of the two species was clearly different (Figure 7.1). The experiment was performed twice for R. grassei UK and low survival occurred on both occasions. The lowest survival for R. santonensis BRE was seen on sand and the highest on Warren, an alluvial gley soil, which had the highest moisture content after vermiculite. Survival of R. santonensis BRE on Warren, a flinty, silty, clay loam, seems to be equivalent to Stackyard, a sandy soil. For R. santonensis BRE, at least, soil type did not seem to be a limiting factor. The significant interaction term between soil and species, and thus the significant term of soil was most probably due to the high variability seen in R. grassei UK.
The survival of the two species was not equivalent to the moisture content of the various substrates. Indeed, the moisture content of the substrates was found to be the same as at the start, so therefore loss of moisture was not an explanation. This
127 Temperature and soil type finding contrasts with the statement that moisture is important in survival, as in fact the substrate with the most humidity, vermiculite, did not show the highest survival.
The lower survival on sand may be because of the greater difficulty in maintaining a stable environment on this substrate. The particles are larger and therefore more difficult to move around for tunnel production by the termites. Sand also has a lower water holding capacity, thus the termites have less water available to them (Williams, 1934) which is limiting, especially for Reticulitermes species (Grube & Rudolph, 1999). Clay soil, although harder to tunnel through, is said to be particularly good for water conservation as the higher surface area means that hygroscopic (water surrounding the particles) and capillary water (water between the particles and that does not move with gravity) available is greater (Williams, 1934; Lee & Wood, 1971). Clay soil is also composed of smaller particles thus making it easier for termites to manipulate (Williams, 1934; Lee & Wood, 1971). The size that the termites use for tunnel formation is determined by the size of the soil particles available and these in turn are determined by the size of the worker (Lee & Wood, 1971). Different proportions of clay have been observed in termite nests, however, these tend to be related to the proportions present in the subsoil (Lee & Wood, 1971).
The low survival seen in R. grassei UK may also have been due to a greater susceptibility of this species to carbon dioxide levels as the jars had relatively little aeration. Smith and Rust (1993) put forward an interesting analogy, that termite tunnels may act as 'insect trachea' thus passively bringing oxygen. Termites may require higher levels of oxygen than expected and this may explain the sensitivity to carbon dioxide levels observed in R. hesperus (Smith & Rust, 1993).
7.4.2 Survival of termites at different temperatures
The results of R. grassei UK for the six-month period were disappointing, survival was minimal (Figure 7.3). In the three-month experiment in all cases, survival was higher in R grassei UK (Figure 7.2). Due to time limitations the six-month experiments could not be repeated. There are two points of interest. Firstly, there is an increase in survival with increase in temperature for both species for the three- month period and for R. santonensis BRE for the six-month period. This is in agreement with Davis and Kamble (1994) who found exposing Reticulitermes
128 Temperature and soil type species to temperatures of 0 and 10 °C cause increased mortality with increase exposure time. The maximal exposure time in their experiment was 30 days. The present data contradict Smith and Rust (1993) who found that R. hesperus had a higher mortality at increased temperature. The same is true for the study by Smythe and Williams (1972) who found an increase in the temperature cause a decrease in survival of R. flavipes. They observed 0 and 73.6 % mortality at 15.6 and 26.7 °C respectively after only 14 days. The higher mortality of the former species was explained by the greater level of metabolism that occurs at higher temperatures. Secondly, higher survival of R. santonensis BRE was seen in the six-month experimental period in contrast to the three-month. Survival at 20 °C was equivalent for both period and, at 25 °C, the survival was higher after three months than after six months. The higher survival observed in the six-month experiment may in part be due to the development of new individuals through reproduction by the neotenics.
An aspect that was not studied was the sensitivity to cold fluctuations. Smith and Rust (1993) suggested that tropical species are better able to survive temperature fluctuations (e.g. rapid rises in temperature) than temperate termites, the latter of which were used to quite stable conditions. Becker (1969b) mentioned that only temperate species were able to survive temperatures below 20 °C. Davis and Kamble (1994) also found that field colonies of Reticulitermes species showed lower supercooling during the colder periods of the year, thus suggesting acclimatisation seems to occur in this genus. Interestingly, the results from Chapter 6 (Table 6.2) suggest that the species studied here survive better at fluctuating temperatures than at constant temperatures.
7.5 Conclusion
Both soil type and temperature were found to be limiting factors for survival of Reticulitermes species. However, the data provided very mixed results and the effect of the different soil types needs to be investigated further. It may be interesting to use OECD (Organisation for Economic Co-ordination and Development) standard soils thus testing the effect of specific changes in soil composition on survival. As Kofoid (1934) stated, moisture and temperature are inseparable. This is also true for soil and moisture as each soil type has a specific moisture holding capacity at a particular temperature.
129 Temperature and soil type
Most temperature experiments on termites, as for most termite experiments, have been rather short-term in relation to the amount of time required for establishment. British summers have been said to provide an insufficient length of time at high temperatures and this is one aspect that is said to limit Reticulitermes species distribution (Harris & Sands, 1965; Nobre & Nunes, 2002). However, survival occurred, at least in R. santonensis BRE, at temperatures as low as 10 °C and reproduction at temperatures of 15 °C over a six-month period. Thus, this species could potentially survive a winter in the United Kingdom and reproduce. This would be even more likely in an urban environment.
Although soil type may be an important factor for establishment in a natural environment, Reticulitermes species are thought not to require soil contact (Weesner, 1965; Forschler, 1998b). Thus, in the absence of soil, establishment could potentially occur if the correct level of moisture and temperature were available.
130 Chapter 8
Summary and General Discussion
This chapter includes a summary of the results of the present study followed by a general discussion about the potential risk that R. grassei and R. santonensis pose on the UK.
8.1 Summary
The following applies to all termite populations tested unless otherwise stated.
• Survival was observed for initial termite numbers as low as five at 25 °C over 12 weeks (4.3.2).
• Increasing initial termite numbers caused a decrease in survival in the laboratory and an increase in survival under field conditions (4.3.1 & 4.3.2).
• Survival decreased with time in laboratory trials (5.3.3).
• The greatest survival in the laboratory was seen in the R. grassei population found in Devon.
• Survival was observed on all wood species tested (oak, beech and pine) (5.3.1).
• The level of consumption in choice experiments depended on other wood species present (5.3.2).
• The results of consumption experiments with living tree seedlings were inconclusive (5.3.4 to 5.3.6).
• Reticulitermes santonensis populations showed a greater activity or movement compared with those of R. grassei (6.3.1 to 6.3.2).
• Reticulitermes grassei UK produced greater numbers of larvae compared with R. santonensis BRE (6.3.3).
• Fluctuating temperature (25 °C to 15 °C and back to 25 °C) gave a more cyclic production of larvae compared with constant temperature (25 °C) and, in R. grassei at least, greater production of larvae (6.3.3).
131 Summary and General Discussion
Soil type was not a significant factor in survival for R. santonensis BRE; results for R. grassei UK were inconclusive (7.3.1).
Termite survival increased with temperature (10, 15, 20 and 25 °C) in three- month laboratory experiments (7.3.2).
Extending the period of exposure to different temperatures (10, 15, 20 and 25 °C) to six months did not affect survival of R. santonensis BRE; results for R. grassei UK were inconclusive (7.3.2).
8.2 Factors affecting term ite establishment
There are a great variety of factors that may have an influence on the establishment of Reticulitermes species in the UK and these are represented in Figure 8.1. The factors examined in this study are in red. Clearly, more work is required to produce an adequate estimate to the risk of establishment of these species. However, it is hoped that the results presented here will aid in pest risk analysis.
The termite numbers required for establishment were found to be quite low however, these results were under ideal conditions. The minimum number of termites required for establishment is likely to be higher under harsher environmental conditions such as low temperatures. Thus the risk of establishment would be much higher in a dwelling where the present results suggest higher ambient temperatures would mean increased termite survival. The wood species tested are commonly used in the construction industry and were all found to sustain termite survival. Termite 'resistant' wood species are rare in the UK as these tend to be tropical hardwoods. The only potential limiting factor that wood species may have is in the rate of development of the termite colony. Soil type did not seem to have a major impact on termite survival and humidity requires further investigation. However, this is unlikely to be a limiting factor in the UK as rainfall is as high and in some cases higher than areas where these species are found in Europe. For alate production, the main concern is the risk of dispersion of established colonies. The present studies suggested that a change in temperature was not sufficient to induce alates and clearly other mechanisms are also involved.
132
Summary and General Discussion
Infestation lest in Climate of point of surrounding area origin Exposure to termites
Presence of elates Likelihood of 4— infested Material elr Presence of Type of storage Humidity 4— attractants e.g. (indoor /outdoor) 4-0 fund
Economic Importance
Likelihood of Level and frequency Mode and length of 4 H import of travel transport
Level of Sento survival Ark's purses
Leal and frequency of travel
Identification by Mode of transport Risk of entry CIA ours A
ot Training of customs officer
• + •
'Nunter of tenviefe and'. Location of °Ate propettion 4-04 destination eurvelng lramenlry
Type and ength el travel from entry tu nnel destration
Risk of 4 Temperature Destination point VP* Establishment
Humidity Field Mae-made structure Substrate I soil type
KEY 0 Season at destination • Season at point of origin • Type of cellulose
Figure 8.1 Conceptual model of factors influencing termites establishment in the UK.
133 Summary and General Discussion
8.3 Climate and termite distribution
One of the aims of this project was to determine potential high-risk areas in the UK. As discussed in the previous section (8.2) one of the important factors in development and establishment of termites is temperature, which in a field situation would be soil temperature.
There are problems with the availability of soil temperature data. Permanent weather stations do not, in general, record daily soil temperatures and if they do, the depth at which temperature is taken varies thus making comparisons virtually impossible (Brown et al., 1999). Models exist to estimate soil temperatures, but, these require parameters which are not always available (Brown et al., 1999).
Soil temperatures can vary depending on the soil type and colour, topography, vegetation cover, as well as precipitation (Edmonds & Campbell, 1984; Walsh et al., 1991; Brown et al., 1999) but air temperature accounts for most of the temperature variation seen in soil (Walsh et al., 1991). Thus, mean air temperature is probably the best available parameter for predicting the effect of temperature on termite distribution.
The 10 °C mean annual isotherm is considered to be the lower temperature limit for termites and thus Reticulitermes species as these are the most northerly distributed species (Light, 1934). This means that potentially parts of South-West England, South Wales and the Republic of Ireland may be at risk of establishment (Figure 8.2). Reticulitermes lucifugus populations (including R. grassei) were reported by Clement (1977a) to be found no further north than the 12 °C isotherm. The species that are found further north are R. santonensis and R. flavipes (Clement et al., 2001).
134 Summary and General Discussion
60N 17
5611 —10 561.1 —15
54N —14 52N —13 50N
—12 4fiN
46N
44N 10
6W AY a 3E 9E 1000mb air (0) Campa$ite M can Jan to Deo: 1950 to 2001 NCEP/N CAR Reanalysis
Figure 8.2 Mean annual air temperature isotherms for Western Europe (from data collected between 1950 and 2001). Image provided by the NOAA-CIRES Climate Diagnostics Center, Boulder Colorado (www.cdc.noaa.gov).
The present study showed that at 10 °C R. santonensis survived for at least six months. However, the soil is likely to act as a buffer to air temperatures and thus termites may be able to escape low winter air temperatures by tunnelling to greater depths is the soil. This clearly seems to be the case in terms of the 'natural' distribution of R. santonensis (Chapter 2 - Fig 2.2) in France which falls between the 6 and 8 °C winter isotherm of the winter temperatures (Figure 8.3). These temperatures are also present in the South-West of England and South Wales, as with the annual mean temperature isotherm, but, in this case Eastern Scotland is also included as is the whole of Ireland. Thus, these are all areas where the termites could potentially survive during the winter.
135 Summary and General Discussion
60N NCIAA.--GIREWCIlnigto Olagnoetie* Urnter 12 5@t1
56N• 19
5.1.N
52N
SON •
48N
46N'
44N
42N
40N .6? SW a 3E 6E 9E 12E 15E 1000mb air (C) Composite Mean Dec to Feb: 1950 to 2001 NCE P/N CAR Re a nalysis
Figure 8.3 Mean winter air temperature isotherms from data between 1950 and 2001. Image provided by the NOAA-CIRES Climate Diagnostics Center, Boulder Colorado (www.cdc.noaa.gov).
The population of R. grassei in Devon was between the 10 and 10.5 °C mean annual temperature isotherm. The infestation was in a domestic environment thus environmental temperature may be less important. However, this was not the case as part of the population was in the soil surrounding the infested properties and thus, the termites' exposure to temperature is likely to have been close to natural environmental conditions.
Climatic factors are only of importance when considering establishment in a natural situation, and thus urban areas are not taken into account (Section 8.2). Thus predicting risk values with modelling systems such as CLIMEX (Sutherst & Maywald, 1985; Sutherst, 1991) is difficult as it would mean having separate parameters for urban and natural environments. Hamburg, for example is found between the 8 and 9 °C mean annual isotherm, yet termite infestations are present.
136 Summary and General Discussion
In France the highest levels of infestation (which includes both R. grassei and R. santonensis) are around the Bordeaux region (Figure 8.4). Areas in the North-East (closer to the UK) are mainly localised and thus probably only urban infestations. The most northerly infestation (in the department of Haute-Normandie) is found, as is Devon, between the 10 and 11 °C mean annual isotherm. The control over accidental import of these species into the UK is made all the more difficult due to the large trade volume (Cannon et al., 1999) and other movement between the UK and France (e.g. second-home owners, tourism etc).
U 75 to 100 % 2 to 10 % • 50 to 75% I 0.5 to 2% 25 to 50 % localised infestation 10 to 25 % ri no infestation
Figure 8.4 Proportion of communes in each department of France reported as being infested with termites. Image provided by the Centre Technique du Bois et de l'Ameublement (CTBA) (www.termites.com.fr).
8.4 Climate change
Species, that are found on the edge of their range are said to be much less stable and thus much more sensitive to environmental change (Cannon, 1998). This is evident in Figure 8.4 where northern France only shows localised infestations. Cannon
137 Summary and General Discussion
(1999) mentions the possibility of glasshouse pests being eventually able to over- winter out-of-doors due to climate change. This may also be the case for termite infestations in urban areas. Higher temperatures due to global warming would be very significant for such populations of Reticulitermes species, that are at the edge of their range. Climate is clearly changing, in Central England the average temperature was 0.5 °C warmer in the 1990s compared to the average from 1961 to 1990 (Hulme et al., 2002). However, the mean annual temperature for the south of England (Figure 8.2) would have to increase by 3 °C to equate to the Bordeaux region where higher termite infestation levels are found (Figure 8.4).
8.5 Conclusion
While the present study is incomplete it is hoped it will aid in future termite risk assessments. Such assessments are difficult because Reticulitermes species are able to establish in urban environments where climatic factors no longer play a key role. The risk posed by termites in natural areas such as forests is much lower although still present in regions which share common environmental conditions (e.g. temperature) to those where termites are common. Despite the Devon infestation, the higher risk seems to be from R. santonensis due to its more northerly distribution in France and its common association with urban areas. Reticulitermes grassei clearly also poses a risk but to a much lesser extent as survival, even in a laboratory environment, is much less stable.
138 Glossary
Term Synonym Definition
Larva White immature First two stages after hatching.
Worker Individual with no wing buds, in the neutral line whose main role it is to feed the colony. Also has role in defence. Pseudergate Individual that has regressed from the sexual line and performs similar functions to the worker. White soldier Pre-soldier, callow soldier, Stage between a worker and a soldier. soldier larva, pseudosoldier, soldier nymph Soldier Individual that develops from the neutral line that defends the colony. Nymph Instars with wing buds, in sexual line which will eventually develop either into a brachypterous neotenic or an alate. Ergatoid Apterous neotenic, third form Reproductive with no wing buds that reproductive develops from the neutral line. Brachypterous Second form reproductive Reproductive with small wing buds neotenic that develops from the sexual line via a 7th instar pre-brachypterous neotenic nymph. Alate Imago Reproductive with wings that develops from the sexual line via a 7th instar long wing bud nymph. Primary First form reproductive Alate which has formed a new colony reproductive after having lost its wings. Secondary Reproductive other than a primary reproductive reproductive. Supplementary Brachyperous neotenic or ergatoid reproductive formed in the presence of a primary reproductive.
139 References
Aber, A. & M. Beltrami. (2002) Reticulitermes lucifugus in urban households in Uruguay (Isoptera: Rhinotermitidae). Sociobiology 40, 153-162.
Aber, A. & L. R. Fontes. (1993) Reticulitermes lucifugus (Isoptera, Rhinotermitidae), a pest of wooden structures, is introduced into the South American continent. Sociobiology 21, 335-339.
Bagnere, A. G., J.-L. Clement, M. Blum, R. F. Severson, C. Joulie & C. Lange. (1990) Cuticular hydrocarbons and defensive compounds of Reticulitermes flavipes (Kollar) and R. santonensis (Feytaud): Polymorphism and chemotaxonomy. Journal of Chemical Ecology 16, 3213-3244.
Baker, J. M. & D. G. Baker. (2002) Long-term ground heat flux and heat storage at a mid-latitude site. Climate Change 54, 295-303.
Banks, N. & T. E. Snyder. (1920) A revision of the nearctic termites. 228 pp. Washington, Washington Government Priniting Office.
Bathelier, J. (1941) Sur le developpement de Leucotermes (Reticulitermes) lucifugus Rossi. Comptes rendus de l'Academie des Sciences 213, 663-665.
Beard, R. L. (1974) Termite biology and bait-block method of control. Connecticut Agricultural Experimental Station Bulletin 748, 1-9.
Becker, G. (1969a) Rearing of termites and testing methods used in the laboratory. pp. 351-385 in K. Krishna and F. M. Weesner (Eds.) Biology of Termites. Vol. 1. London, Academic Press.
Becker, G. (1969b) Termites as laboratory insects. in Termite Symposium, British Wood Preservation Association, p. 81-94.
Becker, G. (1970) Reticulitermes (Ins. Isopt.) in Central and Western Europe. Zeitschrift fur angewandte Entomologie B65/4.3.
Behr, E. A., C. T. Behr & L. F. Wilson. (1972) Influence of wood hardness on feeding by the Eastern subterranean termite, Reticulitermes flavipes (Isopterea: Rhinotermitidae). Annals of the Entomological Society of America 65, 457-460.
140 References
Brown, S. E., K. S. Pregitzer, D. D. Reed & A. J. Burton. (1999) Predicting daily mean soil temperature from daily mean air temperature in four northern hardwood forest stands. Forest Science 46, 297-301.
Buchli, H. (1950a) Recherche sur la fondation et le developpement des nouvelles colonies chez le termite lucifuge (Reticulitermes lucifugus Rossi). Physiologia Comparata et Oecologia 2, 145-160.
Buchli, H. (1950b) La parthenogenese, les rapports entre sexues et leur influence sur la ponte chez Reticulitermes lucifugus Rossi (Rhinotermitidae). Comptes rendus de l'Academie des Sciences 230, 1697-1699.
Buchli, H. (1951) Les potentialites evolutives des castes chez le Reticulitermes de Saintonge (Rhinotermitidae). Comptes rendus de l'Academie des Sciences 233, 206- 208.
Buchli, H. (1956) Le cycle de developpement des castes chez Reticulitermes. Insectes Sociaux 3, 395-401.
Buchli, H. (1958) L'origine des castes et les potentialites ontogeniques des termites Europeens du genre Reticulitermes Holmgren. Annales Science Naturelle, Zoologique 20, 261-429.
Buchli, H. (1961) Les relations entre la colonie maternelle et les jeunes imagos ailes de Reticulitermes lucifugus. Vie et Milieu 12, 627-632.
Butterworth, D., D. Kay & B. J. MacNulty. (1958) Testing for termite resistance. Part II. The dry-wood termite Cryptotermes Havilandi (Sjost), 20 pp. Ministry of Supply, Tropical Testing Establishment, Port Harcourt, Nigeria.
Cals-Uscati, J. & J. de Frescheville. (1963) Perermite d'une colonie de termites a Paris. Cahiers des Naturalistes, Bulletin N. P. 19, 54.
Cannon, R. J. (1998) The implications of predicted climate change for insect pests in the UK, with emphasis on non-indigenous species. Global Change Biology 4, 785- 796.
Cannon, R. J., A. W. Pemberton & P. W. Bartlett. (1999) Appropriate measures for the eradication of unlisted pests. EPPO Bulletin 29, 29-36.
141 References
Carter, F. L. & T. R. Dell. (1981) Screening selected american hardwoods for natural resistance to a native subterranean termites, Reticulitermes flavipes (Kollar), 10 pp. USDA Forest Service, New Orleans, Louisiana.
Carter, F. L., C. A. Stringer & R. V. Smythe. (1972) Survival of six colonies of Reticulitermes flavipes on unfavourable woods. Annals of the Entomological Society of America 65, 984-985.
Castel, G. B. (1934) The experimental determination of caste differentiation in termites. Science 80, 314.
Chaine, J. (1910) Termite et Plantes Vivantes. I Degats occasionnes aux arbres. Comptes rendus de la Societe de Biologie de Bordeaux, 328-330.
Clement, J.-L. (1976) Contribution a la systematique evolutive et biologique des Reticulitermes francais (Isoptere), 97 pp. PhD Dissertation, Universite Pierre et Marie Curie, Paris.
Clement, J.-L. (1977a) Ecologic des Reticulitermes (Holmgren) Francais (Isopteres) position systematique des population. Bulletin de la Societe Zoologique de France 102, 169-185.
Clement, J.-L. (1977b) Speciation in the French populations of Reticulitermes. in Proceedings of the 8th International Congress of IUSSI, p. 297.
Clement, J.-L. (1978a) Nouveaux criteres taxinomiques dans le genre Reticulitermes (Holmgren) [Isoptera] description de nouveaux taxons francais. Annales de la Societe Entomologique de France (N. S.) 14, 131-139.
Clement, J.-L. (1978b) L'agression interspecifique et intraspecifique des especes du genre Reticulitermes (Isoptere). Comptes rendus de PAcademie des Sciences 286, 351-354.
Clement, J.-L. (1979a) Etude biometrique des populations de Reticulitermes (Isopteres) francais (R. lucifugus, R. santonensis et R. banyulensis sp. nov.) et de populations americaines de R. flavipes. Archives de Zoologie Experimentale et Generale 120, 65-87.
Clement, J.-L. (1979b) Hybridation experimentale entre Reticulitermes santonensis Feytaud et Reticulitermes lucifugus Rossi. Annales de Sciences Naturelles, Zoologie 13, 251-260.
142 References
Clement, J.-L. (1981) Enzymatic polymorphism in the european populations of various Reticulitermes species (Isoptera). Systematic Association Special Volume: Biosystematics of Social Insects 19, 49-64.
Clement, J.-L. (1982a) Pheromones d'attraction sexuelle des termites europeens du genre Reticulitermes (Rhinotermitinae). Mechanismes comportementaux et isolements specifiques. Biology of Behaviour 7, 55-68.
Clement, J.-L. (1982b) Les termites Europeens du genre Reticulitermes I. - Statut specifique des population. Comptes rendus des Seances de la Societe de Biogeographie 58, 145-158.
Clement, J.-L. (1984) Diagnostic alleles and systematics in termites species of the genus Reticulitermes in Europe. Experientia 40, 283-285.
Clement, J.-L. (1985) Chimosystematique du genre Reticulitermes (Isopteres) aux U.S.A. et en Europe. Actes des Colloques Insectes Sociaux 2, 123-131.
Clement, J.-L. (1986) Open and closed societies in Reticulitermes termites (Isoptera, Rhinotermitidae): Geographic and seasonal variations. Sociobiology 11, 311-323.
Clement, J.-L., R. Howard, M. Blum & H. Lloyd. (1986) L'isolement specifique des termites du genre Reticulitermes (Isoptera) du sud-est des Etas-Unis. Mise en evidence grace a la chimie et au comportement d'une espece jumelle de R. virginicus = R. malletei sp. nov. et d'une semi-species de R. flavipes. Comptes rendus de PAcademie des Sciences 302, 67-70.
Clement, J.-L., A. G. Bagnere, P. Uva, L. Wilfert, A. Quintana, J. Reinhard & S. Dronnet. (2001) Biosystematics of Reticulitermes termites in Europe: morphology, chemical and molecular data. Insectes Sociaux 48, 202-215.
Crawley, M. J. (1998) Structure of Plant Communities. 2nd edn. 736 pp. Oxford, Blackwell Science.
Crawley, M. J. (2002) Statistical computing: An introduction to data analysis using S-Plus. 761 pp. Chichester, John Wiley & Sons.
Crossley, D. A. & M. P. Hoglund. (1962) A litter-bag method for the study of microarthropods inhabiting leaf litter. Ecology 43, 571-573.
143 References
Culliney, T. W. & J. K. Grace. (2000) Prospects for the biological control of subterranean termites (Isoptera: Rhinotermitidae), with special reference to Coptotermes forrnosanus. Bulletin of Entomological Research 90, 9-21.
Dangerfield, J. M. & D. K. Mosugelo. (1997) Termite foraging on toilet roll baits in semi-arid savanna, South-East Botswana (Isoptera: Termitidae). Sociobiology 30, 133-143.
Davidoff, B., J. W. Lewis & H. M. Selim. (1986) Variability of soil temperature with depth along a transect. Soil Science 142, 114-123.
Davis, R. W. & S. T. Kamble. (1994) Low temperature effects on survival on the Eastern subterranean termite (Isoptera:Rhinotermitidae). Environmental Entomology 23, 1211-1214. de Feytaud, J. (1955) Extension du peril des termites en France. Naturalia 19, 32-38. de Feytaud, J. (1959) L'hisoire veridique du termite de Saintonge. Cahiers de l'Ouest 30, 1-11. de Quatrefages, M. A. (1853) Note sur les termites de La Rochelle. Annales Science Naturelle, Zoologique 30, 16-21.
Edmonds, W. J. & J. B. Campbell. (1984) Spatial estimates of soil temperature. Soil Science 138, 203-208.
Edwards, R. & A. E. Mill. (1986) Termites in buildings: Their biology and control. 261 pp. East Grinstead, Rentokil Limited.
Eggleton, P. (1999) Termite species description rates and the state of termite taxonomy. Insectes Sociaux 46, 1-5.
Esenther, G. R. (1961) The biology of the Eastern subterranean termite, Reticulitermes flavipes (Kollar), in Wisconsin and related experiments with other species, pp. 80. PhD dissertation, University of Wisconsin, Madison.
Esenther, G. R. (1969) Termites in Wisconsin. Annals of the Entomological Society of America 62, 1274 - 1284.
Esenther, G. R. (1977) Fluctuations of survival, growth and differentiation in biweekly samples of Reticulitermes flavipes workers. Material and Organismen 12, 49-58.
144 References
Esenther, G. R. (1980) Estimating the size of subterranean termite colonies by a release-recapture technique. in 11th annual meeting of the International Research Group on Wood Preservation, Raleigh, North Carolina, IRG Secretariat, IRG/WP/1112.
Ferrari, R. & M. Marini. (1999) Subterranean termite Reticulitermes spp. (Isoptera: Rhinotermitidae) baiting control in historical public buildings in Italy. in 3rd International Conference on Urban Pests, Prague, Czech Republic, Graficke Zavody, Hronov, p. 357-365.
Feytaud, J. (1911) Les ravages du termite Lucifuge dans les villes. Bulletin de la Societe Entomologique et Agricole de Bordeaux, 150-160.
Feytaud, J. (1912) Contribution a l'etude du termite Lucifuge (Anatomie - fondation des colonies nouvelles). Archives d'Anatomie Microscopique 13, 481-607.
Feytaud, J. (1920) Sur les jeunes colonies du termite lucifuge. Comptes rendus des seances de ?Academie des Sciences 171, 203-206.
Feytaud, J. (1924) Le termite de Saintonge. Comptes rendus des seances de l'Academie des Sciences 178, 241-244.
Feytaud, J. (1925) Existe-t-il plusieurs races de Reticulitermes lucifugus Rossi? Bulletin de la Societe de Zoologie agricole 24, 161-169.
Feytaud, J. (1946) Le peuple des termites. 127 pp. Paris, Presses Universitaire de France.
Feytaud, J. (1951) A propos de Reticulitermes de France. in Eighth International Congress of Entomology, 1948, Stockholm, 2 pp.
Forschler, B. T. (1994a) Fluorescent spray paint as a topical marker on subterranean termites (Isoptera: Rhinotermitidae). Sociobiology 23, 27-38.
Forschler, B. T. (1994b) Survivorship and tunneling activity of Reticulitermes flavipes (Kollar) (Isoptera: Rhinotermitidae) in response to termiticide soil barriers with and without gaps of untreated soil. Journal of Entomological Science 29, 43-54.
Forschler, B. T. (1996) Incidence of feeding by the Eastern subterranean termite (Isoptera: Rhinotermitidae) in Laboratory Bioassay. Sociobiology 28, 265-273.
145 References
Forschler, B. T. (1998a) Termite baiting: A research perspective. Pest Control Technology, 90,92,94,96,98,99.
Forschler, B. T. (1998b) Part II. Subterranean termite biology in relation to prevention and removal of structure infestation. pp. 31-51 NPCA Research report on Subterranean Termites. Viginia, National Pest Control Association.
Forschler, B. T. & G. Henderson. (1995) Subterranean termite behavioral reaction to water and survival of inundation: Implications for field populations. Environmental Entomology 24, 1592-1597.
Forschler, B. T. & T. M. Jenkins. (1999) Evaluation of subterranean termite biology using genetic, chemotaxonomic and morphometric markers and ecological data: a testimonial for multi-disiplinary efforts. Trends in Entomology 2, 71-80.
Forschler, B. T. & J. W. Lewis. (1997) Why termites can dodge your treatment. Pest Control 65, 42-46,53.
Forschler, B. T. & M. L. Townsend. (1996) Mortality of Eastern subterranean termites (Isoptera:Rhinotermitidae) exposed to four soils treated with termiticides. Journal of Economic Entomology 89, 678-681.
Fougerousse, M. & J. Perlade. (1975) The present status of termites in Europe. in Record of the 1975 Annual Convention of the British Wood Preserving Association, 24th to 27th June 1975 Cambridge, British Wood Preserving Association, p. 35-49.
Gassies, J. B. (1855) De l'introduction des termites dans la ville de Bordeaux. Actes de la Societe Linneenne de Bordeaux 20, 427-428.
Gay, F. J. (1967) A world review of introduced species of termites, 88 pp. Commonwealth Scientific and Industrial Research Organization, Melbourne.
Gay, F. J., T. Greaves, F. G. Holdaway & A. H. Wetherly. (1955) Standard laboratory colonies of termites for evaluating the resistance of timber, timber preservatives,and other materials to termite attack., 62 pp. CSIRO, Melbourne.
Getty, G. M., M. I. Haverty & V. R. Lewis. (2000) Agonistic behaviour between recently collected and laboratory cultured Reticulitermes spp. (Isoptera: Rhinotermitidae) for northern California. Pan-Pacific Entomologist 76, 243-250.
146 References
Getty, G. M., M. I. Haverty, K. A. Copren & V. R. Lewis. (1999) Seasonal foraging behavior of Reticulitermes spp (Isoptera: Rhinotermitidae) in northern California, USA. in 3rd International Conference on Urban Pests, 19th to 22nd July 1999 Prague, Czech Republic, Graficke Zavody, Hronov, p. 347-351.
Grace, J. K. (1989) Northern subterranean termites. Pest Management 8, 14-16.
Grace, J. K. (1996) Temporal and spatial variation in caste proportions in a northen Reticulitermes flavipes colony (Isoptera: Rhinotermitidae). Sociobiology 28, 225- 231.
Grace, J. K. (1997) Influence of tree extractives on foraging preferences of Reticulitermes flavipes (Isoptera: Rhinotermitidae). Sociobiology 30, 35-42.
Grace, J. K. & J. R. Yates. (1999) Termite resistant construction and building materials. in 3rd International Conference on Urban Pests, Prague, Czech Republic, Graficke Zavody, Hronov, p. 399-406.
Grace, J. K., J. R. Yates III, C. H. M. Tome & R. J. Oshiro. (1997) Use of a stainless steel mesh to exclude formosan subterranean termites. in 28th Annual Meeting of the International Research Group on Wood Preservation, Whistler, Canada, IRG Secretariat, IRG/WP 97-10199.
Grasse, P.-P. (1949) Ordre des Isoptere ou Termites. pp. 408-544 Traite de Zoologie IX. Paris, Masson et cie.
Grasse, P.-P. (1954) Origine et repartition geographique des termites francais. Annales de l'Ecole Nationale d'Agriculture de Montpellier 29, 17-21.
Grasse, P.-P. & R. Chauvin. (1946) L'effet de groupe et la survie des neutres dans les societe d'insectes. Revue Scientific 82, 461-464.
Grasse, P.-P., C. Noirot, G. Clement & H. Buchli. (1950) Sur la signification de la caste des ouvriers chez les Termites. in Comptes rendus des seances de l'Academie des Sciences, p. 892-895.
Grassi, B. & A. Sandias. (1893) Constituzione e sviluppo della societa dei Termitidi. The constitution and development of the society of termites; observations on their habits with appendices on the parasitic protozoa and on the Embiidae, translated by T. H. Blandford, Quarterly Journal Microscopical Science, 39, 3, 245-322. Atti Accademia Gioenia dei Scienze Naturali, Catania 6, 7, 151 pp.
147 References
Grube, S. & D. Rudolph. (1999) Water supply during building activities in the subterranean termite Reticulitermes santonensis De Feytaud (Isoptera, Rhinotermitidae). Insectes Sociaux 46, 192-193.
Gupta, S. C., J. K. Radke, W. E. Larson & M. J. Shaffer. (1982) Predicting temperatures of bare- and residue-covered soils from daily maximum and minimum air temperatures. Soil Science 46, 372-376.
Hare, L. (1934) Caste determination with special reference to the genus Reticulitermes (Isoptera). Journal of Morphology 56, 267-293.
Harris, W. V. (1958) Colony formation in the Isoptera. in Tenth International Congress of Entomology, Montreal, p. 435-439.
Harris, W. V. (1962) Termites in Europe. New Scientist 135, 614-617.
Harris, W. V. (1970) Termites of the Palearctic region. pp. 295-313 in K. Krishna and F. M. Weesner (Eds.) Biology of Termites. Vol. 2. New York and London, Academic Press.
Harris, W. V. & W. A. Sands. (1965) The social organisation of termite colonies. Syposium of the Zoological Society 14, 113-131.
Haverty, M. I. (1977) A proportion of soldiers in termite colonies: a list and a bibliography (Isoptera). Sociobiology 2, 199-216.
Haverty, M. I. (1979) Selection of tunneling substrates for laboratory studies with three subterranean termite species. Sociobiology 4, 315-320.
Haverty, M. I. & R. Howard. (1981) Production of soldiers and maintenance of soldier proportions by laboratory experimental groups of Reticulitermes flavipes (Kollar) and Reticulitermes virginicus (Banks) (Isoptera: Rhinotermitidae). Insectes Sociaux 28, 32-39.
Haverty, M. I. & W. L. Nutting. (1974) Natural wood-consumption rates and survival of a dry-wood and a subterranean termite at constant temperatures. Annals of the Entomological Society of America 67, 153-157.
Haverty, M. I., B. T. Forschler & L. J. Nelson. (1996) An assessment of the taxonomy of Reticulitermes (Isoptera: Rhinotermitidae) from the southeastern United States based on cuticular hydrocarbons. Sociobiology 28, 287-318.
148 References
Haverty, M. I., L. J. Nelson & B. T. Forschler. (1999a) New cuticular hydrocarbon phenotypes of Reticulitermes (Isoptera: Rhinotermitidae) from the United States. Sociobiology 34, 1-21.
Haverty, M. I., K. A. Copren, G. M. Getty & V. R. Lewis. (1999c) Agonistic behaviour and cuticular hydrocarbon phenotypes of colonies of Reticulitermes (Isoptera:Rhnotermitidae) from northern California. Annals of the Entomological Society of America 92, 269-277.
Haverty, M. I., G. M. Getty, K. A. Copren & V. R. Lewis. (1999b) Seasonal foraging and feeding behavior of Reticulitermes spp. (Isoptera:Rhinotermiticae) in a wildland
and a residential location in northern California. Environmental Entomology 28, 1077-1084.
Haverty, M. I., M. Page, B. L. Thorne & P. Escoubas. (1991) Cuticular hydrocarbons: Species and population-level discrimination in termites, pp. 15-23. USDA Forest Service.
Hendee, E. C. (1937) Order Isoptera: Laboratory colonies of termites. pp. 275-279 in P. S. Galtsoff, F. E. Lutz, P. S. Welch and J. G. Needham (Eds.) Culture Methods for Invertebrate Animals. New York, Comstock Publishing Company.
Hickin, N. E. (1961) Maintenance of termite colonies in England for research purposes. Pest Technology 3, 84-86.
Hickin, N. E. (1969) An introduction to the study of termites VII: External anatomy, pp. 1-3. British Wood Preserving Association News Sheet.
Howard, R. & M. I. Haverty. (1981) Seasonal variation in caste proportions of field colonies of Reticulitermes flavipes (Kollar). Environmental Entomology 10, 546-549.
Howick, C. D. (1975) Influences of specimen size, test period and matrix on the amounts of wood eaten by similar groups of laboratory termites. in Record of the 1975 Annual Convention of the British Wood Preserving Association, Cambridge, Britsh Wood Preservation Society, p. 51-63.
Howick, C. D. (1978) Influence of crowding during transportation on survival and laboratory feeding of termites (Isoptera). Journal of the Institute of Wood Science 8, 28-31.
149 References
Hrdy, I. (1961) Contribution to the knowledge of european species of the genus Reticulitermes (Holmgren). Acta Faunistica Entomologica Musei Nationalis Pragae 7, 97-107.
Hrdy, I. & J. Krecek. (1972) Development of superfluous soldiers induced by juvenile hormone analogues in the termite Reticulitermes lucifugus santonensis. Insectes Sociaux 2, 105-109.
Hulme, M., G. I. Jenkins, X. Lu, J. R. Turnpenny, T. D. Mitchell, R. G. Jones, J. Lowe, J. M. Murphy, D. Hassell, P. Boorman, R. McDonald & S. Hill. (2002) Climate change scenarios for the United Kingdom: The UKCIP02 scientific report, 120 pp. Tyndall Cantre for Climate Change Research, School of Environmental Sciences, University of East Anglia, Norwich.
Ionescu, M. A. (1959) La repartition geographique et l'ecologie des termites dans le sud-est de l'europe. Revue de Biologie, Bucaristi 4, 107-116.
Jacquiot, C. (1955) Les colonies Parisiennes du termite de Saintonge. La Revue du Bois 10, 15-17.
Jacquiot, C. (1956) Les termites de France. CTBA Document, 7 pp.
Jenkins, T. M., R. E. Dean, R. Verkerk & B. T. Forschler. (2001) Phylogenetic analyses of two mitochondrial genes and a nuclear intron region illumnate european subterranean termite (Isoptera: Rhinotermitidae) gene flow, taxonomy and introduction dynamics. Molecula Phylogenetics and Evolution 20, 286-293.
Jenkins, T. M., C. J. Basten, R. Dean, S. E. Mitchell, S. Kresovich & B. T. Forschler. (1998) Matriarchal genetic structure of Reticulitermes (Isoptera:Rhinotermitidae) populations. Sociobiology 33, 239-263.
Jenkins, T. M., M. I. Haverty, C. J. Basten, L. J. Nelson, M. Page & B. T. Forschler. (2000) Correlation of mitochondrial haplotypes with cuticular hydrocarbon phenotypes of sympatric Reticulitermes species from the southeastern United States. Journal of Chemical Ecology 26, 1525-1542.
Kennedy, M. J., L. L. Dixon & B. C. Peters. (1994) Susceptibility of heartwood of three Pinus species to attack by the subterranean termite Coptotermes acinaciforrnis (Froggatt). in 25th Annual meeting of the International Research Group on Wood Preservation, Bali, Indonesia, IRG Secretariat, IRG/WP 94-20026.
150 References
Kofoid, C. A. (1934) Climatic factors affecting the local occurrence of termites and their geographical distribution. pp. 13-21 in C. A. Kofoid, S. F. Light, A. C. Homer, M. Randall, W. B. Herms and E. E. Bowe (Eds.) Termites and termite control. Berkley, University of California Press.
Krishna, K. (1989) Order Isoptera: Termites. pp. 234-241 in D. J. Borror, C. A. Triplehorn and N. F. Johnson (Eds.) An introduction to the study of insects. Philadelphia, Saunders College Publishing.
Lavers, G. M. (1983) The strength properties of timber, 65 pp. Building Research Establishment.
Le Bayon, I., D. Ansard, C. Brunet, I. Paulmier & A.-M. Pruvost. (1999) Biocontrol of Reticulitermes santonensis by entomopathogenic fungi. in 30th Annual Meeting of the International Research Group on Wood Preservation, Rosenheim, Germany, IRG Secretariat, IRG/WP 99-10324.
Lee, K. E. & T. G. Wood. (1971) Termites and soils. 251 pp. New Nork, Academic Press.
Lenz, M. (1976) The dependence of hormone effects in termite caste determination on external factors. pp. 73-89 in M. Luscher (Ed) Phase and caste determination in insects: Endocrine aspects. Oxford, Pergamon Press.
Lenz, M. & R. A. Barrett. (1982) Neotenic formation in field colonies of Coptotermes lacteus (Froggatt) in Australia, with comments on the roles of neotenics in the genus Coptotermes (Isoptera: Rhinotermitidae). Sociobiology 7, 47-59.
Lenz, M. & S. Runko. (1993) Long term impact of orphaning on field colonies of Coptotermes lacteus (Froggatt) (Isoptera: Rhinotermitidae). Insectes Sociaux 40, 439-456.
Lenz, M. & E. R. Williams. (1980) Influence of container, matrix volume and group size on survival and feeding activity in species of Coptotermes and Nasutitermes (Isoptera: Rhinotermitidae; Termitidae). Material and Organismen 15, 25-46.
Lenz, M., R. A. Barrett & E. R. Williams. (1982) Influence of diet on the survival and wood consumption of Protermes adamsoni (Froggatt) (Isoptera: Termopsidae) at different temperatures. Bulletin of Entomological Research 72, 423-435.
151 References
Lesne, P. (1923) Une station nouvelle du Termite lucifuge. Comptes rendus de l'Academie des Sciences 176, 1507-1508.
Lespes, C. (1856) Recherches sur l'organisation et les moeurs du termite Lucifuge. Annales Science Naturelle, Zoologie 4, 227-282.
Light, S. F. (1934) A world view of termites. pp. 117-126 in C. A. Kofoid, S. F. Light, A. C. Horner, M. Randall, W. B. Herms and E. E. Bowe (Eds.) Termites and termite control. Berkley, University of California Press.
Light, S. F. & P. L. Illg. (1945) Rate and extent of development of neotenic reproductives in groups of nymphs of the termite genus Zootermopsis. University of California Publications in Zoology 53, 1-40.
Light, S. F. & F. M. Weesner. (1947) Methods for culturing termites. Science 106, 131-132.
Lund, A. E. (1962) Subterraneans and their environment: New concepts of termite ecology. Pest Control 30, 30-61.
Luscher, M. (1949) Continuous observation of termites in laboratory cultures. Acta Tropica 6, 161-165.
Luscher, M. (1960) Hormonal control of caste differentiation in termites. Annals of the New York Academy Sciences 89, 549-563.
Mannesmann, R. (1973) Comparison of twenty-one commercial wood species from North America in relation to feeding rates of the formosan termites, Coptotermes formosanus Shiraki. Material and Organismen 8, 107-120.
Matsuura, K. & T. Nishida. (2001) Comparison of colony foundation success between sexual pairs and female asexual units in the termite Reticulitermes speratus (Isoptera: Rhinotermitidae). Population Ecology 43, 119-124.
Miller, M. E. (1969) Caste differentiation in the lower termites. pp. 283-310 in K. Krishna and F. M. Weesner (Eds.) Biology of Termites. Vol. 1. New York and London, Academic Press.
Myles, T. G. (1997) Penetrability of some Ontario construction aggregates by the Eastern subterranean termite (Isoptera: Rhinotermitidae). Sociobiology 30, 277-288.
152 References
Nobre, T. & L. Nunes. (2002) Subterranean termites in Portugal. Tentative model of distribution. in 33rd Annual Meeting of the International Research Group on Wood Preservation, 12-17 May 2002 Cardiff UK, IRG Secretariat, IRG/WP 02-10420.
Noirot, C. (1985) Pathways of caste development in the lower termites. pp. 41-57 in J. A. L. Watson, B. M. Okoi-Kotber and C. Noirot (Eds.) Caste Differentiation in Social Insects. Oxford, Pergamon Press.
Noirot, C. (1990) Sexual castes and reproductive strategies in termites. pp. 5-35 in W. Engels (Ed.) An Evolutionary Approach to Castes and Reproduction. Berlin, Springer-Verlag.
Noirot, C. & J. M. Pasteels. (1987) Ontogenetic development and evolution of the worker caste in termites. Experientia 43, 851-860.
Noirot, C. & J. M. Pasteels. (1988) The worker caste is polyphyletic in termites. Sociobiology 14, 15-20.
Nutting, W. L. (1969) Flight and colony foundation. pp. 233-282 in K. Krishna and F. M. Weesner (Eds.) Biology of Termites. New York, Academic Press.
Nutting, W. L. (1990) Chapter 33. Insecta: Isoptera. pp. 997-1032 in D. L. Dindal (Ed.) Soil Biology Guide. New York, John Wiley & Sons.
0i, F. M., N.-Y. Su, P. G. Koehler & F. Slansky. (1996) Laboratory evaluation of food placement and food types on the feeding preference of Reticulitermes virginicus (Isoptera: Rhinotermitidae). Journal of Economic Entomology 89, 915-921.
Osmun, J. V. (1957) Three entomological laboratory techniques. I. Rearing methods for subterranean termites. Proceedings of the Indiana Academy of Science 66, 141- 143.
Passera, L. (1994) Characteristics of tramp species. pp. 23-43 in D. F. Williams (Ed.) Exotic Ants: Biology, Impact and Control of Introduced Species. Boulder, Westview Press.
Paulmier, I., B. Vauchot, A.-M. Pruvost, C. Lohou, M. Tussac, M. Jequel, J.-L. Leca & J.-L. Clement. (1997) Evaluation of two populations of Reticulitermes santonensis De Feytaud (Isoptera) by triple mark-recapture procedure. in 28th Annual Meeting of the International Research Group on Wood Preservation, 25-30 May 1997 Whistler, Canada, IRG Secretariat, IRG/WP 97-10237.
153 References
Pawson, B. M. & R. E. Gold. (1996) Caste differentiation and reproductive dynamics of the three subterranean termites in the genus Reticulitermes (Isoptera: Rhinotermitidae). Sociobiology 28, 241-25 1 .
Pearce, M. J. (1997) Termite biology and pest management. 172 pp. Oxon, CAB International.
Perez, C. (1907) Les termites dans le sud-ouest de la France. Bulletin de la Societe de Zoologie et Agriculture.
Perez, J. (1895) Le termite Lucifuge. Revue Sciences Naturelles appliquees, 56-62.
Plateaux, L. & J.-L. Clement. (1984) La speciation recente des termites Reticulitermes du complex lucifugus. Revue de la Faculte de Science de Tunis 3, 179-206.
Polizzi, J. M. & B. T. Forschler. (1998) Intra- and interspecific agonism in Reticulitermes flavipes (Kollar) and R. viginicus (Banks) and effects of arena and group size in laboratory assays. Insectes Sociaux 45, 43-49.
Ripa, R., L. Castro, N.-Y. Su & P. Palma. (2002) Laboratory estimate of wood consumption rates by Reticulitermes sp. (Isoptera: Rhinotermitidae) in Chile. Sociobiology 39, 285-290.
Roisin, Y. & M. Lenz. (1999) Caste developmental pathways in colonies of Coptotermes lacteus (Froggatt) headed by primary reproductives (Isoptera, Rhinotermitidae). Insectes Sociaux 46, 273-280.
Sands, W. A. (1998) The identification of worker castes of termite genera from soils of Africa and the Middle East. 497 pp. Wallingford, CAB International.
Sen-Sarma, P. K. (1994) Accelerated anti-termite laboratory tests, simulation of field conditions and assessment of results for service life. in 25th annual meeting of the International Reasearch Group on Wood Preservation, 30 May - 3 June 1994 Bali, Indonesia, IRG secretariat, IRG/WP 94-20034.
Serment, M.-M. & 0. Tourteaux. (1991) Les termites. CTBA Brochure ISSN 0528- 4929.
154 References
Smith, J. L. & M. K. Rust. (1993) Influence of temperature on tunnelling, feeding rates, and oxygen requirements of the western subterranean termite, Reticulitermes hesperus (Isoptera:Rhinotermitidae). Sociobiology 21, 225-236.
Smith, V. K., J. D. Tynes & K. E. Tynes. (1969) Maintain your own termite colony. Pest Control 37, 30,32.
Smythe, R. V. & F. L. Carter. (1969) Feeding responses to sound wood by the Eastern subterranean termites, Reticulitermes flavipes. Annals of the Entomological Society of America 62, 335-337.
Smythe, R. V. & F. L. Carter. (1970) Survival and behavior of three subterranean 'termite species in sawdust of eleven wood species. Annals of the Entomological Society of America 63, 847-850.
Smythe, R. V. & D. F. Williams. (1972) Feeding and survival of two subterranean termite species at constant temperatures. Annals of the Entomological Society of America 65, 226-229.
Snyder, T. E. (1912) Record of the finding of a true queen of Termes flavipes Kol. Proceedings of the Entomological Society of Washington 14, 107-108.
Snyder, T. E. (1915) Insects injurious to forests and forest products. Biology of the termites of the eastern United States, with preventative and remedial measures. U.S. Department of Agriculture Bureau of Entomology Bulletin 94, 13-95.
Snyder, T. E. (1920) The colonizing reproductive adults of termites. Proceedings of the Entomological Society of Washington 22, 109-150.
Snyder, T. E. (1925) The origin of the castes in termites. Biological Society of Washington 38, 57-67.
Snyder, T. E. (1926) The biology of the termite castes. Quarterly Review of Biology 1, 522-552.
Strickland, M. (1950) Difference in tollerance of drying between species of termites (Reticulitermes). Ecology 31, 373-385.
Su, N.-Y. (2002) Novel technologies for subterranean termite control. Sociobiology 40, 95-101.
155 References
Su, N.-Y. & J. P. La Fage. (1984) Comparison of laboratory methods for estimating wood consumption rates by Coptotermes formosanus (Isoptera: Rhinotermitidae). Annals of the Entomological Society of America 77, 125-129.
Su, N.-Y. & J. P. La Fage. (1987) Effects of soldier proportion on the wood consumption rate of the formosan subterranean termites (Isoptera: Rhinotermitidae). Sociobiology 13, 145-151.
Suarez, M. E. & B. L. Thorne. (2000) Rate, amount and distribution pattern of alimentary fluid transfer via trophallaxis in three species of termites (Isoptera: Rhinotermitidae, Termopsidae). Annals of the Entomological Society of America 93, 145-155.
Sutherst, R. W. (1991) Predicting the survival of immigrant insect pests in new environments. Crop Protection 10, 331-332.
Sutherst, R. W. & G. F. Maywald. (1985) A computerised system for matching climates in ecology. Agriculture, Ecosystems and Environment 13, 281-299.
Takahashi, M. & T. Yoshimpra. (2002) Recent development in the control of Japanese subterranean termites. Sociobiology 40, 13-23.
Takematsu, Y. (1992) Biometrical study on the development of the castes in Reticulitermes speratus (Isoptera, Rhinotermitidae). Japanese Journal of Entomology 60, 67-76.
Thompson, C. B. (1917) Origin of the castes of the common termite Leucotermes flavipes Kol. Journal of Morphology 30, 83-153.
Thorne, B. L. (1996) Termite terminology. Sociobiology 28, 253-261.
Thorne, B. L. (1998) Part I. Biology of subterranean termites of the genus Reticulitermes. pp. 1-30 NPCA Research Report on Subterranean Termites. Viginia, National Pest Control Association.
Thorne, B. L. & M. I. Haverty. (1991) A review of intracolony intraspecific agonism in termites. Sociobiology 19, 115-145.
Thorne, B. L., J. F. A. Traniello, E. S. Adams & M. Bulmer. (1999) Reproductive dynamics and colony structure of subterranean termites of the genus Reticulitermes
156 References
(Isoptera Rhinotermitidae): a review of the evidence from behavioural, ecological, and genetic studies. Ethology, Ecology and Evolution 11, 149-169.
Thorne, B. L., E. Russek-Cohen, B. T. Forschler, N. L. Breisch & J. F. A. Traniello. (1996) Evaluation of mark-release-recapture methods for estimating forager population size of subterranean termite (Isoptera: Rhinotermitidae) colonies. Environmental Entomology 25, 938-951.
Vauchot, B., E. Provost, A. G. Bagnere & J.-L. Clement. (1996) Regulation of the chemical signatures of two termite species, Reticulitermes santonensis and Reticulitermes lucifugus grassei, living in mixed experimental colonies. Journal of Insect Physiology 42, 309-321.
Verkerk, R. (1990) Building out termites: an Australian manual for environmentally responsible control. 211 pp. Leichhardt, Pluto Press Australia Ltd.
Verkerk, R. H. J. (1998) Termites and building: assessing the problem from a non- comercial viewpoint. in Proceedings of the 1998 Annual Convention of the British Wood Preservation and Damp-Proofing Association, London, BWPDA, 8 pp.
Verkerk, R. H. J. & A. F. Bravery. (2000) The UK termite eradication programme: justification and implementation. in The 31st Annual Meeting of the International Research Group on Wood Preservation, 14-19 May 2000 Kona, Hawaii, IRG Secretariat, IRG/WP 00 - 10373.
Vieau, F. (1991) Le termite de Saintonge: un danger pour l'Ouest de la France. Penn ar bed, Revue de la Societe pour l'etude de la protection de la nature en Bretagne (SEPNB) 140, 19-32.
Vieau, F. (1993) Le termite de Saintonge Reticulitermes santonensis Feytaud: Termite urbain. Bulletin de la Societe Zoolgique de France 118, 125-133.
Vieau, F. (1994a) Les stades nymphaux 7 et 8 chez Reticulitermes santonensis Feytaud (Isoptera, Rhinotermitidae). Actes des Colloques Insectes Sociaux 9, 61-66.
Vieau, F. (1994b) Formation of the reproductives in Reticulitermes santonensis Feytaud (Isoptera, Rhinotermitidae). in 12th Congress of the International Union for the Study of Social Insects IUSSI, 21-27 August 1994 Paris, Sorbonne, Universite Paris Nord, p. 255.
157 References
Vieau, F. (1996) Les variations saisonnieres des castes chez le termite Reticulitermes santonensis Feytaud (Isoptera: Rhinotermitidae) dans un biotype de l'ouest de la France. Annales de la Societe Entomologique de France (N.S) 32, 207-216.
Vieau, F. (1999) Biologie comparee de Reticulitermes santonensis Feytaud et Reticulitermes lucifugus Rossi en France: differences morphologiques entre les soldats, mode d'implantations urbaine et forestiere, cycles reproducteurs. Insectes Sociaux 12, 151-158.
Vieau, F. (2000) Quelques remarques morphologiques et histologiques sur un individu rarement observe en France: la femelle physogastre des reproducteurs primaires de Reticulitermes lucifugus grassei Clement (Isoptera, Rhinotermitidae). Insectes Sociaux 13, 133-137.
Vieau, F. (2001) Comparison of the spatial distribution and reproductive cycle of Reticulitermes santonensis Feytaud and Reticulitermes lucifugus grassei Clement (Isoptera, Rhinotermitidae) suggests that they represent introduced and native species, respectively. Insectes Sociaux 48, 57-62.
Waller, D. A. (1988) Host selection in subterranean termites: factors affecting choice (Isoptera: Rhinotermitidae). Sociobiology 14, 5-13.
Walsh, S. J., J. D. Vitek & S. E. Panceira. (1991) Variability of soil temperature: a spatial and temporal analysis. Journal of Geography 90, 82-90.
Weesner, F. M. (1956) The biology of colony foundation in Reticuliteremes hesperus Banks. University of California Publications in Zoology 61, 253-314.
Weesner, F. M. (1965) The termites of the United States: A handbook, The National Pest Control Association.
Weesner, F. M. (1970) Termites of the nearctic region. pp. 477-525 in K. Krishna and F. M. Weesner (Eds.) Biology of Termites. Vol. 2. New York, Academic Press.
Williams, 0. L. (1934) Some factors limiting the distribution of termites. pp. 42-49 in C. A. Kofoid, S. F. Light, A. C. Homer, M. Randall, W. B. Herms and E. E. Bowe (Eds.) Termites and termite control. Berkley, University of California Press.
Yamashita, T. & H. Takeda. (1998) Decomposition and nutrient dynamics of leaf litter in litter bags of two mesh sizes set in two dipterocarp forest sites in peninsular Malaysia. Pedobiologia 42, 11-21.
158 Appendix
—4...... 6.04.,
' 414. e.l. ;NO., iite. 1r Z...11pii. ,i—
3:=':7.-
Figure A.1 Field site after two years in place.
Figure A.2 Equipment used to lift Figure A.3 Bucket prepared for buckets. removal. Plate has been hammered under base, metal rods have been hooked into the base and placed through handled lid.
159 Appendix
Figure A.4 Stand for holding the glass plates (designed by Paul Beasely).
Figure A.5 Set-up of experiment 6.2.2 showing tunnelling by Reticulitermes grassei (right).
160 WOBURN EXPERIMENTAL FARM, HUSBORNE CRAWLEY, BEDFORD
Area: 77 ha (190 acres)
Elevation: 79-110 m (260-360 If)
Annual Rainfall: 630 min (25 In)
O Cr' 1-1
fD co . co
P-1 5 rn O
UQ "CS O to X Exparimmtal Farm S Stockyard Hold
01 Far-Field O
0 100 200 300 metrte
0 500 1000 lost
C7% Appendix
A. 1 Soil Analysis Methodology
The following methodology was provided by A. Kelly (University of Plymouth) (Figure A.7). The particle analysis was performed by A. Kelly and the analyses using the Malvern Masteriser were performed by R. Hartley (University of Plymouth).
The soil samples were dried and then sieved through a 2mm round-hole sieve to separate the >2mm faction from the <2mm faction. The two were weighed and the <2mm faction was then treated with Hydrogen peroxide to remove organic material from the sample (see explanatory notes).
Once the organic material had been removed from the samples were put through a Malvern Mastersizer in order to assess the particle size distribution. The samples were passed through two cells, the first to measure particles between 2000 and 4 microns and the second, 80 to 0.1 microns. The results were then blended to produce the percentage clay, sand and silt content.
162 Appendix
DEPARTMENT OF GEOGRAPHICAL SCIENCES 1.4a University of Plymouth
Preparation of samples for analysis with the Malvern Mastersizer X
This sheet deals with the preparation of sand, silt and clay fractions for analysis by the Malvern Mastersizer X.
The preparation involves chemical removal of fine organic particles, and dispersion of the particles.
Removal of organic material is necessary because fine organic matter can distort the particle distribution in three ways: a) small air bubbles can get trapped in the organics which affect the way inwhich the sizer assesses the sample
DUST HAZARD
At several stages in this method dust WILL become airborne in particular pestling, riffling, sieve shaking, sieve cleaning and sediment transfer. This can lead to a build up of airborne dust to high levels which COULD promote respiratory irritation. Reduce this risk by: i) Pestle and riffle samples in the dust extraction unit ii) Transfer sediments in the dust extraction unit iii) DO NOT use split sieves iv) If the weather is fine open the windows v) WEAR A FACE MASK vi) If you suffer from any respiratory or dermal condition which are likely to be brought on by dust, please inform the staff involved - all in strictest confidence. vii) Clean any work area with a damp sponge or cloth.
Also, handling of sediments COULD promote skin rashes or bring on eczema in those who already suffer from such conditions. For such people with sensitive skin it is advised that they wear the gloves provided.
b) the organic particles can themselves be recognised as real sediment grains c) organic particles have a binding effect on fine grains, so affecting the particle distribution.
Dispersion of the particles is necessary to disgregate the grains because the particles are electrically attracted towards each other, and will therefore flocculate e.g. in a river estuary. To the sample is added a dispersant (calgon) which introduces positive ions (Na+) to the normally negatively charged clay lattice. As the clay lattice holds positive ions from the solution, there is a swarm of negatively charged ions surrounding each clay particle. When two particles come close to each other they now repel the other due to the negative ions around them. This way the particles remain dispersed. This dispersive action would be cancelled by adding an electrolyte such as HC1 or NaCl. Here the solution becomes a conductor and the 1-1+ or Na+ ions rob the particle of its negative swarm, allowing them to flocculate. During analysis in the laser sizer, the sample is mechanically dispersed with ultrasound.
Other treatments may include removal of carbonate cements, or removal of ferruginous cements. If you think your sample may require either of these treatments, ask a technician for advice before proceeding any further.
Figure A.7 Method for soil sample preparation provided by A. Kelly (University of Plymouth).
163
Appendix
SAMPLE PREPARATION
1. Collect approximately 6 to 10 grams of sub-sample (2 - 3 teaspoonfuls) of the appropriately sized material; the size of material will depend on your experimental design, but all the sub-sample MUST have passed through a round-hole 2mm sieve to prevent damage to the laser sizer.
2. Tip the sub-sample into a large shallow basin (NOT a 200m1 beaker) and label with a permanent pen.
REMOVAL OF ORGANIC MATERIAL: A t OA. 'tO Oh
YOU MUST READ THE COSHH ASSESSMENT FOR HYDROGEN PEROXIDE BEFORE PROCEEDING TO THE NEXT STAGE. THIS PROCESS MUST BE CARRIED OUT IN A FUME CUPBOARD.
3. Set up a water bath in a fume cupboard. Be sure that the water bath has a trickle of water feeding into it. 4. Add approximately 30 mls of 6% Hydrogen Peroxide (H202) to the sample and place on a water bath. If the sample and peroxide begin to over-react cool the reaction by washing down with distilled water; removal from the water bath may be required. 5. After the initial reaction (in the form of large amounts of bubbling and foaming) has passed, wash the inner sides of the basin with 6% FI202 from a wash bottle. A brief reaction with remaining organic matter may occur — use this to judge how much more H202 to add. 6. Repeat steps 4 and 5 once, though some samples may require a second repetition. 7. When all the organic matter has been digested, any remaining H202 will break down to water and oxygen; this will be evident by a light effervescence. Add distilled water until two-thirds full. Leave on the water bath for the H202 to completely break down (effervescence ceases). Do not allow to dry out. 8. Any floating fibrous material which cannot be broken down by H202 will need to be physically removed with a rough, porous pad or filter (a brillo pad is ideal). 9. Once all effervescence has ceased, transfer the sample from the basin to a 250m1 beaker and dry in an oven.
DISPERSION OF PARTICLES
YOU MUST READ THE COSHH ASSESSMENT FOR CALGON BEFORE PROCEEDING TO THE NEXT STAGE
1. Add a small amount of calgon'* to the dry sample to dampen it. Do not add too much or the coarser particles will settle out from the fines. 2. Mix the sample into an homogenous muddy paste. 3. Present to the laser sizer for analysis.
*33g sodium hexametaphosphate & 7g sodium carbonate per litre of de-ionised water.
RJMH (9/89) 1/00
Figure A.7 continued Method for soil sample preparation provided by the A. Kelly ( University of Plymouth).
164