The effects of earthworm functional group diversity on earthworm community
structure
Cróna Sheehan 1, Laura Kirwan 2, John Connolly2 Thomas Bolger1
1
UCD School of Biological and Environmental Science, University College Dublin,
Belfield, Dublin 4, Ireland
2
UCD School of Mathematical Sciences, University College Dublin, Belfield, Dublin
4, Ireland
1
Summary - A comprehensive understanding of how species/functional group
interactions determine population dynamics, community composition and their effect
on ecosystem functioning is important. This paper presents data from a mesocosm
experiment, based on the simplex design, to examine the effect of interactions
between earthworm functional groups, food supply and initial overall biomass on
community structure. All communities containing anéciques moved towards
domination by anéciques. The survival of anéciques remained constant irrespective of
initial conditions, with no effect of initial community structure, food supply or initial
biomass. The proportional biomass of epigées increased when they were placed in
communities dominated by anéciques. Initial overall biomass had a significant effect
on the survival of endogées, with increased survival at low biomass. Juvenile
production was significantly increased in communities that contained a higher initial
abundance of epigées. The anéciques had significantly increased production of
juveniles at lower levels of initial biomass. Overall, earthworm functional group
diversity had an idiosyncratic effect on earthworm assemblage structure.
Keywords: Functional groups, earthworms, interactions, population dynamics,
community composition, simplex design, initial biomass, food supply, relative growth
rates, survival, number of juveniles
2
Introduction
Earthworms play important roles in determining soil fertility and structure (Darwin,
1881; Barley, 1959b; Lee, 1985; Shaw and Pawluk, 1986; Curry, 1988; Edwards and
Bohlen, 1996, inter alia) therefore it is important to have a comprehensive
understanding of how species/functional groups interact to affect population
dynamics, community composition and the effect of earthworms on ecosystem
functioning. Information on interactions between earthworm species or functional
groups is limited (Abbott, 1980; Elvira et al., 1996; Butt, 1998; Curry, 1998; Dalby et
al. 1998; Lowe and Butt, 1999) but it is obvious that positive and negative interactions
occur.
Both intra- and interspecific/group competition have been shown to affect the
survival, growth, maturation and cocoon production of earthworms and these effects
have even been demonstrated between species from different functional groups
(Elvira et al. 1996; Dalby et al., 1998; Butt 1998; Baker et al 2002; Lowe and Butt,
2002). Hamilton et al (1998) found that the epigeic Eisenia fetida unexpectedly had a
negative effect both on weight gain and cast stability of Lumbricus terrestris and
interpreted this as being due to the rapid growth rates of the epigeic species allowing
it to out-compete the larger anecic species.
Competition also affects burrow systems and burrowing behaviour (Capowiez et al.,
2000). For example, the burrow system of Lumbricus terrestris (anecic) was less deep
in the presence of Aporrectodea caliginosa (endogeic), than in the single species
treatments (Jégou et al., 2001). The burrows of Aporrectodea caliginosa were
significantly deeper in the presence of Aporrectodea giardi compared to the single
species treatment. These changes may have been due to earthworms foraging deeper
into the soil for food but it could also indicate that facilitation was occurring and that
Aporrectodea caliginosa was utilising the burrows made by Aporrectodea giardi for
movement and/or that the below ground casts produced by Aporrectodea giardi might
provide a food resource for Aporrectodea caliginosa.
These effects are usually interpreted as being due to competition for food resources;
however, little research has been done on the influence of food resource availability
on earthworm interactions. In fact, in one of the few studies, which have specifically
3
addressed this issue, it was found that, while growth and maturation of hatchling
Allolobophora chlorotica were negatively influenced by the presence of other
earthworms, a positive interaction was detected between Allolobophora chlorotica
and Lumbricus terrestris, with the latter promoting growth of the former. This
suggested that the anecic Lumbricus terrestris produced a food source (casting) more
palatable than the surrounding soil (Lowe and Butt 2003). In addition, in treatments
where food was milled, growth rates of Lumbricus terrestris hatchlings paired with
adult Lumbricus terrestris and Allolobophora chlorotica were similar. However, in
treatments with unmilled food, the presence of adult Lumbricus terrestris markedly
reduced juvenile growth rates of Lumbricus terrestris over their growth rates in the
presence of adult Allolobophora chlorotica. This suggests that intra-specific
competition that occurred between juvenile and adult Lumbricus terrestris was
reduced when food particles were reduced (milled).
Thus both positive and negative interactions occur among earthworm species and
these are affected by food supply and population density. In this study we use a
mesocosm experiment to examine how interactions between earthworm functional
groups affect earthworm community structure under differing initial densities and
food supply.
Materials and methods
Experimental design
Because of the difficulties identified with many previous experimental designs
(Wardle, 1999, Connolly 1986; Bolger, 2001; Connolly et al., 2001) a novel design,
based on the simplex design, was used (Cornell, 1990, Ramseier et al., 2005) In this
design, communities were set up as either monocultures (endogeic, epigeic or anecic),
centroids, consisting of equal amounts of each functional groups, binary mixtures,
consisting of equal biomass of two functional groups or complete mixtures consisting
of all functional groups, where the proportional contributions of the groups sum to
one (Fig. 1).
The experiment was carried out at two densities of earthworm, 1.7g and 3.4g per
mesocosm and at two levels of food supply (0.32g and 0.64g of ground dried grass
4
per week), giving four combinations of biomass and food supply and a total of 52
communities.
Experimental unit
The experimental units consisted of Plexiglas cylinders (15cm diameter x 30cm
depth) containing 2.65 litres of soil. The base of each soil column was fitted with
2mm mesh to prevent earthworms escaping.
The soil used in this experiment was topsoil obtained on the campus of University
College Dublin. Initially the soil was hand sorted to remove macrofauna, large
particles and plant fragments. It was then sieved through a 4mm mesh followed by a
2mm mesh and screened for any remaining visible organic fragments. After the soil
had been placed into the units described above, 1 litre of distilled water was applied to
each and this was allowed to settle for a day. After this, 2 x 250mls applications of
soil suspension were added to boost the microbial biomass within the soil. Damp
muslin cloth (10cm x 10cm) was placed on the soil surface to act as a non-digestible
refuge for the earthworms particularly the epigées. Dried milled grass (Lolium
perrene) and 300 mls of distilled water were added to the columns every 7 days.
Adult earthworms were obtained from a meadow in University College Dublin and
sorted into three distinct functional groups based on Bouché (1977). The epigées were
Lumbricus rubellus, Lumbricus castaneus and Satchellius mammalis, the anéciques
were Lumbricus friendi and Aporrectodea longa and the endogées were Aporrectodea
caliginosa, Octolasion cyaneum, Allolobophora chlorotica and Aporrectodea rosea.
The earthworms were allocated randomly to experimental units.
Earthworm analysis
The individual weights of the earthworms added to the experimental units were
recorded at the beginning of the experiment and weights were again recorded when
the experimental units were being dismantled. Thus, survivorship and growth could
be examined.
Statistical Analysis
Assessing factors that affect changes in Community Composition
The composition of a community will change if the relative growth rates of functional
groups differ. Therefore, the differences in relative growth rates between functional
5
groups in a community should be analysed in order to determine the factors that affect
changes in community composition (Connolly and Wayne 2005, Ramseier et al,
2005). If there are s functional groups in a composition, and one group is selected as a
reference, there will be s-1 of these relative growth rate differences (RGRD’s). For
this three functional group experiment, anéciques were chosen as the reference and so
the response variables are RGRDAN-EPI and RGRDAN-ENDO.
Each RGRD is
modelled as a linear function of the explanatory variables, which include the
proportions of each functional group in the initial composition (AN, EPI and ENDO),
initial stand biomass and food supply where low levels are coded by 0 and high levels
by 1. (biomass and food are centred in all analyses)
RGRDAn  Epi  a1 AN  a2 EPI  a3 ENDO  a 4 Biomass  a5 Food
RGRDAn  Endo  b1 AN  b2 EPI  b3 ENDO  b4 Biomass  b5 Food
This is the simplest model and could be generalised to include interactions between
biomass and food or proportion of functional groups. These interactions were tested
and none proved to be significant and therefore the simple model as above was used.
Functional group effects are assessed through the coefficients of the functional group
proportions in the model: a1 is the differential (in the limit) between the growth rates
of epigées and anéciques as the assemblage tends towards a monoculture of
anéciques. Thus, a large positive value of a1 means that anéciques are favoured over
epigées in assemblages dominated by anéciques.
Differences in coefficients reflect the effects of neighbours. If a2 = a1 the RGR
difference between epigées and anéciques is unchanged by increasing the initial
proportion of epigées at the expense of anéciques. Under the same change in initial
proportions, a2 > a1 favours anéciques over epigées.
The effect of a change in initial endogeic proportion on the relationship between
anéciques and epigées is measured through a3.
The effects of initial assemblage biomass and of the amount of food added on the
relationships between functional groups are measured through the a4 and a5
coefficients.
6
Survival of Earthworms
Factors that affect the proportion of individuals surviving (Psurv) in each functional
group were investigated using logistic regression (McCullagh and Nelder, 1989). The
proportional biomass of the three functional groups that were present in the initial
community are denoted AN, EPI and ENDO respectively. Model (1) is the simplest
logistic regression model. It assumes that the logit of Psurv , i.e. Ln (Psurv /( 1- Psurv )),
is affected just by the linear effects of initial functional group proportions, the level of
food supplied and the overall initial biomass in the community.
 P

Ln surv    1 AN   2EPI   3 ENDO   4 Food   5 Biomass
 1  Psurv 
(1)
β1, β2 and β3 represent the linear mixing effects of the functional groups. They are the
predicted responses of the three functional groups in monoculture at the low levels of
food and biomass. A linear combination of these gives the response of a mixture of
species. β4 and β5 represent the effects of food and biomass. This is a constant shift in
response at high levels of density and factor. Terms reflecting the pairwise
interactions between functional groups were added to model (1) and tested for
inclusion in a final model for each functional group. Values of the response variable
were predicted from these final models and back transformed to the proportional
scale.
Production of Juveniles
The number of juveniles produced by a community is a non-negative integer and so
cannot be modelled by ordinary least squares regression. It is assumed that the
number of juveniles, Y, has a Poisson probability distribution, and that its mean, μ, is
related to initial community structure (McCullagh and Nelder, 1989).
Poisson
regression was used on these data. The simplest model is as follows: The expected
value of Y is  and
Ln ( )  1 An   2 Epi  3 Endo   4 Food  5 Biomass
(2)
The log link function ensures that the mean number of juveniles for each community
predicted from the fitted model is positive.
7
Results
Biomass and proportions of earthworms
The proportional biomass of the earthworm functional groups changed during the
experiment (Fig. 2). For each of the four combinations of food and biomass, virtually
all communities containing anéciques moved towards domination by that functional
group (Fig 2). The relative growth rates of the different functional groups were
affected by various aspects of community structure and resource availability. The
biomass of anéciques always increased but the higher the proportion of the other two
functional groups present, the greater was their relative growth rate (Table 1).
However, the biomass of epigées declined as the proportions of either epigées or
endogées increased. As the biomass of earthworms in a community increased the
relative growth rates of endogéics declined.
Community composition of earthworms
The parameter estimates for the two RGRD models are given in Table 2. The positive
coefficients for the epigeic (EPI) and endogeic (ENDO) imply that the growth rate of
the anéciques, relative to these groups, increases as the proportions of these functional
groups in a community increase and reflects the greater proportions of anéciques at
the end of the experiment. As the effects of food supply and initial biomass were not
significant, a single graph can be used to predict the changes in community
composition over the period of the experiment (Fig. 3). Almost all communities move
towards domination by anéciques.
Survival
The survival of anéciques remained constant irrespective of initial conditions. There
was no effect of initial community structure, food supply or initial biomass. The
average survival rate was 73%.
Survival rate of epigées was very low in communities dominated by either epigées or
endogées (Table 3, Fig 4). However, increasing the proportion of anéciques in the
8
community increases the rate of survival (Fig. 4). For example, the survival of epigées
in monoculture was 26% while in the community with 80% anéciques it was 86% at
average biomass and food supply. Antagonistic effects of mixing anéciques and
epigées (p=0.031) and anéciques and endogées (p=0.003) ensured a low level of
epigée survival where anecic proportion was less than about 0.5 (Fig 4). These
outcomes were not influenced by initial biomass or food supply.
The survival rate of endogées was significantly higher in communities that contained
a higher initial abundance of endogées (c.f. positive coefficient on the initial
proportion of endogées in the community) (p<0.001) (Table 4). Endogeic survival
was low in communities dominated either by epigées or anéciques but was quite high
when these species were equally represented (Fig. 5) due to a positive synergistic
effect between them (Table 4). The endogées had higher survival rates at lower
overall initial earthworm biomass in the community (ENDO*BIOMASS interaction;
p=0.001) (Fig. 5). For example, the survival of endogées in monocultures at high
biomass was 62% and 90% (high biomass low food supply and high biomass high
food supply respectively) while at low biomass it was 100% (for both low biomass
low food and low biomass high food). Food supply did not have a significant effect on
the survival of endogées (p=0.2289).
Juvenile production
The production of juveniles increased in communities that contained a higher initial
abundance of anéciques as indicated by the positive coefficient for the proportion of
anéciques (p=0.0015) (Table 5). The production of juveniles was significantly
increased when the anéciqueswere mixed with endogées (anéciques*endogées
interaction; p=0.025) (Table 5, Fig. 6).
Food supply significantly affected the production of juveniles. In particular there was
a significant interaction between the proportion of epigées and food supply and large
numbers of juveniles were present in monocultures of epigées at high food supply
(p=0.049) (Fig. 6 and Table 5). There was also a larger production of juvenile epigées
at the lower initial community biomass (p=0.049). With high biomass and low food
supply, a significant difference was detected between the epigées and anéciques
(p=0.0265) and anéciques were predicted to produce 2.86 more than the epigées (Fig.
9
6). At low biomass and high food supply, epigées were predicted to produce 7.3 more
juveniles than the endogées (p=0.0211) (Fig. 6).
Discussion
A comprehensive understanding of how functional group interactions determine
population dynamics, community composition and their effect on ecosystem function
is important. The coexistence of species/functional groups in the same habitat can
lead to interactions, which can have a positive or negative effect upon them. Although
interactions between organisms are generally well documented (Connell, 1983;
Gurevitch et al., 1992), little work has been done on interactions among earthworm
species/groups (Curry, 1998; Dalby et al., 1998; Lowe and Butt, 1999).
In this study all communities containing anéciques, even those initially dominated by
either of the other groups, became more dominated by this group as the experiment
progressed.. In fact the relative growth rate of the anécique population was increased
as the proportion of epigées and endogées increased. However, they became less
dominant in communities containing the large initial biomass of anéciques perhaps
indicating competition for food and/or habitat. The relative growth rates for epigées
on the other hand indicate that they are significantly negatively affected by increased
biomasses of the other two groups.
Similar trends were observed with survival. The survival of the anéciques remained
constant irrespective of initial condition. There was no effect of initial community
structure, food supply or initial biomass on their survival. There was evidence of
intra-group competition or lack of required synergistic effects occurring in
communities dominated by the epigées, as their survival was found to be significantly
lower in these communities. The results imply that the survival of the epigées may be
directly related to optimal habitat availability. As with the proportional biomass, the
survival of the epigées was greatly increased as the initial proportion of the anéciques
in the community increased. This suggests that a positive inter-group interaction was
occurring.
Despite the very poor survival of endogées in communities dominated by either of the
other two groups, the greatly enhanced survival where these species are in balance is
intriguing. The anéciques may have provided a food resource to the endogeic
10
earthworms by depositing faecal organic matter (or casts) into the soil where the
endogées can reach it. The epigées play a major role in organic matter comminution
and reduce the size of organic matter particles during the passage through the
earthworm and in doing so makes organic matter more assessable to digestive action
by other decomposers. Endogées prefer feeding on well decomposed organic matter.
Hence these combined effects may have played a role in the improved survival of
these earthworms. In communities dominated by endogées their survival was reduced
at high biomass suggesting intra-group competition.
The production of juveniles was significantly increased in communities that contained
a higher initial overall abundance of anéciques. This is presumably correlated with the
high rates of growth and survivorship of this group. However, there was also an
increased survivorship of epigées under these circumstances and, although the
biomass of this group may have been low under these circumstances, the smaller adult
sizes of these worms and the fast rate of development of epigeic cocoons could have
contributed to the increased production of juveniles (Senapati & Sahu, 1993;
Bhattacharjee & Chaudhuri, 2002). Increased functional diversity led an increased
production of juveniles, as indicated by the positive interaction between the anéciques
and endogées, but food supply and biomass proved to be important for the epigées,
with increased juvenile production at high levels of food supply and at lower levels of
initial overall biomass suggesting that the higher number of juveniles occurred where
there was reduced competition for food and space or that the additional food supplied
at the surface of the soil provided an appropriate habitat for this group..
It is clear that, increased functional group diversity had an idiosyncratic effect on
earthworm assemblage structure and that certain community compositions led to
increased survival of endogeic and epigeic earthworms . Therefore, identifying the
interactions (both positive and negative) occurring between earthworm functional
groups is important in understanding their effects on community structure. Also,
demonstration of complementary relationships between earthworm functional groups
may help identify the communities necessary for the proper functioning of
ecosystems.
Acknowledgements
This project was partially funded by the Irish Environmental Protection Agency.
11
12
References
Abbott, I. 1980. Do earthworms compete for food? - Soil Biol. Biochem. 12: 523-530.
Baker, G., Carter, P., Barrett, V. et al. 2002. Does the deep-burrowing earthworm,
Aporrectodea longa, compete with resident earthworm communities when
introduced to pastures in south-eastern Australia? - Eur. J. Soil Biol. 38: 3942.
Barley, K.P. 1959. Earthworms and soil fertility. IV. The influence of earthworms on
the physical properties of a red-brown earth. - Aust. J. Agric. Res. 10: 371376.
Bengtsson, J. 1998. Which species? What kind of diversity? Which ecosystem
function? Some problems in studies of relations between biodiversity and
ecosystem function. - Appl. Soil Ecol. 10: 191-199.
Bhattacharjee, G. and Chaudhuri, P.S. 2002. Cocoon production, morphology,
hatchling pattern and fecundity in seven tropical earthworm species – a
laboratory based investigation. – J. Bio. 27: 283-294.
Bouché, M.B. 1977. Stratégies lombriciennes. – In: Lohm, U. and Persson, T. (eds),
Soil Organisms as Components of Ecosystems. pp. 122-132.
Butt, K. R. 1998. Interactions between selected earthworm species: A preliminary
laboratory-based study. - Appl. Soil Ecol. 9: 75-79.
Capowiez, Y., Monestiez, P. and Belzunces, L. 2001. Burrow systems made by
Aporrectodea nocturna and Allolobophora chlorotica in artificial cores:
morphological differences and effects of interspecific interactions. – Appl.
Soil Ecol.16: 109-120.
Capowiez, Y., Pierret, A., Monestiez, P. and Belzunces, L. 2000. Evolution of burrow
systems after the accidental introduction of a new earthworm species into a
Swiss pre-alpine meadow. – Biol. Fert. Soils 31: 494-500.
Connolly, J. 1997. Substitutive experiments and the evidence for competitive
hierarchies in plant communities. - Oikos 80: 179-182.
Connolly, J., Wayne, P. and Bazzaz, F.A. 2001. Interspecific competition in plants:
How well do current methods answer fundamental questions. – Am. Nat. 157:
107-125.
13
Connolly, J. and Wayne, P. 2005. Assessing determinants of community biomass
composition in two-species plant competition studies. Oecologia, 142,450457.
Cornell, J.A. 1990. Experiments with mixtures - New York, NY.
Cotton, D.C.F. and Curry, J.P. 1980. The effects of cattle and pig slurry fertilisers on
earthworms (Oligochaeta, Lumbricidae) in grassland managed for silage
production. - Pedobiologia 21: 181-188.
Curry, J.P. 1988. The ecology of earthworms in reclaimed soils and their influence on
soil fertility. In: Edwards, C.A. and Neuhauser, E.F. (eds), Earthworms in
Waste and Environmental Management. SPB Academic Publishers, The
Hague, The Netherlands. pp. 251-261
Curry, JP. 1998. Factors affecting earthworm abundance in soils. In: Edwards, C.A.
(ed.), Earthworm Ecology. CRC Press, Boca Raton, Florida. pp. 37-64
Dalby, P. R., Baker, G. H. and Smith S. E. 1998. Competition and cocoon
consumption by the earthworm Aporrectodea longa. - Appl. Soil Ecol. 10:
127-136.
Darwin, C. 1881. The Formation of Vegetable Mould through the Action of Worms,
with Observations of their Habits. - Murray, London.
Edwards, C.A. and Bohlen, P.J. 1996. Biology and Ecology of Earthworms. Chapman and Hall, London.
Elvira, C., Dominguez, J. and Mato, S. 1996. The growth and reproduction of
Lumbricus rubellus and Dendrobaena rubida in cow manure mixed cultures
with Eisenia andrei. – Appl. Soil Ecol. 5: 97-103.
Evans, A.C. and Guild, W. J. Mc.L. 1948. Studies on the relationship between
earthworms and soil fertility. IV. On the life cycles of some British
Lumbricidae. – Ann. Appl. Biol. 35: 471-84.
Hamilton, W. E., Dindal, D. L., Parkinson, C. M. and Mitchell, M. J. 1988.
Interaction of earthworm species in sewage sludge-amended soil microcosms:
Lumbricus terrestris and Eisenia fetida. – J. Appl. Ecol. 25: 847-852.
Jégou, D., Capowiez, Y. and Cluzeau, D. 2001. Interactions between earthworm
species in artificial soil cores assessed through the 3D reconstruction of the
burrow systems. - Geoderma 102: 123-137.
14
Lee, K.E. 1985. Earthworms. Their ecology and relationship with soils and land use. Academic Press Sydney.
Lowe, C. N. and Butt, K. R. 1999. Interspecific interactions between earthworms: A
laboratory-based investigation. - Pedobiologia 43: 808-817.
Lowe, C. N. and Butt, K. R. 2002. Growth of hatchling earthworms in the presence of
adults: interactions in laboratory culture. – Biol. Fertil. Soils 35: 204-209.
Lowe, C. N. and Butt, K. R. 2003. Influence of food particle size on inter- and intraspecific interactions of Allolobophora chlorotica (Savigny) and Lumbricus
terrestris (L.) - Pedobiologia 47: 1-4.
McColl, H. P., Hart, P. B. S. and Cook, F. J. 1982. Influence of earthworms on some
soil chemical and physical properties, and the growth of ryegrass on a soil
after topsoil stripping – a pot experiment. – N. Z. J. Agric. Res. 25: 229-243.
McCullagh, P. and Nelder, J.A. 1989. Generalized Linear Models. - Chapman and
Hall.
Monestiez, P. and Kretzschmar, A. 1992. Estimation of the relationship between
structural parameters of simulated burrow systems and their partitioning
effect. - Soil Biol. Biochem. 24: 1549-1554.
Ramseier, D., Connolly, J. and Bazzaz, F.A. 2005. Carbon dioxide regime, species
identity and influence of species initial abundance as determinants of change
in stand biomass composition in five-species communities: an investigation
using a simplex design and RGRD analysis. Journal of Ecology 93:502-511
SAS. 1998. SAS user’s guide: Statistics, Version 8 Edition. SAS Inst. Inc., Cary, NC.
Satchell, J.E. 1967. Lumbricidae. In: Burgess, A. and Raw, F. (eds), Soil biology.
Academic Press, London. pp. 259-322
Senapati, B. K. and Sahu, S. K. 1993. Reproductive biology (cocoon morphology, life
cycle pattern and life table analysis) in earthworms. In: Earthworm resources
and vermiculture. Calcutta: Zoological Survey of India. pp. 79–96.
Shaw, C. and Pawluk, S. 1986. The development of soil structure of Octolasion
tyrtaeum, Aporrectodea turgida and Lumbricus terrestris in parent plant
materials belonging to different textural classes. - Pedobiologia 29: 327-339.
Wardle, D.A. 1999. Is “Sampling Effect” a problem for experiments investigating
biodiversity-ecosystem function relationship? - Oikos 87: 403-407.
15
16
Anecic
(1, 0, 0)
(0.8, 0.1, 0.1)
(0.45, 0.45, 0.1)
(0.45, 0.1, 0.45)
(0.5, 0.5, 0)
(0.5, 0, 0.5)
(0.1, 0.8, 0.1)
(1/3,1/3,1/3)
(0.1, 0.1, 0.8)
(0, 1, 0)
(0, 0.5, 0.5)
(0, 0, 1)
Epigeic
(0.1, 0.45, 0.45)
Endogeic
Fig. 1. Simplex design proportions. For example, (0.1, 0.1, 0.8) indicates a
community consisting of 10% anecic, 10% epigeic individuals and 80% endogeic
individuals.
17
(A)
Epigeic
Anecic
(B)
Anecic
Endogeic
Epigeic
(C)
(D)
Anecic
Epigeic
Epigeic
Endogeic
Anecic
Endogeic
Endogeic
Fig. 2. Proportional biomass of communities at all combinations of biomass and food supply. (A) high biomass low food supply (B) high biomass high
food supply (C) low biomass low food supply (D) low biomass high food supply. Initial compositions are shown by open circles and compositions at
harvest are shown by closed circles.
18
Anéciques
1
0
1
0
0
Epigées
1
Endogées
Fig. 3. Modelled changes in community composition at average food and biomass.
Initial composition is shown by open circles and predicted compositions are shown by
closed circles.
19
1.0
p
Anecic
0.5
1.00
1.00
0.00
0.0
Epigeic
0.00
0.00
1.00
Endogeic
Fig. 4. Response surface plot of predicted survival of epigées (proportions) at average
biomass and food supply.
20
A) Low Biomass
B) High Biomass
1.0
1.0
1.0
1.0
0.5
0.5
pp 0.5
p 0.5
Epigeic
Epigeic
0.00
0.00
0.0
0.0
1.00
1.00
0.00
0.00
1.00
1.00
Endogeic
Endogeic
Epigeic
Epigeic
0.00
0.00
1.00
1.00
Anecic
Anecic
0.00
0.0
1.00
0.0
1.00
0.00
0.00
0.00
1.00
1.00
Endogeic
Endogeic
0.00
0.00
1.00
1.00
Anecic
Anecic
Fig. 5. Response surface plots of predicted proportion of endogées surviving.
Predictions are at average food.
21
Food
10
108
8
6
4
n 5
JUVENILES
2
Anecic
0
Epigeic
0
0.00
1.00
1.00
0.00
1.00
Anecic
1.00
01.00
Anecic
0.00
1.00
Biomass
0
Epigeic
1.00
0.00
0.00
0.00
Endogeic
10
6
4
n 5
2
0
Epigeic
MBC
1.00
n 5
Epigeic
Anecic
0
0
Anecic
1.00
Endogeic
1
8
10
6
JUVENILES
2
1.00
1.00 0.00
0.00
0.00
1.00
4
1.00
0.00
Epigeic
Endogeic
8
0
0.00
Endogeic
6
JUVENILES
1.00
0.00
0.00
0.00
0.00
0.001.00
Endogeic
1.00
Epigeic
1.00
Anecic
1.00
2
Epigeic
0.00
0.00
4
n 5
1
Anecic
1.00
0.00
0
1.00
0.00
1.00
0
0.00
0.00
1.00
0.00
1.00
Endogeic
Endogeic
Epigeic
Anecic
0.00
1.00
Endogeic
Fig. 6. Response surface of predicted juvenile production. (0 = low levels of food or
biomass (0.32g and 1.7g respectively), 1 = high level of food or biomass (0.64g and
3.4g respectively)).
22
Table1. Parameter estimates for average RGR for each functional group. Coefficients
that are significant at the 5% level are bolded.
Average RGR
Epigées Anéciques Endogées
Epigées
-1.87
1.5
-0.19
Anéciques -0.1
0.71
0.79
Endogées -1.95
Food
-0.5
Bio
-0.55
1.62
0.18
-0.3
0.07
0.29
-0.85
23
Table 2. Parameter estimates for RGRD models showing the effects of initial
functional group composition (in proportions), food supply and initial biomass.
Significant values are in bold
Epigée Anécique
s
s
Endogées
Food
Bio
Average
RGRD An - Epi
3.37
0.80
3.56
0.7
0.3
2.58
RGRD An - Endo
1.69
-0.09
1.55
-0.1
0.5
1.05
RGRD Epi - Endo -1.68
-0.89
-2.02
-0.8
0.3
-1.53
24
Table 3. Parameter estimates for epigeic survival
Parameter
estimate
s.e.
t
Pr > t
Epigées
-1.04
0.209
-4.96
<.001
Anéciques
3.98
1.4
2.84
0.004
Endogées
-0.17
0.567
-0.29
0.771
Anéciques*Epigées
-6.31
2.92
-2.16
0.031
Anéciques*Endogées -11.41
3.88
-2.94
0.003
Food
0.10
0.213
0.48
0.634
Biomass
-0.21
0.211
-1
0.317
25
Table 4. Parameter estimates for endogeic survival
Parameter
estimate
s.e.
t
Pr > t
Epigées
-2.442
0.989
-2.47
0.014
Anéciques
-1.18
1.23
-0.96
0.335
Endogées
3.02
0.674
4.48
<.001
Anéciques*Epigées
18.6
7.81
2.38
0.017
Food
0.157
0.432
0.36
0.716
Endogées*Biomass
-3.86
1.14
-3.37
<.001
26
Table 5. Parameter estimates for predicted juvenile production.
Parameter
estimate
s.e.
Epigées
0.65
0.423
2.4
0.125
Anéciques
1.18
0.372
10.0
0.002
Endogées
0.80
0.439
3.4
0.067
Anéciques*Endogées
3.86
1.728
5.0
0.025
Epigées*Food
1.94
0.733
7.0
0.008
Epigées*Biomass
-1.28
0.649
3.9
0.049
27
Chi-Square P > Chi-Square