See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/226812100

Stream community structure in relation to spatial variation: The influence of

mesohabitat characteristics

Article in Hydrobiologia · December 1998

DOI: 10.1023/A:1003519429979

CITATIONS READS

235 698

4 authors, including:

Jean-Nicolas Beisel Philippe Usseglio-Polatera

University of Strasbourg University of Lorraine
136 PUBLICATIONS 2,754 CITATIONS 127 PUBLICATIONS 5,661 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Coevolution of interconnected river infrastructures and biodiversity (INTERCONNECT - ITTECOP Programme) View project

Integrating bio-ecological traits in the biomonitoring of streams View project

All content following this page was uploaded by Jean-Nicolas Beisel on 08 July 2015.

The user has requested enhancement of the downloaded file.

 Hydrobiologia 389: 73–88, 1998.
© 1998 Kluwer Academic Publishers. Printed in the Netherlands.
73

Stream community structure in relation to spatial variation:

the influence of mesohabitat characteristics

Jean-Nicolas Beisel1∗ , Philippe Usseglio-Polatera1,2, Sandra Thomas1 &

Jean-Claude Moreteau1
1 UPRES E.B.S.E., Université de Metz, UFR Sciences, Ile du Saulcy, 57045 Metz Cedex 01, France
2 ESA CNRS 5023 “Ecologie des Eaux Douces et des Grands Fleuves”, Université Lyon I, 69622 Villeurbanne
Cedex, France
(∗ author for correspondence)

Received 19 February 1997; in revised form 19 August 1998; accepted 27 October 1998

Key words: macroinvertebrates, mesohabitat scale, community structure, habitat characteristics, multivariate
analysis

Abstract
Community structure of benthic macroinvertebrates was studied in six first- through fourth-order streams in north-
east France, to elucidate changes in richness, abundance, diversity and evenness of mesohabitat assemblages as
a function of environmental conditions. Patch samples were subjected to multivariate analyses to determine: (i)
relationships among seven indices describing community structure (structure parameters); (ii) relationships among
seven environmental variables; (iii) the relationship between community structure and environmental characteris-
tics of patches. Faunal data showed that indices measuring the distribution of individuals among taxa (evenness,
dominance) and richness are prominent in describing the structure of macroinvertebrate communities of mesohab-
itats. The analysis of environmental data demonstrated a major differentiating ability of current velocity and strong
inter-relations among in-stream hydraulic-dependent parameters in structuring the mesohabitat environment. The
co-structure (= relationship) between community organization and environmental variables indicated that substrate
may be a primary determinant of community structure. Current velocity and water depth emerged as secondary
factors. Trends in community structure were closely related to the spatial variability of mesohabitats. Species
richness increased with habitat heterogeneity. Total abundance increased with trophic potentialities of patches.
Equitability and diversity seemed to increase with patch stability.

Introduction Four processes are often proposed as determinants

of small-scale variation in stream community struc-
Rivers and streams may be regarded as a hierarchical ture: (1) physical factors, a consequence of microto-
system of patches that differ in age, size and envi- pography and hydrodynamics, (2) patchy food sup-
ronmental conditions. Both anthropogenic and natural plies, (3) reproductive and age related effects, and (4)
processes create patchiness across a wide range of biological interactions (Hynes, 1970; Cummins et al.,
spatial scales, resulting in complex biotope structure. 1984; Bournaud & Cogerino, 1986; Fleeger & Decho,
Habitat characteristics vary drastically over small dis- 1987; Statzner et al., 1988; Ward & Palmer, 1994). But
tances in streams. The effect of spatial patchiness many researchers accept that the rigour and variation
on pattern and process in aquatic ecosystems has of the physical (and chemical) environment mainly ac-
long been recognized (Illies & Botosaneanu, 1963; count for patterns in species composition, abundance
Egglishaw, 1964; Hynes, 1970; Pringle et al., 1988; and richness, and leave little scope for species inter-
Townsend, 1989; Townsend & Hildrew, 1994; Wu & actions (Hildrew & Townsend, 1987; Statzner, 1987;
Levin, 1994). Townsend & Hildrew, 1994).
74
Table 1. Environmental variables and modalities (= categories
The distributions or dynamics of many taxa have plus their numbers used as labels in the figures) used in this
been described in relation to their environment. At a analysis
microscale, interest in microdistribution has focused
on physical environmental constraints such as sub- Environmental variable No. Modality
strate composition (Cummins & Lauff, 1969; Minshall 1. Current velocity 1 ≤ 5 cm/s
& Minshall, 1977; Rabeni & Minshall, 1977; Lam- 2 >5–20 cm/s
berti & Resh, 1979; Reice, 1980, 1981; Erman & 3 >20–60 cm/s
Erman, 1984; Brussock & Brown, 1991), hydraulic 4 >60–90 cm/s
conditions (Edington, 1968; Chutter, 1969a; Statzner, 5 >90 cm/s
1981; Statzner et al., 1988; Doeg et al., 1989a; Quinn 2.Water Depth 1 ≤10 cm
& Hickey, 1994), or food availability (Rabeni & 2 >10–20 cm
Minshall, 1977; Hawkins et al., 1982; Winterbourn, 3 >20–30 cm
1982; Drake, 1984; Dudgeon, 1988; Dobson & Hil- 4 >30 cm
drew, 1992). In contrast, few studies have analysed 3. Froude number 1 ≤0.4
community structure itself and its relation to habitat 2 >0.4–1.6
characteristics at a mesospatial scale (Jenkins et al., 3 >1.6–3.2
1984; Sagova & Adams, 1993). Consequently, we at- 4 >3.2–6.4
tempted to examine some aspects of spatial variation. 5 >6.4
4. Substrate 1 Mud
The objective of our study was to relate community
or 2 Silt
epiphenomena (i.e., richness, abundance, diversity or
5. Neighbouring patch(es) 3 Sand
evenness indices) to environmental conditions at the
4 Gravel
mesohabitat level. Based on a sampling program de-
5 Pebbles
signed to collect a wide variety of sampled community 6 Cobbles
structures, this study examined: (i) the relationships 7 Boulder
among 14 basic measures of community structure, 8 Rock
(ii) the relationships among seven variables describing 9 Flagstone
the environmental characteristics of their habitat, (iii) 10 Clay
the relationship between environmental characteristics 11 Organic detritus
and community structure. We chose to ignore variation 12 Filamentous salgae
in community structure between sites or times, even if 13 Helophytes
the spatial and temporal variation in the data allowed 14 Hydrophytes
for its analysis. 15 Bryophytes
16 Roots
6. Patch diversity 1 ≤0.5
2 >0.5−1.0
Materials and methods
3 >1.0−1.5
4 >1.5−2.0
Study areas 5 >2.0
7. Substrate richness 1 ≤2
Field work was conducted at four study sites on 1st 2 >3–4
to 4th order streams in each of three catchments (Fig- 3 >5
ure 1) situated in northeastern France. Sites were
selected: (1) to be easily accessible, (2) to permit a
sampling of all representative habitats of each study
site, (3) to have no or little sustained anthropogenic
land use, (4) to exhibit high mesoscale heterogeneity.
Field campaigns were carried out in spring and autumn
of 1993.
Ranges for physical variables recorded at the sam- to 15‰. pH fluctuated from 6.5 to 8.2; conductivity
pling dates were: mean width of stream, 2.5–11 m; from 48 to 1080 µS cm−1 ; dissolved oxygen from 60
mean depth, 12–35 cm; mean annual discharge, 0.5–5 to 100%. Ten study sites had forest canopies, all others
m3 s−1 . The slopes of the study sites ranged from 0.5 were open.
 75

Figure 1. Design of both the sampling procedure and the analysis strategy to investigate relationships among structure indices or environmental
variables and between community structure and environmental characteristics in 1st–4th order streams of the northeastern part of the France.
Samples n1 and n2 are two given stands that are to be projected on both the environment axis and the fauna axis of the co-inertia analysis. These
projections define new scores of stands that are the most covariant, i.e., these new scores have a maximal correlation and the standard deviation
of the new scores are maximal (for further details, see Dolédec & Chessel, 1994).
76

Sampling design type was classified according to a modified Wasson

et al. (1981) scale. Current velocity was measured
The twelve sites were mapped out at the beginning in front of each sample site at different depths with
of each sampling date to record the substrate mosaic a battery-operated portable water velocity meter (type
to an accuracy of 0.10–0.25 m (see Figure 1). The OTT). For each patch, Froude number, an indicator of
maps permitted us to record seasonal changes in en- near-bed hydraulic conditions (Statzner et al., 1988),
vironmental organization due to growth and recession was inferred from mean velocity and depth.
of macrophytes and erosion or deposition of substrata. Doeg et al. (1989b) demonstrated that 75% of the
Furthermore, they gave an exhaustive list of substrate colonizing fauna of stones in the Acheron River came
types to sample. In each habitat type, we tried to col- from the immediate neighbouring substrata. Conse-
lect different velocity classes to obtain a variety of quently, the maps were used as a posteriori to produce
substrate-velocity-depth associations. Indeed, we de- a description of the neighbouring patch mosaic within
fined patches (or mesohabitats) as combinations of a a radius of two meters round the sample, in terms of
mixture of substrate types (among 16 modalities; Ta- substrate type (16 modalities) and substrate richness
ble 1) and a range of current speed (five classes – see (three classes). In addition, we used the Shannon–
Beisel (1996) for further details). Only patches cover- Weaver diversity index, calculated on relative patch
ing a substantial area (≥ 0.05 m2 ) were sampled. They areas, as a measure of patch diversity of the bottom
were assumed to be little influenced by faunal conta- mosaic within a radius of two meters round each sam-
mination from neighbouring patches, and to support a ple. These supplementary data were recorded to eval-
typical community. uate the potential influence of faunal contagion among
On each sampling occasion, eight to 12 samples neighbouring mesohabitats, focusing on mechanisms
of substrate, each of 0.05 m2 (in terms of projected of dispersal such as crawling or short distance swim-
surface area), were collected within each study site. ming from one patch to another. The available data has
Benthic invertebrates were collected using a modi- been accumulated in a ‘fuzzy coding’ procedure (see
fied Surber sampler (mesh size 500 µm). The sub- below).
strate within the sampler was removed to a depth of
about six cm. To obtain an accurate evaluation of Data expression and analysis
taxa abundance, substrate was collected, brushed or
washed to remove any organisms, and preserved in 4% Seven indices were used to summarize aspects of the
formaldehyde. community structure of our 189 patch samples (Ta-
Within the limitations of time required for sort- ble 3). The environmental analysis of patch samples
ing and identification, a list of invertebrate taxa was must take into account a combination of discrete and
obtained for each sample (Table 2). We used the continuous variables. To standardize the environmen-
family level in structure index calculations to give tal description of samples, we used a ‘fuzzy coding’
the same weight to faunal units of a same system- procedure that assigned a positive score (from 0 = no
atic level, excepted for Hydracarina and Oligochaeta, affinity to 3 = high affinity) to describe the affinity
which were recorded as such. Faunal or phytosoci- of environmental characteristics of patches to the dif-
ological studies in which classification or ordination ferent modalities (or classes) of a given variable (see
techniques were used, showed that ecologically mean- Van Rijckevorsel, 1987; Chevenet et al., 1994). This
ingful patterns could be obtained using family data approach is particularly effective for the description of
(discussed by Furse et al., 1984; Resh & McElravy, variables such as substrate or neighbouring substrates
1993; Usseglio-Polatera, 1994; Bournaud et al., 1996), for which several modalities could be simultaneously
even if species-level identification contains the highest involved for a patch sample. The resulting informa-
information. tion consisted of an array of 189 samples × seven
Before each locality was sampled, conventional environmental variables (Figure 1).
hydraulic variables: current velocity, water depth, and A multivariate approach was used to examine rela-
substrate type were measured or estimated. Depth was tionships between indices and environmental variables
recorded at the top of the mesohabitat rather than the (Figure 1). A first step consisted of conducting sep-
stream bottom because certain habitats (roots of the ri- arate analyses to interpret both the faunal structure
parian vegetation, stumps, branchs for example) were (Figure 2) and the environmental structure (Figure 3).
not always in contact with the river bottom. Substrate The faunal array (a 189 sample by seven biocenotic
 77

Table 2. List of benthic macroinvertebrates sampled on study sites (Families only sampled on one station/date are not presented)

Plecoptera Coleoptera
Chloroperlidae Siphonoperla torrentium Dytiscidae
Leuctridae Leuctra Elmidae Elmis
Nemouridae Amphinemura Esolus parallelepipedus
Protonemura Limnius
Perlodidae Isoperla Oulimnius tuberculatus
Perlodes Riolus subviolaceus
Taeniopterygidae Brachyptera Stenelmis canaliculata
Taeniopteryx Gyrinidae

Ephemeroptera Haliplidae Haliplus

Baetidae Helodidae
Caenidae Caenis Hydraenidae Hydraena riparia
Ephemerellidae Ephemerella ignita Hydrophilidae

Torleya major Diptera

Ephemeridae Ephemera danica Anthomyidae
Heptageniidae Ecdyonurus Atheridicae Atherix
Epeorus Atrichops
Heptagenia Blephariceridae Liponeura
Rhithrogena Ceratopogonidae
Leptophlebiidae Habrophlebia Chironomidae
Habroleptoides Dolichopodidae

Odonata Empididae
Calopterygidae Agrion Limoniidae
Coenagrionidae Psychodidae
Corculegasteridae Cordulegaster Simuliidae
Gomphidae Gomphus pulchellus Stratiomyidae
Gomphus vulgatissimus Tabanidae
Onychogomphus forcipatus Tipulidae

Platycnemididae Platycnemis pennipes Hydracarina

Heteroptera Crustacea
Aphelocheiridae Aphelocheirus aestivalis Asellidae Asellus aquaticus
Veliidae Gammaridae

Megaloptera Mollusca
Sialidae Sialis Ancylidae Ancylus fluviatilis

Lepidoptera Acroloxus
Pyralidae Bithyniidae Bithynia tentaculata

Trichoptera Hydrobiidae
Brachycentridae Micrasema longulum Lymnaeidae
Ecnomidae Ecnomus Neritidae Theodoxus fluviatilis
Glossosomatidae Physidae Physa
Goeridae Silo Sphaeriidae Pisidium

Continued on p.78
78

Table 2. contd.

Hydropsychidae Cheumatopsyche lepida Sphaerium

Hydropsyche Unionidae Anodonta
Hydroptilidae Hydroptila Unio
Ithytrichia Valvatidae Valvata
Lepidostomatidae Lasiocephala basalis Oligochaeta
Lepidostoma hirtum Achaeta
Leptoceridae Adicella Erpobdellidae Erpobdella octoculata
Athripsodes Glossiphoniidae Glossiphonia
Ceraclea Helobdella stagnalis
Mystacides Hemiclepsis marginata
Limnephilidae Piscicolidae Piscicola geometra

Odontoceridae Odontocerum albicorne Turbellaria

Polycentropodidae Dugesiidae Dugesia
Psychomyiidae Planariidae Polycelis
Rhyacophila Rhyacophila
Sericostomatidae

Table 3. References and formulae (plus their labels in the figures) of structure indices calculated on patch samples during this study. pi =
number of individuals of each taxon; S = number of taxa (taxonomic richness); Q = total number of individuals; H0min = minimal diversity;
H0max = maximal diversity. The MacIntosh distance and diversity, Gleason’s index, maximal and minimal diversity, different descriptions of
evenness (Heip, Pielou, MacIntosh, Hill, Atalato, Sheldon), and Menhinick (B) index (Log2 S/Log2 Q) were also envisioned in a preliminary
analysis (Beisel, 1996)

No. Index Formula Reference Code

1 Taxonomic Number of taxa S

richness

2 Total Total number of individuals Q

abundance

3 Margalef’s S−1 Margalef (1958) Margalef

Log2 Q
index
S
4 Menhinick (A) √ Menhinick (1964) Menhinick
Q

6 pi (pi −1)
i
5 Dominance Simpson (1949) λ
Q(Q−1)
index
   
H0
p1 pi
6 Shannon’s −6 Log2 Shannon & Weaver
i Q Q
diversity (1949)
H −H min0 0
7 Hurlbert’s Hurlbert (1971) e Hurlbert
H 0 max −H 0 min
evenness
 79

structure (Bornette et al., 1994; Dolédec & Statzner,

1994; Usseglio-Polatera, 1994; 1996; Bournaud et al.,
1996). Sample scores from the two previous matrices
are computed in order to be the most covariant one to
the other, and at the same time to have the maximal
standard deviation. As a result, both types of sample
scores can be compared directly to the same factorial
map (not shown). In addition, as both the environ-
ment and indice factorial maps are presented (Figure
4), changes between samples can be simultaneously
explained by ordination of indices and by environmen-
tal modality scores. Relationships between indices and
environment variables can be observed (Figure 5). All
calculations and graphics were made using the ADE
(Analyse de Données Ecologiques) Software, a pack-
age for Multivariate Analysis and Graphical Display
for Environmental Data (Chessel & Dolédec, 1994;
Thioulouse et al., 1995).

Results

Faunal structure of patches

Standardized principal components analysis, used to

quantify patterns of covarying indices within benthic
communities, extracted two orthogonal components
accounting for 88.0% of the common variation among
indices (Figure 2). The homogeneous dispersion of
samples in the F1 × F2 factorial plane (Figure 2b)
focused on the wide variety of sampled community
structures and validated the sampling program in the
Figure 2. Results of the standardized principal component analysis sense that a particular faunal structure was not over
of seven structure indices in 189 patch samples. (a) Correlation cir-
cle of active variables (see Table 3 for full labels of indices). (b) F1 or under sampled. Axis F1 was highly correlated with
× F2 factorial plane of patch samples. The percentages of explained Shannon-Weaver’s diversity and Hurlbert’s equitabil-
variance by each axis are given. ity that were opposite Simpson’s index in structuring
the faunal data table (Figure 2a). Axis F1 may be
regarded as a diversity/evenness axis, describing the
index array) was subjected to a standardized prin- distribution of individuals into systematic units. Total
cipal components analysis. The environmental data abundance (log2Q) and taxonomic richness (S) were
(a 189 sample by seven environmental variable ar- taken into consideration by F2 axis. These two para-
ray) were processed by fuzzy correspondence analysis meters were related insofar as the total number of taxa
(Chevenet et al., 1994). The second step consisted clearly increased with the number of sampled individ-
of matching biological and environmental information uals (Frontier & Pichod-Viale, 1993). Thus, the first
to obtain a simultaneous picture of the structure of factorial plane represented the two major concepts of
fauna and environment in samples, by simultaneous taxonomic richness and evenness which diversity in-
ordination of the two previous matrices. This two- dices try to summarize with a single value (Peet, 1974;
table analysis is a co-inertia analysis (Mercier, 1991; Magurran, 1988).
Dolédec & Chessel, 1994; Prodon & Lebreton, 1994),
that is suitable for the simultaneous detection of fau-
nal and environmental features in studies of ecosystem
80

Figure 3. Ordination of environmental variables by fuzzy correspondence analysis. (a) Distribution of modalities (the number in the circle) of
seven environmental variables on the F1 × F2 factorial plane, with correlation ratios for each variable indicated on the axes (see Table 1 for
full labels of environmental variables and variable modalities, for variables that exhibit a gradient, modality no. 1 is at the lowest end of the
gradient). Each modality is positioned at the weighted average of samples representing this modality. The positions of samples are not shown.
(b) Distribution of modalities of mineral substrates along the F1 axis. Each histogram represents the frequency distribution (into ten classes) of
the coordinates, along the F1 axis, of samples belonging to a given substrate modality. The mean value of each substrate modality is represented
by a black rectangle.

Environmental structure of patches environmental variables in structuring the samples for

the two first axes (Figure 3a). Thus, on axis F1, the
Results of the fuzzy correspondence analysis (Chevenet modalities of the variables current velocity, Froude
et al., 1994) indicated eigenvalues for two factors that number, with nearly 80% of explained variance, were
were potentially important to demonstrate the main slightly more separated than those of the variable sub-
structure of the environmental data. Correlation ra- strate. The modalities of the variables patch diversity
tios, representing proportions of the total variance and substrate richness were much more separated on
explained by each axis to depict the separation among axis F2 with a percentage of explained variance equal
variable modalities, allowed ordering the relevance of to 58.4% and 50.8%. The modalities of mineral sub-
 81

Figure 4. Co-structure between structure indices and environmental variables by co-inertia analysis. (a) Ordination of environmental variables
on the F1 × F2 plane with samples positioned according to their community structure (see Table 1 and Figure 3 for further details). Correlation
ratios for each variable are indicated on the axes. (b) Position of the structure indices on the F1 × F2 co-inertia plane (see Table 3 for index
identifiers).
82

Figure 5. Relationships between structure indices and substratum types (= the modalities of the variable substrate) investigated by co-inertia
analysis. Orthogonal projections of substrate modalities along (b) taxonomic richness, (c) abundance and (d) evenness patterns on the F1 × F2
plane of the co-inertia analysis (a) are used. Each modality (a black square) is positioned at the weighted average of samples representing this
modality. Samples were positioned according to their community structure. In each case, groups of substrate modalities are distinguished by
different screens.

strates (Figure 3a, b) were mainly arranged along a for low water depths and moderate (mod. 2 and 3) val-
gradient of granulometry, from mud (modality 1 at the ues of current velocity and/or water turbulence (Figure
positive side of the F1 axis) to cobbles (modality 6 at 3a). It should be noted that substratum modalities were
the negative side of the F1 axis). This corresponded less separated when located by neighbouring patches
to a gradient with current speed (and Froude number) than when located by substrate sensu stricto. Conse-
from the low velocities on the positive side to the high quently, it may be argued that the ‘ambiance’ (i.e.,
velocities on the negative side of the F1 axis (Figure patch mosaic) within a radius of two meters round the
3a). Organic detritus (modality 11) were closely re- sampled mesohabitat, in most cases, displayed only
lated to substrates of fine particle size (mud, silt). The moderate differences. Consequently, differences in
patterns observed are thought to reflect the association community structures among mesohabitats were more
of in-stream hydraulic parameters. The mesospatial often interpreted as reflecting differences in specific
heterogeneity (var. 6 & 7) seemed to reach an optimum
 83

attributes of mesohabitats than in reflecting differences from patches displaying the hardest hydraulic condi-
in their environmental ‘ambiance’. tions (modality 5 for var. 1 and 3) were excluded from
this gradient. Furthermore, organic supports (plants,
Relationship between faunal structure and roots, leaf and twig litters) on the negative side of the
environmental structure of patches F2 axis were opposite the compact mineral substrates
(i.e., clay and flagstone).
The relationship between faunal structure and environ- To show potential relationships between major bio-
mental attributes was examined in a co-inertia analy- cenotic structure attributes and substratum types, we
sis, i.e., by a simultaneous ordination of the structure chose an orthogonal projection of the positions (F1
index array and the environmental parameter array. × F2 co-inertia plane) of the modalities of substrate
According to the eigenvalues (Figure 4a), the F1 × F2 types (positioned according to their faunal structure)
ordination diagram accounted for 92.1% of the total using three indices: richness (Figure 5b), abundance
variability. A permutation test was used to examine (Figure 5c) and evenness (Figure 5d) patterns in co-
the significance of the correlation between the two inertia analysis (Figure 5a). As a result, substrate types
sets of projected coordinates of samples resulting from may be grouped together according to their location
the co-inertia analysis (Dolédec & Chessel, 1994). along each structural pattern. Then, we were able to
This test makes a comparison between the observed debate the composition of these groups in regards to
R2 value for a co-inertia axis and the distribution of bibliographic information.
R2 values obtained from 1000 co-inertia analyses us-
ing random matching of the two data files, which
are produced by randomly permutating the rows. The Discussion
first two axes demonstrated a significant (p<0.001)
co-structure (= relationship) between macrobenthic The habitat preferences of benthic macroinvertebrates
community structure and environmental description. result from the balance of a variety of requirements
Considering the factorial plane F1 × F2 (Fig- of organisms. The major role of near-bed hydraulic
ure 4b), it appeared that Shannon–Weaver’s diver- conditions in influencing their mesodistribution is self
sity, Simpson’s index and Hurlbert’s evenness were evident (Hynes, 1970; Minshall, 1984). Many work-
the most important parameters correlated with the ers have confirmed that substrate type is a useful and
environmental structure of patches and extracted by convenient predictor of the abundance and diversity of
co-inertia analysis. Axis F2 was correlated with abun- benthic macroinvertebrates (Minshall, 1984; Ormerod
dance and richness. On the Figure 4b, each modal- & Edwards, 1987; Jowett & Richardson, 1990). In this
ity was positioned on the F1 × F2 plane, at the study the nature of substratum also appeared to exert a
weighted average of samples representing this modal- strong influence on community structure. On the F1 ×
ity. Samples (not shown) were positioned according to F2 plane of co-inertia analysis, the ordination of patch
their community structure. These positions provided samples by taxonomic richness separated three groups
insights about the particular relevance of substrate, of substrate modalities that seemed to correspond to
Froude number, current velocity and water depth in an increasing support complexity gradient (Figure 5b).
order to study the co-structure between community The first group comprised substrates with low lo-
structure and environment organization. In contrast, cal heterogeneity (i.e., at a meso/microscale), plotted
patch diversity and substrate richness did not seem to on the negative side of the richness axis. This set
be pertinent to explain faunal structure. These vari- of mesohabitats comprised (1) hard substrates which
ables related to the possible impact of faunal contagion were colonizable only in two dimensions, and (2)
on the structure of macrobenthic communities. Their substrates with fine particle size in which compact-
positions on the F1 × F2 plane of the co-inertia ness and/or unstability were also highly selective for
analysis suggested that this biotic phenomenon had no most macroinvertebrates. In contrast, patches display-
significant impact at this analysis level. The first axis ing a high microheterogeneity (e.g. roots, bryophytes)
highlighted a water depth gradient (Figure 4a). Addi- were plotted on the positive side of the richness axis.
tionally, a patch ordination according to their current All other substrates were located in an intermediate
speed and substrate granulometry was found along the position. Therefore microspatial heterogeneity, ap-
first axis from the lentic muddy or silty mesohabitats peared to enhance the taxonomic richness of a faunal
to the lotic strong substrates. However, communities community.
84

Substrates with a high intra-habitat heterogeneity, Our findings (see F1 scores of the co-inertia analy-
such as bryophytes (see Suren, 1991, 1993) or roots sis: Figure 4a) were consistent with the common
(Bournaud & Cogérino, 1986), are assumed to be of hypothesis that benthic macroinvertebrate abundance
considerable importance to invertebrate microdistrib- would increase with substrate size up to cobble size
ution because they provide a wide range of refugia and decrease as the substrate becomes boulder or
which can serve to buffer populations against a variety bedrock (Ward, 1975, Brusven 1984, Minshall, 1984;
of abiotic perturbations or biotic interactions. Then, Jowett & Richardson, 1990; Quinn & Hickey, 1990).
they have a variety of important functions (oviposi- Evenness (described by Hurlbert’s index) was
tion site, protective nursery for small larvae, plant strongly related to the F1 axis (Figure 5d). This axis
detritus or periphyton trapping, refugial space during corresponded to an increasing gradient of hydraulic
spates) which can promote the taxonomic richness of constraints (up to 90 cm s−1 for current speed and/or
patch communities. Many researchers have focused 6.4 for Froude number). We noticed a clear succession
on the low taxonomic richness of mineral bottoms of of three groups of substrate types along the evenness
fine granulometry, especially sands. The size and the axis: fine particle size substrates, organic substrates
mobility of their particles constrain benthic communi- and mineral substrates of coarse particle size. This
ties to species capable of penetrating the substratum have corresponded to a habitat stability gradient that
interstices. mainly refers to a mixture of disturbance vulnerability
Resource limitation is often proposed as a promi- and durational stability of patches. Muddy, silty and
nent factor in community dynamics (Richardson, sandy substrates (Figure 5d) are the most fluctuating
1991; Dobson & Hildrew, 1992; Diehl 1993). In this habitats because hydraulic variations (i.e., floods) will
work, total abundance of invertebrates, which was be more likely to act as disturbances for such mineral
closely related to F2 axis of the co-inertia analysis habitats. Living plants change according to their life
(Figure 5c), seemed to be linked to food availabil- history. During the autumn campaign, some sampled
ity, as demonstrated for benthic species by Egglishaw plants were already in a declining period. In contrast,
(1969), Crowder & Cooper (1982), Drake (1984), mineral substrates, from gravel to flagstone, are more
Gilinsky (1984), Gregg & Rose (1985), and Winter- stable in time. Furthermore, the water depth which
bourn (1990). The amount (and most likely diversity) increased with evenness, also contributes to the sub-
of food increased along this axis. Clay and flagstone strate stability: the more important is the water depth,
may be considered as offering the lowest food availi- the less probable is a periodic desiccation. The best
bility. They cannot act as coarse detritus collectors contribution of this variable on the F1 × F2 co-inertia
and macroinvertebrates (excluding predators) can only plane in comparison to the corresponding CA plane
feed upon the algal layer. The second group was (Figure 3a), supported the relationship between water
composed of mineral substrates that are able to trap depth and faunal structure.
organic detritus (Rabeni & Minshall, 1977). Substrates Diversity, as defined by Shannon–Weaver, was
of this group offer more suitable sites for intensive strongly related to F1 axis (Figure 5a). Located be-
growth rates of benthic algae. The third group mainly tween evenness and taxonomic richness, diversity may
consisted of organic detritus and living plants demon- be regarded as a compromise between mesohabitat
strating rather similar trophic potentialities to mineral complexity and substrate stability in time. Margalef
substrates in addition to being themselves a poten- (1958, 1963) suggested using species diversity as a
tial food for benthic invertebrates. Tree roots sampled measure of stability. High species diversity had of-
on banks near the water surface, appeared to be the ten been related to spatial heterogeneity or predator-
substrate displaying the most important trophic poten- mediated coexistence (Resh et al., 1988). It was
tialities because of their ability to filter and collect not surprising that Simpson’s index, as an evaluation
a wide range of detrital plant material. This trophic of dominance or homogeneity (Kvalseth, 1991) and
gradient may illustrate the relationship between sub- Shannon’s formula as a measure of structural hetero-
strate trophic potentialities and observed invertebrate geneity (Peet, 1974) were in opposition on F1 axis of
numbers. Organisms probably more likely aggregate the co-inertia analysis (Figure 4d).
where they can easily use food supplies. However, not Water depth and current velocity also emerged
only abundance, but also palatability of food resources as factors influencing mesohabitat community struc-
must be taken into account (Dobson, 1994). ture. Water depth was positively related to evenness
and, to a lesser extent, negatively related with abun-
 85

dance (Figure 4). An increase in water level (over a that the size distribution often reflected a past flood-
certain range) may reduce the probability of desicca- ing rather than present flow conditions. Consequently,
tion, increase the patch stability, and then promote especially in low flow conditions, substratum size dis-
invertebrate densities. But an increase in water depth tribution may have little to do with the force of flow
may also reduce the substrate illumination which can at the sampling date. In conclusion, it is difficult to
negatively affect animal distribution (Hugues, 1966a, discriminate the effect of each variable on community
b; Dudgeon, 1988; Sweeney, 1993). An impact on structure of natural patches.
evenness has never been demonstrated. Minshall (1988) and Resh et al. (1988) argued that
The current velocity was positively related to di- the mechanisms controlling the colonization dynamics
versity (Degani et al., 1993) and, to a lesser extent, of lotic macroinvertebrates are important when ex-
with taxonomic richness (Figure 4). Many inverte- plaining benthic community structure. Abiotic factors
brates have an inherent need for current and also ex- are assumed to be most important during early com-
hibit an upper tolerance limit (Hynes, 1970; Minshall, munity development, whereas biotic factors come into
1984). However, mesoscale heterogeneity creates lo- play later (Robinson et al., 1990). But, in this study,
cal variation in velocity about patches, which allows patch diversity and substrate richness did not emerge
invertebrates to find satisfactory mesohabitats in a as factors structuring mesohabitat community organi-
broad range of current speed, regardless of the mean zation. We speculated that in the sampled streams,
water column velocity (Jowett & Richardson, 1990). mesohabitats displayed a nearly similar combination
Consequently, structurally complex habitats are ex- of patches in their close proximity (within a radius
pected to sustain more diverse communities than struc- of two meters round the sampling site) so that the
(PLJUDFLRQ
turally simple ones. Under low flow conditions, low potential pathways of colonization vary slightly be- GLVSRQLELOLG
velocities decrease the competence and the capacity of tween patches. This suggested that abiotic and biotic
flow for sediment transport. As a result, fine deposits characteristics of each patch controlled colonization,
accumulate over the stream-bed. These circumstances and the subsequent community structure of mesohab-
might be thought to simplify the habitat and to remove itats. Kohler (1985) observed that emigration rates by
much of the potential micro-environmental variation. Baetis tricaudatus differed on substrate having either
Opinions vary on the effects of sediment addition on low or high food resource levels, with emigration rates
community structure. Most investigators report that being lower from substrates having high food level.
the density of benthic macroinvertebrates is severely Richards & Minshall (1988) found that residence time
reduced (Nuttall & Bielby, 1973; Lenat et al., 1981; by Baetis tricaudatus on stone was positively corre-
Rutherfold & Mackay, 1986; Cobb & Flannagan, lated with periphyton abundance. In testing variations
1990). Other workers argue that sediment addition has of the initial benthic density in experimental cages
few effects on total density (Hamilton, 1961; Chutter, on the immigration or emigration rates of species,
1969b; Barton, 1977). A change in the community Peckarsky (1979) distinguished in particular density-
composition is not universal (Lenat et al., 1981). dependent dispersers which increased their emigration
These apparent conflicts may illustrate the variable rates with increasing density of patch assemblage and
effects of reduction of water velocity and a complex density-dependent aggregators which increased their
trade-off between flow and food, because flow re- immigration rates with increasing initial density of
duction may be partly responsible for the increased their own species.
deposition of organic material (detritus), a useful food In this study we attempted to evaluate some of
supply for invertebrates (Peckarsky & Penton, 1990). the environmental factors affecting the community
In fact, most of the abiotic attributes of patches are structure in first- through fourth-order streams of
inter-related in complex ways (Minshall & Minshall, three drainage basins. The results and analyses pre-
1977; Lamberti & Resh, 1979; Statzner et al., 1988; sented described clear relationships between commu-
Jowett & Richardson, 1990). The interaction of veloc- nity structure of macroinvertebrate assemblages and
ity and depth should be considered by an exponential environmental factors by a co-inertia analysis. Our
polynominal model (Gore & Judy, 1981). The current results emphasized the importance of substrate het-
speed is correlated with substrate particule size (see re- erogeneity in benthic invertebrate studies. Species
sults of the CA), despite the demonstration by Statzner richness obviously increases with habitat heterogene-
et al. (1988) that substratum size alone was a poor ity, total abundance increased with food availability
descriptor of near-bottom hydraulics, the reason being
86

and both equitability and diversity seemed to increase Chutter, F. M., 1969a. The distribution of some stream invertebrates
with substrate stability. in relation to current speed. Int. Revue ges. Hydrobiol. 54: 413–
422.
In their development of habitat templets and their Chutter, F. M., 1969b. The effects of silt and sand on the invertebrate
possible application to the study of freshwater benthic fauna of streams and rivers. Hydrobiol. 34: 57–76.
communities, Southwood (1977, 1988), Hildrew & Cobb, D. G. & J. F. Flannagan, 1990. Trichoptera and substrate
Townsend (1987), Townsend (1989), and Townsend stability in the Ochre River, Manitoba. Hydrobiol. 206: 29–38.
Cowie, A., 1985. An analysis of changes in the invertebrate commu-
&Hildrew (1994) focused on the premise that the habi- nity along a southern New Zealand montane stream. Hydrobiol.
tat provides the templet on which evolution forges 120: 35–46.
characteristic life history strategies. Therefore, cer- Crowder, L. B. & W. E. Cooper, 1982. Habitat structural complexity
tain combinations of adaptations are assumed to be between bluegills and their prey. Ecology 63: 1802–1813.
Cummins, K. W. & G. H. Lauff, 1969. The influence of substrate
related to different habitats (see for example Dolédec particle size on the microdistribution of stream macrobenthos.
& Statzner (1994) for 548 plant or animal species; Hydrobiol. 34: 145–181.
Usseglio-Polatera (1994) for 264 Insecta species). Cummins, K. W., G. W. Minshall, J. R. Sedell, C. E. Cushing & R.
Consequently, testing relationships between biolog- C. Petersen, 1984. Stream ecology theory. Verh. int. Ver. Limnol.
20: 1818–1827.
ical traits or ecological strategies of taxa and both Degani, G., G. N. Herbst, R. Ortal, H. J. Bromley, D. Levanon,
mesoscale community organization or patch attributes Y. Netzer, N. Harari & H. Glazman, 1993. Relationship between
might provide further insights into problems of com- current velocity, depth and the invertebrate community in a stable
river system. Hydrobiol. 263: 163–172.
munity structure.
Diehl, S., 1993. Relative consumer sizes and the strengths of direct
and indirect interactions in omnivorous feeding relationships.
Oikos 68: 151–157.
Acknowledgments Dobson, M., 1994. Microhabitat as a determinant of diversity:
stream invertebrates colonizing leaf packs. Freshwat. Biol. 32:
565–572.
We thank the Agence Financière de Bassin Rhin- Dobson, M. & A. G. Hildrew, 1992. A test of resource limitation
Meuse for financial support, and Mrs Antoine and among shredding detritivores in low order streams in southern
Mrs Chaumeil for the linguistic corrections of the England. J. anim. Ecol. 61: 69–77.
Doeg T. J., P. S. Lake & R. Marchant, 1989. Colonization of exper-
manuscript.
imentally disturbed patches by stream macroinvertebrates in the
Acheron River, Victoria. Austral. J. Ecol. 14: 207–220.
Doeg T. J., R. Marchant, M. Douglas & P. S. Lake, 1989b. Exper-
References imental colonization of sand, gravel and stones by macroinver-
tebrates in the Acheron River, southeastern Austral. Freshwat.
Barton, B. A., 1977. Short-term effects of highway construction on Biol. 22: 57–64.
the limnology of a small stream in southern Ontario. Freshwat. Dolédec, S. & D. Chessel, 1994. Co-inertia analysis: an alter-
Biol. 7: 99–108. native method for studying species–environment relationships.
Bornette, G., C. Amoros & D. Chessel, 1994. Rejuvenation in for- Freshwat. Biol. 31: 277–294.
mer braided channels of the Rhône River: successional patterns Dolédec, S. & B. Statzner, 1994. Theoretical habitat templets,
and allogenic processes. J. Veg. Sci. 5: 237–246. species traits, and species richness: 548 plant and animal species
Bournaud, M. & L. Cogérino, 1986. Les microhabitats aquatiques in the Upper Rhône River and its floodplain. Freshwat. Biol. 31:
des rives d’un grand cours d’eau: approche faunistique. Ann. 523–538.
Limnol. 22: 285–294. Drake, J. A., 1984. Species aggregation: the influence of detritus in
Bournaud, M., B. Cellot, P. Richoux & A. Berrahou, 1996. Macroin- a benthic invertebrate community. Hydrobiologia 112: 109–115.
vertebrate community structure and environmental characteris- Dudgeon, D., 1988. The influence of riparian vegetation on
tics along a large river: congruity of patterns for identification to macroinvertebrate community structure in four Hong Kong
species or family. J. North Am. Benthol. Soc. 15: 232–253. streams. J. Zool. 216: 609–627.
Brussock, P. P. & A. V. Brown, 1991. Riffle-pool geomorphology Edington, J. M., 1968. Habitat preferences in net-spinning caddis
disrupts longitudinal patterns of stream benthos. Hydrobiol. 220: larvae with special reference to the influence of water velocity. J.
109–117. anim. Ecol. 37: 675–692.
Brusven, M. A., 1984. The distribution and abundance of benthic in- Egglishaw, H. J., 1964. The distributional relationship between the
sects subjected to reservoir release flows in the Clearwater River, bottom fauna and plant detritus in streams. J. anim. Ecol. 33:
Idaho, USA. In A. Lillehammer & S. Saltveit (eds), Regulated 463–476.
Rivers. Universitetsforlaget AS, Oslo: 167–180. Egglishaw, H. J., 1969. The distribution of benthic invertebrates on
Chessel, D. & S. Dolédec, 1994. ADE Software. Multivariate substrata in fast-flowing streams. J. anim. Ecol. 38: 19–33.
Analyses and Graphical Display for Environmental Data (version Erman, D. C. & N. A. Erman, 1984. The response of stream in-
3.7) Vol. 1 User’s Manual, Vol. 2 Examples, Vol. 3 Graphics vertebrates to substrate size and heterogeneity. Hydrobiol. 108:
Documentation. Université Lyon 1, France, 523 pp. 75–82.
Chevenet, F., S. Dolédec & D. Chessel, 1994. A fuzzy coding ap- Fleeger, J. W. & A. W. Decho, 1987. Spatial variability of interstitial
proach for the analysis of long-term ecological data. Freshwat. meiofauna: A review. Stygologia 3: 35–54.
Biol. 31: 295–309.
 87

Frontier, S. & D. Pichod-Viale, 1993. Ecosystèmes: Structure, McAuliffe, J. R., 1984. Competition for space, disturbance, and the
Fonctionnement, Évolution. Masson, Paris, 447 pp. structure of a benthic stream community. Ecology 65: 894–908.
Furse, M. T., D. Moss, J. F. Wright & P. D. Armitage, 1984. The in- Menhinick, E. F., 1964. A comparison of some species-individuals
fluence of seasonal and taxonomic factors on the ordination and diversity indices applied to samples of field insects. Ecology 45:
classification of running-water sites in Great Britain and on the 859–861.
prediction of their macro-invertebrate communities. Freshwat. Mercier, P., 1991. Etude des relations espèces-environnement et
Biol. 14: 257–280. analyse de la co-structure d’un couple de tableaux. Thesis,
Gilinsky, E., 1984. The role of fish predation and spatial hetero- Université Lyon I, 178 pp.
geneity in determining benthic community structure. Ecology 65: Minshall, G. W., 1984. Aquatic insect substratum relationships In
455–468. V.H. Resh & D.M. Rosenberg (eds), The Ecology of Aquatic
Gore, J. A. & R. D. Judy Jr., 1981. Predictive models of benthic Insects. Praeger, N.Y.: 358–400.
macroinvertebrate density for use in instream flow studies and Minshall, G. W., 1988. Stream ecosystem theory: a global perspec-
regulated flow management. Can. J. Fish. aquat Sci. 38: 1363– tive. J. n. am. benthol. Soc. 7: 263–288.
1370. Minshall, G. W. & J. N. Minshall, 1977. Microdistribution of
Gregg, W. W. & F. L. Rose, 1985. Influences of aquatic macro- benthic invertebrates in a rocky mountain (U.S.A.) stream. Hy-
phytes on invertebrate community structure, guild structure, and drobiol. 55: 231–249.
microdistribution in streams. Hydrobiol. 128: 45–56. Nutall, P. M. & G. H. Bielby, 1973. The effects of china-clay wastes
Hamilton, J. D., 1961. The effect of sand-pit washings on a stream on stream invertebrates. Envir. Pollut. 5: 77–86.
fauna. Verh. int. Ver. Limnol. 14: 435–439. Ormerod, S. J. & R. W. Edwards, 1987. The ordination and classifi-
Hawkins, C. P., M. L. Murphy & N. H. Anderson, 1982. Effects cation of macroinvertebrate assemblages in the catchment of the
of canopy, substrate composition, and gradients on the structure River Wye in relation to environmental factors. Freshwat. Biol.
of macroinvertebrate communities in Cascade Range streams of 17: 533–546.
Oregon. Ecology 62: 387–397. Orth, D. J. & O. E. Maughan, 1983. Microhabitat preferences of
Hildrew, A. G. & C. R. Townsend, 1987. 16. Organization in benthic fauna in a woodland stream. Hydrobiol. 106: 157–168.
freshwater benthic communities. In J. H. R. Gee & P. S. Giller Peckarsky, B. L., 1979. Biological interactions as determinants of
(eds), The 27th Symposium of the British Ecological Society. distributions of benthic invertebrates within the substrate of stony
Blackwell Scientific Publication, Oxford: 347–371. streams. Limnol. Oceanogr. 24: 59–68.
Hugues, D. A., 1966a. On the dorsal light response in a mayfly Peckarsky, B. L. & M. A. Penton, 1990. Effects of enclosure on
nymph. Anim. Behav. 14: 13–16. stream microhabitat and invertebrate community structure. J. n.
Hugues, D. A., 1966b. Mountain stream of the Barberton Area, am. benthol. Soc. 9: 249–261.
Eastern Transvaal. Part II, the effect of vegetational shading Peet, R. K., 1974. The measurement of species diversity. Annu. Rev.
and direct illumination on the distribution of stream fauna. Ecol. Syst. 5: 285–308.
Hydrobiologia 20: 439–459. Pringle, C. M., R. J. Naiman, G. Bretschko, J. R. Karr, M. W.
Hurlbert, S. H., 1971. The nonconcept of species diversity: a critique Oswood, J. R. Webster, R. L. Welcomme & M. J. Winterbourn,
and alternative parameters. Ecology 52: 577–586. 1988. Patch dynamics in lotic systems: the stream as a mosaic. J.
Hynes, H. B. N., 1970. The Ecology of Running Waters. Liverpool n. am. benthol. Soc. 7: 503–524.
University Press, Liverpool: 555 pp. Prodon, R. & J. D. Lebreton, 1994. Analyses multivariées des re-
Illies, J. & L. Botosaneanu, 1963. Problèmes et méthodes de la lations espèces-milieu: structure et interprétations écologiques.
classification et de la zonation écologique des eaux courantes Vie et Milieu 44: 69–91.
considérées surtout du point de vue faunistique. Verh. int. Ver. Quinn, J. M. & C. W. Hickey, 1990. Magnitude of effects of sub-
Limnol. 12: 1–57. strate particle size, recent flooding and catchment development
Jenkins, R. A., K. R. Wade & E. Pugh, 1984. Macroinvertebrate- on benthic invertebrates in 88 New Zealand rivers. New Zeal. J.
habitat relationships in the River Teifi catchment and the signifi- Mar. Freshwat. 24: 411–427.
cance to conservation. Freshwat. Biol. 14: 23–42. Quinn, J. M. & C. W. Hickey, 1994. Hydraulic parameters and
Jowett, I. & J. Richardson, 1990. Microhabitat preferences of ben- benthic invertebrate distributions in two gravel-bed New Zealand
thic invertebrates in a New Zealand river and the development rivers. Freshwat. Biol. 32: 489–500.
of in-stream models for Deleatidium spp. New Zeal. J. Mar. Rabeni, C. F. & G. W. Minshall, 1977. Factors affecting microdis-
Freshwat. 24: 19–30. tribution of stream benthic insects. Oikos 29: 33–43.
Kohler, S. L., 1985. Identification of stream drift mechanisms: an Reice, S. R., 1980. The role of substratum in benthic macroinver-
experimental and observational approach. Ecology 66: 1749– tebrate microdistribution and litter decomposition in a woodland
1761. stream. Ecology 61: 580–590.
Kvalseth, T. O., 1991. Note on biological diversity, evenness, and Reice, S. R., 1981. Interspecific associations in a woodland stream.
homogeneity measures. Oikos 62: 123–127. Can. J. Fish. aquat Sci. 38: 1271–1280.
Lamberti, G. A. & V. H. Resh, 1979. Substrate relationships, spa- Resh, V. H., A. V. Brown, A. P. Covich, M. E. Gurtz, H. W. Li, G.
tial distribution patterns, and sampling variability in a stream W. Minshall, S. R. Reice, A. L. Sheldon, J. B. Wallace & R. C.
caddisfly population. Envir. Ent. 8: 561–567. Wissmar, 1988. The role of disturbance in stream ecology. J. n.
Lenat, D. R., D. L. Penrose & K. W. Eagleson, 1981. Variable effects am. benthol. Soc. 7: 433–455.
of sediment addition on stream benthos. Hydrobiol. 79: 187–194. Resh, V. H. & E. P. McElravy, 1993. 5. Contemporary quantitative
Magurran, A. E., 1988. Ecological diversity and its measurement. approaches to biomonitoring using benthic macroinvertebrates.
Princeton University Press, Princeton, 179 pp. In D. M. Rosenberg & V. H. Resh (eds), Freshwater Biomoni-
Margalef, R., 1958. Information theory in ecology. Gen. Syst. 3: toring and Benthic Macroinvertebrates. Chapman & Hall, N.Y.:
36–71. 159–194.
Margalef, R., 1963. On certain unifying principles in ecology. Am.
Nat. 97: 357–374.
 88

Richards, C. & G. W. Minshall, 1988. The influence of periphyton Thioulouse, J., S. Dolédec, D. Chessel & J. M. Olivier, 1995. ADE
abundance on Baetis bicaudatus distribution and colonization in software: multivariate analysis and graphical display of environ-
a small stream. J. North Am. Benthol. Soc. 7: 77–86. mental data. In G. Guariso & A. Rizzoli (eds), Software per
Richardson, J. S., 1991. Seasonal food limitation of detritivorous l’ambiante. Pàtron Editore, Bologna: 57–62.
insects in a montane stream: an experimental test. Ecology 72: Townsend, C. R., 1989. The patch dynamics concept of stream
873–887. community ecology. J. n. am. benthol. Soc. 8: 36–50.
Robinson, C. T., G. W. Minshall & S. R. Rushforth, 1990. Seasonal Townsend, C. R. & A. G. Hildrew, 1994. Species traits in relation to
colonization dynamics of macroinvertebrates in an Idaho stream. a habitat templet for river systems. Freshwat. Biol. 31: 265–275.
J. n. am. benthol. Soc. 9: 240–248. Usseglio-Polatera, P., 1994. Theoretical habitat templets, species
Rutherfold, J. E. & R.J. Mackay, 1986. Patterns of pupal mortal- traits, and species richness: aquatic insects in the Upper Rhône
ity in field populations of Hydropsyche and Cheumatopsyche River and its floodplain. Freshwat. Biol. 31: 417–437.
(Trichoptera: Hydropsychidae). Freshwat. Biol. 16: 337–350. Usseglio-Polatera, P., 1996. Stratégies biologiques et évolution en-
Sagova, M & M. S. Adams, 1993. Aggregation of numbers, size vironnementale d’un hydrosystème: l’exemple des Trichoptères
and taxa of benthic animals at four levels of spatial scale. Arch. (Insecta) du Rhône à Lyon. Geobios (in press).
Hydrobiol. 128: 329–352. Van Rijckevorsel, J., 1987. The Application of Fuzzy Coding and
Shannon, C. E. & W. Weaver, 1949. The Mathematical Theory of Horseshoes in Multiple Correspondence Analysis. DSWO Press,
Communication. University of Illinois Press, Urbana, 117 pp. Leiden, 272 pp.
Simpson, E. H., 1949. Measurement of diversity. Nature 163: 688. Ward, J. A., 1975. Bottom fauna-substrate relationships in a
Southwood, T. R. E., 1977. Habitat, the templet for ecological northern Colorado trout stream: 1945 and 1974. Ecology 56:
strategies. J. anim. Ecol. 46: 337–365. 1429–1434.
Southwood, T. R. E., 1988. Tactics, strategies and templets. Oikos Ward, J. V. & M. A. Palmer, 1994. Distribution patterns of inter-
52: 3–18. stitial freshwater meiofauna over a range of spatial scales, with
Statzner, B., 1981. The relation between ‘hydraulic stress’ and emphasis on alluvial river-aquifer systems. Hydrobiologia 287:
microdistribution of benthic macroinvertebrates in a lowland 147–156.
running water system, the Schierenseebrooks (North Germany). Wasson, J. G., B. Dumont & F. Trocherie, 1981. Protocole de
Arch. Hydrobiol. 91: 192–218. description des habitats aquatiques et de prélèvements des in-
Statzner, B., 1987. Characteristics of lotic ecosystems and conse- vertébrés benthiques dans les cours d’eau. CEMAGREF de Lyon,
quences for future research directions. In E.-D. Schulze & H. Division Qualité des Eaux, Pêche et Pisciculture, 18 pp.
Zwölfer (eds), Potentials and Limitations of Ecosystem Analysis, Winterbourn, M. J., 1982. Food utilization by a stream detritivore
Ecological Studies 61. Springer-Verlag, N.Y.: 365–390. Zelandopsyche ingens (Trichoptera: Oeconesidae). Int. Revue
Statzner, B., J. A. Gore & V. H. Resh, 1988. Hydraulic stream ges. Hydrobiol. 67: 209–222.
ecology: observed patterns and potential applications. J. n. am. Winterbourn, M. J., 1990. Interactions among nutrients, algae and
benthol. Soc. 7: 307–360. invertebrates in a New Zealand mountain stream. Freshwat. Biol.
Suren, A. M., 1991. Bryophytes as invertebrate habitat in two New 23: 463–474.
Zealand alpine streams. Freshwat. Biol. 26: 399–418. Wu, J. & S. A. Levin, 1994. A spatial patch dynamic modeling
Suren, A. M., 1993. Bryophytes and associated invertebrates in first- approach to pattern and process in an annual grassland. Ecol.
order alpine streams of Arthur’s Pass, New Zealand. New Zeal. Monogr. 64: 447–464.
J. Mar. Freshwat. 27: 479–494.
Sweeney, B. W., 1993. Effects of streamside vegetation on macroin-
vertebrate communities of White Clay Creek in Eastern North
America. Proc. Acad. nat. Sci. Philad. 144: 291–340.

View publication stats