BIOTIC AND ABIOTIC INTERACTIONS IN CENTRAL TEXAS

STREAMS WITH SPECIAL REFERENCE

TO NAUCORIDAE (HEMIPTERA)

by

DAVID PATRICK HERRMANN, B.S.

A THESIS

IN

ENTOMOLOGY

Submitted to the Graduate Faculty of Texas Tech University in

Partial Fulfillment of the Requirements for

the Degree of

MASTER OF SCIENCE

Approved

Accepted

May, 1992

ACKNOWLEDGMENTS

I am indebted to my major advisors, Robert Sites and Michael Willig

for support throughout this project. Under their guidance and supervision,

numerous doors of opportunity have been opened for which I will always be

grateful. I also thank Harlan Thorvilson who served on my committee and

provided critical reviews of my work. I gratefully acknowledge the

following people for their instrumental field and laboratory assistance:

Jeffery Back, Leslie Baker, Daniel Brooks, Gerardo Camilo, Steven Cox,

Brian Croyle, Clara Langner, Richard Stevens and Judy York. Only through

their patience and diligence was this project completed successfully.

Anthony Adamo, James Herrmann, Becky Nichols, and James Suh provided

technical assistance for which I am most thankful. I also extend thanks to

Brant Baugh, John Cary, Thomas Doederlein, Dou Dou Faye, Doug Paxton,

Sergio Sanchez-Pena, and Clinton Smith for their friendships and assistance

with various chapters of this manuscript. Timothy Elkner and Dary 1

Moorhead critically reviewed portions of this manuscript. The staff of the

Texas Tech University Center at Junction (C. Len Ainsworth, Bill Durbin,

Martha Richardson, and Betty Street) provided logistic support during the

long summer months of field work. This project was supported by

Advanced Research Program grant 003644-081, State of Texas Higher

Education Coordinating Board.

11

TABLE OF CONTENTS

ACKNOWLEOOMENTS ....................................................................... 11

LIST OF TABLES .................................................................................. IV

LI~T OF ~<:J~E~ ················································································"i

I. IN'TRODUCTION ...................................................................... 1

o"~I1Vi~~ ............................................................................. 1

~tllcl)f Sit~s ........................................................................... ~

Stlldy Organisms ................................................................. 2

Literatllre Cited .................................................................... 4

II. A LABORATORY FLOW TANK WITH VARIABLE CURRENT AND DEPTHS FOR REPLICATIN<:J RIFFLES AND SHALLOW STREAMS ................................... 5

Abstract ................................................................................ 5

Introduction ......................................................................... 5

Materials and Methods ........................................................ 7

D. . ISCUS Sl on ......................................................................... .

Literature Cited ................................................................. .

ill. INFLUENCE OF C~RENT VELOCITY ON SUBSTRATUM SELECTION BY NAUCORIDAE (HEMIPTERA): AN EXPERIMENTAL APPROACH VIA

13

15

STRE-AM SIMULATION ......................................................... 16

Abstract .............................................................................. 16

Introduction ....................................................................... 17

Materials and Methods ...................................................... 19

Results ............................................................................... 21

D. . lSCUSSIOll •••••••••••••••••••••••••••••••••••••••••••••••••••••••••.•.•.•••••••••••• 26

Litemtllre Cited .................................................................. 31

lll

IV. DENSITY-DEPENDENT HABITAT SELECTION AND COMPETITION AMONG THREE SYMPATRIC NAUCORIDS ........................................................................... 35

Abstract .............................................................................. 35

In.troduction ....................................................................... 35

Materials an.d Methods ...................................................... 39

~~Slll~ •••••••••.••..•••••••••••••.••••••••••••••••••••.•••••••••••••••••••••••••••••• ~3

Discussion .......................................................................... 52

Litera.ture Cited .................................................................. 59

V. EFFECfS OF SUBSffiATUM SIZE AND HETE~OGENEITY ON MAC~OBENTHIC DIVE~SITY AND Cc:>~c:>~~c:>l'l .............................................................. ()~

Abstract .............................................................................. 63

In.troduction ....................................................................... 64

Materials an.d Methods ...................................................... 66

~esults an.d Discussion ...................................................... 68

Literature Cited .................................................................. 85

IV

LIST OFT ABLES

3.1--Separate replicated goodness of fit tests at each of three current velocities for the three possible pairs of substratum size (denoted by A, B, and C) offered to Ambrysus circumcinctus, Cryphocricos hun~erfordi, and Limnocoris lutzi .......................... 22

4.1--Likelihood ratio chi-square analyses for each contrast of substratum size for A. circumcinctus, .C. hun~erfordi, and L. l11~i ........................................................................................... ~~

4.2--Likelihood ratio chi-square analyses for the two-way interaction of density and substratum selection at each of three current velocities for experiments in which a significant three-way interaction (velocity x density x substratum) was detected ......................................................................................... 46

4.4--A priori comparisons of substratum selection for different density treatments based upon Z-scores (SPSS Inc., 1988) ........... 4 7

5.1--Abiotic conditions (mean+ S.D.) at South and North Llano River sites (N=25) ............................................................... 69

5.2--Mean density(+ S.E.) of macro invertebrate taxa (per 0.09 m2 found in the South and North Llano Rivers according to substrate size (CG, coarse gravel; SP, small pebble; SC, small cobble; HM, heterogeneous mix) (n=75) ............................ 70

5.3--Comparisons of species diversity between all possible pairs of four substratum types in each of two different rivers South, (SLR) and North, (NLR) Llano River ............................... 80

v

LIST OF FIGURES

2.1--Side view depicting overall format of the artificial stream. Water is cycled through the pump within the lower reservoir into the upper reservoir .......................................... 8

2.2--Plan view showing components (quadrats, weir, and screen frames) within the upper reservoir. Abbreviations are: GV= gate valve, WF=weir frame, S=screen, P=pump ............................. 9

2.2--Frontal and rear views of the upper reservoir showing weir used to raise the water level and screen for retaining organisms in the upper reservoir ..................................................................... 11

2.3--The relationship between the lower weir height and current velocity .............................................................................. 14

V1

CHAPTER I

INTRODUCTION

Overview

Ecology focuses on the factors which influence the structure and

function of ecosystems. In freshwater systems, many species exhibit a

benthic mode of existence (Ward, 1992). Therefore, it is not surprising

that the substratum has been considered a major determinant of the

distribution and abundance of macroinvertebrates (Minshall, 1984). Recent

studies have questioned the importance of the substratum (Statzner et al.,

1988), in tum focusing on hydraulic attributes as primary factors that

influence community function and structure. Nonetheless, current and

substratum are correlated in natural systems (Reice, 1980) and may have

independent effects on species microdistributions. Only rigorous

experimentation and hypotheses testing can elucidate the importance of

each factor separately, thereby facilitating a fundamental understanding of

complex interactions occurring in lotic systems.

This thesis comprises four manuscripts, each of which deals directly

or indirectly with microhabitat selection of stream invertebrates with

particular reference to creeping water bugs (Naucoridae). The first

manuscript (Chapter II) is methodological and concerns the simulation of

lotic environments in laboratory situations. The paper briefly reviews

problems associated with riffle simulation (including current limitations),

introduces potential solutions, and describes a simple yet flexible design for

use in ecological studies. Subsequent experiments within the thesis use this

methodology, at least in part. In the second manuscript (Chapter ill),

several experiments examined the influence of current velocity on

1

substratum selection by three riffle-inhabiting predators (Hemiptera:

Naucoridae) of central Texas streams. The experiments were aimed at

testing the hypotheses that current velocity mediates macrobenthic

microdistributions. The third paper (Chapter IV) builds upon the results

obtained in Chapter m, and describes additional laboratory experiments

that evaluated intra- and interspecific competition for space among

naucorids. Finally, the fourth manuscript (Chapter V) expands the domain

of the thesis to include entire macro benthic assemblages, and evaluates the

importance of substratum size and heterogeneity on species diversity and

taxonomic composition. This thesis illustrates the complexity of biotic and

abiotic interactions occurring in stream communities and exemplifies the

usefulness of a quantitative approach in lotic ecology.

Study Sites

The South and North Llano Rivers are located on the Edwards

Plateau in central Texas. Naucorids for stream simulation were collected

from the former; whereas, both are considered and compared in the

chapter dealing with substratum effects on community organization.

Appropriate details of the study sites are included in the materials and

methods of each chapter.

Study Or&anisms

Creeping water bugs (Hemiptera: Naucoridae) are predacious

aquatic insects common in lentic and lotic habitats of the southwestern U.S.

and tropical regions of the world. Although these insects are considered

keystone consumers in many aquatic systems (Sites and Willig, 1991 ), little

is known of their basic biologies (Gonsoulin, 1973; Constantz, 1974; Stout,

2

1978; Venk.atesan and Cloarec, 1988). As a consequence, the role of biotic

and abiotic interactions are evaluated on substratum selection by the three

common riffle inhabiting species, Ambr.ysus circumcinctus (Montandon,

Cryphocricos hun&erfordi (Usinger}, and Limnocoris lutzi (La Rivers},

and entire macroinvertebrate assemblages.

3

Literature Cited

Constantz, G. D. 1974. The mating behavior of a creeping water bug, Ambr.ysus occidenta}is (Hemiptera: Naucoridae ). Am. Midi. Nat. 92:234-239.

Gonsoulin, G. J. 1973. Seven families of aquatic and semiaquatic Hemiptera in Louisiana. Entomol. News 84:83-88.

Minshall, G. W. 1984. Aquatic insect-substratum relationships, pp. 358-400. In V. H. Resh and D. M. Rosenberg (eds.). The ecology of aquatic insects. Praeger, New York.

Reice, S. R. 1980. The role of substratum in benthic macroinvertebrate microdistribution and litter decomposition in a woodland stream. Ecology 61:580-590.

Sites, R. W. and M. R. Willig. 1991. Microhabitat associations of three sympatric species of Naucoridae (Insecta: Hemiptera). Environ. Entomol. 20:127-134.

Statzner, B., J. A. Gore and V. H. Resh. 1988. Hydraulic stream ecology: observed patterns and potential applications. J. N. Am. Benthol. Soc. 7:307-360.

Stout, R. J. 1978. Migration of the aquatic hemipteran Limnocoris insularis (Naucoridae) in a tropical lowland stream (Costa Rica, Central America). Brenesia 14:1-11.

Venkatesan, P. and A. Cloarec. 1988. Density dependent prey selection in llyocoris (Naucoridae). Aquat. Insec. 10:105-116.

Ward, J. V. 1992. Aquatic insect ecology. 1. Biology and habitat. Wiley and Sons, Inc. New York.

4

CHAPTER IT

A LABORATORY FLOW TANK WITH VARIABLE

CURRENT AND DEPTHS FOR REPLICATING

~ESANDSHALLOWSTREAMS

Abstract

Laboratory streams have been used for ecological studies of lotic

organisms, providing a means to determine the relative importances of

physical factors in structuring communities. Herein, an inexpensive

( <$500) artificial stream is described that has several advantages over other

designs; it offers: (1) accurate current velocities of 0 - 45.5 em/sec, (2) a

large surface area for experimentation, (3) both spatial and temporal

current homogeneity, and ( 4) riffle simulation. This artificial stream

requires approximately 1.8 m2 of floor space and can be used indoors.

Introduction

Studying the behavior of benthic animals is aided by laboratory

studies (Cummins, 1962). For example, the association between behavior

and abiotic factors can be determined by individually varying factors that

are not independent in natural situations. Such experiments are

fundamental to studies of interspecific interaction (competition or

predation) or habitat selection.

Artificial streams provide a means to experimentally modify a

number of parameters which are correlated in nature or would require

extensive modification of natural systems. The options in artificial stream

design include open or closed systems (Gee and Bartnik, 1969). Open

systems require constant input of new water and are useful in observing the

5

spawning behavior of fish (Gee and Bartnik, 1969); however, it is difficult

to suppress environmental influences and this design includes many natural

constraints. The closed system, in which water is internally circulated, is

often used in aquatic insect studies and can be used to determine which

factors affect behavior and distribution of macrobenthos. Closed systems

have been used to rear insect larvae (Sudia, 1951; Mason and Lewis, 1970),

determine the effects of siltation on the microdistributions of stream insects

(Cummins and Lauff, 1969), evaluate the effects of substratum type on

prey vulnerability (Fuller and Rand, 1990), and determine the effects of

population density on the emigration rates of aquatic larvae (Wiley, 1981 ).

Proper simulation in closed systems requires an appropriate pump

mechanism. For example, air-powered (Mackay, 1981) and propeller­

powered designs (Hartman, 1965; Vogel and LaBarbera, 1978) have been

used; however, spatially uniform currents were difficult to obtain.

Techniques can be used to minimize heterogeneous currents and

turbulence. Lauff and Cummins (1964) developed an excellent design that

produced uniform currents; however, velocities were apparently limited

(see Cummins and Lauff, 1969). Thus, design problems are two-fold:

either spatial uniformity is not achieved, or the potential for a realistic

range of current velocities is compromised by the design.

In contrast, a water-fall system provides a swift, homogeneous

current in which water travels along an upper tank, falls into a reservoir,

and is recycled into the upper tank. There has been reluctance to use

water-fall systems because the pumps required for such designs generally

heat the water and produce appreciable turbulence (Vogel and LaBarbera,

1978). The water-heating problem can be minimized by modifying the

design to function as a radiator. This is accomplished by exposing the

6

water within the upper reservoir to a large surface area, in tum, allowing

for the loss of excessive heat. Turbulence can be minimized by the

addition of a weir located near the site of water introduction in the upper

tank.

Materials and Methods

The artificial stream comprises three main components: ( 1) a

rectangular plywood trough (upper reservoir) supported by four wooden

posts, (2) a cylindrical galvanized tin tank (lower reservoir), and (3) a

submersible sewage pump with plumbing. Water is lifted from the lower

reservoir to the upstream end of the upper reservoir, where it then empties

and travels through the upper reservoir. The water then falls into the

lower reservoir and is recycled. Current velocities are manipulated with a

gate valve and stream depths with a lower weir.

Upper reservoir

The upper reservoir (243.8 x 60.1 x 30.5 em) was constructed from

1.9 em (3/4 in.) plywood (Figure 2.1). Galvanized nails and wood glue

were used to secure the sides of the upper reservoir to its base. Silicon

putty prevented leakage from joints and nail holes. The entire upper

resetvoir was coated with a potable epoxy. Fiberglass may be substituted

for the epoxy. If a potable epoxy is used, the upper reservoir might

require a grade-4 plastic lining to prevent possible leakage. The upper

reservoir stream path was divided into eight quadrats (30.5 x 30.5 em)

with 2.5 em (1 in.) comer molding used as a partition (Figure 2.2). Two

7

00

-G

V

1 r

I I

I I

""""

~

I I -

UP

PE

R

RE

S.

I I

LO

WE

R I

I

RE

S.

I~-,

~~-:

Fig

ure

2.1.

--S

ide

view

dep

icti

ng o

vera

ll fo

rmat

of

the

arti

fici

al s

trea

m.

Wat

er is

cyc

led

thro

ugh

the

pum

p w

ithi

n th

e lo

wer

res

ervo

ir i

nto

the

uppe

r re

serv

oir.

\0

WF n

~ Q

UA

DR

AT

Sl

t s t s

WF

n II

II

" II " " \\ \\

Fig

ure

2.2.

--P

lan

view

sho

win

g co

mpo

nent

s (q

uadr

ats,

wei

r, a

nd s

cree

n fr

ames

) w

ithi

n th

e up

per

rese

rvoi

r.

Abb

revi

atio

ns a

re:

GV

= g

ate

valv

e, W

F =

wei

r fr

ame,

S =

scre

en,

P =

pu

mp

.

61.0 x 30.5 em fiberglass screens (1 mm2 mesh) were used to retain

organisms within the quadrats, collect debris, and prevent experimental

materials from entering the lower reservoir (Figure 2.2). Screens were

located 30.5 em from each end of the upper reservoir, and were attached

vertically to frames constructed from 2.5 em (1 in.) corner molding

(Figure 2.2) to provide stabilization.

Water level manipulations and the production of homogeneous

currents were accomplished by the addition of two weirs. The upstream

weir was located 29 em from the farthest upstream position, and the second

was located immediately downstream from the downstream screen frame.

The weir at the upper end (61.0 x 1.9 x 10.2 em) reduced turbulence

caused by the introduction of water at the upstream position; whereas, the

second weir (61.0 x 1.9 x 5.1 em) controlled water depth. Variation in

current depth was controlled by using lower weirs of different heights.

Weirs were stabilized by two wooden frames (60.1 x 5.1 x 2.5 em)

positioned 2.5 em apart. Wooden posts (Figure 2.3) (10.2 x 10.2 x 75.0

em) attached beneath the upper reservoir provided support. The slope of

the stream path could be manipulated by changing the heights of these

supports.

Lower reservoir

The lower reservoir consisted of a galvanized tin tank ( 119.4 em

diam, 61.0 em height) with a volume capacity of 2.73 m3. This reservoir

was the ideal size for the pump described herein; however, more powerful

pumps may be used to increase the current velocity, and in such cases the

10

- -

fi

I

Fig

ure

2.3.

--F

ront

al a

nd r

ear

view

s o

f up

per

rese

rvoi

r sh

owin

g w

eir

used

to

rais

e th

e w

ater

lev

el a

nd

scre

en f

or r

etai

ning

org

anis

ms

in t

he u

pper

res

ervo

ir.

The

site

of

wat

er i

ntro

duct

ion

into

the

upp

er

rese

rvoi

r is

clo

sed

at t

he u

pstr

eam

pos

itio

n, t

here

by f

orci

ng w

ater

to t

rave

l ov

er th

e ex

peri

men

tal

area

to

war

d th

e lo

wer

res

ervo

ir w

here

the

wat

er is

rec

ycle

d.

lower reseiVoir size should change correspondingly.

A 0.5 Hp submersible sewage pump (Goulds, model, 3887, Seneca

Falls, N.Y.) was positioned within the lower reseiVoir (Figure 2.1). A 5.1

em diam (2 in.) PVC pipe, including two elbow fittings (Figure 2.1),

carried water from the pump to the upstream position of the upper

reseiVoir. A brass gate valve was inserted along the horizontal length of

the pipe, and regulated current velocity. All pipe joints were sealed with

PVC pipe cement.

Performance

To obtain accurate current velocities, the lower reseiVoir was placed

behind the upper reseiVoir (opposite the normal position), filled with water,

and depth recorded. Water was then pumped into the upper reseiVoir until

the water level in the lower reseiVoir fell immediately below the intake

level of the pump. Depth was measured (while the pump was running) in

the upper reseiVoir at the position in the stream path for which a known

current velocity was required. The current velocity was then calculated as:

V = m2~hfta

where V =current velocity, r =radius of the lower reseiVoir, ~=change

in the water depth of the lower reseiVoir, t = time the pump was on, and a

= cross-sectional area of the current. After six trials this system produced

an average maximum velocity (+ SD) of 45.45 + 2.73 em/sec . With a 5.1

em lower weir, an average maximum (+ SD) of 13.7 + 0.5 em/sec current

velocity was obtained. This velocity could then be further reduced to 0.0

em/sec without harm to the pump by closing the gate valve.

12

Discussion

This system is inexpensive ($460.00 in 1990) and can be constructed

by one person in two days. The system can be used indoors or outdoors,

but indoor usage minimizes temperature fluctuations. Water temperature

should be monitored with a maximum minimum thermometer. The

average water temperature after 32 hrs of continuous use was 29.1 °C

(constant room temperature of 25°C). The 4.1 oc temperature difference is

attributable to electrical resistance within the pump. This temperature may

not be suitable for some research; however, cooling devices are available

and can be placed within the lower reservoir. Minimal temperature

elevation is obtained by maintaining the lower reservoir at maximum

capacity.

For studies in toxicology or those in which sensitive stream

organisms are used, it is important that the upper reservoir be coated with

a non-toxic (e.g., fiberglass) substance to prevent intoxication of the biota.

In addition, experiments should be designed that simulate depths and

current velocities that correspond to the natural conditions encountered by

the study organism(s). Because current velocity is inversely proportional

to stream depth, relatively high lower weirs decrease the potential for swift

currents (Figure 2.4 ).

13

-u 30 ~ VJ -~ u ->- 20 E-..... u 0 .J ~ 10 >

0+---~~---r--~~~~---r--~~--~

0 1 2 3 4 s 6 7 8 9 10

WEIR HEIGHT (CM)

Figure 2.4.--The relationship between the lower weir height and current velocity. Relatively deep streams compromise swift currents.

14

Literature Cited

Cummins, K. W. 1962. An evaluation of some techniques for the collection and analysis of benthic samples with special emphasis on lotic samples. Am. Midi. Nat. 67:477-504.

Cummins, K. W. and G. H. Lauff. 1969. The influence of substrate particle size on the microdistribution of stream macrobenthos. Hydrobiologia 34:145-181.

Fuller, R. L. and P. S. Rand. 1990. Influence of substrate type on vulnerability of prey to predacious aquatic insects. J. N. Am. Benthol. Soc. 9:1-8.

Gee, J. H. and V. G. Bartnik. 1969. Simple stream tank simulating a rapids environment. J. Fish. Res. Bd. Canada 26:2227-2230.

Hartman, G. F. 1965. An aquarium with simulated stream flow. Am. Fish. Soc. Trans. 94:274-276.

Lauff, G. H. and K. W. Cummins. 1964. A model stream for studies in lotic ecology. Ecology 45: 188-191.

Mackay, R. J. 1981. A miniature laboratory stream powered by air bubbles. Hydrobiologia 83:383-385.

Mason, W. T. and P. A. Lewis. 1970. Rearing devices for stream insect larvae. Prog. Fish-Cult. 32:61-62.

Sudia, W. D. 1951. A device for rearing animals requiring a flowing water environment. Ohio J. Sci. 51:197-202.

Vogel, S. and M. LaBarbera. 1978. Simple flow tanks for research and teaching. Bioscience 28:638-643.

Wiley, M. J. 1981. Interacting influences of density and preference on the emigration rates of some lotic chironomid larvae (Diptera: Chironomidae). Ecology 62:426-438.

15

CHAPTER ill

INFLUENCE OF CURRENT VELOCITY ON SUBSTRATUM

SELECITON BY NAUCORIDAE (HEMIPTERA): AN

EXPERIMENTAL APPROACH VIA

STREAM SIMULATION

Abstract

Microhabitat preferences of three sympatric naucorids

(Ambrysus circumcinctus Montandon, Limnocoris lutzi La Rivers,

and Cryphocricos hungerfordi Usinger) from the South Llano River

of central Texas were evaluated in laboratory experiments. Each

species was individually subjected to all paired contrasts produced

from three substratum size classes (coarse gravel, small pebble, and

small cobble). Within each contrast, three trials were conducted,

each involving a different current velocity (slow, 6.6 + 0.1 em/sec;

intermediate, 9.8 + 0.1 em/sec; and fast, 13.7 + 0.5 em/sec). The

gravel substratum was never preferred; moreover, preference in the

pebble versus cobble contrast, when present, was only shown for the

cobble substratum. Overall, A. circumcinctus exhibited consistent

substratum preferences (cobble > pebble > gravel) independent of

current velocity. Cryphocricos hungerfordi exhibited selectivity in

which cobble was preferred over either gravel or pebble substrata;

however, a synergistic effect between current velocity and

substratum size was detected in the gravel versus pebble contrast.

Finally, b lutzi exhibited the highest degree of velocity-dependent

substratum selection. The laboratory analyses corroborate field

investigations by Sites and Willig (1991) and offer insight into the

16

influence of interacting abiotic variables on the microdistribution of

Naucoridae.

Introduction

The theory of erosion-deposition (Moon, 1939) suggests that

both the substratum and current are directly correlated within the

physical stream environment. Therefore, it is difficult to assess the

individual effects of either current or substratum on a stream fauna

under natural conditions (Minshall, 1984). Nonetheless, both are

important abiotic factors that influence the microdistribution of

stream invertebrates in a variety of habitats (Cummins and Lauff,

1969; Mackay and Kalff, 1969; Harman, 1972; Allan, 1975; de

March, 197 6; Minshall and Minshall, 1977; Lamberti and Resh,

1979; Williams, 1980; Erman and Erman, 1984; Statzner et al.,

1988). Because of the close relationship between substratum type

and current velocity, any interpretation of the effect of one of these

abiotic factors which excludes the other would be unrealistic (Reice,

1980). Substratum characteristics and flow attributes should be

simultaneously controlled to assess the relative role of each in

affecting site selection and population density. Works by Cummins

and Lauff (1969), Wiley (1981), and Fuller and Rand (1990)

produced lotic environments within laboratory settings to observe

the manner in which abiotic variables influence stream biota. These

"laboratory streams" permit the control and manipulation of

important abiotic and biotic parameters while simultaneously

reducing confounding effects. Such simulation is ideal for studies of

habitat selection.

17

Creeping water bugs (Hemiptera: Naucoridae) are predacious

aquatic insects common in lentic and lotic habitats of the

southwestern U.S. and tropical regions of the world. Although these

insects are considered keystone consumers in many aquatic systems

(Sites and Willig, 1991 ), little is known of their biologies

(Gonsoulin, 1973; Constantz, 1974; Venkatesan and Cloarec; 1988).

Most literature concerning naucorids addresses taxonomic (e.g.,

Usinger, 1947; La Rivers, 1971; 1974; 1976; Polhemus and

Polhemus 1988) or biogeographic questions (e.g., La Rivers, 1951;

Davis, 1986). The U.S. contains five naucorid genera (Ambrysus,

Cryphocricos, Limnocoris, Pelocoris, and Usingerina) which

comprise 21 species. The South Llano River in central Texas

harbors eight species (Sites and Willig, 1991 ), the most species rich

assemblage known to occur in temperate regions.

Ambrysus is the most speciose naucorid genus within the U.S.

Ambrysus circumcinctus Montandon is the most abundant naucorid

in the South Llano River where it easily can be found under large

rocks in swift waters (Sites and Willig, 1991). Cryphocricos is

represented in the U.S. by only ,C. hungerfordi Usinger (Polhemus

and Polhemus, 1988); the South Llano River demarks the

northeastern-most known boundary (Sites, 1990). Cuticular

evidence suggests plastral respiration within this genus (Parsons and

Hewson, 197 4 ). This species is morphologically hydrodynamic, and

exploits swift, deep, benthic habitats which contain large stones

(Parsons and Hewson, 1974; Sites and Willig, 1991). In the U.S.,

Limnocoris is represented by only one species, L. lutzi La Rivers.

This species is considered a substratum generalist (Sites and Willig,

18

1991) and occurs in riffle habitats. Although little is known of the

ecology of this species, its abundance is correlated with stream depth

(Sites and Willig, 1991). Limnocoris lutzi is the smallest (adult size)

naucorid inhabiting the South Llano River. The basic microhabitat

associations of each of these three species were described by Sites

and Willig (1991) who stressed the importance of substratum size

and current velocity; however, they could not uncouple the two. The

purpose of this study was to evaluate the influence of current

velocity on substratum selectivity for each of these three naucorid

species.

Materials and Methods

An artificial stream (Figure 2.1 ), described in detail elsewhere

(Herrmann et al., 1992), was used to evaluate substratum preferences

of the three most common riffle-dwelling naucorids (A.

circumcinctus, !:. hungerfordi, and L. lutzi) of the South Llano

River (Sites and Willig, 1991). Each species (n=18) was offered all

paired contrasts of three substratum size classes: (1) coarse gravel

(8-16 mm rocks), (2) small pebbles (16-32 mm rocks), and (3) small

cobble (64-128 mm rocks) [Wentworth (1922) classification scheme

after Minshall (1984)]. Each of these contrasts(= experiments) was

replicated under three current velocities (mean+ S.D.): slow (6.6 +

0.1 em/sec), intermediate (9.8 + 0.1 em/sec), and fast (13.7 + 0.5

em/sec), which each represent different trials. In all cases, water

depth was held constant with a 5.1 em lower weir. Four identically

designed streams were used, and each trial was conducted in a

randomly assigned stream.

19

Each stream contained eight quadrats which were dusted with

a fme layer of washed sand (ca. 1 mm deep) before each trial. The

left upstream-most quadrat was packed loosely with the larger rock

class and the smaller rock class placed in the right upstream-most

quadrat. Thereafter, the two substratum size classes alternated in a

checkerboard pattern. This arrangement reduced the potential of

biased selection related to complex hydraulic features (see Statzner et

al., 1988).

Water from the South Llano River was used in all

experiments, changed biweekly, and replenished periodically to

compensate for evaporation. Before each trial, naucorids were

collected from a riffle in the South Llano River and placed in

holding containers for no more than 1 hr before use. Stream flow

was initiated at least 30 min before each trial to stabilize water

temperature. Eighteen individuals of a given species were released

into each stream and left undisturbed for 32 hrs, after which the

pumps were turned off and the lower weir removed. This allowed

water drainage and minimized naucorid movement. All insects were

then removed and their locations recorded. On average, 4.1% of all

individuals were found on the upper or lower retaining screens, but

these individuals were not included in any subsequent analyses (Lauff

and Cummins, 1964).

A replicated goodness-of-fit test based upon G-Statistics (Sokal

and Rohlf, 1981) was used to evaluate the hypothesis that each

species would occupy the two presented substrata equally in each

trial. For each paired contrast of substratum size, the three different

velocities constituted the replicates. A significant pooled effect in the

20

absence of significant heterogeneity indicated consistent selection of

one substratum size, regardless of current velocity. Significant

heterogeneity indicated substratum selection that was velocity

dependent. Examination of the significance of selection at each

velocity revealed whether the interaction involved only differences

in the magnitude of selection or changes in the direction of selection

(e.g., preference for the large substratum at the slow velocity, no

preference at the intermediate velocity, and preference for the

smaller substratum at the fast velocity). The absence of both

significant heterogeneity and a pooled effect indicated no substratum

preference at any current velocity.

Results

Ambrysus circumcinctus

The larger substratum was consistently preferred, regardless

of current velocity in all paired contrasts (Table 3.1) involving A.

circumcinctus. In particular, 86.8% of individuals preferred small

pebble to coarse gravel; 77.4% preferred small cobble to coarse

gravel and 70.4% of individuals preferred small cobble over small

pebble. Current velocity did not affect the magnitude or direction of

substratum selection in any case.

Cryphocricos hungerfordi

Substratum selection was complex for ~- hungerfordi. In the

gravel versus pebble contrast, significant heterogeneity occurred in

the presence of a significant pooled effect (Table 3.1A); this

21

N

N

Tab

le 3

.1.-

-Sep

arat

e re

plic

ated

goo

dnes

s-of

-fit

test

s at

eac

h o

f thr

ee c

urre

nt v

eloc

itie

s fo

r th

e th

ree

poss

ible

pa

irs

of s

ubst

ratu

m s

ize

(den

oted

by

A, B

, and

C)

offe

red

to A

mbr

.ysu

s ci

rcum

cinc

tus,

Czy

phoc

rico

s hu

nger

ford

i, a

nd L

imno

cori

s lu

tzi.

Per

cent

of

indi

vidu

als

that

sel

ecte

d th

e la

rger

sub

stra

tum

is i

ndic

ated

(%

) · f

or e

ach

spec

ies

A.

circ

umci

nctu

s

Tes

ts

df

oa

o/o

A. C

oars

e G

rave

l vs.

Sm

all P

ebbl

e

Poo

led

1 32

.10*

**

----

Het

erog

enei

ty

2 4.

77

----

Tot

al

3 36

.87*

**

86.8

Slo

w v

eloc

ity

1 15

.96*

**

94.1

Inte

rmed

iate

vel

ocit

y 1

17.2

3***

94

.4

Fas

t vel

ocit

y 1

3.68

72

.2

C. h

unge

rfor

di

oa

%

6.83

**

----

23.3

1 **

* ---

-30

.14

***

68.6

3.48

27

.8

20.3

8***

10

0.0

6.28

* 81

.3

L.

lutz

i

oa

17.1

4***

15.5

3***

32.6

6***

20.3

8***

0.06

12.2

2***

%

79.6

100.

0

47.1

93.3

Tab

le 3

.1

cont

inue

d

A.

circ

umci

nctu

s

Tes

ts

df

oa

%

B.

Coa

rse

Gra

vel v

s. S

mal

l Cob

ble

Poo

led

1 16

.77*

**

----

Het

erog

enei

ty

2 2.

40

----

N

Tot

al

3 19

.15*

**

77.4

U

l

Slo

w v

eloc

ity

1 5.

88*

77.8

Inte

rmed

iate

vel

ocit

y 1

2.04

66

.7

Fas

t vel

ocit

y 1

11.2

5***

. 88

.2

C. h

unge

rfor

di

oa

%

29.9

0***

---

-4.

99

----

34.9

0***

86

.3

17.2

3***

94

.4

14.7

0***

93

.8

2.97

70

.6

L.l

utzi

oa

1.05

0.28

1.33

0.25

1.01

0.07

%

57.4

53.6

62.5

53.3

Tab

le 3

.1

cont

inue

d

A. c

ircu

mci

nctu

s C

. hun

gerf

ordi

Tes

ts

df

a a

%

a a

%

C.

Sm

all P

ebbl

e vs

. Sm

all C

obbl

e

Poo

led

1 9.

23**

---

-11

.05*

**

----

Het

erog

enei

ty

2 5.

39

----

3.79

---

-N

T

otal

3

14.6

2**

70.4

14

.84

**

72.2

~

Slo

w v

eloc

ity

1 8.

73**

83

.3

0.22

55

.6

Inte

rmed

iate

vel

ocit

y 1

0.00

50

.0

5.88

* 77

.8

Fas

t vel

ocit

y 1

5.88

* 77

.8

8.73

**

83.3

asig

nifi

canc

e o

f G:

, P >

0.0

5; *

, 0.0

5 >

P >

0.0

1; *

*, 0

.01

> P

> 0

.001

; **

*, P

< 0

.001

L.l

utzi

a a

2.01

7.32

*

9.33

*

8.73

**

0.07

0.53

%

60.0

83.3

53.3

41.2

suggested that substratum selection depended upon current velocity.

Moreover, the direction of preference differed depending on current

velocity; in particular, the smaller substratum (coarse gravel) was

preferred at the slow velocity; whereas, the larger substratum (small

pebble) was preferred at the intermediate and fast velocities. In the

gravel versus cobble contrast {Table 3.1B), ~. hungerfordi exhibited

a consistent preference for the larger substratum (86.3% of all

individuals occurred in the cobble) over the smaller (gravel).

Finally, C. hungerfordi selected cobble over pebble (72.2% of

individuals) regardless of current velocity (Table 3.1C). Current

affected the direction of substratum selection for C. hungerfordi, but

only in the gravel versus pebble contrast. In the other contrasts,

neither the direction nor magnitude of substratum selection was

affected by current: preference in these latter two experiments was

consistently in the direction of the larger available substratum.

Limnocoris lutzi

Substratum selection varied with current velocity for L. lutzi

in the gravel versus pebble contrast {Table 3.1A). The larger

substratum (small pebble) was preferred at the slow and fast

velocities; whereas, no significant selection was apparent at the

intermediate current velocity. The significant pooled effect was

attributable to strong selection for the pebble substratum only during

two (slow and fast) of the three current trials, (>90%) which

overwhelmed the even distribution of individuals at the intermediate

velocity. The absence of significance for any test in the gravel

versus cobble contrast (Table 3.1B) indicated no substratum selection

25

at any velocity. Current velocity influenced the direction of

substratum selection in the pebble and cobble contrast (Table 3.1 C).

No selectivity could be detected at fast or intermediate velocities;

however, in the slow velocity, small cobble was preferred over small

pebble.

Discussion

Laboratory experiments concerning lotic organisms can

clarify and complement research conducted in field settings

(Cummins, 1962). Sites and Willig (1991) presented quantitative

analyses of microhabitat associations of each of these three naucorid

species; however, they did not uncouple the effects of the substratum

and velocity on microhabitat associations. Because our study

regulates both parameters independently, it refines our

understanding of specific substratum associations as potentially

modified by current speed. These experiments addressed the manner

in which current velocity could alter substratum selection. Velocity

preferences were not evaluated per se.

In each experiment involving A. circumcinctus, selection was

in the direction of the larger available rock class. This resulted in a

hierarchical arrangement of substratum selectivity (cobble > pebble

> gravel), which was unaltered by current velocity. In comparison,

Sites and Willig (1991) found that mean rock size as well as current

speed were significant variables influencing the microdistribution of

A. circumcinctus (mean rock size and mean current velocity in the

quadrats containing this species were found to be 184.7 mm and 37.8

em/sec, respectively). Rock size and current velocity in their field

26

situations exceeded those of our largest rock category (small cobble,

upper range of 128 mm) and the fastest current (13.7 em/sec).

Apparently current speeds up to 13.7 em/sec do not alter substratum

selectivity.

Cryphocricos hungerfordi often inhabits swift, deep waters in

which frequent surface visits to replenish a compressible air store

would cause downstream displacement. It has been shown that adult

C. hungerfordi and C. barozzii Signoret have cuticular modifications

necessary for plastral respiration (Parsons and Hewson, 197 4 ),

thereby eliminating the need for frequent surfacing and enabling

them to exploit habitats such as large rocks in swift currents. These

habitats are generally unavailable to species that utilize compressible

air bubbles. As a consequence, C. hungerfordi may be associated

with large rock sizes and swifter currents because it experiences

ecological release from interspecific competition with other

naucorids. A significant pooled effect existed in all three

experiments involving C. hungerfordi; however, a current­

substratum interaction was observed between pebble and gravel in

which the direction of selection was affected by current. Strong

association with the larger of the available substrata occurs at fast

(81.3%) and intermediate (100%) velocities; however, this

relationship is vitiated by slow velocities at which no substratum

preference was shown. Similarly, Sites and Willig (1991) found that

both rock size and current velocity were important variables in

determining the microdistributions of .C.. hungerfordi in field

situations. Mean rock size in the quadrats containing this species was

189.0 mm, with an average current velocity of 37.2 em/sec. This

27

rock size and current velocity exceeded those tested herein. In our

experiments, preference only for the small cobble could be

delineated over the other substrata. At faster current velocities,

substratum-current interactions may become more prevalent and

might confound field interpretations.

Although L. lutzi has been viewed as a substratum generalist

with sensitivity to current velocity (Sites and Willig, 1991), it did

exhibit some preferential substratum selection in our study. In fact,

current velocity influenced substratum selection in the gravel versus

pebble contrast and in the pebble versus cobble contrast. Clearly,

current velocity plays an important role in modifying substratum

selection for this species, a view which did not emerge from a field

study in which their separate effects could not be evaluated.

Field observations suggest that niche partitioning is based

primarily on flow characteristics which separate the three naucorid

species tested here from the other five that occur in the South Llano

River (see Sites and Willig, 1991). The former three naucorid

species occur in association with rapidly flowing water, whereas the

other five species generally occur in shoreline and backwater

situations, frequently among algae and aquatic macrophytes.

Shoreline algae often grow among moderate to large-sized alluvium

as well as in silt, consequently substratum size exploited by those five

species is variable.

The fmdings in this study strongly paralleled those of Sites and

Willig (1991); however, important differences were observed. For

example, substratum size was relatively unimportant for L. hUti in

field situations; however, laboratory analyses suggested that at least

28

some consistent substratum bias existed. Similarly, field analyses

suggested an important role of current velocity on the

microdistribution of C. hun~erfordi. This was expected to occur

considering the respiratory adaptations of this species for swift

waters. Contrary to expectation, current velocity only minimally

influenced substratum selectivity in the laboratory. Ironically, this

occurred at rock sizes (gravel and pebble), which do not occur in

common habitat for C. hun~erfordi. In reality, a tradeoff may exist

between substratum size and current velocity. At small rocks sizes

the importance of substratum size is dependent upon current.

Conversely, at larger rock sizes, current does not alter the

preference for larger substrata.

Historically, stream ecologists have suggested that substratum

characteristics are primarily responsible for determining

microspatial arrangements of stream invertebrates [see Minshall

(1984) and citations therein]. However, more recent work (e.g.,

Statzner et al., 1988) has redirected attention to flow attributes as

primary factors affecting lotic community structure. Strong

evidence herein suggests that substratum selection by macrobenthic

predators varies interspecifically with respect to rock size and can be

altered by current velocity. The inability to uncouple the effects of

current velocity on substratum selection in natural environments,

where the two may be correlated, could give rise to erroneous

conclusions concerning the relative importance of hydraulic or

substratum attributes. Even in situations when current velocity may

not appear to directly affect the presence or absence of species, it

may have an indirect effect by modifying substratum selection.

29

Thus, the uncoupling of abiotic factors when determining

microhabitat selection of benthic organisms is imperative.

30

Literature Cited

Allan, D. J. 1975. The distributional ecology and diversity of benthic insects in Cement Creek, Colorado. Ecology 56:1040-1053.

Constantz, G. D. 1974. The mating behavior of a creeping water bug, Ambrysus occidentalis (Hemiptera: Naucoridae ). Am. Midi. Nat. 92:234-239.

Cummins, K. W. 1962. An evaluation of some techniques for the collection and analysis of benthic samples with special emphasis on lotic waters. Am. Midi. Nat. 67:477-504.

Cummins, K. W. and G. H. Lauff. 1969. The influence of substrate particle size on the microdistribution of stream macrobenthos. Hydrobiologia 34:145-181.

Davis, J. R. 1986. New distributional records of N aucoridae (Hemiptera) from the United States and Mexico, with morphological and taxonomic notes. Southwest. Nat. 31:535-539.

de March, B.G.E. 1976. Spatial and temporal patterns in macrobenthic stream diversity. J. Fish. Res. Board Can. 33:1261-1270.

Erman, D.C. and N.A. Erman. 1984. The response of stream macroinvertebrates to substrate size and heterogeneity. Hydrobiologia 108:75-82.

Fuller, R. L. and P. S. Rand. 1990. Influence of substrate type on vulnerability of prey to predacious aquatic insects. J. N. Am. Benthol. Soc. 9:1-8.

Gonsoulin, G. J. 1973. Seven families of aquatic and semiaquatic Hemiptera in Louisiana. Entomol. News 84:83-88.

Harman, W. N. 1972. Benthic substrates: Their effect on fresh­water Mollusca. Ecology 53:271-277.

31

Herrmann, D.P., R. W. Sites and M. R. Willig. 1992. A laboratory flow tank with variable current and depths for replicating riffles and shallow streams. Texas J. Sci. 44:89-94.

Lamberti, G.A. and V. H. Resh. 1979. Substrate relationships, spatial distribution patterns, and sampling variability in a stream caddisfly population. Environ. Entomol. 8:561-567.

La Rivers, I. 1951. A revision of the Genus Ambrysus in the United States (Hemiptera: Naucoridae ). Univ. Calif. Pub I. Entomol. 8:277-338.

La Rivers, I. 1971. Studies of Naucoridae (Hemiptera). Memoir 2, Bioi. Soc. Nevada, Verdi.

La Rivers, I. 1974. Catalogue of taxa described in the family Naucoridae (Hemiptera) supplement no. 1: Corrections, emendations and additions, with descriptions of new species. Occas. Pap. 38, Bioi. Soc. Nevada, Verdi.

La Rivers, I. 1976. Supplement no. 2 to the catalogue described in the family Naucoridae (Hemiptera), with descriptions of new species. Occas. Pap. 41, Bioi. Soc. Nevada, Verdi.

Lauff, G. H. and K. W. Cummins. 1964. A model stream for studies in lotic ecology. Ecology 45:188-191.

Mackay, R. J. and J. Kalff. 1969. Seasonal variation in standing crop and species diversity of insect communities in a small Quebec stream. Ecology 50:101-109.

Minshall, G. W. 1984. Aquatic insect-substratum relationships, pp. 358-400. In V. H. Resh and D. M. Rosenberg (eds.). The ecology of aquatic insects. Praeger, New York.

Minshall, G. W. and J. N. Minshall. 1977. Microdistribution of benthic invertebrates in a Rocky Mountain (U.S.A.) stream. Hydrobiologia 55:231-249.

Moon, H. P. 1939. Aspects of the ecology of aquatic insects. Trans. Brit. Entomol. Soc. 6:39-49.

32

Parsons, M. C. and R. J. Hewson. 1974. Plastral respiratory devices in adult Cryphocricos (Naucoridae: Heteroptera). Psyche 81:510-527.

Polhemus, D. A. and J. T. Polhemus. 1988. Family Naucoridae Leach 1815: the creeping water bugs, pp. 521-527. InT. J. Henry and R. C. Froeschner (eds.). Catalogue of the Heteroptera, or true bugs, of Canada and the continental United States. Brill, Leiden.

Reice, S. R. 1980. The role of substratum in benthic macroinvertebrate microdistribution and litter decomposition in a woodland stream. Ecology 61:580-590.

Sites, R. W. 1990. Morphological variations in the hemelytra of Cryphocricos hungerfordi Usinger (Heteroptera: Naucoridae). Proc. Entomol. Soc. Wash. 92:111-114.

Sites, R. W. and M. R. Willig. 1991. Microhabitat associations of three sympatric species of Naucoridae (Insecta: Hemiptera). Environ. Entomol. 20:127-134.

Sokal, R. R. and F. J. Rohlf. 1981. Biometry: the principles and practice of statistics in biological research, 2nd ed. Freeman, San Francisco.

Statzner, B., J. A. Gore and V. H. Resh. 1988. Hydraulic stream ecology: observed patterns and potential applications. J. N. Am. Benthol. Soc. 7:307-360.

Usinger, R. L. 1947. Classification of the Cryphocricinae (Hemiptera: Naucoridae ). Ann. Entomol. Soc. Am. 40:329-343.

Venkatesan, P. and A. Cloarec. 1988. Density dependent prey selection in Dyocoris (N aucoridae ). Aquatic Insects 10:105-116.

33

Wentworth, C. K. 1922. A scale of grade and class terms for cladistic sediments. J. Geol. 30:377-392.

Wiley, M. J. 1981. Interacting influences of density and preference of the emigration rates of some lotic chironomid larvae (Diptera: Chironomidae ). Ecology 62:426-438.

Williams, D. D. 1980. Some relationships between stream benthos and substrate heterogeneity. Limnol. Oceanogr. 25: 166-172.

34

CHAPTER IV

DENSITY-DEPENDENT HABITAT SELECTION

AND COMPETITION AMONG THREE

SYMPATRIC NAUCORIDS

Abstract

The role of competition for space in structuring stream communities

remains unclear. Much of the evidence for competition in streams has been

derived from work concerning sessile or slow moving benthos. Relatively

few studies have focused on mobile predators which may compete in more

complex manners than their sessile counterparts. As such, three naucorid

species (Ambrysus circumcinctus Montandon, Cryphocricos hungerfordi

U singer, and Limnocoris lutzi La Rivers) were subjected to laboratory

experiments to determine the influence of intra- and interspecific

competitive interactions on habitat selection. Intraspecific experiments

evaluated the effects of density per se; whereas, interspecific experiments

evaluated the effects of competition between naucorids beyond those

associated with an increase in density. The results suggest that naucorids

respond to both density and interspecific competitors in terms of habitat

selection; however, these relationships often varied among substratum

comparisons, with current velocity and interspecifically. This variation

illustrates the complex nature of interacting variables in stream communities

and offers evidence that competition is one mediator of invertebrate habitat

associations.

Introduction

Much of contemporary ecology focuses on the factors which facilitate

the coexistence of ecologically similar species. A popular view holds that

35

disturbance and predation can prevent the over-exploitation of space by

suppressing the density of competitive dominants (Connell, 1978; Sousa,

1984 ), which in tum, frees resources for less dominant species (McAuliffe,

1984 ). Alternatively, competition may mediate species distributions and

alter an organism's perception of habitat suitability (Schoener, 1983;

Hemphill, 1991 ). However, the relevance of competition in nature has been

questioned (Connell, 1975; 1978; 1980; Pulliam, 1983) and viewed as an

mtermittent phenomenon. Nevertheless, the mechanistic roles of

competition in affecting the abundance and distribution of species warrant

continued exploration.

Within stream ecosystems, the potential contributions of the

substratum and current to community structure have been intensively

investigated (Minshall, 1984; Statzner et al., 1988; Ward, 1992).

Consequently, both are considered to be primary determinants of the

abundance and distribution of stream invertebrates. Macrobenthic

populations frequently exist at high densities and extensive overlap in

resource utilization occurs between potential competitors (Dudley et al.,

1990). Thus, intra- and interspecific competition for these resources as well

as space in general may contribute to the overall structure and function of

stream communities (Harding and Colbo, 1981; McAuliffe, 1984; Dudley et

al., 1986; Hemphill, 1988; 1991). In theory, the effects of these

contributions on community organization may differ fundamentally. An

increase in a population's density may be conducive to generalistic resource

utilization, whereas, the presence of interspecific competitors may force a

generalistic species to specialize on a limited range of resources.

Comparisons of intra- and interspecific competition may offer insight into

the coexistence and abundance of competitors. For example, if a

36

hypothetical species A is a superior competitor of species B, one might

conclude that A will eventually eliminate B, unless the process is interrupted

by other factors (Connell, 1983). If, however, species A is subjected to

intraspecific pressures greater than the interspecific pressures imposed by

species B, species A may be self-limited to a density which can not eliminate

species B (Connell, 1983).

Intraspecific competition for space may be considered to be density­

dependent. Contemporary habitat selection theory based on Fretwell and

Lucas (1970) predicts that individuals select habitats based on fitness. As

the density of a species increases within a habitat, individual fitness

generally decreases and individuals may select what was initially suboptimal

habitat as a consequence of cost-benefit analyses. Evidence for intraspecific

competition in stream communities is variable. Caddisflies (Glass and

Bovbjerg, 1969) and black flies (Zahar, 1951) are thought to alter their

microdistributions through territorial spacing, whereas, chironomids respond

to elevated density via alteration of emigration rates (McLachlan, 1969).

Wiley (1981) established an interaction between sediment type and

chironomid density, even though emigration rates were unpredictable based

on either factor alone.

Several groups of sessile and slow moving benthos contribute the

majority of information on interspecific competition in stream systems.

Hemphill (1988) established a competitive dominance between Hydropsyche

oslari and Simulium vir~atum on the basis of former's ability to initiate

aggressive encounters and preempt boulder habitat. Competition between

the two species was less prevalent during winter months when floods

lowered the densities of Hydropsyche (Hemphill, 1991). McAuliffe (1983)

established a negative correlation between the colonization rates of Baetis

37

and Simulium and the density of the caddisfly Leucotrichia. Again,

disturbance mediated density reductions of Leucotrichia, thereby preventing

the monopolization of space (McAuliffe, 1984 ).

Many stream benthos that have been shown to compete, do so through

interference mechanisms including overgrowth, crushing, and undercutting

(Hemphill, 1991). Highly mobile species, which do not interact in those

manners have received little attention in terms of competition. Furthermore,

many studies have focused on species for which competition was expected

to occur. For this reason, the prevalence of competition in nature may be

overestimated and represent a strong bias (Connell, 1983). A wider range of

species must be examined to alleviate this concern.

Creeping water bugs (Naucoridae: Hemiptera) are prominent keystone

members of stream ecosystems in the southwestern U.S.; however, little is

known about their basic biology (Gonsoulin, 1973). Most studies of

naucorid ecology have focused on abiotic interactions. Stout (1981)

examined flooding effects associated with two tropical species, Limnocoris

insularis Champion and Czyphocricos latus Usinger. Sites and Willig (1991)

examined abiotic factors that contribute to the microdistributions of riffle­

inhabiting naucorids (see also Chapter III). Relatively few studies have

focused on biotic interactions associated with the group. Of those, two

(V enkatesan and Cloarec, 1988; Camilo and Willig, unpublished data)

examined naucorid prey selection. The former examined density-dependent

prey selection in a European species, llyocoris cimicoides, and the latter

evaluated prey selection by three Texas species. Other ecological efforts

have focused on life histories (Constantz, 1974; McPherson et al., 1987;

Sites and Nichols, 1990); however, none has focused on competition among

naucorids. The South llano River (SLR) of central Texas harbors eight

38

naucorid species (Sites and Willig, 1991), which is the most species rich

assemblage known to occur in the U.S. The SLR lies on Cretaceous

limestone and is characterized by well-defined riffle and pool habitats.

Naucorids can be found within the substratum in these habitats and among

aquatic macrophytes along stream margins. These naucorids feed on a

variety of invertebrates including immature mayflies and caddisflies (Camilo

and Willig, unpublished data). Three species, Ambrysus circumcinctus

Montandon, Limnocoris lutzi La Rivers, and Cryphocricos hungerfordi

Usinger, are common riffle inhabitants which coexist at high densities and

exhibit partial overlap in substratum (Chapter ill), current (Sites and Willig,

1991) and prey (Camilo and Willig, unpublished data) utilization. The

degree to which these insects compete for spatial resources and the

mechanisms which allow them to coexist remains poorly understood. The

following study attempted to evaluate the effects of both density and

interspecific competition on habitat selection by the three naucorid species

and offer potential mechanisms by which they compete for spatial resources.

Materials and Methods

Laboratory Experiments

A method of stream simulation described in detail by Herrmann et al.

(1992) was used in laboratory experiments (see Chapter IT). The design

involved a simple water fall system that could produce current velocities

from 0-45 em/sec. Each stream contained four paired quadrats (30.5 X 30.5

em) in which substratum could be positioned. Each of three naucorid

species (A. circumcinctus, .C. hun~:erfordi and L.lmW was subjected to

every paired contrast of three substratum categories [from the Wentworth

39

(1922) classification scheme after Minshall (1984)]: (1) coarse gravel (CG;

8-16 mm rocks), (2) small pebble (small pebble; 16-32 mm rocks), and (3)

small cobble (SC; 64-128 mm rocks) within the streams. The amount of

substratum in each category was standardized according to volumetric

displacement (CG, 900 ml; SP, 1200 ml, SC, 1500 ml). For each contrast,

three trials were conducted, each at a different current velocity (slow, 6.6

em/sec; intermediate, 9.8 em/sec; and fast, 13.7 em/sec). To reduce the

potential for biased selection related to complex hydraulic features, the two

substratum categories in each replicate were alternated in a checkerboard

pattern. Water used in the experiments was collected from a riffle in the

SLR and current depth was standardized with a 2 em lower weir (see

Chapter II). Each trial was conducted for 32 hrs in one of 6 randomly

assigned streams. At the conclusion of each experiment, the location of each

individual naucorid was determined by visual inspection of the substrata.

Individuals found on the upper and lower retaining screens were discarded

and not included in subsequent analyses.

All naucorids were collected from various riffles in the SLR. Because

of the large numbers of individuals used in experimentation, most were

released after use; however, none was used more than once. To determine

the importance of density-dependent intraspecific interactions on substratum

selection, the aforementioned experiment was conducted at three densities

representing a geometric progression (N=18, N=36, N=54). To determine

the influence of interspecific interactions, experiments were conducted with

all possible pair-wise combinations of species [e.g., A. circumcinctus

(N=18) + k lutzi (N=18) or~- hun&erfordi (N=18) + L. lutzi (N=18)], and

one experiment involving all three species, each at N=18.

40

Statistical Analyses

A series of hierarchical log linear models (SPSS Inc., 1988) was used

to test hypotheses that three- and two-way interactions involving current

velocity, density treatment, and substratum selection were negligible for

each species. If significant three-way interactions occurred, then substratum

selection was density dependent in a manner that varied with current

velocity. Thereafter, the model was decomposed via a series of .a priori

contrasts to evaluate the manner in which competition occurred. If no three­

way interaction was detected, that interaction term was not included in

subsequent models. A series of models then systematically evaluated two­

way interactions. To test for a current velocity-substratum selection

interaction, that two-way interaction term was eliminated from the model.

Similarly, the two-way interaction between density and substratum selection

was tested by evaluating a model with the corresponding two-way

interaction term absent. If both possible two-way interactions were

significant, then substratum selection consistently varied with density

regardless velocity, and the effects of velocity on substratum selection were

consistent at all densities. As a consequence, only one series of .a priori

contrasts would be needed to determine the relationship between density and

substratum selection. If only a current velocity-substratum selection

interaction was significant, current velocity affected substratum selection in

the same way, regardless of density treatment. Thus, a priori contrasts were

not appropriate. If only a density-substratum selection interaction was

significant, it was concluded that density affected the way in which

substratum selection occurred, independent of current velocity.

Subsequently, i priori contrasts were implemented to compare substratum

selection at different densities. If neither three- nor two-way interactions

41

occurred, then substratum selection was considered independent of current

velocity and density regime.

If a density-substratum selection interaction was detected, the two­

way interaction was resolved via non-orthogonal i! priori contrasts involving

Z scores (SPSS Inc., 1988). Two intraspecific comparisons of substratum

selection were conducted between: (1) experiments involving 18 and 36

individuals, and (2) experiments involving 18 and 54 individuals.

Comparisons of substratum selection when interspecific competitors were

present were of two types: those involving two species and those involving

three species. In all cases, density was constant within a particular contrast.

For example, substratum selection by A. circumcinctus when N=36 was

compared to substratum selection when N=18 for A. circumcinctus and

N=18 for ,C. hungerfordi. Similarly, substratum selection by A.

circumcinctus when N=36 was compared to substratum selection by A.

circumcinctus when its density and that of L.lutzi were each 18. In the

three species experiment, substratum selection for a focal species (N=54 ),

was compared to substratum selection by that same species when each of the

three species occurred at equal densities (N=18). For each comparison,

significance implied that doubling or tripling density, or that the presence of

interspecific competitors altered microhabitat associations. For the five D.

priori contrasts, comparison-wise error rate was held at 0.05; and not all

possible contrasts were made in order to reduce experiment-wise error rate.

Intraspecific comparisons evaluated density w g, whereas interspecific

comparisons held density constant but varied species composition.

42

Results

Ambrysus circumcinctus

For all three paired contrasts of substratum sizes, density affected

substratum selection in a consistent fashion, regardless of current velocity

(absence of three-way interaction with a significant density by substratum

interaction; Table 4.1 and Table 4.2). Nonetheless, A. circumcinctus

responded in a complex fashion to the various density treatments. In the

pebble vs. cobble contrast, velocity affected substratum selection in a

consistent fashion regardless of density. More powerful a priori contrasts

revealed the manner in which density affected substratum selection (Table

4.3). Increased density of conspecifics modified substratum selection only

in one of the pebble vs. cobble contrasts (18 vs. 36). The presence of L. lutzi

affected substratum selection by A. circumcinctus in all but the gravel vs.

pebble contrast; conversely, the joint presence of C. hungerfordi and L. lutzi

only affected substratum selection in the gravel vs. pebble contrast.

Cryphocricos hungerfordi

Velocity affected the manner in which density treatment modified

substratum selection in all three paired contrasts of substratum sizes

(presence of three-way interactions in each case; Table 4.1). Nonetheless,

differences in density treatment at least approached significance in

modifying substratum selection in all but the intermediate and fast velocity

experiments involving the gravel vs. cobble contrast (Table 4.2).

Again, patterns of significance related to density treatment are complex

(Table 4.3). Substratum selection was modified by increased numbers of

conspecifics at slow velocities in all but the pebble vs. cobble contrast,

43

t

Tab

le 4

.1.-

-Lik

elih

ood

rati

o ch

i-sq

uare

ana

lysi

s fo

r ea

ch c

ontr

ast o

f sub

stra

tum

siz

e fo

r A

. cir

cum

cinc

tus,

.C..

hung

erfo

rdi,

and

L. l

utzi

. H

a t

hree

-way

inte

ract

ion

(vel

ocit

y x

dens

ity

x se

lect

ion)

was

pre

sent

, sep

arat

e an

alys

is o

f the

den

sity

by

subs

trat

um i

nter

acti

on a

t eac

h cu

rren

t vel

ocit

y w

as r

equi

red.

D

ensi

ty a

nd c

urre

nt

velo

city

are

ind

epen

dent

var

iabl

es, a

nd c

anno

t app

ear

in tw

o-w

ay i

nter

acti

on te

rms

3-W

ay I

nter

acti

on

2-W

ay I

nter

acti

ons

Vel

ocit

y vs

. S

elec

tion

D

ensi

ty v

s. S

elec

tion

Sub

stra

tum

Con

tras

t X

2 p

X2

p X

2 p

A. c

ircu

mci

nctu

s

grav

el v

s. p

ebbl

e 9.

07

0.52

6 9.

10

0.69

4 30

.41

0.02

4*

grav

el v

s. c

obbl

e 11

.83

0.29

7 12

.44

0.41

1 28

.08

0.04

4*

pebb

le v

s. c

obbl

e 15

.78

0.10

6 25

.69

0.01

2**

28.5

6 0.

018*

C. h

une:

erfo

rdi

grav

el v

s. p

ebbl

e 56

.19

< 0

.001

***

grav

el v

s. c

obbl

e 36

.09

< 0

.001

***

pebb

le v

s. c

obbl

e 30

.03

< 0

.001

***

~

VI

Tab

le 4

.1

cont

inue

d

3-W

ay I

nter

acti

on

2-W

ay I

nter

acti

ons

Vel

ocit

y vs

. S

elec

tion

D

ensi

ty v

s. S

elec

tion

Sub

stra

tum

Con

tras

t X

2 p

X2

p X

2 p

L.m

w gr

avel

vs.

peb

ble

34.0

2 <

0.0

01 *

**

grav

el v

s. c

obbl

e 14

.23

0.16

3 28

.31

0.00

5**

29.6

6 0.

013*

pebb

le v

s. c

obbl

e 43

.62

< 0

.001

***

Lev

els

of s

igni

fica

nce:

N

.S, n

ot s

igni

fica

nt;@

, app

roac

hing

sig

nifi

cant

; *,

sign

ific

ant;

**,

high

ly s

igni

fica

nt;

***,

ver

y hi

ghly

sig

nifi

cant

~

0\

Tab

le 4

.2.-

-Lik

elih

ood

rati

o ch

i-sq

uare

ana

lyse

s fo

r th

e tw

o-w

ay i

nter

acti

on o

f den

sity

and

sub

stra

tum

se

lect

ion

at e

ach

of t

hree

cur

rent

vel

ocit

ies

for

expe

rim

ents

in

whi

ch a

sig

nifi

cant

thre

e-w

ay i

nter

acti

on

(vel

ocit

y x

dens

ity

x su

bstr

atum

) w

as d

etec

ted

Sub

stra

tum

Con

tras

t S

low

In

term

edia

te

Fas

t

X2

p X

2 p

X2

.C. h

unge

rfor

di

grav

el v

s. p

ebbl

e 30

.87

< 0

.001

***

23

.15

< 0

.001

***

17

.85

grav

el v

s. c

obbl

e 31

.88

< 0

.001

***

7.

19

0.20

7 5.

33

pebb

le v

s. c

obbl

e 12

.12

0.03

3*

9.44

0.

093@

20

.45

L.h

!W

grav

el v

s. p

ebbl

e 43

.28

< 0

.001

***

8.

61

0.12

6 12

.48

pebb

le v

s. c

obbl

e 25

.02

< 0

.001

***

21

.29

< 0

.001

***

18

.20

Lev

els

of s

igni

fica

nce:

N

.S, n

ot s

igni

fica

nt;@

, app

roac

hing

sig

nifi

cant

; *,

sign

ific

ant;

**, h

ighl

y

sign

ific

ant;

***

,ver

y hi

ghly

sig

nifi

cant

p

< 0

.001

***

0.37

7

< 0

.001

***

0.02

9*

0.00

3**

:!j

Tab

le 4

.3.-

-A p

rior

i com

pari

sons

of

subs

trat

um s

elec

tion

for

dif

fere

nt d

ensi

ty t

reat

men

ts b

ased

upo

n Z

­sc

ores

(S

PS

S I

nc.,

1988

). V

eloc

ity

inde

pend

ence

ind

icat

es th

ose

expe

rim

ents

in

whi

ch v

eloc

ity

did

not

mod

ify

the

man

ner i

n w

hich

den

sity

aff

ecte

d su

bstr

atum

sel

ecti

on.

All

two-

way

int

erac

tion

s fo

cus

on

dens

ity

trea

tmen

t and

sub

stra

tum

sel

ecti

on.

Vel

ocit

y de

pend

ence

inc

lude

s al

l si

tuat

ions

in w

hich

vel

ocit

y af

fect

ed th

e m

anne

r in

whi

ch d

ensi

ty m

odif

ied

subs

trat

um s

elec

tion

. A

s a

resu

lt,

two-

way

inte

ract

ions

wer

e ev

alua

ted

at e

ach

curr

ent v

eloc

ity

Intr

aspe

cifi

c D

ensi

ty

Com

pari

sons

18 v

s. 3

6 18

vs.

54

VE

LO

CIT

Y I

ND

EP

EN

DE

NC

E

A.ci

rcur

nnci

n~tu

s

grav

el v

s. p

ebbl

e -1

.43

-1.2

7

grav

el v

s. c

obbl

e -1

.94@

-0

.23

pebb

le v

s. c

obbl

e -2

.87*

* 0.

90

L.l

utzi

gr

avel

vs.

cob

ble

-0.1

0 -2

.83*

*

Inte

rspe

cifi

c C

ompa

riso

ns w

ith

Tot

al D

ensi

ty C

onst

ant

2 S

peci

es (

N=

36)

Aci

rc

Chu

ng

Llu

tzi

(N=

18)

(N=

18)

(N=

18)

----

-1.4

5 -0

.24

----

-0.9

3 -2

.84*

*

----

-1.5

3 -2

.40*

-0.0

8 1.

50

----

3 S

peci

es (

N =

54)

All

Thr

ee a

t N=

18

-2.9

6**

-1.3

0

0.66

-2.4

6*

Tab

le 4

.3 c

onti

nued

Intr

as.p

ecif

ic D

ensi

ty

Com

pari

sons

18 v

s. 3

6 18

vs.

54

VE

LO

CIT

Y D

EP

EN

DE

NC

E

~

00

C.

hun

gerf

ordi

grav

el v

s. p

ebbl

e

Slo

w

3.24

**

2.36

*

Inte

rmed

iate

-2

.40*

1.

10

Fas

t -0

.42

0.79

grav

el v

s. c

obbl

e

Slo

w

-3.6

8***

3.

68**

*

Inte

rmed

iate

N

S N

S F

ast

NS

NS

Inte

rspe

cifi

c C

ompa

riso

ns w

ith T

otal

Den

sity

Con

stan

t

2 S

peci

es (

N=

36)

Aci

rc

Chu

ng

Llu

tzi

(N=

18)

(N=

18)

(N=

18)

1.18

---

-0.

97

1.16

---

--3

.49*

**

1.76

@

----

-1.3

5

-1.4

2 ---

--1

.42

NS

----

NS

NS

----

NS

3 S

peci

es (

N =

54)

All

Thr

ee a

t N

=18

-0.0

3

0.66

-0.0

6

0.45

NS

NS

Tab

le 4

.3 c

onti

nued

Intr

aspe

cifi

c D

ensi

ty

Com

pari

sons

18 v

s. 3

6 18

vs.

54

pebb

le v

s. c

obbl

e ~

\0

Slo

w

-0.4

7 0.

89

Inte

rmed

iate

0.

08

-1.0

6

Fas

t -3

.19*

* 1.

32

L.m

m

grav

el v

s. p

ebbl

e

Slo

w

-1.2

4 -1

.58

Inte

rmed

iate

N

S N

S

Fas

t -0

.15

-2.0

3*

Inte

rspe

cifi

c C

ompa

riso

ns w

ith T

otal

Den

sity

Con

stan

t

2 S

peci

es (

N=

36)

Aci

rc

Chu

ng

Llu

tzi

(N=

18)

(N=

18)

(N=

18)

-0.9

8 ---

--1

.50

-1.0

7 ---

-1.

04

0.04

---

--3

.44*

**

-1.7

0@

1.97

* ---

-N

S N

S ---

-

-0.0

6 -0

.08

----

3 S

peci

es (

N=

54)

All

Thr

ee a

t N=

18

1.52

-0.2

7

0.04

0.19

NS

-2.1

7*

VI

0

Tab

le 4

.3 c

onti

nued

pebb

le v

s. c

obbl

e

Slo

w

Inte

rmed

iate

Fas

t

Intr

as.p

ecif

ic D

ensi

ty

Com

pari

sons

18 v

s. 3

6 18

vs.

54

-3.1

1 **

0.

55

-0.1

6

-1.0

0

-2.0

2*

0.95

Inte

rspe

cifi

c Comparison~ w

ith

Tot

al D

ensi

ty C

onst

ant

2 S

peci

es (

N=

36)

Aci

rc

Chu

ng

Llu

tzi

(N=

18)

(N=

18)

(N=

18)

0.06

-2

.90*

*

-2.4

9*

2.81

**

0.50

-1

.21

3 S

peci

es (

N -5

4)

All

Thr

ee a

t N

= 18

-0.9

6

-0.3

6

-1.2

1

Lev

elso

fsig

nifi

canc

e: N

S,P

>0

.1;@

,0.1

>P

>0

.05

; *,

0.0

5>

P>

0.0

1;

**,0

.01

>P

>0

.00

1;

**

*,P

<

0.00

1

whereas doubling (but not tripling) the density of .C. hungerfordi affected

substratum selection only in the fast current. Interspecific competition with

A. circumcinctus never significantly modified substratum selection by .C.

hungerfordi. However, the presence of L . .1Y1zi did modify substratum

selection in some substratum size combinations in intermediate and fast

current velocities. The simultaneous presence of all three naucorid species

did not alter substratum selection compared to situations with similarly

elevated densities of C. hungerfordi.

Limnocoris lutzi

Current velocity affected the way in which density treatments

modified substratum selection in all contrasts except those involving gravel

and cobble, where the effects of both density treatment and velocity were

consistent (Table 4.1). In cases where a three-way interaction existed,

substratum selection usually varied among the density treatments, except at

the intermediate velocity and the gravel vs. pebble contrast. Tripling the

number of L. lutzi consistently modified substratum selection in the gravel

vs. cobble contrast, but at only one of the three ve~ocities in each of the

other two substratum combinations. The addition of a single competing

species modified substratum selection, although not at fast current speeds or

in the gravel vs. cobble contrast. In the simultaneous presence of both other

species of naucorid, substratum selection by L. lutzi changed consistently

in the gravel vs. cobble contrast, but modified substratum selection in the

gravel vs. pebble contrast only at fast current velocities.

51

Discussion

Early work in stream ecology emphasized substratum characteristics

as factors largely responsible for the microdistribution of stream benthos.

Recently, researchers have stressed the importance of flow characteristics

(Statzner et al., 1988) and questioned the relevance of the substratum.

Clearly substratum and current attributes are confounded in natural streams

(Moon, 1939; Reice, 1980), yet have important independent implications

(Chapter lll). The premise for these experiments was based on competition

for space. Because naucorids respond to both current velocity and

substratum character (Chapter ill) in terms of habitat selection, variation in

both was included in these experiments.

Velocity effects

The role of current velocity in altering substratum selectivity for each

of the three naucorid species has been previously established (Chapter ill).

The way in which velocity affected substratum selectivity differed among

species. In the extreme case, substratum selection by A. circumcinctus did

not change in response current velocity. Similarly, substratum selection by

C. hun~erfordi, was fairly unresponsive to changes in current speeds. Those

responses that did occur were associated with small rock sizes which

constitute uncommon habitat for this species in the field. In comparison, the

manner in which L.lutzi selected substrata often depended on current

velocity. Given that velocity can alter habitat selection (at least for some

species), does this relationship remain constant over a realm of biotic

interactions? More specifically, does velocity affect habitat associations of a

species in the same manner, regardless of the type and magnitude of

competition that occurs within a community?

52

These results revealed a complex relationship, in which velocity often

affected the manner in which competition altered substratum selectivity. In

biological terms, a species response to an increase in its own density or the

presence of another competitor may vary, depending on current velocity.

Perhaps at fast currents, one species is able to cling to the substratum and

displace (exploitative or interference) other competitors, whereas, at slower

currents, adaptations for clinging or displacement may not suffice to preempt

a microhabitat. Instead, a second species (which is perhaps more mobile at

slow currents) may exhibit a higher degree of exploitative efficiency, and

may dominate the competitive interaction. The idea of environment­

dependent competitive dominance is not new. Park (1954) showed how two

species of Tribolium could vary in their competitive dominance depending

on temperature. Similarly, Connell (1961) established a competitive

dominance between barnacle species which varied with altitude.

Of the three naucorid species tested, two (C. hungerfordi and L.lutzi)

exhibited complex responses to current and competition. For the former, a

three-way interaction between current velocity, density, and substratum

selection was consistently detected for every substratum contrast. For the

latter, a three-way response was detected in two of the three substratum

contrasts, reflecting the intimate relationship between this species and

current.

Density effects

The proponents of density-dependent habitat selection hold that at

high densities, a species will occupy a wider range of habitats (i.e., initially

suboptimal habitat will be selected as density increases). As the number of

individuals increases, the proportion of individuals in preferred to initially

53

unpreferred habitat will gradually decrease. As a consequence, each

naucorid species should respond to an increase in density by exhibiting

diminished substratum associations (see Chapter ill).

In nature, one would expect intraspecific competition to affect habitat

associations of a species more dramatically at high densities. These data

support the presence of intraspecific competition, but they do not advocate

the idea that highest densities affect habitat associations the most. For

example, ~. hun~erfordi was quite responsive to an increase density, but the

majority of these responses were detected when densities were doubled.

Similarly, A. circumcinctus changed its habitat associations in response to

doubled densities, but never responded to tripled densities. In comparison,

responses by L. lutzi to increases in density were less frequent than by C.

hun~erfordi, but tripled densities modified habitat associations more

frequently than at doubled densities. The frequency of responses by A.

circumcinctus and C. hun~erfordi to doubled densities over that of tripled

densities reflects some non-linearity of density effects on habitat selection.

The density of a species in a habitat that was initially suboptimal may reach

a threshold that can not sustain increasing densities without severe

consequences to an individual's fitness. At this point, the initially preferred

habitat offers higher fitness and individuals opt for this habitat secondarily.

This may explain those cases in which doubled densities altered habitat

associations, while tripling a species density failed to alter habitat

associations.

The degree and manner by which species altered their behavior in

response to density varied among species. Cryphocricos hun~erfordi was

quite responsive to an increase in density (in 42.9% of the i priori

intraspecific comparisons, competition was detected); whereas, A.

54

circumcinctus responded to an increase in density in a low 16.7% of the

time. Limnocoris ~was intermediate in response to increases in density;

this species altered its microdistribution in one-third of those l! priori

comparisons. Habitat selection is dynamic; a species perception of habitat

quality may depend on the density of conspecifics in that habitat. The

consequence of increasing population density on habitat selection may not

be a function of linearity as shown by these naucorids. Furthermore, the

relationships between density and habitat selection can vary among species.

Interspecific effects

Interspecific competition in nature appears to be mostly asymmetrical

(Lawton and Hassell, 1981; McAuliffe, 1984; Hemphill, 1988); however,

many studies focus on effects of competition on one target species, in tum,

neglecting reciprocal responses by other species (Schoener, 1983; Connell,

1983 ). Dominant hierarchies of competition and symmetry of responses are

impossible to establish in this manner. Approaches that measure the effects

of interspecific competition on all species concerned may offer insight into

mechanisms by which species coexist.

Responses by A. circumcinctus to the presence of other naucorids was

highly variable. Evidence from the two-species interaction experiments

indicated that A. circumcinctus was virtually unresponsive to~.

hun~erfordi; however, the presence of L . .him frequently altered habitat

associations. Similarly, the joint occurrence of all three species affected

habitat associations of A. circumcinctus. In all likelihood this response was

induced by L. lutzi. rather than the simultaneous presence of all three

species, considering that A. circumcinctus was unresponsive to ~.

hun~erfordi in previous experiments. Limnocoris ~ responded to

55

interspecific pressures in a third of the two-species a priori contrasts. The

majority of those responses were detected in the presence of .C. hungerfordi.

Czyphocricos hungerfordi was by far the least sensitive species to

interspecific competition. Habitat associations were never modified in the

presence of L. ~ or in the sirilultaneous presence of all three-species.

Ecologists frequently measure competitive dominance (in tenns of

competition for space) by the ability of a species to impose habitat

restrictions on other species while retaining its own habitat associations.

Given that the outcomes of competitive interactions are ultimately dependent

upon physical factors (Connell, 1983), such an approach may oversimplify

competitive relationships between species. The relationship between C.

hungerfordi and L. lutzi illustrates this point nicely. Czyphocricos.

hungerfordi was unresponsive to the presence L.lutzi. Conversely, L. lutzi

altered its patterns of habitat associations in the presence of ,C. hungerfordi

at larger rock sizes (i.e., pebble vs. cobble) but not in other contrasts. The

ability of~- hungerfordi to affect the distribution of L. lutzi at large rock

sizes, but not at small sizes illustrates how physical factors can detennine

competitive outcomes. If an attempt was made to determine competition for

smaller rocks between the two species, competition may not have been

detected, when in fact these species compete intensely for larger rock sizes.

Potential Mechanisms and Observations

Few studies isolate competitive mechanisms (interference vs.

exploitation vs. preemption) in stream systems, in part because of the

difficulty to quantify shared resources (Dudley et al., 1990). The difficulty

to detect these resources increases as the mobility of an organism increases.

For example, a sessile (or semi-sessile) benthic inhabitant is limited to a set

56

of resources defined by those it encounters in the three dimensional column

in which it inhabits. As mobility increases, the choices offered to an animal

is not limited to those encountered at the site of attachment. The ability to

forage enables a species to avoid deleterious effects of overcrowding thereby

optimizing a habitats ability to sustain diversity and high population

densities. Naucorids are among the most mobile lotic benthos in

southwestern stream systems. Furthennore the three species tested exhibit

extensive overlap in substratum, current, and prey utilization. This overlap

may explain the complex competitive relationships among the naucorids but

fails to offer mechanisms by which naucorids compete. Through

observation of naucorid behavior in closed containers and at slow velocities

in the artificial streams, this author offers two potential mechanisms by

which naucorids compete for space. The first entails a direct interference

mechanism. Intra- and interspecific encounters by A. circumcinctus and C.

hun~erfordi are not passive. Upon an encounter, two individuals grasp the

pro legs of one another. The individuals remain interlocked (often seconds at

a time) until one individual relinquishes its hold and continues its movement

to an unoccupied substratum. Another mechanism involves avoidance

behavior. When an individual L. ~ encounter conspecifics or other

naucorid species, the individual moves up into the water column and swims

in a random direction to a new site of settlement. If the new site of

settlement was occupied this behavior was frequently repeated until an

unoccupied microhabitat was identified. This avoidance behavior may serve

to segregate the species and explain why L.bmi altered its substratum

associations in the presence of other species.

Naucorids may minimize interspecific interactions by exploiting

different regions of substratum particles. For instance, ~. hun2erfordi was

57

rarely viewed clinging to the top or sides of individual rocks. Instead they

were found associated with under surfaces of larger rocks. In contrast, L.

lutzi frequently existed at the water-substratum interface where it minimized

contact with C. hungerfordi. This "particle partitioning" may minimize

conflicting interactions in tum facilitating the coexistence of the two species.

Competition among naucorids is complex. Naucorids can exhibit

variation in habitat associations depending upon current velocity, substratum

size distributions, density, and interspecific competition. The ability of

substrata to supply food may also influence habitat associations of naucorids

especially if the abundance and distribution of prey species are correlated

with flow and substratum characteristics. Furthermore, density reductions

caused by flooding events suppress intraspecific (Feminella and Resh, 1990)

and interspecific (Hemphill, 1991) competition in stream communities. The

SLR as well as other streams on the Edwards plateau is frequently subjected

to spates and high intensity floods which affect naucorid densities (York et

al., unpublished data). These events may reduce competition among

naucorid species and ensure species richness in the SLR.

58

Literature Cited

Connell, J. H. 1961. The influence of interspecific competition and other factors on the distribution of the barnacle Chthamalus stellatus. Ecology 42:710-723

Connell, J. H. 1975. Some mechanisms producing structure in natural communities: a model and evidence from field experiments. pp. 460-490. in M. L. Ecology and evolution of communities. Belknap Press, Cambridge, Massachusetts, U.S.A.

Connell, J. H. 1978. Diversity in tropical rainforests and coral reefs. Science 199:1302-1310.

Connell, J. H. 1980. Diversity and the coevolution of competitors, or the ghost of competition past. Oikos 35:131-138.

Connell, J. H. 1983. On the prevalence and relative importance of interspecific competition: evidence from field experiments. Am. Nat. 122:661-669.

Constantz, G. D. 1974. The mating behavior of a creeping water bug, Ambzysus occidentalis (Hemiptera: Naucoridae ). Am. Midi. Nat. 92:234-239.

Dudley, T. L., S.D. Cooper, and N. Hemphill. 1986. Effects of macroalgae on a stream invertebrate community. J. N. Am. Benthol. Soc. 5:93-106.

Dudley, T. L., C. M. D'Antonio, & S.D. Cooper. 1990. Mechanisms and consequences of interspecific competition between two stream insects. I. Anim. Ecol. 59:849-866.

Feminella, I. W., & V. H. Resh. 1990. Hydrologic influences, disturbance and intraspecific competition in a stream caddisfly population. Ecology 71:2083-2094.

Fretwell, S.D., and H. L. Lucas, Jr. 1970. On territorial behavior and other factors influencing habitat distribution in birds. I. Theoretical development. Acta Biotheoretica 19:16-36.

59

•

Glass, L.W. and R.V. Bovbjerg. 1969. Density and dispersion in laboratory populations of caddisfly larvae (Cheumatopsyche, Hydropsychidae). Ecology 50:1082-1084.

Gonsoulin, G. J. 1973. Seven families of aquatic and semiaquatic Hemiptera in Louisiana. Entomol. News 84:83-88.

Harding, J. and M. H. Colbo. 1981. Competition for attachment sites between larvae of Simuliidae (Diptera). Can. Ent. 113:761-763.

Hemphill, N. 1988. Competition between two stream dwelling filter­feeders, Hydropsyche oslari and Simulium virgatum. Oecologia 77:73-80.

Hemphill, N. 1991. Disturbance and variation in competition between two stream insects. Ecology 72:864-872.

Herrmann, D.P., R. W. Sites and M. R. Willig. 1992. A laboratory flow tank with variable currents and depths for replicating riffles and shallow streams. Texas J. Sci. 44:89-94.

Lawton, J. H., and M.P. Hassell. 1981. Asymmetrical competition in insects. Nature 289:793-795.

McAuliffe, J. R. 1983. Competition, colonization patterns and disturbance in stream benthic communities. pp. 137-156. in J. R. Barnes and G. W. Minshall ( eds. ). Stream ecology: application and testing of general ecological theory. Plenum, New York, New York.

McAuliffe, J. R. 1984. Competition for space, disturbance, and the structure of a benthic stream community. Ecology 65:894-908.

McLachlan, A. J. 1969. Substrate preferences and invasion behaviour exhibited by larvae of Nilodorum breyibucca Freeman (Chironomidae) under experimental conditions. Hydrobiologia 33:237-249.

60

McPherson J. E., R. J. Packauskas, and P. P. Korch, ill. 1987. Life history and laboratory rearing of Pelocoris femoratus (Hemiptera: Naucoridae), with descriptions of immature stages. Proc. Entomol. Soc. Wash. 89:288-295.

Minshall, G. W. 1984. Aquatic insect-substratum relationships, pp. 358-400. in V. H. Resh and D. M. Rosenberg (eds.). The ecology of aquatic insects. Praeger, New York.

Moon, H.P. 1939. Aspects of the ecology of aquatic insects. Trans. Brit. Entomol. Soc. 6:39-49.

Park, T. 1954. Experimental studies of interspecific competition. II. Temperature, humidity and competition in two species of Tribolium. Physiol. Zool. 27:177-238.

Pulliam, H. R. 1983. Ecological community theory and the coexistence of sparrows. Ecology 64:45-52

Reice, S. R. 1980. The role of substratum in benthic macroinvertebrate microdistribution and litter decomposition in a woodland stream. Ecology 61 :580-590.

Schoener, T. W. 1983. Field experiments on interspecific competition. Amer. Nat. 122:240-285.

Sites, R. W. and B. J. Nichols. 1990. Life history and descriptions of immature stages of Ambrysus lunatus lunatus (Hemiptera: Naucoridae). Ann. Entomol. Soc. Am. 83:800-808.

Sites, R. W. and M. R. Willig. 1991. Microhabitat associations of three sympatric species of Naucoridae (Insecta: Hemiptera). Environ. Entomol. 20:127-134.

Sousa, W. P. 1984. The role of disturbance in natural communities. Ann. Rev. Ecol. Sys. 15:353-391.

SPSS Inc. Advanced statistics guide. 2nd Edition. pp. 527. Chicago n.

61

Stout, R. J. 1981. How abiotic factors affect the distribution of two species of tropical predaceous aquatic bugs (Family: Naucoridae ). Ecology 62:1170-1178

Venkatesan, P. and A. Cloarec. 1988. Density dependent prey selection in llyocoris (Naucoridae). Aquatic Insects 10:105-116.

Ward, J. V. 1992. Aquatic insect ecology. 1. Biology and habitat. Wiley and Sons, Inc New York.

Wentworth, C. K. 1922. A scale of grade and class terms for cladistic sediments. J. Geol. 30:377-392.

Wiley, M. J. 1981. Interacting influences of density and preference of the emigration rates of some lotic chironomid latvae (Diptera: Chironomidae). Ecology 62:426-438.

Zahar, A. R. 1951. The ecology and distribution of black-flies (Simuliidae) in southeast Scotland. J. Anim. Ecol. 20:33-62.

62

CHAPTER V

EFFECTS OF SUBSTRATUM SIZE AND HETEROGENEITY

ON MACROBENTHIC DNERSITY AND

COMPOSffiON

Abstract

Studies concerning diversity and taxonomic composition associated

with substratum characteristics in temperate streams have received little

attention despite the ascendancy of biodiversity as a central theme in

contemporary ecology. This study evaluated the effects of substratum

characteristics on species diversity and taxonomic composition. In

particular, colonization of four rock size classes (gravel, pebble, cobble and

a heterogeneous combination of all three) by macrobenthos was quantified in

the South and North Llano Rivers of central Texas during the summer of

1991. In each of six experimental treatments, the effect of substratum type

on species diversity and composition was evaluated by placing 25 pairs of

trays into riffle habitat within each river. Each treatment was defmed by one

paired contrast between two of the four rock classes. Trays were left

undisturbed for three days after which all macroinvertebrates were extracted

and preserved. The Shannon index of diversity was used to estimate the

diversity of all macroinvertebrates within each tray, and a paired-t test was

used to evaluate if statistically significant differences in species diversity

existed between substrata. Similarly, compositional differences between

substrata within each river were evaluated by log-linear models (SPSS Inc,

1988). Highly significant differences in taxonomic composition between

substrata occurred for all rock contrasts in both rivers. However, substratum

size and heterogeneity had little or no effect on species diversity.

63

Introduction

Biodiversity is a central theme in contemporary ecology. As a

consequence, ecologists have developed a variety of methods to measure

diversity. These measures have potential application to conservation, land

management, and environmental monitoring (Magurran, 1988). In fact,

considerable debate surrounds the appropriateness of diversity measures. In

conseiVation, measurements are based largely on species richness, with little

regard to relative abundance of constituent taxa. In environmental

monitoring, measures are used that include both relative abundances and

species richness. Nevertheless, studies that examine multiple aspects of

community complexity and structure are few. The isolation of factors that

contribute to species diversity, and taxonomic composition is desirable

because these factors reflect the mechanisms contributing to community­

level processes.

A thorough approach for the description of diversity data involves

species abundances models [e.g., log-nonnal model (Sugihara, 1980),

logarithmic series model (Magurran, 1988), and random niche boundary

hypothesis (MacArthur, 1957)]; however, these approaches can involve

tedious model fitting, and problems arise when two communities that are

being compared do not fit the same model. A simpler and more popular

approach involves indices that are based on proportional abundances. These

measures are desirable because they not only account for eveness (a measure

of the distribution of abundances) but also account for species richness.

Nevertheless, examination of only this type of measure may not be

conducive to a fundamental understanding of diversity because of their

generality. The Shannon (H') infonnation index (Krebs, 1985) has become

64

one of the most widely used indices. If the Shannon index is applied to a

number of samples, the measures themselves are nonnally distributed

(Magurran, 1988). This allows for the use of parametric statistics. Although

information indices such as H' are popular they remain taxonomically

insensitive. For example, the Shannon index will render the same value

given that the number of species and their proportional abundances remain

constant (Magurran, 1988) even though constituent taxa are completely

different. Approaches that are sensitive to taxonomic composition may offer

additional insight into species diversity.

Many stream species exploit benthic modes of existence (Ward,

1992). As a consequence, it is not surprising that substratum characteristics

are major determinants of the distribution and abundance of stream

organisms (Minshall, 1984 ). Particle size and heterogeneity are important

components of the substratum (Cummins and Lauff, 1969, Minshall and

Minshall, 1977, Lamberti and Resh, 1979, Fuller and Rand, 1990, and many

others). Nonetheless, the effects of particle size and heterogeneity on stream

benthic species composition are often confounded by factors such as current

association, food habits, respiratory requirements, and other substratum

characteristics (Ward, 1992). As such, careful observations and analyses are

required to interpret substratum effects on community organization.

The idea that complex habitats offer resources that are unavailable to

species in simple habitats is almost universally accepted (Minshall, 1984 ).

In stream environments, a positive correlation may exist between particle

heterogeneity and species richness (Allan, 1975; de March, 1976; Williams,

1980). Nevertheless, a direct correlation between substratum heterogeneity

and density is speculative (Williams, 1980; Ennan and Ennan, 1984).

65

Mean particle size also affects the abundance and distribution of

stream benthos, but size and heterogeneity are often correlated. As the

median particle size of sediments increases, sediment complexity can

increase as well (Reice, 1974). For example, cobble substrata have

propensities for tiny gravel to become lodged in interstitial spaces. The

resulting mixture of particles minimizes gravel-current interactions, thereby

reducing the ability of current to displace gravel. Consequently, cobble is

often considered a conglomeration of particle sizes (Ward, 1992). Although

particle size and heterogeneity are positively correlated, many investigations

have failed to control for their confounding effects on species diversity and

composition. As a result, attempts (e.g., Erman and Erman, 1984) have

focused on elucidating the separate effects of particle size and complexity on

macro benthic assemblages. The effects of heterogeneity on species diversity

may indeed be confounded with particle size (Erman and Ennan, 1984 ). The

following study was aimed at determining the effects of substratum size and

heterogeneity on species diversity and taxonomic composition in two similar

river ecosystems.

Materials and Methods

Study Sites

The South (SLR) and North (NLR) Llano Rivers are located on the

Edwards Plateau of central Texas. The NLR originates in Sutton Co. and

flows eastward ca. 80 km to its confluence with the SLR in Kimble Co. The

SLR originates in Edwards Co., and meanders ca. 17 5 km NE at which point

it joins the NLR. At the confluence, the Llano River emerges and drains into

the Colorado River basin. The SLR and NLR lie on Cretaceous limestone

and are typical of streams on the Edwards Plateau in having gravel and

66

cobble substrata with well-defined riffle and pool habitats. The watersheds

of each river receive substantial quantities of rainfall in the spring and early

fall as a result of moisture-laden air derived from the tropical Gulf storm

tract and convective thunderstorm activity (Baker, 1977). Although streams

in close proximity often exhibit similar flow regimes, the SLR and NLR

vary in both point of origin and flow direction. Consequently, flow regimes

and floods vary with respect to frequency and magnitude in accordance with

the location and prevalence of thunderstorm activity.

Strikingly different habitat structure characterizes channel vs.

marginal habitats in each river, and contributes to contrasting biota in bank

and mid-channel areas. Emergent angiosperms, algae, and silt are common

components in marginal areas. In contrast, exposed cobble and gravel

substrata dominate riffle habitat. Riffles are often vegetatively

impoverished; however, a plethora of invertebrates inhabit these areas.

Colonization

Substratum trays (30.5 x 30.5 x 2.5 em) were constructed from

aluminum flashing, and were used in field experiments as colonization

platforms. Pairs of trays filled with substratum were placed into both the

SLR and NLR during the Summer of 1991. Six treatments were defmed by

all paired contrasts between four substratum classes: (1) coarse gravel (CG;

8-16 mm rocks), (2) small pebble (SP; 16-32 mm rocks), (3) small cobble

(SC; 64-128 mm rocks) and ( 4) a heterogeneous mix (HM) [the first three

size categories from the Wentworth (1922) classification scheme after

Minshall (1984)]. Volumetric displacement standardized the amount of

substrata used in each category (CG, 900 ml; SP, 1200 ml; SC, 1500 m1 and

HM, 1/3 of each homogenous category). Trays were left undisturbed for

67

three days, after which all macroinvertebrates were extracted and preserved

in 70% ethanol. The Shannon index of diversity (Krebs, 1985) estimated the

diversity of all macroinvertebrates within each tray, and a paired-t test

(Sokal and Rohlf, 1981) evaluated differences in species diversity between

substratum types (25 paired samples). Similarly, differences in taxonomic

composition between paired substrata were evaluated for each rock contrast

with log-linear models (SPSS Inc., 1988).

Results and Discussion

Site comparison

Mean temperature, depth, and current velocity were similar at sites in

the SLR and NLR (Table 5.1); however, statistically significant differences

in habitat characteristics were detected by both multivariate (P<0.001) and

univariate tests (temperature, P<0.001; depth, P=0.007; current velocity,

P=0.005). These differences may contribute to different faunal compositions

between rivers.

Riffle faunas

The riffle faunas within the SLR and NLR sites are characteristically

litho- and psephophilic (Table 5.2). Nonetheless, riffle faunas differ

dramatically with respect to the abundances of several species. For instance,

a caddisfly, Marilia sp., was virtually absent (only one specimen collected)

from the SLR; however, it constituted 4.8% of all individuals collected in the

NLR. Conversely, a mayfly Traverella presidiana, was scarce in the NLR

(only two specimens collected); however, it comprised 1.7% of all

individuals collected within the SLR.

68

Table 5.1.--Abiotic conditions (mean+ S.D.) at South and North Llano River sites (N=25)

River Temperature

North Llano 83.8 + 0.4

South Llano 82.3 + 0.5

Depth (em)

69

21.6 + 4.8

18.1 + 3.8

Velocity (em/sec)

90.0

65.0

...J

0

Tab

le 5

.2--

Mea

n de

nsit

y(+

S.E

.) o

fmac

roin

vert

ebra

te ta

xa (

per 0

.09

m2)

fou

nd i

n th

e S

outh

and

Nor

th

Lla

no R

iver

s ac

cord

ing

to s

ubst

ratu

m ty

pe (

CG

, coa

rse

grav

el;

SP

, sm

all p

ebbl

e; S

C, s

mal

l cob

ble;

HM

, he

tero

gene

ous

mix

) (N

=75

). N

F in

dica

tes

that

no

indi

vidu

als

of a

giv

en ta

xon

wer

e fo

und

in th

e co

rres

pond

ing

subs

trat

um ty

pe

Tax

a S

outh

Lla

no R

iver

N

orth

_Lla

no R

ive_

r__

CG

S

P

SC

H

M

CG

S

P

SC

H

M

Oli

goch

aeta

N

F

0.01

±0.

01

NF

N

F

NF

N

F

NF

0.

01±

0.01

Pla

tyhe

lmin

thes

0.

04±

0.03

0.

03±

0.02

0.

03±

0.02

N

F

NF

0.

01±

0.01

0.

03±

0.02

0.

03±

0.02

Odo

nata

Cal

opte

rygi

dae

Het

aeri

na s

p

0.37

±0.

07

0.45

±0.

08

0.19

±0.

05

0.32

±0.

07

0.23

±0.

05

0.28

±0.

06

0.19

±0.

05

0.17

±0.

04

Gom

phid

ae

Spe

cies

A

0.03

±0.

02

NF

N

F

0.03

±0.

02

0.01

±0.

01

0.05

±0.

03

0.01

±0.

01

0.03

±0.

02

Lib

ellu

lida

e

Br~hmQrhQ&il

men

dax

0.24

±0.

05

0.13

±0.

04

0.03

±0.

02

0.12

±0.

04

0.15

±0.

04

0.07

±0.

03

0.03

±0.

02

0.05

±0.

03

Tab

le 5

.2 c

onti

nued

Tax

a So

uth

Lla

no R

iver

N

orth

Lla

no R

iver

CG

SP

SC

HM

CG

SP

SC

H

M

Coe

nagr

ioni

dae

Spec

ies

A

0.13

±0.

05

0.29

±0.

06

0.19

±0.

05

0.12

±0.

04

0.19

±0.

05

0.13

±0.

04

0.09

±0.

03

0.15

±0.

05

Spec

ies

B

0.04

±0.

02

NF

0.

01±

0.01

0.

09±

0.03

0.

08±

0.04

0.

07±

0.03

N

F

0.11

±0.

04

Eph

emer

opte

ra

Bae

tidae

-..

J B

aetis

sp

0.07

±0.

03

0.05

±0.

03

0.16

±0.

05

0.08

±0.

04

0.03

±0.

02

NF

0.

03±

0.02

0.

03±

0.02

-

Da~~

lob~

&eti

s

mex

ican

us

0.13

±0.

06

0.08

±0.

03

0.28

±0.

08

0.16

±0.

05

0.23

±0.

06

0.12

±0.

05

0.73

±0.

27

0.16

±0.

05

Fa.Il

eceo

n

quill

eri

0.23

±0.

07

0.21

±0.

05

0.68

±0.

17

0.53

±0.

10

0.61

±0.

13

1.07

±0.

22

1.27

±0.

21

1.00

±0.

15

Lep

toph

lebi

idae

Ihra

ulod

es

2onz

ales

i 1.

48±

0.19

1.

89±

0.22

0.

61±

0.10

1.

43±

0.20

2.

99±

0.33

2.

64±

0.28

0.

93±

0.14

2.

23±

0.24

Tab

le 5

.2 c

onti

nued

Tax

a S

outh

Lla

no R

iver

N

orth

Lla

no R

iver

CG

S

P

SC

H

M

co

SP

S

C

HM

Tra

yere

lla

pres

idia

na

0.49

±0.

12

0.40

±0.

08

0.31

±0.

08

0.71

±0.

11

0.01

±0.

01

NF

N

F

0.01

±0.

01

Spe

cies

C

0.01

±0.

01

0.03

±0.

02

NF

0.

04±

0.02

0.

32±

0.10

0.

37±

0.08

0.

11±

0.04

0.

44±

0.09

Oli

gone

urid

ae

Isonychia~

7.48

±0.

86

15.0

4±1.

18

7.41

±0.

88

13.4

0±1.

27

4.40

i0.6

0 8.

23±

0.88

1.

64±

0.28

5.

53±

0.74

-..

.J T

rico

ryth

idae

N

wtohmh~s

spA

0.

69±

0.17

0.

33±

0.07

0.

43±

0.10

0.

63±

0.10

0.

19±

0.05

0.

08±

0.03

0.

08±

0.04

0.

21±

0.06

Le~1

2hmh

~s

spB

0.

68±

0.17

0.

39±

0.10

0.

85±

0.20

0.

75±

0.16

0.

27±

0.08

0.

45±

0.11

0.

48±

0.12

0.

32±

0.09

Tri~

QQth

od~~

sp

0.09

±0.

04

0.16

±0.

11

0.11

±0.

04

0.19

±0.

05

0.61

±0.

17

0.36

±0.

11

0.29

±0.

06

0.28

±0.

13

Tri

chop

tera

Hel

icop

sych

idae

Hel

icoo

svch

e sp

0.

40±

0.12

0.

24±

0.06

0.

15±

0.05

0.

29±

0.12

0.

81±

0.19

0.

41±

0.10

0.

28±

0.09

0.

56±

0.12

Tab

le 5

.2 c

onti

nued

Tax

a S

outh

Lla

no R

iver

N

orth

Lla

no R

iver

co

SP

S

C

HM

co

S

P

SC

H

M

Hyd

rops

ychi

dae

Cheumatop~~b~

sp

0.05

±0.

03

0.12

±0.

04

0.24

±0.

07

0.27

±0.

08

0.12

±0.

06

0.04

±0.

02

0.20

±0.

05

0.16

±0.

05

Sm

icri

dea

21·

0.29

±0.

08

0.35

±0.

16

0.40

±0.

10

0.51

±0.

16

0.20

±0.

05

0.04

±0.

02

0.13

±0.

06

0.08

±0.

03

Odo

ntoc

erid

ae

....,J

M

aril

ia s

p 0.

01±

0.01

N

F

NF

N

F

1.84

±0.

29

0.85

±0.

20

0.13

±0.

05

1.03

±0.

18

Vol

Phi

lopo

tam

idae

Cbi

mar

ra s

p 0.

05±

0.03

0.

09±

0.04

0.

12±

0.04

0.

12±

0.04

0.

23±

0.07

0.

04±

0.02

0.

09±

0.03

0.

21±

0.06

Pol

ycen

trop

odid

ae

PQI!~~ntropys

sp

0.01

±0.

01

NF

0.

01±

0.01

0.

01±

0.01

0.

13±

0.05

0.

19±

0.05

0.

17±

0.05

0.

21±

0.07

Polm

l~~t

tQPY

S

sp

0.05

±0.

03

0.05

±0.

03

0.08

±0.

04

0.09

±0.

04

0.31

±0.

03

0.31

±0.

07

0.20

±0.

06

0.17

±0.

05

Rhy

acop

hili

dae

Ato

osvc

he s

p N

F

0.01

±0.

01

0.03

±0.

02

0.01

±0.

01

0.03

±0.

02

NF

0.

03±

0.02

0.

01±

0.01

Tab

le 5

.2 c

onti

nued

Tax

a So

11th

Lla

no R

iver

N

orth

Lla

no_R

iver

CG

SP

SC

HM

CG

SP

SC

H

M

Spec

ies

A

NF

N

F

NF

0.

01±

0.01

0.

01±

o.01

0.

01±

o.01

0.

04±

0.02

N

F

Hem

ipte

ra

Nau

cori

dae

Am

bt:y

sus

-..J

ci

rc:u

mcin

~tus

0.

92±

0.13

1.

00±

0.12

0.

53±

0.09

1.

07±

0.13

1.

16±

o.13

1.

12±

0.14

0.

76±

o.13

1.

33±

0.16

~

Am

bt:y

sus

luna

tus

0.07

±0.

03

0.09

±0.

04

0.13

±0.

05

0.09

±0.

05

0.03

±0.

03

0.07

±0.

04

0.04

±0.

02

0.01

±0.

01

Ccyp

hocr

i~QS

bYn&

erfQ

rdi

0.33

±0.

07

0.17

±0.

06

0.12

±0.

04

0.37

±0.

08

NF

N

F

NF

N

F

Lim

noco

ris

hlW

0.

51±

0.09

0.

17±

0.07

0.

07±

0.03

0.

37±

0.08

0.

01±

o.01

0.

01±

0.01

N

F

NF

Tab

le 5

.2 c

ontin

ued

Tax

a

Vel

iidae

Rha

gove

lia

dist

inct

a

Neu

ropt

era

Cor

ydal

idae

CG

NF

Sout

h L

lano

Riv

er

SP

SC

HM

NF

N

F

NF

......,J

v.

C

ot:y

dalu

s

corn

utus

Col

eopt

era

Dry

opid

ae

Hel

ie bu

s

sutu

ral i

s

Elm

idae

0.08

±0.0

3 0.

04±0

.02

0.08

±0.0

4 0.

19±0

.05

0.15

±0.0

5 0.

09±0

.04

0.04

±0.0

3 0.

11±0

.04

Dub

irap

hia

sp

0.12

:t.i).

05

0.07

±0.0

5 0.

08±0

.04

0.12

±0.0

5

Nor

th L

lanQ

RiY

er

CG

SP

SC

HM

0.08

±0.0

6 1.

08±0

.72

0.28

±0.2

1 0.

60±0

.45

0.03

±0.0

2 0.

05±0

.03

0.03

±0.0

2 N

F

0.16

±0.0

5 0.

12±0

.05

0.01

±0.0

1 0.

13±0

.04

NF

N

F

0.03

±0.0

2 0.

01±0

.01

Tab

le 5

.2 c

onti

nued

Tax

a S

outh

Lla

no R

iver

N

orth

Lla

no R

iver

CG

SP

S

C

HM

CG

S

P

SC

H

M

H~x~

~!ll

oe.a

:us

ferr

u&in

ius

14.1

5±1.

40

5.12

±0.

65

2.44

±0.

29

6.37

±0.

69

2.20

±0.

32

0.79

±0.

14

0.35

±0.

07

1.28

±0.

32

Mac

relm

is

texa

nys

0.01

±0.

01

NF

0.

03±

0.02

N

F

0.03

±0.

02

0.01

±0.

01

NF

N

F

Mi~mc!lloe,ay:i

-...J

pusi

llus

3.

37±

0.42

2.

21±

0.37

2.

40±

().4

2 2.

99±

0.45

0.

72±

0.15

0.

31±

0.07

0.

39±

0.08

0.

37±

0.10

0

'\

Neo

elm

is ~

0.33

±0.

11

0.08

±0.

04

0.05

±0.

04

0.09

±0.

04

0.19

±0.

07

0.04

±0.

02

NF

0.

07±

0.03

Ste

nelm

is s

p 0.

20±

0.05

0.

04±

0.02

0.

01±

0.01

0.

03±

0.02

0.

01±

0.01

N

F

NF

N

F

Lut

roch

idae

Lut

roch

:us

lyte

ys

0.11

±0.

04

0.08

±0.

04

0.07

±0.

03

0.07

±0.

03

0.01

±0.

01

NF

0.

07±

0.03

0.

04±

0.03

Pse

phen

idae

Tab

le 5

.2 c

onti

nued

Tax

a So

uthL

lano

Riv

er

Nor

th L

lano

.RiY

er

CG

S

P

SC

H

M

CG

S

P

SC

H

M

Psm

>hen

us

texa

nus

NF

N

F

NF

N

F

0.01

±0.

01

NF

N

F

NF

Hyd

roph

ilid

ae

Spe

cies

A

0.05

±0.

05

NF

N

F

NF

N

F

NF

N

F

NF

Spe

cies

B

NF

N

F

0.01

±0.

01

NF

N

F

NF

N

F

NF

...

.,J

Dip

tera

...

.,J

Chi

rono

mid

ae

0.04

±0.

02

0.01

±0.

01

0.07

±0.

03

0.05

±0.

03

0.11

±0.

04

0.01

±0.

01

0.03

±0.

02

0.08

±0.

03

Sim

ulii

dae

0.07

±0.

03

0.11

±0.

04

0.23

±0.

06

0.08

±0.

04

0.48

±0.

15

0.64

±0.

28

0.33

±0.

10

0.08

±0.

04

Pup

a N

F

NF

0.

01±

0.01

N

F

0.01

±0.

01

0.01

±0.

01

NF

N

F

Lep

idop

tera

Pyr

alid

ae

Pet

ro.p

hila

sp

0.01

±0.

01

0.01

±0.

01

0.05

±0.

03

0.01

±0.

01

0.03

±0.

02

0.04

±0.

03

0.01

±0.

01

NF

Mol

lusc

a 0.

05±

0.03

0.

08±

0.05

0.

09±

0.05

0.

11±

0.04

0.

01±

0.01

0.

01±

0.01

N

F

NF

Gas

trop

oda

NF

N

F

NF

N

F

0.01

±0.

01

NF

N

F

NF

Dryopoid beetles were well-represented within the S:...R (8 species)

and NLR (9 species) sites; however, the waterpenny CPsephenus texanus),

was represented by only one immature specimen in the NLR. Caddisflies

were represented by six families and eight genera, most prominently by

hydropsychids. Mayflies comprised the most individuals from each site;

taxonomic composition included common species (8 genera and 4 families)

of the Edwards Plateau. The SLR harbors eight naucorid (Hemiptera)

species (Sites and Willig, 1991) which is most species-rich temperate

naucorid fauna. Four naucorids were found in the SLR riffle site and three

in the NLR site.

The most obvious difference in riffle biota between the NLR and SLR

involved the total density of invertebrates. Macroinvertebrate densities in

the SLR site were twice that of the NLR site. This difference was

accentuated by the presence of several common species (including two

beetles and one mayfly), whose densities were consistently higher in the

SLR. In addition, conspicuous differences in certain predator densities were

detected between sites. For instance, five invertebrate predators appear to

dominate the SLR. The hellgrammite, Cozydalus comutus (Corydalidae), is

the top invertebrate consumer, feeding on all consumer trophic levels

(Camilo and Willig, unpublished data). Similarly, four naucorids (A.

circumcinctus, A. lunatus, C. hun~erfordi and L.lutzil are prevalent

consumers that feed at more than one trophic level (Camilo and Willig,

unpublished data). A contrasting scenario existed at the NLR, wherein only

one predator (A. circumcinctus) occurred in substantial numbers. This

species existed at lower densities in the SLR than in the NLR. Consistently

lower densities in the SLR may be a result of interspecific interactions

(Chapter IV), physical differences between the riffles (Table 5.1 ),

78

differences in time since perturbation (York et al., unpublished data) or a

combination of each. Nevertheless, the faunas in the SLR and NLR are

diverse and constituent species are abundant. Overall the faunas are

strikingly similar; however, the biotic structure of each site differs in various

respects.

South Llano River

Forty-six taxa were collected from samples in the SLR (Table 5.2).

The mayfly, Isonychia sicca clearly dominated the riffle by contributing

37.9% of the total individuals. Four of the 46 species a. sicca,

Hexacylloepus ferruginius (24.6%), Microcylloepus pusillus (9.6%) and

Thraulodes gonzalesi ( 4. 7%) comprised 76.8% of all individuals.

North Llano River

Forty-seven taxa were collected from the NLR (Table 5.2), 44 of

which co-occurred in the SLR. Again, I. sicca dominated the riffle

comprising 29.9% of all individuals. Four species[! . .s,im, T. gonzalesi

(13.3%), H. ferruginius (7.0%), M. pusillus (2.7%)] accounted for 52.9% of

all individuals collected.

Taxonomic composition

Substratum size

For every paired contrast of substratum size, taxonomic composition

varied very highly significantly (P<O.OOl) between the SLR and NLR (Table

5.3). Within the gravel vs. pebble, strong association by I. sicca with the

pebble substratum and by H. ferru~inius for the gravel substratum

contributed to much of the compositional variation between substrata at both

79

00

0

Tab

le 5

.3--

Com

pari

sons

of s

peci

es d

iver

sity

and

taxo

nom

ic c

ompo

siti

on b

etw

een

all p

ossi

ble

pair

s o

f fou

r su

bstr

atum

type

s in

eac

h o

f tw

o di

ffer

ent r

iver

s, S

outh

, (S

LR

) an

d N

orth

, (N

LR

) L

lano

Riv

er.

For

eac

h co

mpa

riso

n, 2

5 pa

ired

sam

ples

wer

e ta

ken

in e

ach

rive

r. S

peci

es

dive

rsit

y w

as e

stim

ated

for

eac

h sa

mpl

e vi

a th

e S

hann

on d

iver

sity

inde

x; c

ompa

riso

ns w

ere

base

d up

on p

aire

d-t t

ests

(S

okal

and

Roh

lf,

1981

). C

ompo

siti

onal

dif

fere

nces

bet

wee

n su

bstr

ate

type

s fo

r ea

ch c

ontr

ast w

ere

eval

uate

d vi

a lo

g-li

near

ana

lyse

s (S

PSS

Inc.

, 19

88),

and

list

ed a

s li

keli

hood

rat

io c

hi-s

quar

e st

atis

tics

Con

tras

t D

iver

sity

Com

pari

sons

C

ompo

siti

onal

Com

pari

sons

' Pai

red-

t p

X2

p

Gra

vel v

s. P

ebbl

e SL

R

1.66

0.

11N

S

320.

66

P<0.

001 **

* N

LR

4.

97

<0.0

01 **

* 13

2.08

P<

0.00

1 **

* G

rave

l vs.

Cob

ble SL

R

1.56

0.

13N

S

214.

38

P<0.

001 **

* N

LR

1.

89

0.07

NS

13

6.27

P<

0.00

1 **

* P

ebbl

e vs

. Cob

ble SL

R

-1.3

3 0.

20N

S

106.

53

P<O

.OO

l ***

N

LR

-0

.35

0.73

NS

23

3.33

P<

O.O

Ol *

**

Tab

le 5

.3 c

onti

nued

Con

tras

t D

iver

sity

Com

pari

sons

C

ompo

siti

onal

Com

pari

sons

Pai

red-

t p

X2

p

Gra

vel v

s. M

ixtu

re

SL

R

0.97

0.

34N

S

138.

32

P<

0.00

1 **

*

NL

R

2.60

0.

02*

107.

62

P<

0.00

1 **

* 0

0

Peb

ble

vs.

Mix

ture

-

SL

R

-2.3

7 0.

03*

90.4

2 P<

O.O

Ol *

**

NL

R

-0.3

4 0.

74N

S

91.4

6 P<

O.O

Ol *

**

Cob

ble

vs.

Mix

ture

SL

R

-0.9

6 0.

35N

S

102.

17

P<0.

001

***

NL

R

-0.4

4 0.

66N

S

101.

87

P<

0.00

1 **

*

Lev

els

of

sign

ific

ance

: *,

sig

nifi

cant

; **

*, v

ery

high

ly s

igni

fica

nt

sites. In the gravel vs. cobble contrast in the NLR, Leptohyphes sp.B. and

Dactylobaetis mexicanus occurred more frequently in the cobble over the

gravel; this significantly contributed to the overall difference in taxonomic

composition between the substrata. Within the SLR, a strong affmity by H.

ferru~inius for gravel compared with cobble was again evident and

significantly contributed to compositional variation between substrata. In

the pebble vs. cobble contrast, strong substratum associations by four species

of mayflies significantly contributed to the overall variation in compositional

differences between substrata. Influences by I. sicca (pebble) and D.

mexicanus (cobble) in the SLR and Falleceon quilleri (cobble) and

Leptohyphes sp. B (cobble) in the NLR were most evident.

Hetero~eneity

In the gravel vs. heterogeneous mixture contrast, much of the

variation involved with compositional differences at each site was affected

by strong associations of I. sicca with the heterogeneous substratum. The

strong associations of Marilia sp. and H. ferru~inius for the gravel within the

NLR and SLR, respectively, also contributed to the variation. A caddis fly,

Helicopsyche sp. had little effect on compositional variation between

substrata in previous contrasts; however, it significantly contributed to

differences in the pebble vs. heterogeneous mixture contrast in the NLR as

evidenced by its strong association with cobble. In the SLR, the mayfly, T.

gonzalesi (pebble), and a trichopteran, Smicridea sp. (heterogeneous

mixture), each exhibited distinct habitat associations and contributed

substantially to overall differences between substrata. In the cobble vs.

heterogeneous mixture, a strong association between H. ferruginius and the

82

heterogeneous substratum in both rivers accentuated compositional

differences between substrata.

Diversity comparisons

Although highly significant compositional differences associated with

substratum size existed in all cases, species diversity differed between

substratum types in only one instance within the SLR (Table 5.3). This

difference occurred between the pebble and the heterogeneous mixture; the

latter supporting significantly (P=0.03) higher diversity than the former.

Results obtained in the NLR differed from those in the SLR. Two instances

were detected in which a particular substratum supported higher species

diversities than its paired counterpart. In the gravel vs. pebble contrast,

species diversity was significantly higher (P<O.OOl) in the gravel

substratum. Similarly, in the coarse gravel vs. heterogeneous mixture

contrast, the gravel substratum yielded a higher species diversity than did the

heterogeneous mixture.

Although three instances were detected in which substratum type

affected species diversity, no pattern existed in both rivers. Moreover, those

instances in which species diversities differed between substrata provided

little support for hypotheses that favor heterogeneity as a contributor to

species diversity in benthic communities. In fact, in one instance,

significantly higher species diversity was detected for a homogeneous

substratum (gravel) over the heterogeneous mixture in the NLR. Substratum

size and heterogeneity influence stream biota in many fashions. Most

notable are those involved with taxonomic composition. This study

illustrates how differences in habitat structure may support equally diverse

assemblages yet may lead to highly significant differences in taxonomic

83

composition. Approaches that examine species diversity from multiple

perspectives including taxonomic considerations may facilitate a

fundamental understanding of biodiversity and help elucidate community­

level processes.

84

Literature Cited

Allan, _D. J. 1 ?75. The distributional ecology and diversity of benthic msects tn Cement Creek, Colorado. Ecology 56:1040-1053.

Baker, V. R. 1977. Stream-channel response to floods, with examples from central Texas. Geol. Soc. Am. Bull. 88:1057-1071.

Cummins, K. W. and G. H. Lauff. 1969. The influence of substrate particle size on the microdistribution of stream macrobenthos. Hydrobiologia 34:145-181.

de March, B.G.E. 1976. Spatial and temporal patterns in macrobenthic stream diversity. J. Fish. Res. Board Can. 33:1261-1270.

Erman, D.C. and N.A. Ennan. 1984. The response of stream macroinvertebrates to substrate size and heterogeneity. Hydrobiologia 108:75-82.

Fuller, R. L. and P. S. Rand. 1990. Influence of substrate type on vulnerability of prey to predacious aquatic insects. J. N. Am. Benthol. Soc. 9:1-8.

Krebs, C. J. 1985. Ecology: the experimental analysis of distribution and abundance. Harper and Row, New York.

Lamberti, G.A. and V. H. Resh. 1979. Substrate relationships, spatial distribution patterns, and sampling variability in a stream caddisfly population. Environ. Entomol. 8:561-567.

MacArthur, R. H. 1957. On the relative abundance of bird species. Proc. Nat. Acad. Sci. 43:293-295.

Magurran, A. E. 1988. Ecological diversity and its measurement. Princeton University Press, Princeton, New Jersey.

Minshall, G. W. and J. N. Minshall, 1977. Microdistribution of benthic invertebrates in a Rocky Mountain (U.S.A.) stream. Hydrobiologia 55:324-355.

85

Minshall, G. W. 1984. Aquatic insect-substratum relationships, pp. 358-400 . .m.v. H. Resh and D. M. Rosenberg (eds.). The ecology of aquat1c msects. Praeger, New York.

Reice, S. R. 197 4. Environmental patchiness and the breakdown of leaf litter in a woodland stream. Ecology 55:1271-1282.

Sites, R. W. and M. R. Willig. 1991. Microhabitat associations of three sympatric species of Naucoridae (Insecta: Hemiptera). Environ. Entomol. 20: 127-134.

Sokal, R. R. and F. J. Rohlf. 1981. Biometry: the principles and practice of statistics in biological research, 2nd ed. Freeman, San Francisco.

SPSS Inc. Advanced statistics guide. 2nd Edition. pp. 527. Chicago n.

Sugihara, G. 1980. Minimal community structure: an explanation of species abundance patterns. Amer. Nat. 116:770-787.

Ward, J. V. 1992. Aquatic insect ecology. 1. Biology and habitat. Wiley and Sons, Inc. New York.

Wentworth, C. K. 1922. A scale of grade and class terms for cladistic sediments. J. Geol. 30:377-392.

Williams, D. D. 1980. Some relationships between stream benthos and substrate heterogeneity. Limnol. Oceanogr. 25: 166-172.

86