Documente Academic
Documente Profesional
Documente Cultură
Michael Lynch
Stable URL:
http://links.jstor.org/sici?sici=0012-9658%28197821%2959%3A3%3C552%3ACIBNCA%3E2.0.CO%3B2-Z
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained
prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in
the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/journals/esa.html.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academic
journals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers,
and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community take
advantage of advances in technology. For more information regarding JSTOR, please contact support@jstor.org.
http://www.jstor.org
Thu Aug 30 12:15:28 2007
E < o b ~ ?' .h 7 ) 1978. pp. 552-564
i 1978 b) the E c o l o g ~ c aSoc~etb
l of A r n e r ~ c a
MICHAELL Y N C H
Department o f Ecology. Ethology, and Evolution, University of Illinois,
Abstract. This study concerns a seasonal succession of a large herbivore, Daphnia pule.^, by a
smaller one. Crrioduphniu reticulata, in a Minnesota pond. Based on a description of events in the
pond and results of several field experiments, one can explain this replacement in terms of a coex-
ploitative interaction between these 2 species. The field experiments. in which pairs of cladocerans
were isolated for 4 wk, gave several results which are inconsistent with the assumptions of contem-
porary competition theory. In the 1st experiment. Daphnia survived when solitary but went extinct
when paired with Cerioduphnia, regardless of starting densities or age structure. At the same. Daphnia
was declining in the pond while Ceriotluphniu was increasing. Four wk later, Duphniu coexisted in
pairs with both Cerioduphnia and B o ~ m i n ulongiro.stris, and at much greater densities than in its
controls. Not only did the competitive abilities of these species shift over this very short time period.
but the densities which they attained in control enclosures dramatically declined. Thus, static mea-
sures of competition (such as alphas) and carrying capacities are inappropriate for these species.
Furthermore. these interactions can be understood only in the context of a framework which considers
age structure. Generally, juveniles were more sensitive to coexploitative interactions while adults
were unaffected and sometimes even benefited. Since their resources interact, coexploitative inter-
actions between these species are not simply a result of resource depression but of resource alteration.
Despite the fact that they overlap extensively in resource utilization, an appropriate modification of
the resource base by I species can lead to enhanced conditions for a 2nd. I suggest that such facili-
tation may be an important process in these communities which merits further investigation.
Key words: age-specijic inteructions; Ceriodaphnia; coexploituti~~e interactions; community le\'c.l
evolrrtion; community ~trrrctrrre;competition; Crrr~tuceu;Daphnia;facilitation; intercrcting resources;
.Minnesota; :ooplunkton.
quently attributed to competition (Cody and Diamond predict such things as the limiting similarity of species,
1975). However, hardly any of the myriad of studies invasibility of habitats, and evolutionary convergence
mechanism, i.e. the relation between resources and I emphasize 3 shortcomings of contemporary com-
the coexploiters' population processes (behavior and petition theory. ( 1 ) Conclusions drawn from this the-
demography). In the absence of sufficient information ory are strictly dependent on the linearity assumptions
on the biology of this interaction, contemporary com- of the Lotka-Volterra equations. In the past, most in-
petition theory relies heavily on a simple set of math- vestigators have set aside this caveat, stating that the
ematical constructs believed to mimic the outcome of equations should be suitable for communities close to
competition. equilibrium. However, as Armstrong and McGehee
Almost all of these formulations follow the Mac- (1976) pointed out, it may be entirely inappropriate to
Arthur-Levins extrapolations (MacArthur 1970, 1972, analyze communities in terms of point equilibria. (2)
MacArthur and Levins 1964, 1967) of Volterra's (1928) Models of interacting populations have traditionally
equations (but see Grenney et al. 1973, Smith et al. ignored age structure. Neil1 (19754) demonstrated with
1975, Titman 1977). Overlap in resource utilization is laboratory populations of cladocerans that different
usually equated with competition, and several esti- age classes are not equally sensitive to resource alter-
mates of "competition coefficients" based on utiliza- ation, and that the outcome of competition between
tion functions have been proposed (Levins 1968, 2 species may be based on a few critical age classes.
MacArthur 1972, Schoener 1974, May 1975) and ap- (3) The theory is probably least appropriate for spe-
plied to natural communities (Culver 1974, Lane 1975, cies whose resources interact. Most studies assume a
constant resource base or simply address the problem
' Manuscript received 1 l August 1977; accepted 6 March of resource depression (MacArthur 1972). However,
1978. since coexploitative interactions also often lead to a
' Contribution No. 170 from the Limnological Research qualitative shift in resources, it is necessary to under-
Center, University of Minnesota. stand the implications of resource alteration as well as
Late Spring 1978 CLADOCERAN INTERACTIONS 553
resource depression. As I will demonstrate below, this ethylene bags suspended in the pond were closed at
is a particularly serious problem because coexploita- the bottom and excluded all potential predators (the
tion of interacting resources may actually lead to im- salamander, Necturrrs, some walleye (Stizostedion \.it-
proved conditions for some exploiters. reutn) fry. and many arthropods). These enclosures
The relevance of each of these problems became were filled with surface water so that zooplankton den-
apparent in a study I did on the interactions between sities were approximately equal to those in the pond.
3 species of filter-feeding cladocerans, L)(rphnicr prrle.u, In both the control enclosure and in 2 enclosures con-
Ceriodophnia retic,ulata and Bostninrr longirosrris, in taining an adult Necrrtrrr.7 (gut analyses indicated that
a small Minnesota pond. Since the importance of com- they did not consume zooplankton), Ceriodrrphnicl re-
petition in structuring natural zooplankton communi- mained numerically dominant over Dophnia through-
ties has recently been questioned (Dodson et al. 1976). out the 4-wk experiment. The average densities in all
the results I present are also of special significance in 3 enclosures across all sampling dates were 2.0 Daph-
their implications for plankton community theory. nirrilitre and 11.8 Ceriod~lphnicrilitre.Dnphnin re-
mained below a density of 3 individualsilitre except on
one occasion in 1 enclosure; this corresponds closely
to the density of Dophnicr in the pond at that time.
Pleasant Pond lies in a small basin (50.25 ha, max- Since all other zooplankton species in the pond were
imum depth = 2 m) located =16.1 km north of St. rare during this period (Fig. I), it is unlikely that they
Paul, Minnesota. It has no inlet o r outlet. The pond were involved in the replacement of Daphnicr by
was divided in half with polyethylene sheeting in 1976 Ceriodrrphnia. This leaves the possibility that a direct
as part of a large-scale fish manipulation experiment interaction between these 2 species was responsible
(Lynch 1978). The studies I consider here are confined for their succession. However, before invoking ex-
to the northern (control) half of the pond where ver- ploitative competition as an explanation for this re-
tebrate planktivores were rare. placement, it is necessary to demonstrate ( 1 ) the ef-
fects which these species have on their resource base,
and (2) the significance of such resource alteration to
Zooplankton was sampled weekly by taking dupli- the coexploiting populations.
cate vertical hauls through the whole water column
Alteration oj'thr resortrce h ~ l s e
using a Wisconsin net with a 15-cm diameter opening
and 64 p m netting. Samples were anesthetized with In Pleasant Pond North, grazing by both D~lphnirr
carbonated water before fixing to prevent evacuation and Ceriod~lphnirrhas a significant impact on the abun-
of the guts, and then preserved with formalin and su- dance of phytoplankton. As shown in Fig. 2. several
crose (Haney and Hall 1973) to prevent loss of eggs. algal species declined to below detectable levels in
At least 2 I-ml subsamples were counted in their en- mid-June when D~phnicr became abundant; they did
tirety under lOOx magnification to determine zooplank- not reappear until August after D ~ p h n i ~declined.
l
ton abundance. Size-frequency distributions were ob- Several other species, which were never noted early
tained for the cladoceran populations by measuring in the summer, appeared immediatelv fnllowing the
50 random individuals from the anterior margin of the decline of D ~ p h n i ~ ~ .
head to the base of the tail spine. Measurements Associated with the dominance of D ~ p h n i ain early
were made to the nearest .01 mm using an ocular summer was an intense bloom of the filamentous blue-
micrometer. Mean clutch sizes were determined for green Aphanizornenon jos-aqurre. In Pleasant Pond
size classes in which at least 3 separate measurements (and other Minnesota ponds which have large h ~ p h -
were made. nirr) this species forms large colonies (up to 2 x 15
Phytoplankton was sampled weekly at 0.3 m depth mm) which cannot be ingested. I have evidence which
and preserved with Lugol's solution. suggests that Dophnicr actually maintain these bloom5
by removing all potential algal competitors (Lynch,
The Daphnia-Ceriodaphnia succession
person~lloh5 er\,crrion).
The seasonal fluctuations of the zooplankton species Division of the phytoplankton into size classes re-
in Pleasant Pond North are illustrated in Fig. 1. The veals a general pattern which is related to the relative
most dramatic event in 1976 was the replacement of abundance of D ~ l p h n i ~and
l Ceriodaphnirr (Fig. 3).
Drlphnia pules by the smaller Ceriodrrphni~lrericulrrtcr Bacteria and 4-63 p m 3 particles (primarily small fla-
at the end of July. The rarity of Ceriodcrphni~lin early gellates) declined to very low densities just before the
summer was most likely a result of Chaohorus (midge D ~ l p h n i ~population
l peaked, and remained low after
larvae) predation (Lynch 1978). However, predators Ceriodrlphnirr became dominant. No significant trend
were not responsible for D ~ p h n i a ' smidsummer de- is apparent with the next largest size class (consisting
cline. of a variety of flagellates, diatoms, green and blue-
An enclosure experiment which I initiated on 3 Au- green algae). The largest size class (1,024-16,383 p m j
gust substantiates this. Several I-m diameter poly- particles) consists of desmids, trachelomonads and
Late Spring 1978 CLADOCERAN INTERACTIONS 555
O '[.-
A_ _ _ _ _ -&
-- - ]'0 dphonizamenon flos-oquoe
~ ~ c i ~ ~ o ti eonrui io~
lo T~ochelarnonorvo~ionr
'[,-"-
-- -
A.&-_--]~~
--
Chroococcus d i r p e ~ s u s
N,bsch,o Sp
.
F
02
-~----------
[--
--..-----.-._-- &
chiorn~dornonors p
a G i o e o c y ~ t ivesicoioso
l0
?\ 02
6 C......... ........& Heferochromoflos qioboso
Qq ~osrnoriurnsexnototurn
? I C.
=t
.......--.-&-. .
...
.
'0
0
'
2-
___--I2 ~ i o t h r i xosciiiormo
04
:Lfi
__-_-A
--_- ]04 Ochrornonor voiiesioco I
I
08
C. ....................... A sphoerocyslis schroeteri
1 1024 - 16,383 pm3
_M,
, *----
dcm
I6 socterlo
I Particles
1 I 1
I I
FIG.'2 . Major phytoplankton species in Pleasant Pond
North. 1976.
0
obtain a full complement of the large herbivores when FIG.3. Abundance of 4 size classes of bacteria and phy-
toplankton capable of being ingested by Drrphnirr and Cf,rio-
filling the control chambers; these were generally most daphnia. Vertical dashed line denotes date of maximum den-
abundant at a depth of .= 1 m. Two differences between sity of Duphniu after which it sharply declined and was
treatments were consistent in all experiments. Chroo- replaced by Ceriodaphnia.
coccus mitlut~ts( a small coccoid blue-green) was al-
ways most abundant in chambers with zooplankton,
supporting Porter's (1976) contention that such gelat- I analyzed gut contents by squashing individuals
inous forms may accrue a competitive advantage by gently under a cover slip and examining the extruded
absorbing nutrients as they viably pass through her- bolus. Gut contents consisted largely of a homoge-
bivore guts. Bacteria were consistently lowest in neous brown material of unknown origin (bacteria,
chambers with zooplankton. detritus. o r partially digested algae). Those phyto-
The density of zooplankton in the pond was very plankton cells which were intact were identified and
low during experiment 2 (29 June-3 July), and few enumerated. and notes were made regarding remnants
additional differences were noted between treatments. of broken and partially digested cells. Although bac-
However, during experiment 3 (18 July-22 July), the teria and detritus were likely to have been important
Dapht1ia population was at its peak and most phyto- components of the diet, they could not be quantified.
plankton species increased upon removal of zooplank- The proportions of intact algal types in the foreguts of
ton. Most significant was the dramatic increase of a juveniles and adults of both species are presented in
tiny species of Srlrr~u.~trirm w hich occurred upon re- Table 1. All of these algae, except Truchelornot1rt.r and
moval of zooplankton; this species was only rarely the gelatinous forms. were frequently found partially
noted in the pond. Daphtlia were scarce during ex- digested. suggesting that they are important compo-
periment 4 (19 August-26 August), while Ceriodaph- nents of the food supply.
niii were at their maximum density. With the excep- All sized individuals of both species were capable
tion of the gelatinous forms, all species increased in of ingesting the whole spectrum of algae found in the
the absence of herbivores. guts; only the largest Cosrnari~trn species were not
Examination of the food ingested by Dapht1iu and found in the guts of immature Ceriodaphnia. In July,
Ceriodaphrlia verified that they share the same re- adult Ceriodaphr~iuand adult and juvenile Drtphnirt
source base, i.e., that they are potential competitors. had very similar diets, while immature Ceriodapht1ia
MICHAEL LYNCH Ecology. Vol. 59. No. 3
EXPERIMENT 2 (JUN 2 9 - JUL 3 , 1976) Thus, Driphniri and Ceriodriphnici share the same
' O O C N -----
02
---a-.-
Llphonizomenon flos-aquae resource base, and both are able to limit the abun-
L Q-.@ Characium omb~guum dance of these resources in Pleasant Pond. Both spe-
08
C- ---------- m%, Chroococcus minutus cies are capable of depressing the small food particles
08
C ----.
, m--a
-----. Heterochromonas globosa to very low levels. However. Cerioduphtliu cannot re-
02
02
C B! Monochrysis Sp duce the abundance of large phytoplankton to the ex-
I--- R-.m
--rr;s--..--- Ochromonas valles~aca tent that Daphtliu can.
4
,E. 6X.N Oscillatoria tenuis
-6
C.n..im..~-~ Rhodomonas minuto The mec,huni.t rn of' competition
02
C-N.-m-.rn--m!- Schroderia setigera The reduction in small food particles throughout the
08
CH . ..---. ,-: Trachelomonas varians summer was reflected in the amount of food observed
08
b..,- -
H.;-P.- Bacter~a in the guts of both species. On 8 July. the guts of all
individuals were visibly full. When the Driphnici pop-
EXPERIMENT 3 ( J U L 1 8 - 2 2 , 1976) ulation peaked on 22 July, the guts of adult Duphnici
8
F,. m.-rr;s Aohan~zomenonflos -aouae
were all full, whereas those of immatures averaged
I
02
C--- R-- Chroococcus minuta <ZW full; the guts of all Ceriodapht~iuwere < 2 5 9
L------- ------- Cryptomonas obovata full, and many were empty. On 13 August, after Cerio-
01
--------- 6Q-.m..m Gomphonemo Sp daphtliri had become abundant, guts of immature
04
C----. , Heterochromonas olobosa Daphtliri averaged 20% full, those of adult Duphtliu
I
-----my5 Nilzschio Sp 60%. and those of Cerioduphnici 15%. The Crriodaph-
08
02
[-- m-.m Oscillatoria tenuis niu population had become very dense on 20 August.
L - ~ - - -.----.
I z ~ m . . ~Rhodomonas minufa and the guts of all individuals of both species were
04
-,- 6i-B - Sch ,r o d e- r ~ as e t- near1
~ g e r a-- - y empty.
-
08
C Selenastrum ~p While the quantity of food ingested declined for both
I
C m!.--?rn Bacter~a species throughout the summer, Ceriodriphtlia were
apparently less sensitive to these low food levels as
EXPERIMENT 4 (AUG 19-26, 1976) evidenced by their high rate of population growth
2
[ ax,
- - .. Chroococcus mlnutus throughout this period (Fig. 1). High clutch sizes were
maintained in mid-July when Driphtliu were most
abundant and only began to decline in August after
2
C .--.-.._- -.:-- .YH Cosmar~umsexnotatum Ceriodriphtlia had reached high densities (Fig. 5).
rr;s H-H Gloeocystls vesiculosa Although Duphtlici appear to have had little effect
08
------.--.,-. m..m Heterochromonos alobosa on Crriodaphtliu, significant changes occurred in the
04 Daphtliri population when Ceriodciphtliri were abun-
L-----,-.--: - ---Y N ~ o c y s t i sI U C U S ~ ~ I S
04
4
----.. rn-.m-.H
..--
Scenedesmus denticulatus dant. Both the size at 1st reproduction and the maxi-
---m-.m Sphoerocystis schroeter~ mum adult size of Dapht~ia declined throughout the
2
- - - - a . Stourastrum p a r a d o ~ u m
summer (Fig. 5). This is characteristic of all food-lim-
ited populations of cladocerans that have been studied
(Ingle et al. 1937, Richman 1958, Hall 1964, Weglenska
1971). It also tends to equalize the number of eggs
carried by equal-sized (but different-aged) individuals
exposed to different nutritional conditions. Thus, the
relation of clutch size to body size does not vary great-
FIG. 4. Response of phytoplankton to zooplankton re- ly throughout the summer (Fig. 5).
moval experiments. First 3 columns are counts from pond
samples; last 2 columns are means of duplicate treatments. The most significant effect of food depression is its
Significant differences between columns are denoted by 1 influence on the survival of immature Daphniri. This
( P > . O O or 2 inverted diamonds (P > .01). Experiment conclusion arises from the observation that newborn
2: Daphnirr and Cerioduphniu rare: experiment 3: Daphnin Daphnici were often very rare on days when adults
abundant. Crriodaphniu rare; experiment 4: Dnphnia rare. were carrying many eggs (for instance, on 29 June and
Ceriodnphnin abundant.
13 August). It also follows from an index I constructed
ingested higher proportions of diatoms and delicate as a relative estimate of survival for newborn (0.6-0.8
flagellates. There was some difference in resource uti- mm) Driphnici.
lization between the 2 species in August. Cerioduph- The expected influx of newborns (N,,,) on day x
niu feeding more intensively on diatoms. Duphnici was calculated according to:
more o n delicate flagellates and Cosmariirm sps.
Nonetheless, these types were shared to some extent.
and all other categories were shared considerably.
Late Spring 1978 CLADOCERAN IN TERACrIONS
Ceriadaphnio reticulafa
where E,.., = average number of eggs carriedlfemale borns from ephippia (resting eggs). Since food levels
on day x - 1 . declined throughout the summer, it is likely that
? ,_ ,= population size of females on day x - 1 . growth rates also declined. causing the late summer
D,.., = duration of egg development on day ratios to be the most overestimated ones. Nonethe-
x - 1 ( a function of temperature, taken less. these ratios declined dramatically throughout the
from Knutson 1970). summer (Fig. 6).
Two additional points emphasize the vulnerability
Animals also leave this size class as they grow larger. of juvenile Dcipiznirr to competitive interactions with
Since the shortest period of growth to 0.8 mm is 1 day Ceriodciphn~a:
for Dciphnici pulex, then the minimum expected den- 1. The restriction of juvenile Daphrlia to feeding on
sity of 0.6-0.8 mm Daphnirr can be estimated with the same small particles which Crriodaplznirr use
equation 1. Dividing this estimate into the actual num- poses a developmental problem for Dciphnici since the
ber of 0.6-0.8 mm individuals on day x gives a relative smaller species are more efficient at gathering these
estimate of the survival of newborns. Values > 1 occur small particles (Gliwicz 1969. Neill 197%). Neill
both because of the maximum growth rate assumption ( 1 9 7 5 ~ f)ound in laboratory microcosms that survival
and also because of an unaccounted influx of new- and reproduction of adult Daphnici rncignu were un-
TABLE1. Average proportion of intact phytoplankton biomasq in the foreguts of Duphfricr and Ceriodaphniu for major
groups of algae. Immature Daphoicr < 1.4 mm long: immature Ceriodaphnia < .56 mm. Sample size denoted by N
O
"zr:
400 (mat. D.)
19 ,
-
; : : ; , /8[
(mat. D.)
Dl
?h
--- . -**o--.= //
10 D . 1 0 C
/
I,,_:
15 (mat. D)
0.4
Daphnlb pu/ex
Single
04r -
o- -0 Daphnia ( 3 , 5 )
I
Ceriodaphnio reficulata
P, Single Food Size Classes
(pm3)
FIG. 9. Mean density of algae and bacteria in different
size classes for different treatments in competition experi-
ment 1 (14 August).
T A B L E2. Mean final densities ( 2 . 9 5 confidence limits) of cladocerans in the 2 competition experiments. Numbers in
parentheses under D a p h n i a + Ceriodaphnia in Experiment 1 are the final densities in the one chamber in which the 2
coexisted. Number of replicates denoted by N
EXPERlMENT l
( 14
.4ugujt)
'V = 3 3 9
(193)
EXPERIMENT 2
( 10 September)
h'= 3 3 2 2 3 3 4
Dcrpl~nia 19.0 -+ 10.8 48.5 -+ 14 38.3 -+ 15.4 17.5 -+ 6.7
C(,riod(~pl~nici 262.0 t 40.2 61.5 -+ 16 137.0 -+ 29.1 2.8 -+ 2.6
Bosmintr 202.0 2 127.6 6.3 -t 6.2 89.7 -+ 23.5 12.2 -+ 5.6
very different impacts on the food supply (Fig. 9). sults. Both Daphnia and Ceriodaphnia persisted
Very small particles were rare in all chambers. Large alone, but at considerably lower densities than in the
particles were scarce in all chambers containing Daph- 1st experiment (Table 2). Furthermore, Daphnia not
nia, but were abundant in those containing only Cerio- only survived when enclosed with Ceriodaphnia, but
daphnia. These large particles remained rare even af- it reached densities twice those in the Daphnia con-
ter Daphnia had gone extinct, and were scarcest in trols; in these same enclosures Ceriodaphnia declined
the one chamber in which Daphnia and Ceriodaphnia significantly relative to their controls. Similarly, when
coexisted. Thus, it appears that together the 2 species Daphnia and Bosmina were paired, Daphnia attained
depressed the food supply to a lower density than twice their control densities, while Bosnlina nearly
either could d o alone-to the point at which Cerio- went extinct. Densities of all 3 species remained low
daphtlia outlasted Daphnia. when they were enclosed together, although the abun-
dance of Daphnia was not significantly different from
E.xperiment 2 its control density.
The 2nd experiment was initiated immediately fol- The enhancement of conditions for Daphnia in the
lowing the 1st but yielded significantly different re- presence of either Ceriodaphnia or Bosmina in this
T A B L E3. Mean and range of major food groups within treatments in Experiment 2, 10 September a s lo5 CLm:l/ml.Flagellates
include all such species without hard outer walls: Chlorococcales include Scer~edesrnlisand Pedicrstrrm~species: desmids
include Cosmaririrn species and S t a ~ r r u s t r ~ i rpna r u o ' o x ~ i m :diatoms include Coc,coneis sp., Cyrnhellu sp.. Gornphot~rmcr
sp.. ,Vcr~.iclilasp. and Nitzschia sp.: gelatinous forms are Chrooc~occrisdisperslis, Gloeocystis ~ ~ r s i c . r i l o sand
a Spl~arroc~.stis
.scl~roetrri
Mean and range of major food groups (10" p d i m l ) on final sampling date, Experiment 2
TREATMENT
Taxon D C B D + C D + B C +B D + C +B Control
Flagellates .58
(0-.91)
Chlorococcales 1.02
(.52-1.45)
Desm~ds 9.84
(4.02-15.92)
D~atoms 2.44
(1.563.76)
Gelat~nousForms 1.51
(.40-2.10)
Bacter~a 1.04
( . ? G I .89)
Late Spring 1978 CLADOCERAN INTERACTIONS 6 1
,
arose for Ceriodaphnia populations under different
treatments (Fig. 10). The reduced population size in
+ Bosrnino ( N = 104 ) + Ceriodaphnia +Bosrnlna the presence of Daphnia appears to be largely a result
2 I 2 ~ : , :N=89) of the reduced clutch sizes. However, the few Cerio-
0.2 Q, 0.2 oo daphnia which were present in chambers containing
0
all 3 species were carrying large clutches. The con-
spicuous absence of juveniles under this treatment
again emphasizes the importance of juvenile mortality
in maintaining these populations at low densities.
Finally, the density of Bosmina was significantly
depressed by the presence of other species in all cases.
Cerlbdaphnlb reticulafa Its reproductive capacity was much more severely af-
I Solitary (N=143) + Daphnia (N=85)
fected in paired situations than was that of Daphnia
P, 2 r 2r
or Ceriodaphnia.
The precise mechanisms leading to the outcome of
this experiment are less clear than for the 1st experi-
ment. However, it is clear that the impact of the her-
bivores on their food supply in this experiment was
+ Bosmlna (N=142) + Daphnia + Bosrnina one of resource alteration rather than simple depres-
2 1 , ;o (Y=,12) sion. Although the food-size distribution was very
similar in all treatments on the final sampling date (Fig.
l l ) , there were significant differences in the abun-
dances of specific food groups (Table 3).
The enhancement of conditions for Daphnia when
paired with Ceriodaphnia o r with Bosmina was asso-
ciated with an increase in the abundance of flagellates,
Chlorococcales and desmids and a decrease in bacteria
Solitary (N-91) and gelatinous forms relative to Daphnia controls.
These algae are significant contributors KO the food
supply of Daphnia, and could explain the increased
clutch sizes and population densities for Daphtlia in
0.2 paired chambers. Flagellates, Chlorococcales, and
desmids were also more abundant in the 3 species
treatments where Daphnia clutch sizes were high, yet
+ Ceriodaphnio ( N = 9 1) + Daphnia + Ceriodaphnia the data are not sufficient to explain the much lower
0,4~[N~001 , 0,4~5[oO~o I N=50) abundance of Daphnirc in these chambers.
Significant shifts in the quality of the food supply
also occurred between treatments with Ceriodaphnia
(Table 3). Bacteria and diatoms, which are significant
components of Ceriodaphnia's diet (Gophen et al.
1974, Smyly and Collins 1975), were lower in the
Carapace Length ( m m ) Daphtlia + Ceriodaphnia treatments than in Cerio-
FIG. 10. Size-frequency distributions and mean clutch dapht~irccontrols and may be responsible for reduced
sizes for individuals of different sizes in experiment 2 (pooled clutch sizes in these chambers. Abundances of most
data from replicates). Abscissas of insets start at carapace
lengths directly below their ordinates. Sample sizes for size- +
food types in Ceriodaphtlia B o s m i t l ~and Daphnirc
frequency distributions are denoted by N. Mean clutch sizes + Ceriodaphnia + Bosmitla treatments were nearly
are for sample sizes 3 3 , and usually between 10 and 30. equal to those in the Ceriodrcphnia controls, and clutch
sizes approximated those in the controls. Again, this
experiment is further supported by the larger clutches information is not sufficient to explain the lower num-
carried by females in paired chambers (Fig. 10). N o bers of Ceriodaphtlia (a probable result of reduced
Daphnia were carrying eggs in the controls at the end juvenile survival) in the paired chambers.
of the experiment. However, despite the fact that fe- The lack of explanation for several population char-
562 MICHAEL L Y N C H Ecology. Vol. 59, No. 3
Daphnia and Ceriodaphnia involves several complex- the removal of 1 prey species often results in quanti-
ities which are not generally appreciated in theoretical tative, qualitative or distributional changes in other
investigations of competition: (1) seasonality, (2) age prey (Brooks and Dodson 1965, Paine 1966, Harper
structure, and (3) resource interactions. 1969). Such results occur when the presence of a pred-
I . There was a shift in the competitive abilities of ator alleviates competition between prey species by
Daphnia and Ceriodaphnia over a very short time pe- removing the competitive dominant. In those cases,
riod. Over the same period there was a dramatic de- coexploitative interactions are more accurately de-
cline in the densities to which each species could grow scribed in terms of resource alteration than simple
in the absence of competitors. Such disparities be- depression.
tween experiments show the artificiality of considering When the restriction of resource independence is
"competition coefficients" and "carrying capacities" relaxed, the predictions drawn from various compe-
as time invariant. tition formulations may be altered significantly (Lev-
The shift in the competitive ability of Daphnia be- ine 1976). Under such conditions it becomes theoret-
tween experiments may be related to a seasonal de- ically possible for one species to alter the resource
cline in water temperature. Daphnia pulex has a max- base in a way that directly benefits a 2nd coexploiter
imum feeding efficiency (food intake divided by (e.g. Charnov, Orians and Hyatt, 1976). The results of
metabolic demands) at a temperature of =20C; feed- these experiments clearly demonstrate the potential
ing efficiency drops off very sharply above this tem- for such facilitation to occur in natural communities.
perature, declining to very low values at 25'C (Lynch The simple depression of all resources in the 1st
1977). Temperatures during the 1st experiment often experiment may have been related to the relatively
exceeded X C , and Daphnia always went extinct high levels of nutrients present then (Lynch and Sha-
when paired with Ceriodaphnia. Temperatures had piro, personal observation). If the algae in the Daph-
dropped below 25C by the 2nd experiment and ap- nia coritrol chambers were not competitively con-
proached 20C on 6 September; Daphnia increased at strained, the activity of a 2nd grazer would be unlikely
the expense of Ceriodaphnia in this experiment. to benefit any alga. The fact that almost all algal spe-
2. Because different age classes are not equally cies increased in Ceriodaphnia controls relative to
prone to coexploitative interactions, the interactions Daphnia controls provides further evidence that these
of these species can be understood only in the context algae were not tightly constrained by nutrients.
of a framework which considers age structure. While Phosphorus (dissolved and total) and dissolved in-
certain age classes were severely affected by resource organic nitrogen (ammonia, nitrate, nitrite) levels de-
alteration, others were unaffected and sometimes even clined very significantly in August, resulting in a sharp
benefited. The competitive displacement of most spe- decline in species specific algal productivities (Lynch
cies in the experiments appeared to be not s o much a and Shapiro, personal observation). These changes
result of any adverse conditions for adults as a con- may be responsible for the decline in the carrying ca-
sequence of reduced juvenile survival. pacities for Daphnia and Ceriodaphnia in the 2nd ex-
3. The frequent enhancement of conditions for cer- periment. This also suggests that the algae were more
tain age classes in the presence of coexploiters further competitively constrained than in the 1st experiment,
emphasizes the complexity of these interactions. and that some species would have benefited from the
When paired with Ceriodaphnia or Bosmina in the 2nd depression of their competitors, conditions necessary
experiment, Daphnia attained a density twice as great for the establishment of facilitation between 2 exploit-
as in their controls. However, an enhancement of con- ers.
ditions for adults is not sufficient for an increase in It is entirely plausible that the facilitation noted in
population size. In the presence of both Ceriodccphnia this study was simply a transient phenomenon result-
and to.smina in experiment 2, Daphnia females car- ing from environmental vagaries. However, such fa-
ried many more eggs than in Daphnia control cham- cilitation between species may in some cases have an
bers. but their final population levels were similar. evolutionary basis. There would be a distinct advan-
Thus, the improved conditions for adults in that case tage for an organism which benefits from the resource
must have been balanced by a reduced survival at an alteration mediated by a coexploiter.
earlier age. The results of this study emphasize the importance
Such enhancement effects mediated by coexploiters of studying species in the context of their natural en-
are not consistent with our present concept of com- vironment. While facilitation appears to be widespread
petition. The assumption that resources are indepen- in nature ( M . Lynch, personal observation), it has not
dent of each other prevents a direct enhancement of generally arisen in laboratory investigations, probably
conditions for one species by a coexploiter from aris- due to the simplified communities used in such studies.
ing in theoretical investigations. Thus, an assumption Consequently, contemporary ecological theory em-
implicit in most studies of competition is that ex- phasizes the selective nature of negative interactions
ploitative interactions are based sloelq on resource between species-competition and predation. 1 stress
depression. However, many studies have shown that that our present understanding of community structure
564 MICHAEL LYNCH Ecology. Vol. 59. No. 3