The Effect of Grazing on Marine Littoral Diatom Populations

Author(s): Richard W. Castenholz

Source: Ecology , Oct., 1961, Vol. 42, No. 4 (Oct., 1961), pp. 783-794
Published by: Wiley on behalf of the Ecological Society of America

Stable URL: https://www.jstor.org/stable/1933507

REFERENCES
Linked references are available on JSTOR for this article:
https://www.jstor.org/stable/1933507?seq=1&cid=pdf-
reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Ecological Society of America and Wiley are collaborating with JSTOR to digitize, preserve and
extend access to Ecology

This content downloaded from

91.203.32.150 on Wed, 12 Jun 2024 06:57:18 +00:00
All use subject to https://about.jstor.org/terms
Autumn 1961 TIIE EFFECT OF GRAZING ON MARINE LITTORAL DIATO.NM POPUL.ATIONS 783

tatum i and A. dese rtorUm-14 and the other seeded in somie Ar-tcfm-isiai types oI northern Harniey apld
Lake counties, Oregon. 1-h.D. thesis. Oregon State
species reacted differently on each site.
College, Corvallis.
The restlits have several implicatiolns to ranige Gardner, R. A. and J. L. Retzer. 1949. Interpretive soil
maniagement. A. cristatum and A. desertorutm classification. Soil Sci. 67: 151-157.
are better adapted to the sites described than are Hay, W. D. 1939. Identification of stalndard and fair-
ainy of the other species seeded. Also, since A. way strains of crested wheatgrass. Agron. J. 31
620-624.
cristati-ou-i 1iow dominates the lower site, apparently
Heinrichs, D. H. and J. L. Botlon. 1950. Studies on
it is better adapted to more miiesic conditionis thani the competition of crestecl wheatgrass with perenniial
is 1. dcsertorunib and is mnore competitive than native species. Sci. Agr. 30: 428-443.
A. dcser-torit-1z on a miiesic site. Conlversely, since Kittock, D. L. and J. K. Patterson. 1959. Measure-
meint of relative root penetrationi of grass seedlings.
A. dcscrtoront now dominates the more xeric
Agron. J. 51: 512.
tipper site, apparelitly it is more drotnght tolerant Kittridge, J., Jr. 1927. The use of soil surveys in for-
and iimore competitive thain 1i. cristatuint oni a dry estry. Proc. of the First Internat. Cong. of Soil Sci.
site. Since most seedinlg in the IJltermotnlitain 5: 562-565.

W\'est is (lolne oni the more xeric soils, A. deser- Knowles, R. P. 1955. A study of variahility in crested
wheatgrass. Canadian J. Bot. 33: 534-546.
torrumt would appear to be the species of choice Konstantinov, P. N. 1922. Zlhitlnyak i ego kultura na
from the standpoint of adaptability and life of yugovostoke Evropeiskoi RZossii. Plokrovsk, Zlhurnal
stand. Unsere WlVirtschaft. 66 p. (Wheatgrass and its culti-
vation in Southeastern European Russia. Translated
ACKNOWLEDGMENT
by Lowell L. Jones, Univ. of Nev., Reno, Nev.).
The authors are indebted to Douglas Dewey, Agri-
--. 1923. Zhitnyak (Wheatgrass). Krassnokutsk
cultural Research Service, Logan, Utah, for making the
Agr. Expt. Sta. Moscow, New Lands. 68 p. Trans-
chromosome counts.
lated by T. K. Pavlychenko.
REFERENCES Peto, F. H. 1929. Chromosome numbers in the Agro-
Billings, W. D. 1945. The plant associations of the pyronis. N'ature 124: 181-182.
Carson Desert region, western Nevada. Butler Univ. 1930. Cytological studies in the genus Agro-
Bot. Studies 7: 89-123. pvron. Caniadiani J. of Res. 3: 428-448.
. 1949. The shadscale vegetation zone of Nevada Poulton, C. E. 1955. Ecology of the non-forested
and eastern California in relation to climate and soil. vegetation in Umatilla and Morrow counties, Oregon.
Am. Midland Naturalist 42: 87-109. Ph.D. thesis. State College of Wash., Pullman,
Bouyoucos, G. J. 1951. A recalibration of the hy- Wash.
drometer method for making mechanical analyses of Sarkar, P. 1956. The crested wheatgrass complex.
soils. Agron. J. 43: 424-438. Canadialn J. Bot. 34: 328-345.
Clements, F. E. 1928. Plant succession and indicators. Salinity Laboratory Staff. 1954. Diagnosis and im-
Carnegie Institutioni of Washington. H. W. Wilson provement of saline and alkali soils. Agr. Handbook
Company, New York. No. 60. USDA. 160 p.
Daubenmire, R. F. 1942. Vegetation of southeastern Spillsbury, R. H. and E. W. Tisdale. 1944. Soil-plant
Washington and adjacent Idaho. Ecological Monog. relationships and vertical zonation in the southern
12: 53-80. interior of British Columbia. Sci. Agr. 24: 395-436.
. 1952. Forest vegetation of northern Idaho and Tisdale, E. W. 1947. The grasslands of the southern
adjacent Washington and its bearing on vegetation interior of British Columbia. Ecology 28: 346-382.
classificaion. Ecological Monog. 22: 301-330. Weislander, A. E. and R. E. Storie. 1952. The vege-
Dyksterhuis, E. J. 1951. Use of ecology on range tation-soil survey in California and its use in the
land. J. Ranige Management 4: 319-322. management of wild lands for yield of timber, forage
Eckert, R. E., Jr. 1957. Vegetationi-soil relationiships and water. J. Forestry 50: 521-526.

THE EFFECT OF GRAZING ON MARINE LITTORAL DIATOM POPULATIONS'

RICHARD WV. CASTENHOLZ

Deparitennt of Biology, Un-ivtersity of Oregon, Eugene

I NTRODUCTION diatoms form on the rock substrate of littoral

During the colder seasons on the southern slopes and tide pools. During summer these
Oregon coast dark-brown carpets of attached covers seldom appear except in the sublittoral.
1 The author is very grateful to Mr. Warren E. Wilson Frequently "bare" areas of various sizes occur in
for excellent field and laboratory assistance. The author the upper half of the littoral zone and in mid-
would like to thank Dr. J. Connell for suggestions per- level and high-level tide pools. Th-ese areas lack
taining to closure construction. The study was sup-
visible algal covers but are usually densely popu-
ported by Grant G-9057 from the National Science Foun-
dation. The facilities of the Oregon Institute of Marine
lated by herbivorous littorines and limpets. Such
Biology were used through most of the study. undisturbed intertidal surfaces are almost devoid

This content downloaded from

91.203.32.150 on Wed, 12 Jun 2024 06:57:18 +00:00
All use subject to https://about.jstor.org/terms
784 RICHARD W. CASTENHOLZ Ecology, Vol. 42, No. 4

of a measurable diatom population. During the

summers of 1958-60, near Cape Arago (Lat. 430
20' N), a dark diatom slick developed rapidly on
most of these areas and on most other littoral
substrate after all organisms had been removed
by scraping and sterilization. The main purpose
of the study in 1960 was to demonstrate the prime
importance of grazers in maintaining this anom-
alous summer situation of high diatom reproduc-
tion rate and low standing crop.
This study attempted to measure quantitatively
. a *.

the diatom crop which developed in ungrazed

areas and in areas where the number of grazers
was controlled. In both cases screen closures 'S
were constructed for the experimental plots. This
was part of a longer study of seasonal production
changes in marine littoral and sublittoral diatoms.
The effect of grazing on sessile microvegetation
has rarely been studied, either in marine or fresh-
water situations. Douglas (1958) noted an inverse A~~~A
relationship between the number of caddis larvae
and numbers of the freshwater diatom Achnan-
thes. Brook (1952, 1954, 1955) showed that
seasonal reduction of the freshwater diatom popu-
lation on sand filter beds and macrophytes may be
caused by the grazing activities of various insect
larvae and even ciliate protozoans.
Moore (1938) and others described the ability
FIG. 1. Artificial tide pools carved in a sandstone shelf
of limpets to clear a "feeding" patch in an algal
at Yoakam Point. a. View of the 6 pools and the sur-
turf. Fischer-Piette (1948) studied the range rounding "bare" area. Pools are numbered 1-6 from
and "prosperity" of Patella in relation to the right to left; Pools 5 and 6 are screened. Trough-like
abundance of algae. Test (1945) commented on natural pools may be seen in the foreground. Rhodomela
the feeding habits of West coast species of "turf" and Mytilus beds appear in the background. b.
Pools 6 and 5 on July 25, after removal of screens (Expt.
Acmaea. North (1954) studied the erosive ac- 2, Table 1). Note the bare rock with littorines well dis-
tivities and metabolic efficiency of West coast persed in Pool 6. The density was about 7.1/dM2. Pool
Littorina but did not investigate the sources of5, the control, is completely covered by a diatom slick
food. about 3 mm in thickness..

STUDY AREA
The study area is Yoakam Point, which ex- shallow tide pools (usually less than 10 cm deep)
were scattered amid mussel beds and Rhodomela
tends northwestward as intertidal spurs alter-
nating with long slot-like troughs. The sandstone "turf." Odonthalia, Prionitis and Cladophora
promontory is less than one mile southwest of the were the most abundant algae in the pools. The
"bare" rock between plants were dominated by
Coos Bay entrance. The work area is a broad
Littorina scltulata and 6Acmaea spp. Mytilus
flat shelf at about the 6-8 ft tidal level and a nearly
was commol a in many pools. The lip of the
vertical wall extending from the +8 to -2 ft
levels.
shelf was exposed to severe wave action during
The Yoakam shelf was half covered by beds of most high tides; the broad flat top usually re-
ceived only the wash.
Mytilus californianus. The inner half was large-
The vertical wall used at Yoakam Point faces
ly covered by the red alga Rhodomela larix and
northeast and is exposed to the heavy wave action
some Cladophora. These algae were confined to
moist depressions and conspicuous "bare" areas of combers running parallel to the wall. The
separated the algal "turfs" (Fig. la). The "bare" zonation pattern is shown in Figure 2. Natural
areas were usually populated with Littorina scutu- "bare" areas inhabited mainly by A. digitalis,
lata and small individuals of Acmaea spp. The Balanus, and tufts of Endocladia lie above the
littorines were generally clustered in depressions mussel bed from the 7-8 ft level and higher.
or cracks except on overcast days. Numerous Directly below the Mytilus is a narrow fringe of

This content downloaded from

91.203.32.150 on Wed, 12 Jun 2024 06:57:18 +00:00
All use subject to https://about.jstor.org/terms
Autumn 1961 THE EFFECT OF GRAZING ON MARINE LITTORAL DIATOM POPULATIONS 785

erally covered by a crust of Lithothamnion, these

areas were often clothed by patchy or continuous
carpets of attached diatoms at any season.
Navicula (Schizonemca) ramosissima Ag., N.
(S.) grevillei Ag. and Licmophora abbreviata Ag.
x r were generally the dominant species. Relatively
few herbivorous gastropods were present at
these lower levels. A. mitra was regular but
Inot common. Diatoms were also abundant as
epiphytes in the lower littoral most of the year
in some regions between Yoakam Point and Cape
Arago. Isthmia, Coscinodiscus, and Biddulphia
often completely clothed the branching red algae
Rhodomela and Odonthalia when the latter were
located in flat beds protected from heavy wave
force. Very few grazers were present on these
beds.

METHODS

Tide Pools
Six artificial tide pools, approximately 24 x 24
x 4.5 cm with a capacity of about 2.2 liters, were
carved in the sandstone shelf at about the 6.5-7.0
ft tidal level, above the Mytilus zone (Fig. la).
The pools were made with hammer and cold
chisel; the bottom was smoothed with a wide-
blade wood chisel. Eight to 10 holes (d4 in.
diam) were drilled around the lip of each pool.
FIG. 2. The vertical transect at Yoakam Point. a.
The bare transect on June 23, 1960, soon after mounting Plastic screw fasteners were set in each, and a
the baskets. The upper plots (U) and lower plots (L) stainless steel screen (8 mesh sq. in.) was secured
are numbered 1-4 from left to right. The approximate over each pool by nickel-plated wood screws with
tide levels are indicated in feet above O tide mark. Some
thick rubber faucet washers. Each screen was
of the natural "bare" areas are marked with an x. The
area marked with a Y is part of a vertical strip cleared molded around the edges with a light hammer to
more than a year before. b. The transect 16 days later. seal all openings. The mesh was fine enough to
The dark diatom cover has become apparent from the exclude almost all littorines. The screens could
base to the top of the upper baskets. Note the cover on be rapidly removed and reset using the same
the lower baskets. Two 6 x 6 inch squares have been
cleared below these baskets. The bare streak on the
mounts.
right-hand side was caused by the abrasive action of algal Various numbers of Littorina scutulata were
fJronds. introduced into bare, sterilized pools; pools with-
out littorines served as controls. Sometimes lit-
Odonthalia followed from the 5-1 ft level by torines were introduced into pools that already
Iridaea, Hedophyllum, and Corallina, all super- had a diatom cover. All pools were screened
imposed on a crust of Lithothamnion. "Bare" and collections were usually made after 14 days.
areas which interrupt the mussel beds are com- At that time littorines were removed, counted
mon (Fig. 2), and smaller "bare" patches often and their displacement volume determined. In
occur in the Iridaea-Hedophyllum area. A4. most cases the dry weight was also determined.
digitalis dominates the "bare" areas above the 4 Any diatom cover was removed with a fine edge
or 5 ft level. A4. pelta and A. scutumn are also carpenter's scraper after the pool was drained.
common from below the 6 ft level to as low as The remaining diatoms from bottom and sides
1 ft. Shotwell ( 1950) described the vertical were removed with absorbent cotton. The pools
zonation of Acmaea in a nearby cove on the were then resterilized with 95% ethanol or ab-
Oregon coast. solute methanol, rinsed, and reused.
Rock substrate not covered by macrophytes The abundance of diatoms was estimated by
was also common in the sublittoral zone and the determining the absorbance of a methanol extract
lowest parts of the littoral from Yoakam Point in a Beckman Model DU spectrophotometer at
to Cape Arago (about 4 miles). Although gen- 665 m,u. The absorption maximum of chlorophyll

This content downloaded from

91.203.32.150 on Wed, 12 Jun 2024 06:57:18 +00:00
All use subject to https://about.jstor.org/terms
786 RICHARD W. CASTENHOLZ Ecology, Vol. 42, No. 4

a in the red is close to 665 m,u. Chlorophyll c TABLE I. Tide pool experiments in which littorines were
is the onlly other pigment contributing to the introduced into bare, sterilized pools. All data are based
on 14 day periods. See text for additional explanation
absorption at this wavelengtlh. All samples were
frozen sooln after collectioni, stored in darkness Vol. of Dry wt.
and extracted later. Comparison of fresh and Absorbance No. of litt. litt. per of litt. per No. of No. of
Pool of diatoms per pool & dM2 dM2 (g) lit". per dM2 per
frozeln portions of the same material inidicated @ 665 my type (cm3) dM2 litt.

that freezing had no observable effect on the ab-

Expt. 1 Mid-date 6-30-60
sorption properties of the extract. Absolute
methanol with 0.5% dimetlhylanililne was used
6 as
1.29 0 - - -
the solvent (Strain aind Malnining 1942). Etlhanol 5 .80 65 "shelf" .2) - 8 3 .12

or acetone did not effect complete extraction.

The sample was extracted for 24-30 hr in dark- Exlpt. 2 Mid-date 7-16-60 _

ness. The material was always damp Nwhen the 1'1 1.29 0 - - _ - -
solvenit was added. Extractioni was not complete 5 1.19 0 _ _
3 1 07 0 _ _
if the sample was dry wl-hen sol-v-ent was added.
6 .02 62 "pool" .35 .401 7.1 .14
Differelnt volumes of solvenit were uised for- 4 .01 126 "pool" .70 .787 14. 4 .07
samples of various size anid dilutions were made 2 .03 122 "pool" .70 .814 15.0 .07

when necessary so that lno absorbaiice valuie was

over 0.8. Serial dilutions of some extracts Expt. 3 MIid-date 8-2-60

showed that Beer's Law applied to absorbance 5 2.22 0 - - - -

values of about 1.1 and below. Final relative 6 1. 83 12 "pool" .09 .108 1.4 . 73
4 1.52 6 "pool" .05 .051 .7 1. 45
absorbance calculations were based on a non-
3 .97 25 "pool" .17 .172 2 9 .34
diluted 30 ml extraction. Corrections were made 2 .33 36 "pool" .21 .228 4.4 . 23
for the slight differences in area among the 6 I .10 50 pool" .27 .286 5.9 .17

pools. The results of a few tests indicate that a

chlorophyll absorbance value of 1.0 may be equal Expt. 4 Mid-date 8-15-60

to about 30 mg of diatom "organic matter" 4 1. 48 0 - - - -

(weight lost on ignition). This ratio may change 1 .61 .12"Pool" .09 .091 1.4 .71
2 .17 24 "pool" .21 .210 2.9 .34
with season, species, or physiological condition of
3 .04 36 "pool" .28 .288 4.1 .24
the plants.
Chlorophyll was used as the criterion of diatom Expt. 5 Mid-date 9-11-6(1 (pools u_
mass because the collections included sand and
4 .01 Mixed 1.06 - 32.7 .03
cotton, which made it impractical to use weight.
5 .01 Mixed .54 - 17. 4 .06
Cell counts could have been used, but the differ- 3 .02 Mixed .87 - 27.5 .04
ence in cell volumes made such a technique too 1 .02 Mixed 1.08 - 32 9 .03
6 .02 Mixed .32 - 8.6 .12
cumbersome. 2 .03 Mixed .68 - 22. 8 .04

Vertical Transect
Expt. 6 Mid-date 9-25-60
Enclosures were prepared on the vertical wall
4 1 66 0 - - - -
to measure the effect of limpet grazing. A strip2 .22 36 "Pool" .25 - 4.4 .23
1.2 m wide was cleared of all organisms from
the one foot tide level to the top of the rock at
about the 8 ft level (Fig. 2a). The rock was bodies. A plot was resterilized with 95%o ethanol
cleared by scraping and was sterilized with con- at the beginning of each experiment.
centrated formalin.
Two rows of 4 stainless steel baskets were MAINTAINING DIATOM-FREE AREAS
mounted at different levels. Each basket covered Littorine Grazing
an area approximately 15 x 15 cm. The upper
In the tide pool experiments littorines were
row (U) was at the 5 ft level; the lower row
(L) was at the 3 ft level (Fig. 2). During all initroduced into bare, newly sterilized pools. The
results appear in Table I and Figure 3.
the experiments 2 plots were run as limpet-free
Control pools invariably developed a thick
controls and 2 were used for different numbers of
A. digitalis, A. pelta, or A. sctituli. carpet (0.5 cm) of diatoms on both bottom and
The same method of collecting and processing sides in 12-16 days (Fig. lb). The main source
was used. The volume of whole limpets was area of the diatoms was almost certainly the
sublittoral region. In one experiment (Expt. 3,
taken as well as the dry weight of the shell-free

This content downloaded from

91.203.32.150 on Wed, 12 Jun 2024 06:57:18 +00:00
All use subject to https://about.jstor.org/terms
Autumni 1961 TIIE EFFECT OF GRAZING ON MARINE LITTORAL DIATOM POPULATION.3 787

100 out of the pool entirely or clustered above the

water line.
-j
It became clear that littorilnes naturally ini-
80
o \\
Hz\
801\\ habiting tide pools were necessary to obtain
efficient grazing in a permanently submerged sit-
uation. In all cases when "pool' littorines were
introduced into a bare pool they dispersed widely
060, over the bottom and sides and remained spread
throughout each experiment or until a diatom
cover appearedl. Besides the behavioral difference
ui ~~A there was a definite size difference between the
>40
0 "shelf' and "pool" types. The meain volume of
the 'shelf" type was about .035 cmn"; that of the
0
pool" was about .050 cmi. Very little overlap
?20 occurred and no morphological (liffere3lces were
observed. Nortth (1954) demiionstrated that
smaller littorines were able to hang on wave
00 ~ ~ ~~~~A 0. %- w-ashed rock better thani larger ones; thus, the
a ) .1 .2 3 4 .7
larger sniails could be selected against, except in
LITTORINE VOL. CMYDM2
the greater protection of a tide pool.
FIG. 3. The relationship between total littorine volume
(not individual volumes) and the amount of diatom cover The rest of the tide pool experiments involved
developed. In all cases littorines were introduced into the use of various numbers of "pool" littorines.
bare, sterilized pools. The dashed line is for guidance. In the second experiment (Table I) "pool" lit-
O = Expt. 2, o = Expt. 3, A - Expt. 4 F = Expt. 6
torines of about the same volume and number as
(Table I). See text for further explanation.
the "shelf" types used in the first experinment were
Table I) the screens of Pools 5 and 6 had not able to keep Pool 6 essentially clear of diatolms.
been clealned and sterilized. As a result, the The great difference betwveen Pool 6 and the con-
inoculation alnd appearance of a diatom cover was trol Pool 5 is shown in Fig. lb. The results of all
greatly accelerated. The flora in the ungrazed the experiments using "pool" types are combined
and grazed pools was similar and presumably no and presented graphically in Figure 3. A pattern
selective grazing occurred. Stauroneis constricta is easily recognized even though experiments are
(Ehr.) Cleve, Navicula ramosissima, Coscino- separated by almost 3 months. The relative ab-
discus spp., and Fragilaria striatula Lyngbye com- sorbance values of diatom extract are given as
posed the bulk of the material in decreasing order percentages of the control in each case. Control
of abunldance. Other diatoms made up the rest, values did not differ greatly throughout the study
almost to the exclusioln of other groups of algae. (Table I). In Figure 3 the total volume of
Nonle of the cells was v7ery large anid presumably littorines/din2 is used rather than the nu
all could be ingested or at least removed easily by individuals/dmi2. With respect to diatom eli
Littorina. These species typified shallow tide tion, there appeared to be closer agreement be-
pools at high levels where temperature fluctuations tween littorine volumes than between numbers

were great. The pools were usually flushed twice (Table I). A plot using dry weight of littorines
daily with 10-14? C water; however, at mid-day would produce a curve similar to that using
pool temperatures usually rose to 18-23?C unless volume. At present little is known about littorine
the area was under fog. In deeper and con- range, feeding rate or scraping rate relative to
sistenitlv colder pools the diatom flora often anlimal size.
differed conisiderably. A total combined littorine volume of about
The first experiment in late June used about 8 0.2 cm3 /dm2 (or about 0.2g/dm2) seems to be a
littorines /din2 collected from the exposed shelfcritical value for the summer period (Fig. 3). A
rock. After 2 weeks the rather dense population littorine volume above this value was invariably
of littorines did not nearly succeed in maintaining able to maintain a pool free of a visible diatom
a diatom-free area (Table I). Most of the snails cover, at least over 2 weeks. Volumes below this
in Pool 5 were aggregated along the walls and in value usually resulted in a patchy but quite ap-
the corners throughout the 2 weeks. Only the parent brown diatom carpet with relatively high
periplhery was cleanly grazed. "Shelf" littorines chlorophyll values. In the latter case littorines
were placed in tide pools and observed for short were able to disperse throughout the pool during
periods at other times. Invariably they moved the first week of the experiment, but the entire

This content downloaded from

91.203.32.150 on Wed, 12 Jun 2024 06:57:18 +00:00
All use subject to https://about.jstor.org/terms
 788 RICHARD W. CASTENHOLZ Ecology, Vol. 42, No. 4

area was not covered frequently enough by the TABLE II. Transect experiments in which limpets were
grazers, since small patches of diatoms developed introduced on bare, sterilized plots. All data (except
Expt. 4) are based on 14 day periods. See text for addi-
during the second week. As these patches en-
tional explanation
larged, the range of the littorines appeared to be
correspondingly reduced. After 2 weeks each Vol. of
limp. Dry wt. of No. of No. of
littorine was restricted to a cleared area of not Plot Absorbance No. of limp. per shell-free limp. dM2
of diatoms per plot & dM2 limp. per per per
more than a few cm2; elsewhere a relatively thick @ 665 mwL species (cm3) dm2 (g) dM2 limp.
diatom cover was present. Littorines were in-
variably within their small cleared areas. Ap- Expt. 1 Mid-date 7-2-60
parently they did not traverse a diatom carpet
U-i 1.12 0 - - - -
but worked along its edges. Thus the feeding U-2 .04 10 A. dig. 1.51 - 4.3 .23
range (or total range of movement) was inverse- U-3 2.03 0 - - -

ly related to the food supply and there was a vast U-4 .08 10 A. dig. 1.42 _ 4.3 .23
L-1 1.58 0 _ - -
difference between range on an essentially bare L-2 1.25 8 A. dig. .97 - 3.4 .29
substrate and range in a carpet of diatoms. A L-3 1.51 0 - - - -
L-4 1.47 9 A. dig. 1.18 - 3.9 .26
"pool" littorine volume of over 0.2 cm3/dm2 (over Unscreened
3 individuals/diM2) was apparently sufficient to plot 3.37 0 - - -

range the entire area of the pool frequently

enough to prevent the growth of diatom patches. Expt. 2 Mid-date 7-19-60

In mid-September the 6 sterilized pools were

U-1 3.10 0 _ - - -
unscreened for 2 weeks, during which time there U-2 2.00 2 A. dig. .54 .083 .9 1.16
was free entry and exit of littorines and small U-3 2.79 0 _ - - -
U-4 2.08 2 A. dig. .43 .063 9 1.16
limpets from the shelf and adjacent natural pools.
No visible cover of diatoms developed in any
Expt. 3 Mid-date 8-2-60
pools. Chlorophyll values were similar to those
obtained with littorine volumes of 0.35 to 0.70 U-1 1.14 0 - - - -
1U-2 .50 5 A. dig. .56 .073 2.1 .47
cm3/dm2 in controlled experiments (Table I). U-3 .85 0 - - - -
The littorine volumes in the open pools were as U-4 .13 5 A. dig. .69 .093 2 1 .47
high as 1.0 cm3/dm2 at the time of collection.
In June and early July the density of littorines Expt. 4 Mid-date 7-25-60 (28 days

on the Yoakam shelf was fairly low (< 1/dm2). L-1 .32 0 _ - - -
Pools 1-4 were unscreened at that time, and a L-2 .07 2 A. pelta 1.85 _ 9 1.16
heavy diatom cover developed in each within 2 L-3 .28 0 - - - -
L-4 .03 2 A. pelta 1.81 .317 .9 1.16
weeks. Littorines were removed from the pools
every second day but there were never more than
Expt. 5 Mid-date 8-15-60
6 per pool. Several natural pools which had
grazers removed in June developed a diatom cover U-1 1.97 0 - - - -
U-2 .04 3 A. pelta 1.16 .202 1.3 .77
before being invaded and eventually cleared.
U-3 3.29 0 - - - -
The general lack of diatom cover in tide pools U-4 .09 3 A pelta 1.03 - 1.3 .77
I-1 3.41 0 - - - -
of the Yoakam shelf may thus be explained by
L-2 .02 4 A. scutuml 1.29 .232 1.7 .58
the natural density of littorines in these pools, L-3 3.18 0 - - - -
densities which generally exceeded 6/dm2. How- L-4 .02 4 A. scutum 1.20 .195 1.7 .58

ever, the question of the "bare" areas on the shelf

itself remains. Occasionally in early summer,
during periods of fog, a few diatom patches did by water for only a few hours per day, some-
develop in relatively moist areas. This was dur- times by only one high tide. Even if the settled
ing a time of relatively low littorine density in diatoms could withstand the hours of exposure,
the particular area (1/dm2). If the "shelf" lit- often to full sunlight, it is almost certain that
torines were as effective grazers as the "pool" their growth period would be limited to those few
type it would be difficult to imagine a diatom hours of high moisture as contrasted with the full
cover developing during late July, August, Sep- daylight period in the pools. Probably relatively
tember, or October. The density during this few effective grazers would be required to main-
period was usually over 12/dm.2 Even with few tain a clear surface under these conditions.
grazers an extensive diatom cover should probably However, during late October and early No-
not be expected during summer. The "bare" vember (1960), extensive diatoms patches (pri-
areas were at the 6-8 ft levels and were covered marily Fragilaria) developed on the "bare" areas

This content downloaded from

91.203.32.150 on Wed, 12 Jun 2024 06:57:18 +00:00
All use subject to https://about.jstor.org/terms
Autumn 1961 THE EFFECT OF GRAZING ON MARINE LITTORAL DIATOM POPULATIONS 789

of the shelf, particularly near the artificial pools lowed a pattern observed several times in past
with their abundant source of inoculum. This de- years at the same location and in other intertidal
velopment occurred during almost continually stations in the Cape Arago region. In less than
overcast conditions. At the time of observation a week a dark diatom film appeared at the lowest
large numbers of littorines were grazing at the levels and seemingly spread gradually upwards.
edges of the patches. Whether the littorine pop- After 2 weeks the dark slime cover had invariably
ulation was low during the establishment of the spread to at least the 5.5 ft level (Fig. 2b.) and
patches or whether the "shelf" littorines were not eventually extended to the top of the transect.
effective enough to cope with the increased diatom The apparent spread, however, was simply a re-
production could not be determined. The winter flection of differences in growth rate at different
situation will be discussed later. levels. Several days after sterilization diatom
Littorina scutulata ranged to at least the 11 ft settlement had occurred at almost all levels. The
level on the side of the Yoakam shelf. These last lower portion of the transect was covered by water
few feet of vertical rise were only occasionally for longer periods allowing a longer growth period
covered by water or spray. The littorines of this per day. Mortality rate during low tide ex-
"supra-littoral" fringe seldom contained diatom posure may have been higher in the upper por-
cells. It is unlikely that a diatom cover could tion.
develop at this level in summer even in the com- Licmophora abbreviata and Navicula ramosis-
plete absence of grazers. sima were the first colonizers of the lower two-
thirds of the transect; Fragilaria striatula was
Limpet Grazing
the successful slower growing diatom which cov-
The results of the limpet grazing experiments ered the upper one-third. After 3 weeks various
are presented in Table II and Figure 4. De- changes took place that may be referred to as
pendence on favorable low tides limited the num- "successional" events. These will be discussed
ber of experiments on the Yoakam vertical tran- later.
sect. Tides after August were not adequate for Acmaea digitalis was used at both the upper
continued work. (U) and lower (L) levels for the first experiment
The development of a diatom cover after (Fig. 2). The lowest natural occurrence of A.
scraping and sterilizing the entire transect fol- digitalis was about at the level of the upper series
(5 ft). As a result, only the upper experiments
0
demonstrated what was probably the grazing
0
ability of these limpets (Table II and Fig. 4).
In Figure 5a the upper plots with limpets are
z8 contrasted with the control plots. Figure 5b
0
o shows the limpets of a lower plot aggregated at

o 2 \~ the top. Only the upper portion of the lower

plots had been grazed. It appears that a tendency
on the part of the limpets to move upward allowed
the development of a heavy diatom cover over
w the lower two thirds of the plots.
In the majority of the experiments there was
0
closer agreement between total volumes than
between total number of individuals with respect
to diatom removal (Table II). Moore (1938)
found that the size of the feeding area and the
~~~~2O ~ ~ ~ ~ ~
volume of the limpet Patella showed a more or
less straight line relationship. Nevertheless, the
C .4 .8 12 I 2.0 2 limpets in Expt. 2, which were almost equal in
LIMPET VOL. CMYtb total volume to the A
5 limpets in Expt. 3, were not
FIG. 4. The relationship between total limpet volume as effective (Table II), indicating some limita-
(not individual volumes) and the amount of diatom cover
tions of large size.
developed. In all cases limpets were introduced on bare,
sterilized plots. The two upper points which are not in The curve of Figure 4 indicates that a total
agreement represent the results of A. digitalis used below combined volume of about 0.8 cm3/dm2 would be
its normal vertical range. The dashed line is for needed to maintain an area essentially clear of
guidance. 0 = Expt. 1, * = Expt. 2, El = Expt. 3,
A = Expt. 5 (Table II). See text for further explana-
diatoms. However, only A. digitalis was used
tion. in volumes lower than this.

This content downloaded from

91.203.32.150 on Wed, 12 Jun 2024 06:57:18 +00:00
All use subject to https://about.jstor.org/terms
 '790 RICHARD W. CASTENIIOLZ Ecology, Vol. 42, No. 4

almost the full length of the transect, there is

reason to believe that the natural "bare" areas
described earlier were kept free of diatom covers
by limpets. In fact, the density of Acmaea on the
upper "bare" areas was greater than any used in
the experimeints. Many limpets on1 the natural
areas were much smaller, however.

CLEARIN G A DIATOM1 COVER

'-
The effectiveness of littorines and limpets in
mainitaininlg diatom-free areas has been dem-
onistrated. How effective are these herbivores in

f~~~~
AV"~~~~~~~~
clearing a well established diatom cover?

TABLE III. Tide pool experiments in which littorines

were introduced into a 2 week diatom cover. They were
allowed 7 days of grazing in Expt. 1 anid 16 days in Expt.
2. See text for additional explanation

Absorbance No. of litt. Volume of

A Pool of diatoms per pool & litt. per dM2 No. of litt. No. of dM2
@665 m,u type (cm3) per dM2 per litt.

Expt. 1 Mid-date 7-2-60

1..... 1.18 0 - - -
2..... 1.50 64 "pool" .40 7.9 .13
3..... 1.14 0 - - -
4 .... .91 64 "shelf" .25 7.3 .14

Expt. 2 Mid-date 9-18-60

FIG. 5. Plots on the vertical transect onl July 8,
1..... 4.81 0 - - -
16 days after sterilization and introduction of limpets.
5..... 5.76 18 "pool" .11 2.3 .44
The baskets have j ust been removed. a. The upper
3..... 4.72 36 "pool" .29 4.2 .24
series of plots. The control areas with a heavy diatom 6 .... 3 83 56 "pool" .33 6.5 .15
cover are outlined ill white. T'en A. digitalis may be seen
,on U-2 and U-4. There is an almost complete absence of
diatom cover. b. Two of the lower plots, L-3 and L-4.
'The control and the partly grazed plots are outlined in In experiments summllarized in Table III lit-
white. The 9 A. digitalis oni L-4 are clustered near the torines were introduced into Pools 1-4 after a
top and only the upper third of the plot has been grazed. diatom cover had developed for 2 weeks in the
'The black backgrouiid is a a continuous cover of diatoms.
absence of grazers. The addition of a large num-
ber of littorines ("pool" and "shelf" types) was
The lower series was not as well used as the
ineffective in checking the increase of the diatom
-Lpper, since there was a period in mid-summer mat over 7 days. In one pool there was more
wlhen diatom growth was poor even in the con- diatom material than in the controls (Table III).
'trol plots. Because of these difficulties the lower In September littorines were again introduced
plots were run for 4 weeks in mid-summer into pools with a diatom cover and left for 2
( Table II) . weeks; even then they had little effect (Table
The last set of experiments be-gan on August 7. III) but rather striking inroads were seen at
A4. pelta was used at the upper level and A. the time of collection (Fig. 6). Cleared patches
-scutumi at the lower. The volume and number occurred around the edges, seldom in the center
were quite sufficient to preserve a clear area. It of a pool. The diatom carpet was very thick after
would have been desirable to use also a smaller 4 weeks and it was obvious that the littorines
-volume and number of these species. Poor low grazed at the edges only. Whether the same lit-
tides during the remainder of the summer and torines would have eventually eliminated the di-
fall prevented access. In all experiments except atom cover in these pools is not known. Diatom
the first the limpets were collected from tidal covers were allowed to develop in 6 natural pools
levels equivalent to the level at which they were on the Yoakam shelf in early summer. Within
-used. 4 weeks all pools had been cleared by the un-
Since a diatom cover was able to develop on controlled invasion of littorines and limpets. The

This content downloaded from

91.203.32.150 on Wed, 12 Jun 2024 06:57:18 +00:00
All use subject to https://about.jstor.org/terms
Autumn 1961 THE EFFECT OF GRAZING ON MARINE LITTORAL DIATOM POPULATIONS 791

- U.? . ' activity of grazers, the disappearance of the cover

ib~~~~~4 in the lower littoral zone may occur in the com-
N
plete absence of grazers, even during high diatom
production. The normal sequence of events on
the bare, sterilized vertical transect was described
abov-e. A cover first b)ecame evidcent on the
lowest part of the transect, followed by the ap-
parent upward extensioni tlhrough a 3 week period.
However, after 4 weeks the lower portion of the
transect was againi bare from about the 3.5 ft
level to the base at 1.0 ft. The 2 week cover in
the lower control plots was also adversely affected.
The disappearance occurred over 5 days. The
cov-er of trailinig Nazicila rainosissintia and L
pliora. abbrezviata disapp)eared except for
isolated tufts of the formner. The same puzzling
l)heiiomelnon occurred in the sutmmer of 1958 at
the samiie locationi and( was observ-ed onl other
tranisects at Gregory Point in the summers of
1959 and(l 19('0. Sometimes different diatom spe-
cies were involved. In each case, there was no
4t evidelnce of grazing.
It was first thought tllat the conisiderable wave
action at high tides may lhave removed the heavy
growth after it lha(l reached aln unstable thickness,
yet this growth did niot form a bound mat which
could easily peel to leave bare rock behind. The
cover at Gregory l'oiint, a rather tlhin film of uni-
cellular formls, was exposed only to light wave
action, yet disappeared silmiilarly. The even thick-
cover has developed over a 4 week period and consists er carpets of Navicila rain osissinia on sublittoral
mainly of N. raiiiosissima. After a two week cover had rocks exposed to very heavy wave action was
developed 18 littorines were placed in Pool 5 and 56 in further evidence against the hypothesis.
Pool 6. Conspicuous inroads may be seen, particularly
The eliminationi of the lower diatom carpet was
in Pool 6. A diatonm patch has also developed on the
shelf rock in the foreground. It has been considerably best correlated with the appearance of young Ulva
reduced by littoriiie grazing. Large numbers of "shelf" thalli oni the samiie substrate. Ulva appeared com-
littorines may be seen along the left-haiid margin and monly from about the 3-1 ft level. The density in
elsewhere.
this zone was about 1 plant/cm2. Smaller thalli
of Ulva also appeared from the 3.0-4.5 ft levels.
familiar periplheral iinvasioti aind constrictioin of the
The diatom cover in this area also lessened grad-
diatom mat was followed through this period.
ually through the succeeding 2 weeks. However,
No controlled experiments of a limpet attack
after maximutim bareness in the lower portion of
on an established diatom mat were conducted;
the tranisect at 4 weeks, there was a gradual re-
bowever, sev,eral observations were 1miade at the
colonization and reestablishment of a diatom
Yoakam wall. The 1natural abundance of limpets
cover, different in aspect from the original. The
increased steadily with increasing height. An
new cover was a patchy thini film, composed pri-
itlvasion of limpets onto the transect from the
marily of Fr-agilaria striatula together with Ulva.
uindisturbed sides was followed during July and
There were also conspicuous increases in the scat-
August. After 5 or 6 weeks, the diatom cover
tered tufts of NTavic ua ran1iosissi;na. which sur-
above the 4 ft level was eliminated by invading
vivedl the first clearing at the 1-3 ft levels. The
limpets, primarily A. digitalis and young, unidenti-
general successionial pattern was Licmophora
fied limpets.
abbrev.-Navzicula raw o. followed by Ulva and
T H EEFFECT OF "SUCCESSION" O N Fragilaria. This appeared not to be a type
DIATOM COVER of seasonal progression, since the whole sequence
Altlough the elimination of a diatom cover in could be repeated by resterilization at different
t .e upper littoral zone may be explaicted by the periods during summer and fall.

This content downloaded from

91.203.32.150 on Wed, 12 Jun 2024 06:57:18 +00:00
All use subject to https://about.jstor.org/terms
792 RICHARD W. CASTENHOLZ Ecology, Vol. 42, No. 4

Similar "successional" sequences have been ficially even steeper, since there is a reverse gradi-
observed in other temperate parts of the world ent in grazer density.
(e.g., Wilson 1925, Bokenham and Stephenson The situation in winter was somewhat puzzling
1938). It was common in the literature to find at first. The same rocks of the upper half of the
reports of a diatom cover succeeded by Entero- littoral zone that were bare in summer were often
morpha or Ulva, and these followed by the array largely covered by extensive diatom slicks. Dia-
of red, brown, and green algae which normally tom production rates on the horizontal shelf at
inhabit the region. Yoakam Point were, nevertheless, not as great as
The "successional" events were followed on suitable intertidal substrate in summer when
through the fall and winter of 1958-59. It became grazers were removed. Yet, winter rates on the
obvious that seasonal progression was being ob- shelf were generally higher than contemporaneous
served as much as succession. At many times rates on permanently submerged substrates in
during these months the lower part of the transect Coos Bay and on vertical littoral transects at
was free of diatom cover; at other times a patchy Gregory Point (Castenholz unpublished). The
cover developed among the young macroscopic rates in the last 2 areas were considerably lower
algae colonizing the surface. There have been in winter than in summer during 2 successive
several reports of algal inhibitory factors, often years. On horizonital exposed substrates, such as
quite specific, which affect algae other than those the high Yoakam shelf, the overcast winter condi-
producing them. Studies of planktonic algae tions were probably responsible for a great re-
have shown that an antibiotic principle may ac- duction in diatom mortality and a lengthening of
cumulate to toxic levels in the mass of water the daily growth period, allowing growth where
or in the culture medium (e.g., Sieburth 1960; little or none was possible during the desiccating
Proctor 1957). If the increase in Ulva inhibited conditions of summer.
the existing diatoms, it might have been caused The winter diatom covers appeared in areas
by a "leachate" derived from the Ulva which with generally large populations of littorines and
spread through the diatom mat during low tide limpets. In summer, areas of potentially high
exposure. If metabolites are involved, more prob- diatom growth, are normally kept free of diatoms
ably an autoinhibitor produced by the diatoms by grazer activity. Why, then, are these same
accumulated to toxic levels in the thick carpet. areas and also higher areas not kept diatom-free
Pratt and Fong (1940) and others have dem- in winter?
onstrated autoinhibition in aging algal cultures. A few possiblities exist. Littorines and limpets
may simply be less active in winter. Their range
DIsCUSSION of movement may be considerably reduced. There
These experiments and observations show that is no evidence for or against this hypothesis. As
a heavy diatom cover develops in summer on another possibility, littorine numbers may be
rock substrate through most of the littoral zone greatly reduced in local areas for a short time by
when grazers are absent or reduced in numbers. the frequent winter storms. If this were true,
However, on high littoral rocks exposed to direct diatom patches could begin to develop. It was
shown above that once a diatom cover is estab-
insolation little diatom production could be ex-
lished, littorines have considerable difficulty in
pected regardless of gastropod density. This
removing it. There is some evidence to support
would be particularly true on horizontal surfaces,
this explanation. On a few occasions, after
such as the Yoakam shelf. Experiments (Casten-
storms, areas of reduced littorine density were
holz, unpublished) concerned with zonation at
observed. At three times during the winter of
Gregory Point have shown that the upward ex-
1960-61 (Jan., Feb., March) a large area of the
tension of diatoms is most severely restricted on
Yoakam shelf (about 3 m2) was sterilized, effec-
rock faces which receive direct insolation during
tively removing the large littorine population.
low tide exposure (on south, east and west faces). Initially there was no measurable diatom cover.
Much depends on whether cloudless or overcast It is presumed that at least several days passed
days are prevalent and on the time of low tides. before a high littorine density was reestablished.
In regions of potential diatom growth there is an After 2 weeks there was a sizable littorine popula-
apparent production gradient which shows in- tion but only a barely perceptible diatom cover.
creasing rate with decreasing tidal level. This Nevertheless, after 4 weeks, a thick diatom carpet
is presumably due to differences in number of (Melosira sp.) covered the area and the littorines
growth hours per day and in mortality rate. The were restricted to the edges of this mat. It is
gradient in undisturbed areas would appear super- possible that the few days of low littorine density

This content downloaded from

91.203.32.150 on Wed, 12 Jun 2024 06:57:18 +00:00
All use subject to https://about.jstor.org/terms
Autumn 1961 THE EFFECT OF GRAZING ON MARINE LITTORAL DIATOM POPULATIONS 793

were sufficient for diatoms to become established obtained on the actual feeding habits of littorines
and that thereafter the littorines were not able to and limpets. The effect of gastropod activity was
suppress the development and extension of this the removal of diatom cells. Whether most of
cover. Similar results were obtained on all 3 these cells were simply removed by the radula and
occasions. The frequent storms of late fall and lost or whether they were ingested and utilized
winter could provide similar circumstances. was not determined. The digestive tracts of both
Something of a double anomaly exists in parts limpets and littorines in all experimental plots
of the intertidal: in summer a high diatom repro- were packed with diatoms to the exclusion of
duction rate with a small standing crop, in winter other material. Limpets and littorines collected
a somewhat lower reproduction rate but with a from areas of potential diatom cover also con-
large standing crop. This type of situation should tained primarily diatom cells, but other unidenti-
interest those workers who describe situations by fied material was present. One may presume,
assuming that the magnitude of the standing crop that the primary food of these gastropods is dia-
is a simple reflection of production rate. Aleem tom material. Only in the "supra-littoral" fringe
(1950) described the seasonal populations of were the littorines and limpets devoid of diatoms.
British marine littoral diatoms. His frequent Blue-green algae may be the principal food at this
references to the disappearance of diatoms from level.
rocks in summer lead me to suspect that grazing
may have been invx olved; lhowever, lie made no SUM MARY
reference to the aniimals inihabiting the study 1. A virtual absence of epilithic diatoms in the
areas. upper and mid littoral of the southern Oregon
The significance of scraping and sterilizing in coast was noted during 3 successive summers;
addition to the removal of grazers was not thor- however, on bare and sterilized littoral rock sub-
oughly treated in this study. Most of these ex- trate thick diatom slicks developed within 2 or 3
periments included this treatment. Is there any- weeks. Diatom production rate appeared very
thing initransically inihibitory to diatom growth high in summer but most rock areas were kept
on an unscraped, unsterilized surface? The prob- diatom-free by herbivorous gastropods.
lem of "succession" is again raised. It was ob- 2. Six artificial tide pools of equal dimensions
served that a diatom cover in the lower littoral carved in a high intertidal sandstone shelf were
may eventually succumb to "successional" events screened and used as closures for controlling
in the absence of grazers, leaving the rock sub- numbers of Littorina. Stainless steel baskets
strate "bare" for various lengths of time. Also, wel e anchored against a vertical wall in the
if other algae are able to "replace" diatoms, then, littoral for similar experiments with Acmaea.
are the natural "bare" areas amid a "mature" Most experiments introduced known numbers
algal assemblage mailltained in that state by the and volumes of littorines or limpets into bare,
activity of grazers or the presence of algae, or sterilized pools or baskets to measure their rela-
both? It was observed many times that an area tive abilities in maintaining diatom-free areas.
within the sweep of a larger algal thallus was Other experiments measured the effect of lit-
torines on established diatom covers. Diatom
easily kept free of cover by simple abrasive action.
These are lnot the areas in question. In late crop, in all cases, was estimated by quantitative
June, 1960, 8 shallow tide pools near the seaward measurements of chlorophyll.
tip of Yoakam shelf were treated as follows: 2 3. Results in summer show that a total littorine
were scraped and sterilized with formalin after thevolume greater than 0.2 cm3/dm2 (or about 3
removal of ,grazers; 2 were merely sterilized afterlittorines/dM2) was adequate to keep an area
the removal of grazers; 2 were denuded of nearly free of diatoms for 2 weeks. Normal
grazers, and 2 were left undistured. Within 3 densities in natural tide pools were several times
weeks all but the unidisturbed pools had developed this value. Volumes of less than 0.2 cm3/dm2
a complete diatom cover. During this time in- allowed (liatom patches to develop. As these
vading grazers were removed periodically. In patches grew in size, each littorine's grazing range
the summer of 1958 the removal of limpets from was progressively diminished until only the
an upper portion of the vertical wall at Yoakam edges of the patches were grazed. A much
also resulted in the development of a diatom larger volume of littorines and many weeks would
cover. These trials support the hypothesis that be required to clear a previously established dia-
"bare" areas may be maintained by grazers alone. tom cover. Limpet volumes over about 0.8
This study has emphasized the effect of grazing cm3/dm2 were required to maintain a diatom-free
oln diatom populations. Little information was area. Numbers depended on the size and species.

This content downloaded from

91.203.32.150 on Wed, 12 Jun 2024 06:57:18 +00:00
All use subject to https://about.jstor.org/terms
794 THADIS W. BOX Ecology, Vol. 42, No. 4

The normal limpet population on "bare" areas in of protozoa on freshwater algae. Hydrobiologia 4:

the upper littoral may equal or exceed this critical 281-2,913.

. 1954. The bottom-living algae of slow sand
value.
filter beds of waterworks. Hydrobiologia 6: 333-
A diatom cover may disappear from rocks of
- . 1955. The aquatic fauna as an ecological factor
351.
the lower littoral without the activity of grazers.
"Successional" events may be involved. in studies of the occurrence of freshwater algae. Rev.

4. Results of general seasonal observations in- Algolog. (N.S.) 1: 141-145.

Douglas, Barbara. 1958. The ecology of the attached
dicate that a heavy summer diatom population is
diatoms and other algae in a small stony stream.
held in check by littorines and limpets in most of J. Ecol. 46: 295-322.
the upper half of the littoral, and by unfavorable Fischer-Piette, E. 1948. Sur les elements de prosperite
factors of exposure in the very high littoral or des patelles et sur leur specificite. J. Conchyl. 88:
45-96.
"supra-littoral" fringe. The potential for diatom
Moore, H. B. 1938. Algal production and the *food
production increases as vertical level decreases.
requirements of a limpet. Proc. Malacol. Soc. 23:
Large diatom populations in winter in the upper 117-118.
littoral and "supra-littoral" may be due to fre- North, W. J. 1954. Size distribution, erosive activities,
quent storms causing temporary reductions in and gross metabolic efficiency of the marine inter-
tidal snails, Littorina plaptaxis and L. scutulata Biol.
littorine densities at which time diatom covers
Bull. 106: 185-197.
may become established. It is also possible that Pratt, R. and J. Fong. 1940. Studies on Chiorella vul-
littorines and/or limpets are less active during garis. II. Further evidence that Chlorella cells form
winter. a growth-inhibiting substance. Am. J. Bot. 27: 431-
436.
5. Diatoms probably constitute the principal
Proctor, V. W. 1957. Studies of algal antibiosis using
food material of Littorina scutulata and most
Haeinatococcus and Chlamiydomonas. Limnology and
epilithic species of Acmaea in the littoral zone. Oceanog. 2: 125-139.
However, the amounit of diatom material removed Sieburth, J. M. 1960. Acrylic acid, an "antibiotic"
from rock may reflect a considerable amount lost principle in Phaeocystis blooms in Antarctic waters.
Science 132: 676-677.
by radular scraping as well as that ingested.
Shotwell, J. A. 1950. The vertical zonation of Acmaea,
REFERENCES the limpet. Ecology 31: 647-649.
Aleem, A. A. 1950. Distribution and ecology of Brit- Strain, H. H. and W. M. Manning. 1942. Chloro-
ish marine littoral diatoms. J. Ecol. 38: 75-106. fucine (chlorophyll y), a green pigment of diatoms
Bokenham, N. A., and T. A. Stephenson. 1938. The and brown algae. J. Biol. Chem. 144: 625-636.
colonization of denuded rock surfaces in the inter- Test, A. R. 1945. Ecology of California Acmaea.
tidal region of the Cape Peninsula. Ann. Natal Ecology 26: 395-405.
Museum 9: 47-81. Wilson, 0. T. 1925. Some experimental observations
Brook, A. J. 1952. Some observations on the feeding of marine algal succession. Ecology 6: 303-311.

RELATIONSHIPS BETWEEN PLANTS AND SOILS OF FOUR RANGE

PLANT COMMUNITIES IN SOUTH TEXAS'

THADIS W. BOX2
Departmnent of Range and Forestry, Agricultural and Mechanical College of Texas

Distinct plant communities occur in South range plants is needed for the management of
Texas and the narrow transitional zones between these native lands. Unfortunately, basic eco-
them suggest that local edaphic conditions may be logical data are not available for most of these
limiting factors in vegetational distribution. The rangelands.
South Texas plant communities not only differ The purpose of this study was to evaluate the
in their floristic composition, but also in produc- vegetational resources of 4 important South Texas
tivity, and in their economic returns to land- plant communities and to isolate factors that may
owners. An understanding of the factors regu- contribute to plant distribution and abundance.
lating potential production and distribution of Plant communities selected were representative of
1 This paper is contribution number 53, Welder large
Wild- areas of South Texas rangeland. These
life Foundation, Sinton, Texas. were (1) a mesquite-buffalo grass community,
' Present address, Range Management Department, (2) a chaparral-bristlegrass community, (3)
Utah State University, Logan, Utah. Formerly of the
a bunchgrass-annual forb community, and (4) a
Rob and Bessie Welder Wildlife Foundation, Sinton,
Texas. prickly pear-short grass community.

This content downloaded from

91.203.32.150 on Wed, 12 Jun 2024 06:57:18 +00:00
All use subject to https://about.jstor.org/terms