A multivariate analysis of the niches of plant populations in raised

advertisement
A multivariate analysis of the niches of plant populations in raised bogs.
11. Niche width and overlap
E. A. JOHNSON'
Depnrtmetlt of Botnt~y,University of New Hrrt?lpshire, Durham, N H , otld The Mnrinc Biologicnl Loborcitory,
Woods Hole, M A , U.S.A.
Received July 19, 1976
E. A. 1977. A multivariate analysis of the niches of plant populations in raised bogs. 11.
JOHNSON,
Niche width and overlap. Can. J. Bot. 55: 1211-1220.
The results of a principal components analysis are used to define the ecological niches of the
plant populations in raised bogs. This paper examines the niche parameters of width and overlap.
Dominance is positively related to niche width and inversely related to the number of species
(richness). Richness is shown to be positively related to increased environmental complexity and
predictability. Dominants appear to decrease in environments that are more complex and predictable because as aclass these organisms are less opportunistic. Niche differentiation measured
as overlap is greater in the richer. more environmentally complex and predictable parts of raised
bogs.
JOHNSON,
E. A. 1977. A multivariate analysis of the niches of plant popi~lationsin raised bogs. 11.
Niche width and overlap. Can. J. Bot. 55: 121 1-1220.
L'auteur utilise les rksultats de I'analyse par composante principale pour definir les niches
Ccologiques des populations vegttales dans les tourbikres Clevees. Ce travail examine les
paramktres de la niche quant B leurs amplitudes et 2 leur recouvrement. La dominance se relie
positivement B I'amplitude de la niche et inversement au nombre d'especes (richesse floristique).
Cette richesse floristique apparait comme positivement reliee B ilne augmentation de la complexite et de la predictabilite de I'environnement. Les dominants semblent diminuer dans les
milieux qui sont plus complexes et previsibles parce qu'une classe d'organismes sont moins
opportunistes. La differenciation de la niche, mesuree en termes de superposition est plus
importante dans les regions des tourbiereselevees, plus riches, plus complexes et plus previsibles
comme milieu.
[Traduit par le journal]
Introduction
part 1 of this pair of papers, principal
components analysis was used to elaborate the
ecological niches of plant popu~ations in 90
sample plots in raised bogs in Maine. The niche
parameter of dimension was related to two
environmental factors: 'salinity and snow cover'
and 'ombrotrophy-mesotrophy.' This paper will
be devoted t o examiliation of niche width and
overlap. Niche width or tolerance of a particula,.
population can be measured separately on each
niche dimension or combined for all dimensions.
Niche width measures a population's position
in the niche hypervolume and also its population
size at that position. Simply it gives an indication
of how wide a range on an environmental
factor(s) a population can exist and where a
population's optimal response is located. Niche
overlap indicates the degree of ecological
'Present address: Department of Plant EcoIogy and
Institute for Northern Studies, University of Saskatchewan, Saskatoon, Sask., Canada.
similarity and therefore also differences in the
use of the same factor(s) by different populations.
Questions
procedures and
specifics on the principal components analysis
are referred to part I,
Niche Parameter: Width
Niche width is the measure of the position
and degree of performance of a species population along its niche dimensions. It is usually
measured by a variance equation. Several
different ways of measuring niche width have
been proposed (Simpson 1949; Moisita 1959;
Levins 1968; McNaughton and Wolf 1970;
Colwell and Futuyma 1971). The most widely
used width nleasure has been the ShannonWiener information equation (Shannon and
Weaver 1949). For our purpose the width
equation (B) of McNaughton and Wolf (1970)
is preferred because besides measuring variance
along the dimension it weights the positions.
Data were prepared for calculation of B in
the following manner. Components were each
1212
CAN. I. BOT. VOL. 5 5 . 1977
divided into eight equally spaced intervals
numbered from 1 at the far positive end to 8 a t
the far negative end. Since the intervals contained
different numbers of sample plots, the frequencies
in each interval were summed and divided by
the number of sample plots in that interval. These
values were then used to calculate B. Components based on species variables involve a
linear assumption that is known from the
transformation analysis not t o be completely
met. This is recognized as a limitation to some
kinds of interpretation of niche width. Furthermore, width is based on frequency as a measure
of species population behavior, which also
limits its interpretation.
Ab~nzl/uticerind Niche Width
The conventional definition of dominance is
confusing (Greig-Smith 1964) but generally
dominance implies a situation where a few
species are able t o usurp more of the resources
of the environment and have greater influence
on other species (cf. Oosting 1959).
Niche width allows us to test the first part
of the dominance definition. If dominance is
interpreted as having a positive relation t o niche
width, then there should also be a positive relation between niche width (B) and maximum
(modal) abundance (A/?INx).The results of this
determination are shown in Figs. I and 2. The
fitted equations of these graphs are as follows:
component 1, B
=
2.48
+ 1.84Amnx**,
component 2, B
=
3.24
+ 1.88Amrrs**,
,. = 0 81:k*
-0.5
-5
I
13
7
19
25
WIDTH
FIG. 1. Graph of the relation between maximum
abundance of species and their niche width on component
one of the vegetation study.
,. = 0.92**,2
It would seem from these results that species
with the largest niche width d o in fact have the
largest maximum abundance.
Other studies of niche width and maximum
abundance have given similar results in tropical
birds (Klopfer and MacArthur 1960), in fruit
flies (Pico et a/. 1965), and in several different
types of vegetation (McNaughton and Wolf
1970).
Species Richness
Since dominance and the number of species
are intricately related (Whittaker 1965), the
changes in the numbers of species is best
discussed now. The number of species per sample
ZConvention:*, significant at 5%;
**, significant at 1%.
-5
3
II
9
27
WIDTH
FIG. 2. Graph of the relation between maximum
abundance of species and their niche width on component
two of the vegetation study.
plot will be called areal species density, or for
short, richness. Other more elaborate measures
of species number density have been suggested
but are a t best confusing (Hurlbert 1971) or are
in fact measures of dominance (Whittaker 1965,
1972).
Richness (R) increases toward the positive
end of the first component (C,) (Fig. 3), as
described by the equation R = 13.21 f
0.002C1**, 7 = 0.62**. By height class the fitted
equations are as follows:
JOHNSON
+ 0.004C1**, r = 0.29**,
herbs, R = 3.46 + O.OOICl**, r = 0.63**,
mosses, R = 3.53 + O.OOlC,**,r = 0.46**.
shrubs, R = 6.03
The richest sample plots are the ones with
high positive values on the salinity - snow-cover
factor. Inspection of the equations reveals that
the herb and moss height classes account for the
increase in richness since the shrub height class
is fairly constant.
Richness (R) on component 2 (C,) increases
towards the positive end then decreases slightly
(Fig. 4). The linear equation is R = 13.21 +
0.001C2**, r = 0.29**. The richest salnple plots
are in the medium-dry moisture segment of the
gradient.
We will now examine component I in more
detail to see if the reasons for the increase in
richness can be ascertained. Increased richness
can result from two possibilities. (I) It is possible
that the areas studied are not all in equilibrium
as far as species number, i.e., invasion is still
occurring. (2) The environment could be
changing in such a way along the gradient that
more species can use certain parts of it.
An invasion effect seems unlikely on the basis
of geological evidence. Most of the richer
sample plots are closer to the coast which has
only recently been exposed (Milliman and
Emery 1968; Grant 1968) by the receding sea
as a result of postglacial rebound of the coast.
Poorer sample plots are inland in areas not
FIG.3. Graph ofsample plots on component one of the
vegetation study compared with species richness.
FIG.4. Graph of sample plots on component two of the
vegetation study compared with species richness.
covered by the sea. However, it could be argued
from ignorance of the ages of origin of these
raised bogs that coastal bogs are actually older
than inland bogs. Thus an ecological test of a
possible invasion effect would be appropriate.
Preston (1962) has proposed a resemblance
equation based on the distribution of abundance
among species and its relation to area. The
equation is as follows:
N , and N , are the numbers of species in two
areas 1 and 2, N,,, is the numbers of species
in common between the two areas 1 and 2, Z
equals the resemblance of two areas. The
equation is solved for Z,which can range from
0 to 1.0. Preston defines equilibrium to lnean
a steady state in which species are freely exchanged between two areas. That is, the species
in one area have had a chance to colonize the
other area. A zero Z means that the two areas
are not only identical but are subsamples of the
same area. These two areas are in equilibrium
by definition. If two areas are isolated but still
able to exchange species so that their species
numbers are still in equilibrium, the Z value is
theoretically determined to be 0.27 (Preston
1962). If the two areas have no species in
common and thus are in complete disequilibrium, the Z is equal to 1.0. Geographically
1214
CAN. J . BOT. VOL. 5 5 . 1977
related changes in Z can thus give a good indication of immigration.
Z values for all combinations of 19 raised
bogs were calculated using an algorithm incorporating the suggestions of Preston (1962,
p. 418). The mean Z value is 0.37, standard
deviation is *0.10, and the range is from 0.1 1
to 0.59. The high mean compared with the
expected Z of 0.27 would seem to indicate areas
fairly isolated from each other. Although this
may be true it should be remembered that the
assumption that the species abundance is
distributed lognormally is being violated since
abundance is distributed more in a geometric
series (Johnson 1972, unpublished results). The
important point is that the Z values show no
consistant relation to an invasion gradient
relative to the coast.
FIG. 5. Graph of the biologically defined carrying
The second possible means of increasing capacity (K) versus richness (R) by interval in component
richness along the component seems to be more one of the vegetation study.
promising. If in fact the environment is changing, All environmental nieasures were means of all
the capacity of the environment along the the values from the sample plots in a particular
component to 'carry' more species should be interval. T h e standard error of the means of
increasing with increasing species number.
intervals 1 and 2 are larger than the rest because
Carrying capacity has been defined by Mc- of fewer sample plots. All means were stanNaughton and Wolf (1970) as the sum of the dardized to a 0-10 scale since the different
niche widths in an interval along the component. measures are on different scales. This standardiBy comparing this sum of niche widths (K) zation was done by setting the maximum mean
with richness (R) a very good linear relation is at 10.
described by the equation K = -7.86 +
Since the number of environmental measures
-.-76R**, I . = 0.97**, and Fig. 5. Although remains constant in each interval, we are actually
7
richness was used as the independent variable measuring what is called evenness or equitability
in the above equation, we d o not really know (Lloyd and Chelardi 1964). Evenness is usually
which variable, richness or carrying capacity, calculated by the formula J = HIlnE, E being
is forcing the other. They are both probably the number of measures in the interval. The H
dependent on an environmental change.
and J values here are perfectly related, I. = 1.0.
MacArthur and MacArthur (1961) and MacThe graph of the environmental complexity
Arthur (1957, 1965) were the first to point out H against richness (Fig. 6) indicates a positive
that in fairly similar environments new species relation of increasing environmental complexity
are added by increasing the structure of the en- and increasing species number density (richness).
vironment. By structure is meant the arrange- Values for intervals 1 and 2 deviate markedly
ment in patterns of the variables of the environ- from the rest of the points. This is expected from
nient. T o measure changes in structure Bril- the large standard error associated with the
louin's formula was used (Pielou 1966). This environmental measures for these intervals.
equation describes the total information content
From these results it is apparent that species
of an interval:
are added to the system's environmental factor
'salinity - snow cover' by increasing the carrying
capacity (the structural complexity) of the
environment. A similar result was found for
H was calculated for interval i on coniponent I, conlponent 2 'ombrotrophy-nlesotrophy.' These
N was the total values of 14 environmental results could have been surmised from what we
(peat) measures in interval i, and N E was the already knew about these factors from the niche
value of a particular environniental measure E. dimension section. Conclusions similar to this
JOHN SON
FIG. 6. Relation of environmental complexity (H)
versus species richness ( R ) on component one of the
vegetation study.
FIG.7. The relation of dominance index (Dl)values
versus species richness ( R ) on component one of the
vegetation study.
have been arrived at in other studies by Whittaker (1960, 1965) and MacArthur (1965).
TABLE1. The percentage of species involved in
calculation of the total niche width which
have frequencies less than 1.0
Dominance and Richness
Theoretical discussions of ecological systems
have often led to the belief that richness and
dominance are inversely related (Margalef 1963).
To test this, an abundance definition of dominance of McNaughton (1967) was adopted:
Dl is calculated for the ith interval on component 1, Y ( , + ~ , is the sum of the two most
abundant specles in the ith interval, and Yi
is the total abundance in the ith interval.
The suggested inverse relation of richness and
dominance is demonstrated in Fig. 7. It should
be remembered that richness is defined here so
that it is logically independent of dominance.
The implication of decreasing dominance with
increasing richness is that the dominant species
are losing abundance as richness increases and
the total niche width is distributed among more
species.
Table 1 shows this implication not to be true
since the percentage of very rare species (i.e.,
less than 1.0 frequency) increases as the number
of species decreases and the frequency of the
dominants remains similar. It appears that the
Dl values are being influenced not by a decrease
in the numerator of the Dl formula, as is often
implied, but by a faster increase in the denominator than the numerator.
% species less than
Interval
1.0 frequency
Dominance
index
Dominant organisms do not decrease in
abundance. They simply are not able to increase
in abundance as fast as some of the smallerniched species. This implies that dominants are
less opportunistic in not being able to expand
as a group as the carrying capacity of the environment increases. Further, the class of species
which increases is the narrow-niched group
which was very rare when the carrying capacity
was lower. In fact, at the lower carrying capacity,
these species had very low population sizes and
probably suffered a corresponding high extinction rate. As the carrying capacity increased,
these species as a group showed a tendency to be
able to occupy this increasing carrying capacity
(i.e., to be opportunistic).
Enuirontnental Predictability
It has been suggested by several persons
1216
C A N . J. BOT. VOL. 55, 1977
(Skellam 1951 ; Hutchinson 1953, 1957; MacArthur and Levins 1964, 1967) that along an
environmental gradient the ability of an organism to occupy particular parts of that
gradient is in direct relation to its predictability
(i.e., the change from one segment of the gradient
t o the next). T o measure the change in predictability along the first component, the rates of
change of both the environmental structure and
the biologically determined carrying capacity
were determined.
The changes in slope between intervals of the
environmental structure (H,) and biologically
determined carrying capacity (K) were determined (Fig. 8). The predictability in environmental structure and in carrying capacity as
shown in Fig. 8 are relatively similar.
Thus the gradient on the component runs
from a more predictable environment to a less
predictable environment. Levins (1968) has
suggested on theoretical grounds that wideniched species are better adapted to uncertain
environments. This he explains on the basis
that species of more generalized tolerance are
needed in an environment which has low predictability. Also he shows that in less predictable
environments there should be fewer numbers of
small-niched species and these would be subject
to high turnover rates. Slobodkin and Sanders
(1969) have suggested a similar explanation.
They contend that low predictability determines
the properties of an organism that occupies that
kind of environment. They believe that low
predictability leads to organisms with high
extinction rates and selection for wide tolerance.
Wide tolerance allows the organism to better
cope with unpredictable changes by .being able
to adiust to a wide set of conditions.
he results presented here seem t o support
these ideas. In a less predictable environment
the wide-niched species (dominants) are more
important and there is a large number of very
small-niched species which have presumed high
turnover rates. These results suggest that a
plant on a gradient which is less predictable
would find it more advantageous to adjust to a
larger part of the gradient than to a smaller
part. Hence, in the terminology of MacArthur
and Levins (1964), the plant would respond to
the gradient as if it were 'fine-grained,' that is,
as if the positions along the gradient were so
finely divided that the plant would not respond
t o them differentially.
FIG.8. (Upper) Change in environmental complexity
(AH,) by intervals on component one of the vegetation
study. (Lower) Change in biologically determined
carrying capacity (AK) by intervals on component one
of the vegetation study.
Niche Parameter: Overlap
Niche overlap is a measure of that part of a
species niche width which is simultaneously
occupied by other species. This does not necessarily imply simultaneous use of the same resource in a competitive relationship. N o distinction is made between types of interaction in
this parameter.
Several measures of niche overlap have been
suggested (Moisita 1959; Horn 1966; Levins
1968). All of these measures have in common
that they are forms of interspecific association
and community resemblance. Niche overlap will
be calculated by interval using the formula
suggested by Horn (1966). From the discussion
of niche width, wide-niched species are not as
efficient at increasing in abundance in a pre-
JOHl
FIG. 9. Relation of total niche overlap (SO) versus
intervals along component one of the vegetation study.
dictable environment as narrow-niched species.
Ecologically this implies that wide-niched species
are not as efficient in interactions in these
environments. If in fact these interpretations of
niche width are reasonable, then the total overlap
per interval should be lower in the predictable
environments and higher in the unpredictable
environments. Whittaker (1960) has pointed
out that more overlap would mean more complete intermixing of the flora and less niche
differentiation. Figure 9 demonstrates these
ideas to be acceptable.
Conclusion
Most of the raised bogs in Washingtoil
County, Maine, are located in a narrow band
along the coast; they are never found more than
40 to 48 km inland. Most of the raised bogs
studied are located in sandy outwash. These
bogs probably developed from shallow marshes
or by paludification resulting from the formation
of an iron pan in flat outwash areas (McKeague
et al. 1968).
Raised bogs are characteristically considered
a 'severe' ecological system. The tree height
class is all but eliminated. The severity is not
caused by fire or other destructive influences.
Raised bogs can be viewed as ecological systems
which 'leak' nutrients. Raised bogs are supplied
mineral ions mainly froin atmospheric sources.
This supply is very scanty and in addition is
'leaked' from the system owing to the nearly
inoperable detritus cycle (Damman 1971), which
normally would recycle this supply.
The sensitivity of the system to very low
mineral ions is reflected by a general lack of
trees and the higher abundance of shrubs and
mosses (but not herbs) adapted to low mineralion budgets, e.g., Ericaceae and Sphagnum. It is
generally believed (see Whittaker 1954; Woodwell 1970) that reduction in number of strata
in vegetation is due to the disadvantage of large
height in severe systems. This is because the
greater respiratory demands of the live parts
of these stems cannot be met by the limited
efficiency of photosynthesis.
Niche analysis has shown that the two most
important environmental factors are related to
mineral-ion concentration: (1) atmospheric input difference~owing to proximity to the ocean,
and (2) a mineral-soil groundwater influence.
It has been further shown that these two
environmental factors represent two important
gradients in the surface water quality in the state
of Maine. Comparison of the structure of the
physical environment and the vegetation within
the raised bogs has demonstrated their similar
but non-lineal relationship.
Thus the plant composition in raised bogs
seems to be primarily queued to the mineral-ion
budget not only for its existence as a system but
for the variation and organization within that
system.
The concept of dominance has a long and
controversial history in ecology. Most recently,
McNaughton and Wolf (1970) have attempted
to show the importance of dominance in ecological systems. Adopting an abundance definition of dominance they hypothesize that it
arises out of unequal efficiencies at the interactions of the inter- and intra-population level.
Their hypothesis is not violated by any conclusions here. However, the concept of dominance
becomes excess baggage. If niche width and
dominance (high abundance) are positively
correlated, and if niche width is determined by
environmental complexity and predictability,
then dominance is a logical outcome of a wider
niche. It would seem more informative to speak
of niche width than of dominance.
From niche analysis it was learned that a
wide-niched species has a higher abundance and
should thus interact more with other individuals
of its population than with individuals of another
species. This interaction between individuals of
the same population will cause 'internal' (i.e.
within the population) divergence of the itzdividual's niches. This divergence between in-
CAN. J. BOT. VOL. 5 5 . 1977
dividuals will cause a widening of the population's niche. Thus the individuals of the population act as if they could, teleologically speaking,
distinguish parts of the environment, that is, as
if the environment were coarse grained. On the
other hand, the population would appear to be
unable to differentiate parts of the environment.
The population thus reacts to the environment
as if it were fine grained.
In the more complex and predictable parts
of the system, there are more narrow-niched
species and consequently more species. Narrowniched species have smaller populations and thus
individuals will have a greater chance of interacting with individuals of different species
populations. Niche differentiation will then occur
in response to other populations and niche
width will not increase. Narrow-niched populations appear to react to the environment as if
it were able to differentiate parts, that is, as if
it were coarse grained. The individual in these
populations on the other hand will appear to
be reacting to the environment with indifference
or as if the environment were fine grained.
Acknowledgments
I thank S. J. B. Johnson and W. D. Valentine
for their help in collecting the data in the field
and R. B. Pike and the late S. T. Pike for their
generous hospitality in Lubec. I thank A. R.
Hodgdon for checking my plant identification.
I also thank P. H. Allen, B. F. Chabot,
A. R. Hodgdon, S. J. B. Johnson, Y. T. Kiang,
W. B. Leak, J. W. Sheard, and W. D. Valentine
for their valuable critical suggestions on the
project and manuscript.
A grant from the Computer Cen'ter of the
University of New Hampshire made the necessary computer time available. E. A. Joslin
provided valuable help and advice on programming.
I acknowledge the facilities provided by the
Marine Biological Laboratory of Woods Hole,
MA, during the analysis of the data.
A grant from the National Cancer Institute
of Canada to V.S. Gupta aided in the preparation of the manuscript.
AHMAVAARA,
Y. 1954a. The mathematical theory of factorial invariance under selection. Psychometrika, 19:
27-38.
1954b. Transformation analysis of factorial data.
Ann. Acad. Sci. Fenn. 88: 1-150.
1957. On the unified factor theory of mind. Ann.
Acad. Sci. Fenn. 106: 1-176.
ALLEN,S. E., A. CARLISK,
E. J. WHITE,and C. C. EVANS.
1968. The plant nutrient content of rain water. J. Ecol.
56: 497-504.
AYALA,F. J. 1970. Competition, coexistence and evolution. In Essays in evolution and genetics in honor of
Theodosis Dobzhansky. Edited b y M. K. Hecht and
W. S. Steere. Appleton-Century-Crofts, New York. pp.
121-158.
BARKHAM,
J . P., and J. M. NORRIS.1970. Multivariate
procedures in an investigation of vegetation and soil
relations of two beech woodlands, Cotswold Hills, England. Ecology, 51: 630-639.
BRAUN,
E. L. 1950. The deciduous forest of eastern North
America. Hafner Publ. Co., New York.
BRAY,
J. R. 1961. A test for estimating the relative informativeness of vegetation gradients. J. Ecol. 49: 631-642.
BRAY,J. R.. and J . T . CURTIS.1957. An ordination of the
upland forest communities of southern Wisconsin. Ecol.
Monogr. 27: 325-349.
BROWN,I. C. 1943. A rapid method of determining exchangeable hydrogen and total exchangeable bases of
soils. Soil Sci. 56: 353-356.
CATTELL,
R . B. 1965. Factor analysis: an introduction to
essentials. I. The purpose and underling models. 11. The
role of factor analysis in research. Biometries, 21:
19&215,405-435.
CHAPMAN,
C. A. 1962. Bay-of-Maine igneous complex.
Geol. Soc. Am. Bull. 73: 883-888.
COLWELL,
R. K., and D. J. FUTUYMA.
1971. On the measurement of niche breadth and overlap. Ecology, 52:
567-576.
COTTAM,
G., and R. P. MCINTOSH.
1966. Vegetation continuum. Science, 152: 546-547.
CRUM,H. 1973. Mosses of the Great Lakes Forest. Univ.
Herb.. Univ. Mich., Ann. Arbor. MI.
CURTIS,J . T. 1956. A prairie continuum in Wisconsin.
Ecology. 36: 558-566.
DAMMAN,
A. W. H. 1971. Effect of vegetation changes on
the fertility of Newfoundland forest site. Ecol. Monogr.
41: 253-270.
R. B. 1966. Spruce-fir forests of the coast of Maine.
DAVIS,
Ecol. Monogr. 36: 79-94.
DIX,R. L., and J. E. BUTLER.1960. A phytosociological
study of a small prairie in Wisconsin. Ecology, 41:
316-327.
FARNHAM
R., S., and H. R. F I N N E Y1965.
.
Classification
and propertiesoforganicsoils. Adv. Agron. 17: 115-162.
FERNALD,
M. L. 1950. Gray's manual of botany. 8th ed.
American Book Co., New York.
GIBBS,R. J. 1970. Mechanisms controlling world water
chemistry. Science, 170: 1088-1090.
GOFF, F. G. 1964. Structure and composition of two oak
woods in the University of Wisconsin Arboretum. M.Sc.
Thesis. Univ. Wisconsin, Madison, WI. Unpublished.
GOODALL,
D. W. 1954. Objective methods for the classification of vegetation. 111. An essay in the use of factor
analysis. Aust. J. Bot. 2: 304-324.
1969. A procedure for recognition of uncommon
species combination in sets of vegetation samples. Vegetatio, 18: 19-35.
GORHAM,
E. 1955. On the acidity and salinity of rain.
Geochim. Cosmochim. Acta, 7: 231-239.
JOHNS O N
-1957. The development of peatlands. Q. Rev. Biol.
32: 145-166.
-1961. Factors influencing supply of major ions to
inland waters, with special reference to the atmosphere.
Bull. Geol. Soc. Am. 72: 795640.
-1967. Some chemical aspects of wetland ecology.
Proc. 12th Annu. Muskeg Res. Conf.. Tech. Mem. No.
90. pp. 2638.
GRANT,D. R. 1968. Recent coastal submergence of the
Maritime Provinces, Canada. Can. J. Earth Sci. 7:
676-689.
GREIG-SMITH,
P. 1964. Quantitative plant ecology. Butterworth and Co., London, England.
HEINSELMAN,
M. L. 1963. Forest sites. bog processes and
peatland types in the Glacial Lake Agassiz Region,
Minn. Ecol. Monogr. 33: 327-374.
-1970. Landscape evolution, peatland types, and the
environment in the Lake Agassiz Peatland Natural Area,
Minn. Ecol. Monogr. 40: 235-261.
HORN,H. S. 1966. Measurement of "overlap" in comparative ecological studies. Am. Nat. 100: 419424.
HURLBERT,
S. H. 1971. The nonconcept of species diversity: a critique and alternative parameter. Ecology, 52:
577-586.
HUTCHINSON,
G. E. 1953. The concept of pattern in ecology. Proc. Acad. Nat. Sci. Philadelphia, 105: 1-12.
-1957. Concluding remarks. Cold Spring Harbor
Symp. Quant. Biol. 22: 415-427.
IVIMEY-COOK,
R. B., and M. C. F. PROCTOR.
1967. Factor
analysis of data from an east Devon heath: a comparison
of principal component and rotated solutions. J. Ecol.
55: 405413.
JUNGE,
C. E., and P. E. GUSTAFSON.
1957. On thedistribution of sea salt over the United States and its removal by
precipitation. Tellus, 9: 164-173.
JUNGE,C. E.. and R. T. WERBY.1958. The concentration
of chloride, sodium, potassium, calcium, and sulfate in
rain water over the United States. J. Meteorol. 15:
417425.
KENDALL.M., and A. STUART.1961. The advanced
theory of statistics. Vol. 11. Inference and relationship.
Hafner, New York.
KLOPFER,
P. H., and R. H. MACARTHUR.
1960. Niche size
and fauna diversity. Am. Nat. 94: 293-300.
LAMBERT,
J. M., and M. B. DALE.1964. The use of statistics in phytosociology. Adv. Ecol. Res. 2: 59-99.
LAUTZENHEISER,
R. E. 1960. Climatography of the U.S.
climatological summary. U.S. Dep. Commerce,
Weather Bureau, Washington, DC.
LEVINS,R. 1968. Evolution in a changing environment.
Monograph in population biology. 2. Princeton Univ.
Press, Princeton, NJ.
LEYTON,L., E. R. C. REYNOLDS,
and F. B. THOMPSON.
1968. Iqterception of rainfall by trees and moorland vegetation. In The measurement of environmental factors in
terrestrial ecology. Edirrd by R. M. Wadsworth.
Blackwell Scientific Publ., Oxford, England. pp. 90-108.
LLOYD,M., and R. J. CHELARDI.
1964. A tableforcalculating the equitability component of species diversity. J.
Anim. Eco1.33: 217-225.
L o u c ~ s 0.
, L. 1962. Ordinating forest communities by
means of environmental scalars and phytosociological
indices. Ecol. Monogr. 32: 137-166.
MACARTHUR,
R. H. 1957. On the relative abundance of
I219
bird species. Proc. Natl. Acad. Sci. (U.S.A.), 43:
293-295.
1965. Patterns of species diversity. Biol. Rev. 40:
516533.
1968. The theow of the niche. In Pooulation biology and evolution. ~ i i r r dby R. C. ~ e w o n i i nSyracuse
.
Univ. Press, Syracuse, NY. pp. 159-176.
MACARTHUR,
R. H., and R. LEVINS.1964. Competition,
habitat selection and character displacement in a patchy
environment. Proc. Natl. Acad. Sci. (U.S.A.), 51:
1207-1210.
1967. The limiting similarity, convergence and divergence of coexisting species. Am. Nat. 101: 377-385.
MACARTHUR,
R. H., and J. W. MACARTHUR.
1961. On
bird species diversity. Ecology. 42: 594-598.
MARGALEF,
R. 1963. On certain unifying principles in
ecology. Am. Nat. 97: 357-374.
MCINTOSH,
R. P. 1958. Plant communities. Science, 128:
115-120.
MCKEAGUE,J. A., A. W. H . D A M M A Nand
,
P. K.
HERINGER.
1968. Iron-manganese and other pans in
some soils of Newfoundland. Can. J. Soil Sci. 48:
243-253.
MCNAUGHTON,
S. J. 1967. Relationships among functional
properties of California grassland. Nature, 216: 168-169.
MCNAUGHTON,
S. J.. and L. L. WOLF. 1970. Dominance
and the niche in ecological systems. Science, 167:
131-139.
MILLIMAN
J ., D., and K. 0 . EMERY.1968. Sea levels
during the past 35,000 years. Science, 162: 1121-1 123.
MOISITA,M. 1959. Measuring of interspecific association
and sim~laritybetween communities. Mem. Fac. Sci.,
Kyushu Univ., Ser. E (Biol.), 3: 64-80.
MORRISON,
D. F. 1967. Multivariate statistical methods.
McGraw-Hill, New York.
OOSTING,
H. 1959. The study of plant communities. W.H.
Freeman Press, San Francisco.
OSBURN.
W. S.. JR. 1969. Accountability of nuclear fallout
deposited in a Colorado high mountain bog. In Symp.
Radioecology. Edited by D. J. Nelson and F. C. Evans.
AEC-CONF 670503. pp. 578-586.
Prco, M. M., P. MALDONADO,
and R. LEVINS.1965. Ecology and genetics of Puerto Rican Drosophila. 1. Food
preferences of sympatric species. Caribb. J. Sci. 5:
29-37.
PIELOU,
E. C. 1966. The use of ~nformationtheory in the
study of the divers~tyof biological populations. Proc. 5th
Berkeley Symp. Math. Stat. Probab. 4: 164-177.
PRESTON,
F. W. 1962. The canonical distribution of commonness and rarlty. Parts I and 11. Ecology, 43: 185-215,
410-432.
1966. The soils of
ROURKES,
R. V., and J. S. HARDESTY.
Maine. Maine Agric. Exp. Stn. Misc. Publ. 676.
SEAL, H. 1964. Multivariate statistical analysis for
biologists. Methuen and Co. Ltd., London, England.
SHANNON
C., E., and W. WEAVER.
1949. The mathematical theory of communication. Univ. Illinois Press, Urbana, Illinois.
SIMPSON,
E. H. 1949. Measurement of diversity. Nature,
163: 688.
SKELLAM,
J . G. 1951. Random dispersal in theoretical
populations. Biometrika, 38: 196-218.
SLOBODKIN,
L. B., and H. L. SANDERS.1969. On the
1220
C A N . J . BOT. VOL. 5 5 . 1977
contribution of environmental predictability to species
diversity. Brookhaven Symp. Biol. 22: 82-95.
SONNESON,
M . 1969. Studies on mire vegetation in the
Tornestrask area, northern Sweden. 11. Winter conditions of the poor mires. Bot. Not. 122: 481-51 I.
1970. Studies on mire vegetation in the Tornestrask
area, northeln Sweden. 111. Communities of the poor
mires. Opera Bot. 26.
SPARLINC,
J . H. 1967. The occurrence of S ~ ~ h o e tnigri~~rs
cans L . in blanket bogs. I. Environmental conditions
affecting the growth of S. tligricnns in blanket bog. 11.
Experiments on the growth of S. nigriccrns under controlled conditions. J. Ecol. 55: 1-14, 15-31.
S W A NJ.
, M. A. 1970. An examination of some ordination
problems by use of simulated vegetational data. Ecology, 51: 89-102.
U.S. Department of Commerce. 1964. Climatography of
the U.S. No. 8 6 2 3 . Climatic summaly of the U.S. Supplement for 1951 through 1960. New England. U.S. Gov.
Print. Off., Washington, DC.
WALKER.
B. H., and R. T. COUPLAND.
1968. An analysis
of vegetation-environment relationships in Saskatchewan sloughs. Can. J. Bot. 46: 509-522.
WALKER,
B. H.. and C. F. WEHRHAHN.
1971. Relationships between derived vegetation gradients and measured environmental variables in Saskatchewan wetlands. Ecology, 52: 85-95.
WHICKEN,
F. W., and C. M. LOVELESS.1968. Relationships of physiography and microclimate to fallout deposition. Ecology, 49: 363-366.
WHITTAKER,
R. H. 1954. The ecology of serpentine soil. I.
Introduction. IV. The vegetation response to serpentine
soils. Ecology. 35: 258-259,275-288.
-1960. Vegetation of the Siskiyou Mountains, Oregon and California. Ecol. Monogr. 30: 279-388.
-1965. Dominance and diversity in land plant communities. Science, 147: 266290.
1967. Gradient analysis of vegetation. Biol. Rev.
49: 207-264.
-1972. Evolution and measurement of species diversity. Taxon, 21: 213-251.
WOODWELL,
G. 1970. Effects of pollution on the structure
and physiology of ecosystems. Science, 168: 429433.
Download