Page et al 1985
The reaction of some sand-dune plant species to experimentally imposed environmental
change: a reductionist approach to stability*
R. R. Page, S. G. da Vinha & A. D. Q. Agnew
Department of Botany and Microbiology, University College of Wales, Aberl,stwyth, Wales, S Y23 3 DA,
U.K.
Keywords: Coastal sand dune, Galium verum, Grazing, Onions repens, Stability, Trampling
Abstract
Coastal dunes in western Europe have long been under grazing and recreational pressure, but the
vegetative cover appears to be remarkably resilient. In a series of experiments on Ynyslas Dunes, Cardigan
Bay (Wales, U.K.), trampling, protection, fertiliser, and herbicide treatments were monitored using non-
destructive and destructive sampling methods. This paper discusses the behaviour of 12 important species
after these treatments on three major dune habitats: Agrostis tenuis, Ammophila arenaria, Bellis perennis,
Festuca rubra, Galium verum, Leontodon taraxacoides, Ononis repens, Plantago lanceolata, Poa subcaeru-
lea, Thymus praecox, Homalothecium lutescens, Tortula ruraliformis.
Although stability is usually thought of as a feature of the community, a reductionist approach suggests
that grasses react quickly but temporarily, dicotyledonous forbs often show slower reaction times, each
species reacting in its own way to changes in the environment. A tentative classification of our study species is
possible.
Introduction the older dunes where Rhytidiadelphus triquerus is
an indicator of such conditions.
Ynyslas Dunes (Fig. 1) are a western dune system The area is a National Nature Reserve but is not
associated with the mouth of the Dovey Estuary in closed to the public and research was set up in 1974
Cardigan Bay, Wales. The dunes are calcareous with the aim of providing information helpful to the
(CaCO 3 content 3-5%) and have grown extensive- management of increasing recreation pressure on
ly in the last 100 years. They show a series of classi- the dune system. Visitors are on foot, often using
cal dune vegetation types (Ranwell, 1972) of strand the dunes as a passageway to the open beaches, and
plants, embryo dunes, mobile dunes with much often as an educational facility. Over 500 000 per-
open sand between Ammophila arenaria tussocks, son-days were recorded in 1974-75, and it was clear
semi-fixed mossy dunes with Tortula ruralis ssp. that damage was being done. At the same time the
ruraliformis, Homalothecium lutescens and Pelti- rabbit population recovered from the effects of
gera canescens/rufescens, and fixed grey dunes myxomatosis and grazing was increasing.
where plant cover reaches 100% and Festuca rubra The first research objective was to separate and
is often dominant. There is no dune heath since the quantify the effects of human trampling and rabbit
dunes are too young and calcareous but some acidi- grazing by means of exclosures and experimental
ty does develop in the winter-flooded slacks and on treatments. The second was to investigate the ef-
fects of the management tools of fertilisers and
*Nomenclature follows Smith (1978), for bryophytes and selective herbicides. This is not a new approach,
Clapham, Tutin & Warburg (1962). for angiosperms. having been extensively reported (Blom, 1977; Hyl-
Vegetatio 61, 105-114 (1985).
© Dr W. Junk Publishers, Dordrecht. Printed in the Netherlands.
106
further trampling. For these four series of investi-
gations human passage was monitored using pres-
sure sensitive counters in paths, as well as trip pins
(Bayfield, 1971)and direct observation, while plant
species occurrence was estimated as cover percent-
age in small permanent quadrats, never larger than
30 X 30 cm. Path trampling was measured in units
of person per metre transect across the path. For
example, 30 people using the central 25 cm cross
section of a path are said to be 120 people m i. In
this way various problems of standardisation were
overcome. Paths were monitored for 40 summer
days in the study of established paths and all year
for the experiment on increased trampling of paths.
For the experimentally increased use of estab-
lished paths trampling (in soft soled walking shoes,
exerting some 150 g m cm 2) was applied at multi-
ples of current recreational use of the paths. The
experiments lasted one year starting in the summer
of 1974, with vegetation cover being estimated at
the start and finish and soils analysed upon comple-
tion.
Exclosures of nearly 20 m 2 area were constructed
in 1973, half inside rabbit-proof wire mesh and the
rest only restricting human access. Cover estima-
tion in small quadrats was again used to monitor
the effect of protection. These areas were after-
Fig. I. The location of Ynyslas Dunes, Wales, with dune habitat wards (April 1977) used to follow the reaction of
types diagrammatically represented. Experimental sites are species to release from competition (by a single
shown as black blocks. application of selective weedkillers) combined with
fertiliser treatments. Here the standing crop was
harvested from random quadrats of 15 X 15 cm.
gaard, 1981; Liddle & Greig-Smith, 1975; Liddle & The subdivision of such fragmentary material into
Moore, 1974; Streeter, 1971 are the more recent species contributions only proved possible for the
references on this topic). Nevertheless, there is a more abundant species, and this part of the record
need to repeat experiments and validate predictions is incomplete.
for specific areas. This paper summarises some as- In field experiments natural population fluctua-
pects of reports by Page (1980) and Da Vinha tions may easily override treatment effects. In these
(1981) on research undertaken from 1975-80. studies control areas were designated and experi-
mental results related to concurrent changes in the
controls. Clearly, this assumes that the latter are
Methods due either to the current weather (for instance 1975
and 1976 had dry summers), or to the continuation
The study of this dune system was started with an of general vegetation change in the area of the
investigation of trampling impact in four ways: 1) experiment. Although open to criticism, we feel
the comparison of soil and species differences on that this approach is better than using a starting
long-established paths and undisturbed areas, 2) population as the only control value against which
experimental increase of the trampling on existing to measure change.
paths, 3) the experimental trampling of undis- S oils were analysed by taking replicated cylindri-
turbed areas, and 4) the protection of paths from cal cores of 4.7 cm diameter and 5 cm depth. Com-
107
bustion and oven drying were used to measure or- LEONTODON TARAXAC01DES [
TORTULA RURALIFORMI S [
ganic matter and soil moisture content respectively, AW0PHILA AP~ARL~
and dry bulk density by displacement with acetone H O M A L O T ~ IUM LVfESC~S [
POA Sg~CAm~gL~A ..~
(Gupta, 1933). Using these data percentage satura- FESTUCA RUBRA
tion of pore space was also calculated. All cores THYMUS PRAECOX , [
were taken after a period of heavy rain followed by ONONIS 2EPm~S [
GALIUM VERUM |
twelve dry hours in which state they should be near pLANTAGO LANCEOLATA ' I
field capacity (Warkentin, 1971). AGROSTIS ~]~VUIS
BELLIS P E R i l S I
19 , 2p , 3,o
VOLUMETRIC PERCENTAGE W T R
AE
Species performance LEONTODON TARAXACOIDES
~0~ULA RURALIFORMIS
HOMALOTHECIUM LUTESCENS I
Great quantities of data were gathered during the
six years of the investigation. T w o species, Ononis A ~ O P H n a A~m~mZA I
P0A SUBCAERULEA
repens and Galium verum, are taken as examples FESTUCA RUBR~
because data on these two are relatively complete 0NONIS REPENS [
BELLIS P E ~ I S --~
and are reported on in detail.
GALI UM VERUM I
pLANTAGO LANCEOLATA I
Ononis repens AGROSTIS TENUIS q .... 5. . I. . . IQ
PERCENTAGE LOSS ON IGNITION
TOHTULA RURALIFORMIS --~
General distribution LEONTODO~ TARAXACO I ~ S
The mean % pore space, organic matter and HOMALOTHECIUM LUTESC~S ~.~
BELLIS PEF~NNIS -.~
moisture levels at which this species occurs are in I
zcn~s PVaEOOX
the middle of the range of species reported on FESTUOA R U ~ --~
POA SUBCAERULEA
(Fig. 2), reflecting its widespread distribution. On
ONONIS H~PE~S
untrampled soils, a l t h o u g h c o m m o n , its percentage AF~IOPHILA A ~ A R I A |
cover is negatively correlated with soil moisture
PLANTAGO LANCEOLATA I
when all fixed and semi-fixed sites are included in AGROSTIS TENUIS
the analysis (r = -0.62, Table 1) and with a greater
50 ~ -+-7o
PERCENTAGE PORF SPACE
J ~
correlation coefficient (r = -0.82) when fixed dune
sites are analysed separately. N o w on paths there is Fig. 2. Ranked species occurrence along ranges of the various
more Ononis on the drier semi-fixed dunes, even soil factors which are associated with trampling, from 40 moni-
t h o u g h path soils are moister than untrampled tored sites which do not all correspond with experimental areas
soils, and there is no correlation between Ononis used in Figures 3 and 4. the mean value and one standard
deviation either side of it is represented.
cover and soil moisture on paths. However, it is
only on the fixed dunes that there is a negative
relationship between trampling and soil moisture,
and the most trampled semi-fixed dune paths are one year. In the absence of any treatment, cover
used by fewer people than the most trampled fixed changed to 156% and 77% of starting (i 974) values
dune paths. Since Ononis repens is apparently un- for semi-fixed and fixed dunes respectively. 1974
able to resist high levels of trampling(see below), its had 29% more days with over 2 m m of rain than
cover is low on the driest fixed-dune soils and high 1973, and the increase in cover of Ononis probably
on the inherently dry semi-fixed dunes. reflects its better performance during moist weather
in the open vegetation of the semi-fixed dune.
Experimental changes in trampling and grazing Thus, on two fixed dune paths monitored during
pressure the same year, soils were as dry as in the dry semi-
There are a n u m b e r of suggestions, arising f r o m fixed dune site used for the trampling but grasses
field experiments, that competititve effects are im- (mainly Festuca rubra and Ammophila arenaria)
p o r t a n t in determining cover of this species. For were c o - d o m i n a n t with Ononis repens which de-
example, in Figure 3 is shown Ononis cover after clined significantly during the year.
trampling was imposed on 2 undisturbed sites over These observations and experiments suggest that
108
Table 1. Significant species correlations in random quadrats on trampling was intermediate in its response. Relative
and off paths with soil factors. LOI, loss on ignition; PS, per- to the control here, there was no significant change
centage pore space; M FS, percentage moisture at field capacity;
in total cover of all species or grasses taken alone.
NS, no significant correlation; +, correlation positive; , correla-
tion negative. However, cover of some species such as Festuca
rubra whose mean cover in the exclosure at the start
Untrampled soils Path soils of the experiment was 30.9% increased relative to
Agrostis tenuis +: LOI, PS, MFS +: LOI, MFS
the control in the rabbit-grazed but not in the
A mmophila rabbit-proof parts, whilst others such as Agrostis
aretlaria +: LOI, PS NS tenuis, Ammophila arenaria and Holcus lanatus
Bellis perennis NS NS (whose total mean cover in the exclosures at the
Festuca ruhra +: LOI, PS, MFS NS
start of the experiments was 6.1%) declined in both
Galium v e r u m NS +: LOI, PS, MFS
Leontodon exclosures, but only significantly in the rabbit-
taraxacoides NS NS proof treatment. Soil moisture content and com-
Ononis repens : MFS NS paction levels were slightly, but significantly, lower
Plantago lanceolata NS +: LOI, PS, MFS in the rabbit-proof than the rabbit-grazed treat-
Poa suhcaerulea +: PS +: LOI, PS, MFS
ment and control soils. Broadly speaking then, the
Th.l'mUS praecox : LOI, PS NS
Tr(/~)lium repens NS +: LOI, MFS changes in cover of Ononis repens could be ex-
Tortula plained by assuming that trampling was so impor-
rural(lormis : LOI, MFS : LOI, MFS tant that release from its suppressive effect allowed
Ononis to compete successfully with grasses, but
the greater resistance to trampling on the fixed
competition is related to moisture supply in deter- dunes (Fig. 3) may relate to reduced competition
mining cover of Ononis repens, but the summer of from grasses combined with the more equable
1975 was particularly dry after April and the tram- microenvironment and soil conditions.
pling experiments were terminated earlier than the When trampling was applied at a range of fre-
exclosure experiments, thus making prediction dif- quencies, but the same intensity (150 passages m i
ficult. Table 2 shows changes in Ononis cover dur- year I), the least damage was caused on both the
ing 1974/1975 in an exclosure on a fixed dune area fixed and semi-fixed dunes by spreading the tram-
previously little used by the public. Ononis fared pling over one year at weakly rates that were a
worst in the control, increased in the: rabbit-proof percentage of actual use made of a path by the
part and in the part simply protected from human public (Fig. 3). On the fixed dunes there was no
Table 2. Differences in percentage cover after 2 years of protection from trampling and grazing or trampling alone, in 3 sites. Control is
an area which continued to receive use. Asterisk marks differences from control at 5% or better.
Fixed dune high-use site Fixed dune low-use site Trampled slack site
Con- Trampling Grazing & Con- Trampling Grazing & Con- Trampling Grazing &
trol prevented trampling trol prevented trampling trol prevented trampling
prevented prevented prevented
Agrostis tenuis +3.0 -5.0" 1.0" +4.5 + 1.0 1.0"
Ammol]hila arenaria +1.5 +0.5 -2.5
Poa subcaerulea + 1.5 0.5 + 1.5 4.5 I 1.0 1.5
Festuca rubra 7.5 -2.5 6.0 -32.0 15.0' 25.0
Bellis perennis -2.0 2.0 4.0*
Galium verum +4.5 +7.5 +5.0 12.0 15.0 1.5
Leontodon
taraxacoides +2.0 11.0" +3.0
Ononis repens 12.5 -6.5* +12.0'
Plantago lanceolata 15.5 -5.5 8.5*
Thymus praecox +4.0 + 17.5 2.0 +7.5 +2.0 2.0
109
A B C A B C D E
C N R L 01~ ~
O T O CONTROL 01 i.o I I---7
A! i .I ! 1
B I**,, I 4.14.*, I
200
IJJ
Z 150 '~" O ~ o
"r"
W 100
O
t,.)
LU
Z
50
\/°\
°
W O
n A
I
B C
I
A B c ? E,
74 296 1184 0 74 148 296 592 1184
passages m-1 yr-1 passages m-1 yr-1
O T Y O T Y
CI t t t
150 150~
i,i
W~
c.oz 100 100
I---r
r,-~
Ww 50 5O
C O T Y C 0 T
TET ET
RAMN TREATMENT
Fig. 3. Cover changes expressed as percentage of the starting value in Ononis repens populations after trampling treatments in previously
undisturbed sites. Above are shown the effect of trampling increases on (left) a semi-fixedand (right) a fixed dune site. Beloware shown
the effects of different frequencies of application of trampling equivalent to 150 passages m i year 1 as: C no trampling control; D, all in
one day; T, all in one week; Y, distributed throughout the year. Triangular matrices give significance of t tests between means
(+, 0.5 > p > 0.1 ; ++, 0.1 > p > 0.01 ; +++, p < 0.01 ) because heterogeneity of the populations precluded the use of variance analysis.
significant difference in relative cover from the con- of drought. U n d e r more average weather condi-
trol. tions results may have been different. W h e n all the
A p p l i c a t i o n of the t r a m p l i n g over ten weeks or t r a m p l i n g was applied on one day the c o n t e m p o -
all in one day resulted in some reduction in relative rary soil moisture c o n d i t i o n was also p r o b a b l y very
cover in c o m p a r i s o n with the c o n t r o l on both sites, i m p o r t a n t . However, it was possible to find a rela-
but this was not significant in the fixed dunes. O n t i o n s h i p between the changes in O n o n i s in these
the semi-fixed d u n e s the ' t e n - w e e k ' t r e a t m e n t was e x p e r i m e n t s a n d changes in measured soil factors,
the most d a m a g i n g . It must also be repeated that of which c o m p a c t i o n was the only one to be affect-
the s u m m e r of 1974 in which the e x p e r i m e n t started ed by the e x p e r i m e n t s o n the fixed dunes. Cover of
was particularly h u m i d a n d 1975 included a period O n o n i s r e p e n s was not correlated with percentage
110
pore space on paths or u n t r a m p l e d soils. The lack r e m o v a l alone resulted in a p p a r e n t l y p e r m a n e n t
of any change in moisture levels was u n e x p e c t e d d a m a g e to Ononis p o p u l a t i o n s . Willis & Y e m m
because a significant c o r r e l a t i o n between c o m p a c - (1961) f o u n d that Ononis repens was e r a d i c a t e d
tion a n d m o i s t u r e levels was f o u n d in b o t h long f r o m turf t r a n s p l a n t s in the greenhouse (to which
established p a t h s and u n t r a m p l e d soils. The results, either c o m p l e t e nutrients, N + P + K or nitrate only
therefore, suggested that not all soil factors are in were a d d e d ) by the dense g r o w t h of Festuca rubra,
e q u i l i b r i u m with t r a m p l i n g pressure after one years' but we have not c o n f i r m e d this in the field. Here,
trampling. fertiliser t r e a t m e n t a l o n e had little effect but there
was a chance a b s e n c e of Ononis u n d e r this treat-
Field manipulation o f nutrients and competition ment in the semi-fixed d u n e site.
D i s t u r b a n c e resulted in c o n s i d e r a b l y m o r e
c h a n g e in the semi-fixed site ( m o s s y dunes) t h a n in S u m m a r y o f O n o n i s repens performance
the fixed d u n e site ( T a b l e 3), as w o u l d be expected, Ononis repens is a c o n s p i c u o u s plant, easily ob-
a n d this was reflected in the b e h a v i o u r of Ononis. served a n d o f wide d i s t r i b u t i o n in W e s t e r n Europe.
On b o t h sites fertiliser a d d i t i o n c o m b i n e d with se-. T h e r e is m u c h i n f o r m a t i o n a n d yet the picture of its
lective d i c o t y l e d o n o u s r e m o v a l resulted after two e c o l o g y a n d b e h a v i o u r in the face of a c h a n g i n g
years in e n h a n c e d p o p u l a t i o n s of Ononis (which e n v i r o n m e n t r e m a i n s r a t h e r confused. F r o m our
r e i n s t a t e d itself very q u i c k l y f r o m rhizomes). O n l y analyses we can say that p o p u l a t i o n s fluctuate
on the semi-fixed dunes was there a significant widely, p r o b a b l y c o n t r o l l e d by seasonal weather
c h a n g e in this species when m o n o c o t y l e d o n s were variables, a n d a rather c o m p l e x interaction be-
r e m o v e d but an excessive v a r i a n c e in the fixed d u n e tween soil moisture, t r a m p l i n g and c o m p e t i t i o n .
site m a y m a s k a real effect here. In b o t h sites dicot But the species is an o p p o r t u n i s t i c one, c a p a b l e of
r a p i d e x p l o i t a t i o n of t e m p o r a r i l y f a v o u r a b l e sites.
Table 3. Ononis repens above ground dry weight mean g • m 2
C o m p e t i t i o n seems to play a key role in switching
after treatment in April 1977 as follows: M, monocotyledon
selection herbicide and M + F, with NPK fertiliser; -D, the species to o c c u p y resources, as is indicated by its
dicotyledon herbicide and -D + F, with NPK fertiliser; F, NPK increase u n d e r p r o t e c t i o n f r o m t r a m p l i n g in one
fertiliser alone; C, control. Asterisks mark significances better site a n d when p r o t e c t e d f r o m c o m p e t i t i o n with
than 5%. NP, species not present in this treatment. o t h e r d i c o t y l e d o n s after herbicide treatment.
A g a i n , g r a z i n g s h o u l d be c o n s i d e r e d as an interac-
A. Grey dunes, untrampled
tive effect.
Treatments June August June F ratio
1977 1977 1978 Galium verum
-M 18.4 23.5 45.3 1.3
-M + F 26.2 54.7 76.1 3.6 General distribution
-D 11.3 17.4 8.0 0.6
This species is c o n f i n e d to the fixed dunes. T h e
-D + F 7.3 40.4 53.7 5.7*
F 68.4 73.7 92.5 0.7 significant c o r r e l a t i o n s between percentage cover
C 13.6 50.9 47.3 4.0* of this species a n d soil moisture, measured volu-
F ratio 4.7* 2. I 5.5* metrically, (r = 0.58) o r g a n i c m a t t e r (r = 0.61) a n d
p o r e space (r = 0.50) when all p a t h sites, b o t h on the
fixed a n d semi-fixed d u n e s are included in the anal-
B. Semi-fixed dunes, previously trampled ysis is t h e r e f o r e to be e x p e c t e d ( T a b l e 1). T h e r e was,
however, no significant c o r r e l a t i o n between per-
Treatments June August June F ratio
1977 1977 1978 centage cover a n d a n y m e a s u r e d soil f a c t o r when
fixed d u n e paths were a n a l y s e d separately. Galium
-M 12,.1 11.7 88.2 4.1
-M + F 7.7 NP 17.2 2.6 verum was present at t o o low a p e r c e n t a g e cover on
-D NP 1.0 5.8 0.8 u n t r a m p l e d fixed d u n e sites to allow regression
-D + F 46.6 38.0 147.1 8.4* analysis to be u n d e r t a k e n ,
F NP NP NP NP
C 66.4 60.4 106.2 1.3 Experimental trampling change
F ratio 4.9* 6.3* 14.9"
Galium verum is p r e s e n t on the most highly
111
1°°
/ The difficulty of dealing with data from c o n t r o l
sites is seen in the changes in G a l i u m v e r u m on
A B C 0
0J I***1~* I *
soF c o n t r o l areas a d j a c e n t to enclosures (Table 2). O n a
moderately t r a m p l e d fixed d u n e site percentage
o~ jo cover r e m a i n e d statistically u n c h a n g e d between
1973 a n d 1975 whilst it declined over this period on
8854 I'7"/06 26562 35416 442?0
0 A B C O a lightly t r a m p l e d site. The difference in changes in
passages rn-1 yr-1
percentage cover on controls is also seen in an
Fig. 4. Cover changes expressed as percentages of starting values e x p e r i m e n t in which levels of t r a m p l i n g on paths is
in Galium verum after increasing the trampling of a previously increased. O n the heavily used path (Fig. 4) cover of
used path. Significancesshown as in Figure 3. G a l i u m v e r u m declined significantly over the ex-
p e r i m e n t a l year, but remained statistically un-
changed on a moderately trampled path in an
t r a m p l e d path recorded on the dunes, used by e x p e r i m e n t not reported here.
8 854 people m i y e a r l , a n d a l t h o u g h the mean Relative to changes on the control, there was a n
percentage cover was only 1.7%, only three other increase in cover of G a l i u m v e r u m on a previously
d i c o t y l e d o n o u s species had a greater cover. O n this little trampled enclosure site (68% grass), only when
path there was an a p p a r e n t l y steady decline in rela- rabbits as well as h u m a n s were barred access (Ta-
tive cover with e x p e r i m e n t a l l y increased t r a m p l i n g ble 2), but on a previously more trampled site (20%
until a 300% in use was applied (Fig. 4). grass), cessation of t r a m p l i n g and rabbit grazing
A p a r t from a significantly higher moisture level had no effect relative to changes on the control.
in the e x p e r i m e n t a l area suffering a 300% increase Given the occurrence of G a l i u m v e r u m in rela-
in t r a m p l i n g , which was p r o b a b l y not related to the tion to soil factors, it is p r o b a b l e that the relative
t r e a t m e n t , there were no significant differences be- increase (or rather failure to decline) on the pre-
tween measured soil parameters on the trampled viously little t r a m p l e d site is related directly to ces-
area a n d the control, a n d the effect of increased sation of grazing, a n d the difference between the
t r a m p l i n g on the species in this e x p e r i m e n t is most changes on the two enclosure sites relates to compe-
likely a direct one. tition from grasses.
Table 4. Galium verum above ground phytomass dry weight in g • m 2 in a fixed dune enclosure after 4 years of protection from
trampling. Treatments as in Table 2. Below is shown ranges of values within which there are no significant differences at 5% or better.
Asterisks mark significanceat 5% or better.
Treatments June 1977 August 1977 June 1978 F ratio
-M 4.4 15.6 15.3 2.1
-M + F 21.7 89.2 62.8 10.4"
D 4.9 10.6 9.8 0.9
-D + F 4.1 21.8 21.5 4.5
F 20.4 27.5 5.6 5.6*
C 14.2 21.1 21.0 1.0
F ratio 4.1" 10.1' 12.9"
June 1977 -M + F F C -D -M -D + F
August 1977 M+ F F -D+ F C M D
June 1978 -M+ F -D + F C -M D F
112
Field manipulation of nutrients and competition While Homalothecium lutescens and Tortula ru-
Galium was present only in one site in sufficient raliformis are semi-fixed dune moss species which
quantities for analysis of these manipulations. This are associated with paths on the fixed dunes, there
is an area of heavily trampled and grazed fixed- is no large group of species which is distributed in
dune vegetation which resembled a close cropped this way, and, apart from a short list of winter
lawn in 1973 when enclosed. Considerable phyto- ephemerals, it is difficult to suggest others, except
mass had accumulated by 1977 when treatments for Leontodon taraxacoides and Plantago lan-
were applied to a totally protected section, although ceolata, where evidence is less clear. This merely
it was still more uniform and less tussocked than confirms that paths, although developed through
undamaged fixed-dune vegetation. Table 4 shows disturbance of fixed dune vegetation, are not com-
that it is only the fertilised treatment in which parable with a successional stage.
monocotyledons had been killed which showed en- The fast recovery of Festuca rubra is of interest.
hanced growth of G. verum, but there is little sign of This is a dominant species of fixed dunes. Willis
a return to normal level after 18 months. (1963) has documented the same sort of reaction to
increased NPK as we have here, but in this case the
Summary of Galium verum performance fertiliser inputs were sustained and the fast rever-
There is much less evidence here than in the case sion to lower production was not observed. Our
of Ononis repens. This species is particularly asso- pulsed treatments allow us to distinguish between
ciated with paths on the fixed dunes, and is resistant grasses and most dicotyledons in their rates of re-
to trampling. It reacts very positively indeed when version to original biomass. Some dicotyledonous
monocotyledons are removed, so that we suspect it species are stress tolerant in the sense of Grime
is sensitive to competition with grasses as its mor- (1979) (Ononis, Galium, Thymus) and their slow
phology would suggest. Its performance fluctuated recovery from perturbation is no surprise, but the
from year to year in the control of some experi- instability of their resident populations is unex-
ments, yet in the perturbed fixed-dune site it was pected in this respect as is their response to the
remarkably stable and recovery in the latter site was removal of grasses, taken to be competitors. We
rather slow. may attempt to categorise the species reported
upon here as follows.
A. Those which have constant cover resistant to
Discussion change and which reinstate quickly after release
from stress. The grasses are examples of this type.
The accounts of the two species' reaction to en- B. Those having cover varying from year to year
vironmental change illustrate the complexity of the with fast reaction to environmental change and,
data base from which we work, and the difficulty of paradoxically, slow equilibration to the environ-
interpretation. A detailed survey of the reactions of ment thereafter. The latter feature is possibly a
every species would not only take a lot of space but matter of slow migration rates. Ononis, Galium and
would obscure an overview of the situation. For Thymus exemplify these.
this reason, Table 5 has been prepared to display C. Those with a particular advantage in certain
the overall features of the species' behaviour. Not types of environmental perturbation which over-
all of the data base for this survey table has been rides other characteristics. Here perhaps the re-
figured in this paper, and reference must be made to mainder of our species belong: Bellis, Leontodon,
Page (1980) and Da Vinha (1981) for the original Homalothecium and Tortula.
analyses. Finally, we turn to the community concept and
The semi-fixed funes, having a lower cover of stability. Our analysis so far has been of individual
plants, lower soil organic matter and drier soil, are species, a reductionist approach. Competition for
conventionally considered to be less stable than the space must be taken into account in explanations of
fixed dune areas. Here species should be opportu- a species' behaviour following experimental treat-
nistic, capable of fast reaction to change and fast ment, as we have tried to show for Ononis repens,
recovery from disturbance. Table 5 does not con- but competition and species' response is the link
firm this supposition. between species and the community. Is the behav-
113
"0 ,.~
0 ,-I=
8 . ~0
r. "
~ ~
o ~ -~,~
z
d
K.~.~ .-~ .-~
~.~,
z ~ ~ ~ -~
-~.
~'~
Z .~ ~ ..~
~.~ ~ . ~. ~
-
121 o '
05
¢- o
o ~ _..N
~8
~..~
~u
~ • ,"
- ~ ~ ~ ~ ~ ~ ~ -~
• ~ o~ o..~
.: 0 0 0
0 0
~- =~ =~ _~-
~.~ "~ . a ~ -
h~
~.~
8
II
8
i14
Table 6. Variance ratio between dry weights in g • m 2 under treatments as itemised in Table 3 for various categories of species.
Significances of 5% or better are given by variance ratios greater than 2.62.
Semi-fixed Fixed dune
dune 4 year Fixed dune 4 year
Mobile dune enclosed path untrampled enclosed path Slack
June '77 "78 "77 '78 "77 '78 "77 '78 '77 '78
Total lower plants 1.5 3.1 4.0 3.6
Total dicots 1.9 2.1 5.3 43.6 10.8 7.1 5.0 15.8 17.1 18.9
Total monocots 3.2 2.4 3.8 1.5 48. I 4.3 6.6 2.7 I 1.4 2.9
Total green 4.4 4.5 3.7 9.5 35.8 1.3 1.9 7.4 6.7 8.8
Litter 1.8 5.2 1.1 1.5 2.5 10.7 1.9 5.3 8.7 1.6
Total 3.4 5.0 1.8 3.5 6.6 I 1.3 1.3 1.9 4.4 7.2
iour of the community as a whole simply the sum of Clapham, A. R., Tutin, T. G. & Warburg, E. F., 1962. Flora of
its parts, so that the species themselves are the units the British Isles (2nd ed.), Cambridge.
Da Vinha, S . G . , 1981. Productivity and succession in sand
for which predictions can be made? Table 6 gives a
dunes. Thesis, University of Wales.
summary of perturbation effects on five habitats of Grime, J. P., 1979. Plant strategies and vegetation processes.
the dunes, and here there is evidence of overall J o h n Wiley & Sons, Chichester.
community reactions to perturbation effects. Gupta, S. P., 1933. The reaction plants to density of s0il. J~ Ec01.
These variance ratios are between treatments 2 I: 452-474.
Hylgaard, T., 1981. Recovery of plant communities on coastal
which include a current control harvest, and show
sand dunes disturbed by h u m a n trampling. Biol. Conserv.
the intensity of the effect of treatments and the 19: 15-25.
likely duration of these effects. The greatest effect Liddle, M. J. & Greig-Smith, P., 1975. A survey of tracks and
of perturbation was found in the grey dunes, but paths in a sand dune ecosystem. J. Appl. Ecol. 12: 893-930.
this was almost entirely through grass growth, and Liddle, M. J. & Moore, K.J., 1971. The microclimate of sand
dune tracks: the relative contribution of vegetation removal
there was fast recovery. Dicotyledons showed slow
and soil compression. J. Appl. Ecol. I1: 1057-1068.
reaction (except on the grey dunes) and this was Mellinger, M.V. & McNaughton, S.J., 1975. Structure and
long maintained, while grasses showed the fastest function of successional vascular plant communities in
recovery throughout. It is interesting that both path coastal New York. Ecol. Monogr. 45:161 182.
and semi-fixed dune vegetation seem to react sim- Page, R . R . , 1980. The effects of trampling on a sand dune
system. Thesis, University of Wales.
ilarly, while slacks and unfixed yellow dunes again
Ranwell, D. S., 1972. Ecology of salt marshes and sand dunes.
behave in nearly the same way. These results seem C h a p m a n & Hall, London.
much more confusing than those of Mellinger & Smith, A. J. E., 1978. The moss flora of Britain and Ireland.
MacNaughton (1975) after similar experimenta- Cambridge.
tion, and we present our data in the hope that Streeter, D. J., 197 I, The effects of public pressure on the vegeta-
tion of chalk downland at Box Hill, Surrey. In: E. Duffey
generalisations can eventually be made about them.
(ed.), The Scientific management of animal and plant com-
munities for conservation, pp. 459-468. Blackwell, Oxford.
Warkentin, B. P., 197 I. The effect of compaction on content and
transmission of water in soil. In: Compaction of agricultural
References soils, pp. 125-133. American Society of Agricultural Engi-
neers.
Bayfield, N. G., 1971. A simple method for detecting variations Willis, A. J. & Yemm, E. W., 1961. Braunton Burrows: mineral
in walker pressure laterally across paths. J. Appl. Ecol. 8: nutrient status of the dune soils. J. Ecol. 49: 377-390.
533-536. Willis, A. J., 1963. Braunton Burrows: the effects on the vegeta-
Blom, C. W. P. M., 1976. Effect of trampling and soil compac- tion of the addition of mineral to the dune soils. J. Ecol. 5 I:
~'fi'ffn on the occurrence of some Plantago species in coastal 353-374.
sand dunes. 1. Soil compaction, soil moisture and seedling
emergence. Oecol. Plant. 11: 225-241. Accepted 12.9.1984,
change: a reductionist approach to stability*
R. R. Page, S. G. da Vinha & A. D. Q. Agnew
Department of Botany and Microbiology, University College of Wales, Aberl,stwyth, Wales, S Y23 3 DA,
U.K.
Keywords: Coastal sand dune, Galium verum, Grazing, Onions repens, Stability, Trampling
Abstract
Coastal dunes in western Europe have long been under grazing and recreational pressure, but the
vegetative cover appears to be remarkably resilient. In a series of experiments on Ynyslas Dunes, Cardigan
Bay (Wales, U.K.), trampling, protection, fertiliser, and herbicide treatments were monitored using non-
destructive and destructive sampling methods. This paper discusses the behaviour of 12 important species
after these treatments on three major dune habitats: Agrostis tenuis, Ammophila arenaria, Bellis perennis,
Festuca rubra, Galium verum, Leontodon taraxacoides, Ononis repens, Plantago lanceolata, Poa subcaeru-
lea, Thymus praecox, Homalothecium lutescens, Tortula ruraliformis.
Although stability is usually thought of as a feature of the community, a reductionist approach suggests
that grasses react quickly but temporarily, dicotyledonous forbs often show slower reaction times, each
species reacting in its own way to changes in the environment. A tentative classification of our study species is
possible.
Introduction the older dunes where Rhytidiadelphus triquerus is
an indicator of such conditions.
Ynyslas Dunes (Fig. 1) are a western dune system The area is a National Nature Reserve but is not
associated with the mouth of the Dovey Estuary in closed to the public and research was set up in 1974
Cardigan Bay, Wales. The dunes are calcareous with the aim of providing information helpful to the
(CaCO 3 content 3-5%) and have grown extensive- management of increasing recreation pressure on
ly in the last 100 years. They show a series of classi- the dune system. Visitors are on foot, often using
cal dune vegetation types (Ranwell, 1972) of strand the dunes as a passageway to the open beaches, and
plants, embryo dunes, mobile dunes with much often as an educational facility. Over 500 000 per-
open sand between Ammophila arenaria tussocks, son-days were recorded in 1974-75, and it was clear
semi-fixed mossy dunes with Tortula ruralis ssp. that damage was being done. At the same time the
ruraliformis, Homalothecium lutescens and Pelti- rabbit population recovered from the effects of
gera canescens/rufescens, and fixed grey dunes myxomatosis and grazing was increasing.
where plant cover reaches 100% and Festuca rubra The first research objective was to separate and
is often dominant. There is no dune heath since the quantify the effects of human trampling and rabbit
dunes are too young and calcareous but some acidi- grazing by means of exclosures and experimental
ty does develop in the winter-flooded slacks and on treatments. The second was to investigate the ef-
fects of the management tools of fertilisers and
*Nomenclature follows Smith (1978), for bryophytes and selective herbicides. This is not a new approach,
Clapham, Tutin & Warburg (1962). for angiosperms. having been extensively reported (Blom, 1977; Hyl-
Vegetatio 61, 105-114 (1985).
© Dr W. Junk Publishers, Dordrecht. Printed in the Netherlands.
106
further trampling. For these four series of investi-
gations human passage was monitored using pres-
sure sensitive counters in paths, as well as trip pins
(Bayfield, 1971)and direct observation, while plant
species occurrence was estimated as cover percent-
age in small permanent quadrats, never larger than
30 X 30 cm. Path trampling was measured in units
of person per metre transect across the path. For
example, 30 people using the central 25 cm cross
section of a path are said to be 120 people m i. In
this way various problems of standardisation were
overcome. Paths were monitored for 40 summer
days in the study of established paths and all year
for the experiment on increased trampling of paths.
For the experimentally increased use of estab-
lished paths trampling (in soft soled walking shoes,
exerting some 150 g m cm 2) was applied at multi-
ples of current recreational use of the paths. The
experiments lasted one year starting in the summer
of 1974, with vegetation cover being estimated at
the start and finish and soils analysed upon comple-
tion.
Exclosures of nearly 20 m 2 area were constructed
in 1973, half inside rabbit-proof wire mesh and the
rest only restricting human access. Cover estima-
tion in small quadrats was again used to monitor
the effect of protection. These areas were after-
Fig. I. The location of Ynyslas Dunes, Wales, with dune habitat wards (April 1977) used to follow the reaction of
types diagrammatically represented. Experimental sites are species to release from competition (by a single
shown as black blocks. application of selective weedkillers) combined with
fertiliser treatments. Here the standing crop was
harvested from random quadrats of 15 X 15 cm.
gaard, 1981; Liddle & Greig-Smith, 1975; Liddle & The subdivision of such fragmentary material into
Moore, 1974; Streeter, 1971 are the more recent species contributions only proved possible for the
references on this topic). Nevertheless, there is a more abundant species, and this part of the record
need to repeat experiments and validate predictions is incomplete.
for specific areas. This paper summarises some as- In field experiments natural population fluctua-
pects of reports by Page (1980) and Da Vinha tions may easily override treatment effects. In these
(1981) on research undertaken from 1975-80. studies control areas were designated and experi-
mental results related to concurrent changes in the
controls. Clearly, this assumes that the latter are
Methods due either to the current weather (for instance 1975
and 1976 had dry summers), or to the continuation
The study of this dune system was started with an of general vegetation change in the area of the
investigation of trampling impact in four ways: 1) experiment. Although open to criticism, we feel
the comparison of soil and species differences on that this approach is better than using a starting
long-established paths and undisturbed areas, 2) population as the only control value against which
experimental increase of the trampling on existing to measure change.
paths, 3) the experimental trampling of undis- S oils were analysed by taking replicated cylindri-
turbed areas, and 4) the protection of paths from cal cores of 4.7 cm diameter and 5 cm depth. Com-
107
bustion and oven drying were used to measure or- LEONTODON TARAXAC01DES [
TORTULA RURALIFORMI S [
ganic matter and soil moisture content respectively, AW0PHILA AP~ARL~
and dry bulk density by displacement with acetone H O M A L O T ~ IUM LVfESC~S [
POA Sg~CAm~gL~A ..~
(Gupta, 1933). Using these data percentage satura- FESTUCA RUBRA
tion of pore space was also calculated. All cores THYMUS PRAECOX , [
were taken after a period of heavy rain followed by ONONIS 2EPm~S [
GALIUM VERUM |
twelve dry hours in which state they should be near pLANTAGO LANCEOLATA ' I
field capacity (Warkentin, 1971). AGROSTIS ~]~VUIS
BELLIS P E R i l S I
19 , 2p , 3,o
VOLUMETRIC PERCENTAGE W T R
AE
Species performance LEONTODON TARAXACOIDES
~0~ULA RURALIFORMIS
HOMALOTHECIUM LUTESCENS I
Great quantities of data were gathered during the
six years of the investigation. T w o species, Ononis A ~ O P H n a A~m~mZA I
P0A SUBCAERULEA
repens and Galium verum, are taken as examples FESTUCA RUBR~
because data on these two are relatively complete 0NONIS REPENS [
BELLIS P E ~ I S --~
and are reported on in detail.
GALI UM VERUM I
pLANTAGO LANCEOLATA I
Ononis repens AGROSTIS TENUIS q .... 5. . I. . . IQ
PERCENTAGE LOSS ON IGNITION
TOHTULA RURALIFORMIS --~
General distribution LEONTODO~ TARAXACO I ~ S
The mean % pore space, organic matter and HOMALOTHECIUM LUTESC~S ~.~
BELLIS PEF~NNIS -.~
moisture levels at which this species occurs are in I
zcn~s PVaEOOX
the middle of the range of species reported on FESTUOA R U ~ --~
POA SUBCAERULEA
(Fig. 2), reflecting its widespread distribution. On
ONONIS H~PE~S
untrampled soils, a l t h o u g h c o m m o n , its percentage AF~IOPHILA A ~ A R I A |
cover is negatively correlated with soil moisture
PLANTAGO LANCEOLATA I
when all fixed and semi-fixed sites are included in AGROSTIS TENUIS
the analysis (r = -0.62, Table 1) and with a greater
50 ~ -+-7o
PERCENTAGE PORF SPACE
J ~
correlation coefficient (r = -0.82) when fixed dune
sites are analysed separately. N o w on paths there is Fig. 2. Ranked species occurrence along ranges of the various
more Ononis on the drier semi-fixed dunes, even soil factors which are associated with trampling, from 40 moni-
t h o u g h path soils are moister than untrampled tored sites which do not all correspond with experimental areas
soils, and there is no correlation between Ononis used in Figures 3 and 4. the mean value and one standard
deviation either side of it is represented.
cover and soil moisture on paths. However, it is
only on the fixed dunes that there is a negative
relationship between trampling and soil moisture,
and the most trampled semi-fixed dune paths are one year. In the absence of any treatment, cover
used by fewer people than the most trampled fixed changed to 156% and 77% of starting (i 974) values
dune paths. Since Ononis repens is apparently un- for semi-fixed and fixed dunes respectively. 1974
able to resist high levels of trampling(see below), its had 29% more days with over 2 m m of rain than
cover is low on the driest fixed-dune soils and high 1973, and the increase in cover of Ononis probably
on the inherently dry semi-fixed dunes. reflects its better performance during moist weather
in the open vegetation of the semi-fixed dune.
Experimental changes in trampling and grazing Thus, on two fixed dune paths monitored during
pressure the same year, soils were as dry as in the dry semi-
There are a n u m b e r of suggestions, arising f r o m fixed dune site used for the trampling but grasses
field experiments, that competititve effects are im- (mainly Festuca rubra and Ammophila arenaria)
p o r t a n t in determining cover of this species. For were c o - d o m i n a n t with Ononis repens which de-
example, in Figure 3 is shown Ononis cover after clined significantly during the year.
trampling was imposed on 2 undisturbed sites over These observations and experiments suggest that
108
Table 1. Significant species correlations in random quadrats on trampling was intermediate in its response. Relative
and off paths with soil factors. LOI, loss on ignition; PS, per- to the control here, there was no significant change
centage pore space; M FS, percentage moisture at field capacity;
in total cover of all species or grasses taken alone.
NS, no significant correlation; +, correlation positive; , correla-
tion negative. However, cover of some species such as Festuca
rubra whose mean cover in the exclosure at the start
Untrampled soils Path soils of the experiment was 30.9% increased relative to
Agrostis tenuis +: LOI, PS, MFS +: LOI, MFS
the control in the rabbit-grazed but not in the
A mmophila rabbit-proof parts, whilst others such as Agrostis
aretlaria +: LOI, PS NS tenuis, Ammophila arenaria and Holcus lanatus
Bellis perennis NS NS (whose total mean cover in the exclosures at the
Festuca ruhra +: LOI, PS, MFS NS
start of the experiments was 6.1%) declined in both
Galium v e r u m NS +: LOI, PS, MFS
Leontodon exclosures, but only significantly in the rabbit-
taraxacoides NS NS proof treatment. Soil moisture content and com-
Ononis repens : MFS NS paction levels were slightly, but significantly, lower
Plantago lanceolata NS +: LOI, PS, MFS in the rabbit-proof than the rabbit-grazed treat-
Poa suhcaerulea +: PS +: LOI, PS, MFS
ment and control soils. Broadly speaking then, the
Th.l'mUS praecox : LOI, PS NS
Tr(/~)lium repens NS +: LOI, MFS changes in cover of Ononis repens could be ex-
Tortula plained by assuming that trampling was so impor-
rural(lormis : LOI, MFS : LOI, MFS tant that release from its suppressive effect allowed
Ononis to compete successfully with grasses, but
the greater resistance to trampling on the fixed
competition is related to moisture supply in deter- dunes (Fig. 3) may relate to reduced competition
mining cover of Ononis repens, but the summer of from grasses combined with the more equable
1975 was particularly dry after April and the tram- microenvironment and soil conditions.
pling experiments were terminated earlier than the When trampling was applied at a range of fre-
exclosure experiments, thus making prediction dif- quencies, but the same intensity (150 passages m i
ficult. Table 2 shows changes in Ononis cover dur- year I), the least damage was caused on both the
ing 1974/1975 in an exclosure on a fixed dune area fixed and semi-fixed dunes by spreading the tram-
previously little used by the public. Ononis fared pling over one year at weakly rates that were a
worst in the control, increased in the: rabbit-proof percentage of actual use made of a path by the
part and in the part simply protected from human public (Fig. 3). On the fixed dunes there was no
Table 2. Differences in percentage cover after 2 years of protection from trampling and grazing or trampling alone, in 3 sites. Control is
an area which continued to receive use. Asterisk marks differences from control at 5% or better.
Fixed dune high-use site Fixed dune low-use site Trampled slack site
Con- Trampling Grazing & Con- Trampling Grazing & Con- Trampling Grazing &
trol prevented trampling trol prevented trampling trol prevented trampling
prevented prevented prevented
Agrostis tenuis +3.0 -5.0" 1.0" +4.5 + 1.0 1.0"
Ammol]hila arenaria +1.5 +0.5 -2.5
Poa subcaerulea + 1.5 0.5 + 1.5 4.5 I 1.0 1.5
Festuca rubra 7.5 -2.5 6.0 -32.0 15.0' 25.0
Bellis perennis -2.0 2.0 4.0*
Galium verum +4.5 +7.5 +5.0 12.0 15.0 1.5
Leontodon
taraxacoides +2.0 11.0" +3.0
Ononis repens 12.5 -6.5* +12.0'
Plantago lanceolata 15.5 -5.5 8.5*
Thymus praecox +4.0 + 17.5 2.0 +7.5 +2.0 2.0
109
A B C A B C D E
C N R L 01~ ~
O T O CONTROL 01 i.o I I---7
A! i .I ! 1
B I**,, I 4.14.*, I
200
IJJ
Z 150 '~" O ~ o
"r"
W 100
O
t,.)
LU
Z
50
\/°\
°
W O
n A
I
B C
I
A B c ? E,
74 296 1184 0 74 148 296 592 1184
passages m-1 yr-1 passages m-1 yr-1
O T Y O T Y
CI t t t
150 150~
i,i
W~
c.oz 100 100
I---r
r,-~
Ww 50 5O
C O T Y C 0 T
TET ET
RAMN TREATMENT
Fig. 3. Cover changes expressed as percentage of the starting value in Ononis repens populations after trampling treatments in previously
undisturbed sites. Above are shown the effect of trampling increases on (left) a semi-fixedand (right) a fixed dune site. Beloware shown
the effects of different frequencies of application of trampling equivalent to 150 passages m i year 1 as: C no trampling control; D, all in
one day; T, all in one week; Y, distributed throughout the year. Triangular matrices give significance of t tests between means
(+, 0.5 > p > 0.1 ; ++, 0.1 > p > 0.01 ; +++, p < 0.01 ) because heterogeneity of the populations precluded the use of variance analysis.
significant difference in relative cover from the con- of drought. U n d e r more average weather condi-
trol. tions results may have been different. W h e n all the
A p p l i c a t i o n of the t r a m p l i n g over ten weeks or t r a m p l i n g was applied on one day the c o n t e m p o -
all in one day resulted in some reduction in relative rary soil moisture c o n d i t i o n was also p r o b a b l y very
cover in c o m p a r i s o n with the c o n t r o l on both sites, i m p o r t a n t . However, it was possible to find a rela-
but this was not significant in the fixed dunes. O n t i o n s h i p between the changes in O n o n i s in these
the semi-fixed d u n e s the ' t e n - w e e k ' t r e a t m e n t was e x p e r i m e n t s a n d changes in measured soil factors,
the most d a m a g i n g . It must also be repeated that of which c o m p a c t i o n was the only one to be affect-
the s u m m e r of 1974 in which the e x p e r i m e n t started ed by the e x p e r i m e n t s o n the fixed dunes. Cover of
was particularly h u m i d a n d 1975 included a period O n o n i s r e p e n s was not correlated with percentage
110
pore space on paths or u n t r a m p l e d soils. The lack r e m o v a l alone resulted in a p p a r e n t l y p e r m a n e n t
of any change in moisture levels was u n e x p e c t e d d a m a g e to Ononis p o p u l a t i o n s . Willis & Y e m m
because a significant c o r r e l a t i o n between c o m p a c - (1961) f o u n d that Ononis repens was e r a d i c a t e d
tion a n d m o i s t u r e levels was f o u n d in b o t h long f r o m turf t r a n s p l a n t s in the greenhouse (to which
established p a t h s and u n t r a m p l e d soils. The results, either c o m p l e t e nutrients, N + P + K or nitrate only
therefore, suggested that not all soil factors are in were a d d e d ) by the dense g r o w t h of Festuca rubra,
e q u i l i b r i u m with t r a m p l i n g pressure after one years' but we have not c o n f i r m e d this in the field. Here,
trampling. fertiliser t r e a t m e n t a l o n e had little effect but there
was a chance a b s e n c e of Ononis u n d e r this treat-
Field manipulation o f nutrients and competition ment in the semi-fixed d u n e site.
D i s t u r b a n c e resulted in c o n s i d e r a b l y m o r e
c h a n g e in the semi-fixed site ( m o s s y dunes) t h a n in S u m m a r y o f O n o n i s repens performance
the fixed d u n e site ( T a b l e 3), as w o u l d be expected, Ononis repens is a c o n s p i c u o u s plant, easily ob-
a n d this was reflected in the b e h a v i o u r of Ononis. served a n d o f wide d i s t r i b u t i o n in W e s t e r n Europe.
On b o t h sites fertiliser a d d i t i o n c o m b i n e d with se-. T h e r e is m u c h i n f o r m a t i o n a n d yet the picture of its
lective d i c o t y l e d o n o u s r e m o v a l resulted after two e c o l o g y a n d b e h a v i o u r in the face of a c h a n g i n g
years in e n h a n c e d p o p u l a t i o n s of Ononis (which e n v i r o n m e n t r e m a i n s r a t h e r confused. F r o m our
r e i n s t a t e d itself very q u i c k l y f r o m rhizomes). O n l y analyses we can say that p o p u l a t i o n s fluctuate
on the semi-fixed dunes was there a significant widely, p r o b a b l y c o n t r o l l e d by seasonal weather
c h a n g e in this species when m o n o c o t y l e d o n s were variables, a n d a rather c o m p l e x interaction be-
r e m o v e d but an excessive v a r i a n c e in the fixed d u n e tween soil moisture, t r a m p l i n g and c o m p e t i t i o n .
site m a y m a s k a real effect here. In b o t h sites dicot But the species is an o p p o r t u n i s t i c one, c a p a b l e of
r a p i d e x p l o i t a t i o n of t e m p o r a r i l y f a v o u r a b l e sites.
Table 3. Ononis repens above ground dry weight mean g • m 2
C o m p e t i t i o n seems to play a key role in switching
after treatment in April 1977 as follows: M, monocotyledon
selection herbicide and M + F, with NPK fertiliser; -D, the species to o c c u p y resources, as is indicated by its
dicotyledon herbicide and -D + F, with NPK fertiliser; F, NPK increase u n d e r p r o t e c t i o n f r o m t r a m p l i n g in one
fertiliser alone; C, control. Asterisks mark significances better site a n d when p r o t e c t e d f r o m c o m p e t i t i o n with
than 5%. NP, species not present in this treatment. o t h e r d i c o t y l e d o n s after herbicide treatment.
A g a i n , g r a z i n g s h o u l d be c o n s i d e r e d as an interac-
A. Grey dunes, untrampled
tive effect.
Treatments June August June F ratio
1977 1977 1978 Galium verum
-M 18.4 23.5 45.3 1.3
-M + F 26.2 54.7 76.1 3.6 General distribution
-D 11.3 17.4 8.0 0.6
This species is c o n f i n e d to the fixed dunes. T h e
-D + F 7.3 40.4 53.7 5.7*
F 68.4 73.7 92.5 0.7 significant c o r r e l a t i o n s between percentage cover
C 13.6 50.9 47.3 4.0* of this species a n d soil moisture, measured volu-
F ratio 4.7* 2. I 5.5* metrically, (r = 0.58) o r g a n i c m a t t e r (r = 0.61) a n d
p o r e space (r = 0.50) when all p a t h sites, b o t h on the
fixed a n d semi-fixed d u n e s are included in the anal-
B. Semi-fixed dunes, previously trampled ysis is t h e r e f o r e to be e x p e c t e d ( T a b l e 1). T h e r e was,
however, no significant c o r r e l a t i o n between per-
Treatments June August June F ratio
1977 1977 1978 centage cover a n d a n y m e a s u r e d soil f a c t o r when
fixed d u n e paths were a n a l y s e d separately. Galium
-M 12,.1 11.7 88.2 4.1
-M + F 7.7 NP 17.2 2.6 verum was present at t o o low a p e r c e n t a g e cover on
-D NP 1.0 5.8 0.8 u n t r a m p l e d fixed d u n e sites to allow regression
-D + F 46.6 38.0 147.1 8.4* analysis to be u n d e r t a k e n ,
F NP NP NP NP
C 66.4 60.4 106.2 1.3 Experimental trampling change
F ratio 4.9* 6.3* 14.9"
Galium verum is p r e s e n t on the most highly
111
1°°
/ The difficulty of dealing with data from c o n t r o l
sites is seen in the changes in G a l i u m v e r u m on
A B C 0
0J I***1~* I *
soF c o n t r o l areas a d j a c e n t to enclosures (Table 2). O n a
moderately t r a m p l e d fixed d u n e site percentage
o~ jo cover r e m a i n e d statistically u n c h a n g e d between
1973 a n d 1975 whilst it declined over this period on
8854 I'7"/06 26562 35416 442?0
0 A B C O a lightly t r a m p l e d site. The difference in changes in
passages rn-1 yr-1
percentage cover on controls is also seen in an
Fig. 4. Cover changes expressed as percentages of starting values e x p e r i m e n t in which levels of t r a m p l i n g on paths is
in Galium verum after increasing the trampling of a previously increased. O n the heavily used path (Fig. 4) cover of
used path. Significancesshown as in Figure 3. G a l i u m v e r u m declined significantly over the ex-
p e r i m e n t a l year, but remained statistically un-
changed on a moderately trampled path in an
t r a m p l e d path recorded on the dunes, used by e x p e r i m e n t not reported here.
8 854 people m i y e a r l , a n d a l t h o u g h the mean Relative to changes on the control, there was a n
percentage cover was only 1.7%, only three other increase in cover of G a l i u m v e r u m on a previously
d i c o t y l e d o n o u s species had a greater cover. O n this little trampled enclosure site (68% grass), only when
path there was an a p p a r e n t l y steady decline in rela- rabbits as well as h u m a n s were barred access (Ta-
tive cover with e x p e r i m e n t a l l y increased t r a m p l i n g ble 2), but on a previously more trampled site (20%
until a 300% in use was applied (Fig. 4). grass), cessation of t r a m p l i n g and rabbit grazing
A p a r t from a significantly higher moisture level had no effect relative to changes on the control.
in the e x p e r i m e n t a l area suffering a 300% increase Given the occurrence of G a l i u m v e r u m in rela-
in t r a m p l i n g , which was p r o b a b l y not related to the tion to soil factors, it is p r o b a b l e that the relative
t r e a t m e n t , there were no significant differences be- increase (or rather failure to decline) on the pre-
tween measured soil parameters on the trampled viously little t r a m p l e d site is related directly to ces-
area a n d the control, a n d the effect of increased sation of grazing, a n d the difference between the
t r a m p l i n g on the species in this e x p e r i m e n t is most changes on the two enclosure sites relates to compe-
likely a direct one. tition from grasses.
Table 4. Galium verum above ground phytomass dry weight in g • m 2 in a fixed dune enclosure after 4 years of protection from
trampling. Treatments as in Table 2. Below is shown ranges of values within which there are no significant differences at 5% or better.
Asterisks mark significanceat 5% or better.
Treatments June 1977 August 1977 June 1978 F ratio
-M 4.4 15.6 15.3 2.1
-M + F 21.7 89.2 62.8 10.4"
D 4.9 10.6 9.8 0.9
-D + F 4.1 21.8 21.5 4.5
F 20.4 27.5 5.6 5.6*
C 14.2 21.1 21.0 1.0
F ratio 4.1" 10.1' 12.9"
June 1977 -M + F F C -D -M -D + F
August 1977 M+ F F -D+ F C M D
June 1978 -M+ F -D + F C -M D F
112
Field manipulation of nutrients and competition While Homalothecium lutescens and Tortula ru-
Galium was present only in one site in sufficient raliformis are semi-fixed dune moss species which
quantities for analysis of these manipulations. This are associated with paths on the fixed dunes, there
is an area of heavily trampled and grazed fixed- is no large group of species which is distributed in
dune vegetation which resembled a close cropped this way, and, apart from a short list of winter
lawn in 1973 when enclosed. Considerable phyto- ephemerals, it is difficult to suggest others, except
mass had accumulated by 1977 when treatments for Leontodon taraxacoides and Plantago lan-
were applied to a totally protected section, although ceolata, where evidence is less clear. This merely
it was still more uniform and less tussocked than confirms that paths, although developed through
undamaged fixed-dune vegetation. Table 4 shows disturbance of fixed dune vegetation, are not com-
that it is only the fertilised treatment in which parable with a successional stage.
monocotyledons had been killed which showed en- The fast recovery of Festuca rubra is of interest.
hanced growth of G. verum, but there is little sign of This is a dominant species of fixed dunes. Willis
a return to normal level after 18 months. (1963) has documented the same sort of reaction to
increased NPK as we have here, but in this case the
Summary of Galium verum performance fertiliser inputs were sustained and the fast rever-
There is much less evidence here than in the case sion to lower production was not observed. Our
of Ononis repens. This species is particularly asso- pulsed treatments allow us to distinguish between
ciated with paths on the fixed dunes, and is resistant grasses and most dicotyledons in their rates of re-
to trampling. It reacts very positively indeed when version to original biomass. Some dicotyledonous
monocotyledons are removed, so that we suspect it species are stress tolerant in the sense of Grime
is sensitive to competition with grasses as its mor- (1979) (Ononis, Galium, Thymus) and their slow
phology would suggest. Its performance fluctuated recovery from perturbation is no surprise, but the
from year to year in the control of some experi- instability of their resident populations is unex-
ments, yet in the perturbed fixed-dune site it was pected in this respect as is their response to the
remarkably stable and recovery in the latter site was removal of grasses, taken to be competitors. We
rather slow. may attempt to categorise the species reported
upon here as follows.
A. Those which have constant cover resistant to
Discussion change and which reinstate quickly after release
from stress. The grasses are examples of this type.
The accounts of the two species' reaction to en- B. Those having cover varying from year to year
vironmental change illustrate the complexity of the with fast reaction to environmental change and,
data base from which we work, and the difficulty of paradoxically, slow equilibration to the environ-
interpretation. A detailed survey of the reactions of ment thereafter. The latter feature is possibly a
every species would not only take a lot of space but matter of slow migration rates. Ononis, Galium and
would obscure an overview of the situation. For Thymus exemplify these.
this reason, Table 5 has been prepared to display C. Those with a particular advantage in certain
the overall features of the species' behaviour. Not types of environmental perturbation which over-
all of the data base for this survey table has been rides other characteristics. Here perhaps the re-
figured in this paper, and reference must be made to mainder of our species belong: Bellis, Leontodon,
Page (1980) and Da Vinha (1981) for the original Homalothecium and Tortula.
analyses. Finally, we turn to the community concept and
The semi-fixed funes, having a lower cover of stability. Our analysis so far has been of individual
plants, lower soil organic matter and drier soil, are species, a reductionist approach. Competition for
conventionally considered to be less stable than the space must be taken into account in explanations of
fixed dune areas. Here species should be opportu- a species' behaviour following experimental treat-
nistic, capable of fast reaction to change and fast ment, as we have tried to show for Ononis repens,
recovery from disturbance. Table 5 does not con- but competition and species' response is the link
firm this supposition. between species and the community. Is the behav-
113
"0 ,.~
0 ,-I=
8 . ~0
r. "
~ ~
o ~ -~,~
z
d
K.~.~ .-~ .-~
~.~,
z ~ ~ ~ -~
-~.
~'~
Z .~ ~ ..~
~.~ ~ . ~. ~
-
121 o '
05
¢- o
o ~ _..N
~8
~..~
~u
~ • ,"
- ~ ~ ~ ~ ~ ~ ~ -~
• ~ o~ o..~
.: 0 0 0
0 0
~- =~ =~ _~-
~.~ "~ . a ~ -
h~
~.~
8
II
8
i14
Table 6. Variance ratio between dry weights in g • m 2 under treatments as itemised in Table 3 for various categories of species.
Significances of 5% or better are given by variance ratios greater than 2.62.
Semi-fixed Fixed dune
dune 4 year Fixed dune 4 year
Mobile dune enclosed path untrampled enclosed path Slack
June '77 "78 "77 '78 "77 '78 "77 '78 '77 '78
Total lower plants 1.5 3.1 4.0 3.6
Total dicots 1.9 2.1 5.3 43.6 10.8 7.1 5.0 15.8 17.1 18.9
Total monocots 3.2 2.4 3.8 1.5 48. I 4.3 6.6 2.7 I 1.4 2.9
Total green 4.4 4.5 3.7 9.5 35.8 1.3 1.9 7.4 6.7 8.8
Litter 1.8 5.2 1.1 1.5 2.5 10.7 1.9 5.3 8.7 1.6
Total 3.4 5.0 1.8 3.5 6.6 I 1.3 1.3 1.9 4.4 7.2
iour of the community as a whole simply the sum of Clapham, A. R., Tutin, T. G. & Warburg, E. F., 1962. Flora of
its parts, so that the species themselves are the units the British Isles (2nd ed.), Cambridge.
Da Vinha, S . G . , 1981. Productivity and succession in sand
for which predictions can be made? Table 6 gives a
dunes. Thesis, University of Wales.
summary of perturbation effects on five habitats of Grime, J. P., 1979. Plant strategies and vegetation processes.
the dunes, and here there is evidence of overall J o h n Wiley & Sons, Chichester.
community reactions to perturbation effects. Gupta, S. P., 1933. The reaction plants to density of s0il. J~ Ec01.
These variance ratios are between treatments 2 I: 452-474.
Hylgaard, T., 1981. Recovery of plant communities on coastal
which include a current control harvest, and show
sand dunes disturbed by h u m a n trampling. Biol. Conserv.
the intensity of the effect of treatments and the 19: 15-25.
likely duration of these effects. The greatest effect Liddle, M. J. & Greig-Smith, P., 1975. A survey of tracks and
of perturbation was found in the grey dunes, but paths in a sand dune ecosystem. J. Appl. Ecol. 12: 893-930.
this was almost entirely through grass growth, and Liddle, M. J. & Moore, K.J., 1971. The microclimate of sand
dune tracks: the relative contribution of vegetation removal
there was fast recovery. Dicotyledons showed slow
and soil compression. J. Appl. Ecol. I1: 1057-1068.
reaction (except on the grey dunes) and this was Mellinger, M.V. & McNaughton, S.J., 1975. Structure and
long maintained, while grasses showed the fastest function of successional vascular plant communities in
recovery throughout. It is interesting that both path coastal New York. Ecol. Monogr. 45:161 182.
and semi-fixed dune vegetation seem to react sim- Page, R . R . , 1980. The effects of trampling on a sand dune
system. Thesis, University of Wales.
ilarly, while slacks and unfixed yellow dunes again
Ranwell, D. S., 1972. Ecology of salt marshes and sand dunes.
behave in nearly the same way. These results seem C h a p m a n & Hall, London.
much more confusing than those of Mellinger & Smith, A. J. E., 1978. The moss flora of Britain and Ireland.
MacNaughton (1975) after similar experimenta- Cambridge.
tion, and we present our data in the hope that Streeter, D. J., 197 I, The effects of public pressure on the vegeta-
tion of chalk downland at Box Hill, Surrey. In: E. Duffey
generalisations can eventually be made about them.
(ed.), The Scientific management of animal and plant com-
munities for conservation, pp. 459-468. Blackwell, Oxford.
Warkentin, B. P., 197 I. The effect of compaction on content and
transmission of water in soil. In: Compaction of agricultural
References soils, pp. 125-133. American Society of Agricultural Engi-
neers.
Bayfield, N. G., 1971. A simple method for detecting variations Willis, A. J. & Yemm, E. W., 1961. Braunton Burrows: mineral
in walker pressure laterally across paths. J. Appl. Ecol. 8: nutrient status of the dune soils. J. Ecol. 49: 377-390.
533-536. Willis, A. J., 1963. Braunton Burrows: the effects on the vegeta-
Blom, C. W. P. M., 1976. Effect of trampling and soil compac- tion of the addition of mineral to the dune soils. J. Ecol. 5 I:
~'fi'ffn on the occurrence of some Plantago species in coastal 353-374.
sand dunes. 1. Soil compaction, soil moisture and seedling
emergence. Oecol. Plant. 11: 225-241. Accepted 12.9.1984,