Biological Invasions (2005) 7: 733–746 Ó Springer 2005

DOI 10.1007/s10530-004-1196-3

Invasive plants and their escape from root herbivory: a worldwide comparison
of the root-feeding nematode communities of the dune grass Ammophila
arenaria in natural and introduced ranges

W.H. van der Putten1,*, G.W. Yeates2, H. Duyts1, C. Schreck Reis3 & G. Karssen4
1
Netherlands Institute of Ecology (NIOO-KNAW), Department of Multitrophic Interactions, P.O. Box
40, 6666 ZG Heteren, The Netherlands; 2Landcare Research, Private Bag 11-052 Palmerston North, New
Zealand; 3IMAR, Departamento de Botânica, Universidade de Coimbra, 3000 Coimbra, Portugal; 4Plant
Protection Service, Nematology section, P.O. Box 9102, 6700 HC Wageningen, The Netherlands; *Author
for correspondence (e-mail: putten@nioo.knaw.nl)

Received 27 February 2004; accepted in revised form 7 July 2004

Key words: Ammophila arenaria, Ammophila breviligulata, biotic resistance hypothesis, enemy escape
hypothesis, feeding specialist, invasive plant, root herbivore, soil pathogen

Abstract

Invasive plants generally have fewer aboveground pathogens and viruses in their introduced range
than in their natural range, and they also have fewer pathogens than do similar plant species native
to the introduced range. However, although plant abundance is strongly controlled by root herbivores
and soil pathogens, there is very little knowledge on how invasive plants escape from belowground
enemies. We therefore investigated if the general pattern for aboveground pathogens also applies to
root-feeding nematodes and used the natural foredune grass Ammophila arenaria as a model. In the
late 1800s, the European A. arenaria was introduced into southeast Australia (Tasmania), New Zea-
land, South Africa, and the west coast of the USA to be used for sand stabilization. In most of these
regions, it has become a threat to native vegetation, because its excessive capacity to stabilize wind-
blown sand has changed the geomorphology of coastal dunes. In stable dunes of most introduced
regions, A. arenaria is more abundant and persists longer than in stabilized dunes of the natural
range. We collected soil and root samples and used additional literature data to quantify the taxon
richness of root-feeding nematodes on A. arenaria in its natural range and collected samples from the
four major regions where it has been introduced. In most introduced regions A. arenaria did not have
fewer root-feeding nematode taxa than the average number in its natural range, and native plant spe-
cies did not have more nematode taxa than the introduced species. However, in the introduced range
native plants had more feeding-specialist nematode taxa than A. arenaria and major feeding specialists
(the sedentary endoparasitic cyst and root knot nematodes) were not found on A. arenaria in the
southern hemisphere. We conclude that invasiveness of A. arenaria correlates with escape from feeding
specialist nematodes, so that the pattern of escape from root-feeding nematodes is more alike escape
from aboveground insect herbivores than escape from aboveground pathogens and viruses. In the nat-
ural range of A. arenaria, the number of specialist-feeding nematode taxa declines towards the mar-
gins. Growth experiments are needed to determine the relationship between nematode taxon diversity,
abundance, and invasiveness of A. arenaria.
734

Introduction involving release of root-feeding insects, for exam-

ple to control Knapweed (Centaurea maculosa) in
Invasive species are a major threat to biodiversity the USA (Clark et al. 2001a, b), although such
and ecosystem processes in native communities introduced enemies also can exert negative effects
(Rejmanek and Richardson 1996) and they may on other plants (Callaway et al. 1999). However,
cause significant costs to the environment, econ- we know of no programs involving the release of
omy and public health (Pimentel 2002). Biolo- soil pathogens or root-feeding nematodes in order
gical invaders include plants and animals to control plant invasions; the few Anguinidae
(Liebhold et al. 1995) that have been introduced (Nematoda: Tylenchida) tested parasitise above-
to novel regions intentionally or unintentionally. ground plant structures (Robinson et al. 1978).
Relatively few introduced species become inva- Few studies on biological invasions have tested
sive, and although there are a number of features how these effects may have arisen (Levine et al.
that characterize potential invaders, invasiveness 2003). Klironomos (2002) showed that five plant
is difficult to predict (Williamson 1996). Some species introduced to North America did not
strategies to control invasive species are based on accumulate soil pathogens as rapidly as did rare
the enemy escape hypothesis, which assumes plants when grown in repeated monocultures.
invasiveness to be due to the release of invasive However, this study did not include the response
species from their natural enemies (Keane and of the introduced plants in their natural soils.
Crawley 2002). Recent overviews on invasive Wide spacing of North American Prunus serotina
plants (Mitchell and Power 2003) and animals (Black cherry) trees in North American forests
(Torchin et al. 2003) have demonstrated that in correlates with high seedling mortality in soil
their introduced ranges exotic species indeed have from underneath parent trees (Packer and Clay
fewer pathogens, parasites, or viruses than in 2000, 2002). Such mortality did not occur in the
their natural range. The introduced species also non-native range, where P. serotina has much
have fewer of these natural enemies than do simi- higher abundance, suggesting that invasiveness of
lar, native species in the introduced range. How- this tree is due to escape from soil pathogens
ever, there are few studies on plant escape from (Reinhart et al. 2003).
root-feeding insects and even fewer on escape There are few studies on the nematode com-
from soil pathogens or root-feeding nematodes. munities of invasive plants. Densities of root-
Root herbivores and soil pathogens are impor- feeding nematodes were higher on the roots of
tant regulators of spatial and temporal changes in the woody legume mesquite (Prosopis glandulosa)
the composition of natural vegetation (Yeates in its original range than in recently invaded de-
1999; Wardle 2002; Bever 2003; van der Putten sertified perennial grasslands in the Chihuahuan
2003). Evidence is accumulating that soil patho- desert, USA (Virginia et al. 1992). In New Zea-
gens and root herbivores play important roles in land the invasive weed Tradescantia fluminensis
controlling plant abundance (Klironomos 2002), had seven additional taxa of nematodes com-
plant species diversity (Bever 1994; Packer and pared to reference locations without this invading
Clay 2000, 2002; De Deyn et al. 2003), and vege- species (Yeates and Williams 2001). In South
tation succession (Brown and Gange 1992; van African coastal foredunes, introduced marram
der Putten et al. 1993; De Deyn et al. 2003). grass (Ammophila arenaria) shared a number of
Root-feeding insects may cause dramatic decline root-feeding and non-root-feeding nematode spe-
of plant populations (Blossey and Hunt-Joshi cies with the indigenous beach grass Elymus disti-
2003), while effects of root-feeding nematodes chus and the indigenous foredune grass Erharta
vary from marked, generalized reduction in plant villosa, while all root-feeding species were rather
production (Stanton 1988), to localized damage unspecialized (Knevel et al. 2004). None of these
patches (Verschoor et al. 2001), and their effects studies, however, included a comparison with
may depend on interactions with, for example, nematodes on the introduced plant species in
pathogenic soil fungi (De Rooij-van der Goes their entire natural and introduced ranges.
1995; van der Putten and van der Stoel 1998). We compare the root-feeding nematode com-
There have been biological control programs munity of the natural foredune grass A. arenaria
 735

in its natural range with that in areas where the than in the introduced range (Jobin et al. 1996;
grass has been introduced for the stabilization of Memmott et al. 2000; the feeding-specialist
coastal sand dunes. A. arenaria occurs naturally hypothesis), and that (3) in its introduced range,
in north-western Europe and along the Mediter- A. arenaria has fewer root-feeding nematode taxa
ranean coast (Huiskes 1979). Because A. arenaria than do similar native species (Mitchell and
is intensively planted for sand stabilization in Power 2003; the native species hypothesis). In
coastal dunes, the history of introduction and addition, since the observed nematode diversity
spread has been well recorded. In the second half may depend on sampling intensity, we also inves-
of the 19th century it was introduced to the USA tigated the effect of sampling intensity in one nat-
(1868, where it had been introduced from Aus- ural and in one introduced region.
tralia; Wiedemann and Pickart 1996), South
Africa (1870s; Hertling and Lubke 2000), south-
eastern Australia/Tasmania (before 1868; Wiede- Materials and methods
mann and Pickart 1996), and New Zealand
(1873; Owen 1996). The plant was introduced as Terminology
seeds (Hertling and Lubke 1999) which may have
enabled escape from root-feeding nematodes, Geography
since these are usually not vertically transmitted. We distinguish the natural range (northwestern
In southeast Australia (Tasmania) and the USA, and Mediterranean Europe) and the introduced
A. arenaria has become an invasive weed (Hey- range consisting of four regions: one on the
ligers 1985, Buell et al. 1995), in New Zealand northern hemisphere (USA) and three on the
A. arenaria is considered a weed (Owen 1996), southern hemisphere (South Africa, Australia/
while in South Africa, the invasive potential is Tasmania, and New Zealand). Within the native
supposed to be limited by water supply (Hertling range and each introduced region, we have col-
and Lubke 2000). In the USA, A. arenaria has lected samples from a number of locations.
been responsible for changing the geomorphol- Within each location, five samples were collected
ogy of the coastal dunes, which is a major reason (only three from South Africa), which were
of its negative effects on native plant species pooled because these were pseudoreplicates; the
(Wiedemann and Pickart 1996). In stabilized locations were the true replicates. Locations
non-native dunes, A. arenaria is more abundant (abbreviations) examined in the present study
and more persistent than in stabilized dunes of are: Kampinos dunes Poland (Pl), Tentsmuir
its natural range. Point, Fife, Scotland (S), Voorne, Netherlands
The similar history, date, and mode of intro- (N), São Jacinto, Portugal north (Pn), Sado
duction of A. arenaria to all these new territories Estuary, Portugal south (Ps), two sites in the
in temperate regions of the northern and south- Camarque, Mediterranean France (Fa and Fb) in
ern hemispheres makes it a suitable model for Europe, Sunset beach road, Oregon (SR), Sunset
comparing the taxon richness of root-feeding beach road, Oregon (SB), South Salmon creek,
nematode communities. We grouped nematodes California (SC), Humboldt bay, California (HB)
as feeding specialists and feeding generalists in USA, De Mond (D), Blue bay (Bb), Sedgefield
(Yeates et al. 1993), which allows comparison (Sf), Klein Brak (Kl), Koeberg (Ko), Kleinmonde
with aboveground feeding-specialist invertebrate field (Kf), Kleinmonde slack (Ks), Table View
herbivores on invasive plants. Jobin et al. (1996) (TV) in South Africa, Patea (P), Castlecliff (C),
and Memmott et al. (2000) observed less feeding- Himatangi (H), Sumners (S), Taylors (T), Bir-
specialist aboveground insects on invasive plant dlings (B) in New Zealand, and prograding
species. We test the hypotheses that in its natural (P) and degrading (D) sites in Tasmania.
range (1) A. arenaria has more taxa of root-feed-
ing nematodes than in its introduced range Nematodes
(Mitchell and Power 2003; the taxon diversity In case of literature data, we used the list of
hypothesis), (2) A. arenaria has more specialist nematode genera or species provided. We distin-
feeders (i.e. sedentary, endoparasitic nematodes) guished sedentary endoparasites, migratory
736

endoparasites, and ectoparasites according to of nematodes in the samples, so that we decided

Yeates et al. (1993), and considered semi-endo ecto- not to perform quantitative comparison of abun-
parasites as ectoparasites. Sedentary endoparasites dance among sampling locations. Upon arrival,
were considered as feeding specialists, because of the samples were stored at 4 °C until nematodes
their ability to initiate feeding structures within were collected from the sand using an Oosten-
plant roots. brink elutriator and from the roots by wet funnel
extraction. Roots were also inspected for the
Collection of samples and identification of presence of cysts (females of Heterodera spp.)
nematodes and root knots (induced by females of Meloido-
gyne spp.).
Soil samples were collected from locations in The level of nematode identification depended
Europe (the natural range) and from regions on the availability of suitable keys or the pres-
where A. arenaria had been introduced. In all ence of appropriate life stages (e.g. cysts). For
regions, the nematode community was assessed each location, we counted all root-feeding nema-
in the active growth season. In Europe and the tode taxa. Identifications that have been made to
USA, we sampled in August, and in South Africa species level have been mentioned as such; In the
and Tasmania in October, and for New Zealand cases where we did not identify to species, we
in May–June. The locations in the Netherlands have aggregated by genus for the different loca-
and at Himatangi Beach in New Zealand were tions and regions, unless we were sure that the
sampled year-round, and we used these data to species should be different. In the latter case, we
determine possible underestimation arising from designated them as Genus a, b, etc. This proce-
a single sampling. dure was designed to have the least impact on
From each sampling location, 300–500 g sam- our analyses, and to reduce the size of tables.
ples of soil and roots were collected from the
1-year old root layer of five A. arenaria plants. Data analysis
When there were similar native plant species
around, soil and root samples from these native To compare nematode species richness associated
species were collected similar as from A. arenaria. with A. arenaria in natural and introduced
In mobile (yellow) dunes, A. arenaria plants are regions (hypothesis 1), we compared the number
buried each year by windblown beach sand, fol- of nematode taxa present in a region by one-way
lowed by shoot elongation and the formation of ANOVA with unequal number of replicates, after
new roots (de Rooij-van der Goes et al. 1995). testing homogeneity of variances. The numbers of
The 1-year old root layer is usually well colo- root-feeding nematode taxa present at sampling
nized by nematode taxa (van der Stoel et al. locations (for example Scotland, Poland, The
2002). At each sampling location, the five repli- Netherlands, Portugal north, Portugal south, and
cates (three in South Africa) were collected from southern France) within region (Europe) were
the outer edge of single tussocks, 25 m apart used as replicates. In order to test if A. arenaria
along a transect parallel to the shoreline. Addi- has more specialist feeders (i.e. sedentary, endo-
tional data were obtained from the literature: parasitic nematodes) than in the introduced range
Scotland (Wall et al. 2002), Poland (Kisiel 1970), (hypothesis 2) we performed the same analysis for
The Netherlands (van der Stoel et al. 2002), and feeding specialist (sedentary endoparasites) and
New Zealand (Yeates 1967, 1968). Root-feeding generalist (ectoparasites and migratory endopara-
nematode species were categorized into feeding sites) feeding types using the classification by
groups as was done for the sites that we have Yeates et al. (1993).
sampled ourselves and the data from each indi- To test if in its introduced range, A. arenaria
vidual survey (one in Scotland and Poland and has fewer root-feeding nematode taxa than do
six in New Zealand) were considered as repli- similar native species (hypothesis 3), t-tests were
cates. applied to determine if the difference between the
The soil/root samples were transported by air. number of root-feeding nematode taxa on the
This procedure may have reduced the abundance introduced A. arenaria and numbers of taxa on
 737

the native plant species would be less than zero detected in the root zone or on the roots of
(P < 0.05). If the test resulted in a non-signifi- A. arenaria. On average the number of feeding
cant P-value A. arenaria had equal or higher specialist nematodes on A. arenaria in the USA
number of nematode taxa than similar native was not different (P > 0.05) from that in the
plant species, so that hypothesis 3 should be natural range, Europe. However, within the
rejected. In that case, the average of the data native range, there were fewer specialist root fee-
points would end up on or above the 1:1 line in der taxa in the margins (Portugal and Mediterra-
a graph where number of nematode taxa of nean France in the south and Poland in the
native plant species (x-axis) are plotted against north) than in the centre (The Netherlands and
the numbers on A. arenaria (y-axis). We per- Scotland; see Table 2). In the Netherlands and
formed this test for all nematode taxa, as well as Scotland, there were two genera of feeding-spe-
for number of taxa of specialist and generalist cialist (sedentary endoparasitic) nematodes pres-
feeding types, and did a similar test for the native ent (Heterodera and Meloidogyne), while in
region. Portugal and southern France, there was only
one genus of sedentary endoparasites (Meloidogy-
ne in Portugal and Heterodera in Mediterranean
Results France), and sedentary endoparasites were not
detected on A. arenaria in Poland (Table 2).
Comparing nematode taxon diversity on The number of feeding generalist nematodes
Ammophila arenaria between natural and were significantly higher (P < 0.05) in Europe
introduced ranges and South Africa than in New Zealand, while
numbers in the USA and Tasmania were inter-
The plant-feeding nematodes identified from the mediate (replicates listed in Table 2). Distinct
root zone of A. arenaria in natural and introduced from the patterns for feeding specialists, some
regions are listed in Table 1. There were 6 subor- feeding generalists were present at all locations of
ders, 13 families, and 17 genera in the total data the native range, but the identity of the species
set. The number of species could not be accurately representing that genus, for example Helicotylen-
determined, as it was not possible to identify all chus, differed among locations; in Poland, Heli-
genera to species. The number of nematode taxa cotylenchus multicinctus was detected, in The
differed significantly among ranges (Figure 1; one- Netherlands Helicotylenchus pseudorobustus, and
way ANOVA F ¼ 5.44, df ¼ 4, P ¼ 0.0034). In in Portugal we found a species of Helicotylenchus
New Zealand, the total number of nematode taxa that has not yet been described. Therefore, the
was significantly lower (P < 0.05) than in Europe, data on feeding specialists and feeding generalists
the USA and South Africa, while there were no are in support of feeding-specialist hypothesis
significant differences in number of nematode taxa (2), but the lower number of feeding-specialist
among Europe, USA, South Africa, and Tasmania nematode taxa in the margins of the native range
(P > 0.05). Therefore, based on our level of iden- reduces the strength of the comparison.
tification the diversity of root-feeding nematodes
on A. arenaria was not particularly higher in its Comparison of nematodes of Ammophila arenaria
natural range (Europe) than in most introduced and similar plant species in introduced ranges
regions (except New Zealand), not supporting
the taxon diversity hypothesis (1). The comparison between the total number of
There were significantly fewer taxa of feeding root-feeding nematode taxa on native, similar
specialist nematodes in South Africa, New Zea- plant species and on A. arenaria showed data
land and Tasmania than in Europe and the USA points that were scattered around the 1:1 line (Fig-
(P < 0.05; one-way ANOVA F ¼ 15.03, df ¼ 4, ure 2a). In the introduced range, there were not
P < 0.0001; means not shown). This result sup- significantly fewer nematode taxa on A. arenaria
ports the feeding-specialist hypothesis (2), but than on similar native species (t8 ¼ 0.45,
only for the non-native regions in the southern P ¼ 0.667). This result is identical to the native
Hemisphere, where no feeding specialists were range, where A. arenaria did not have fewer
738

Table 1. Plant-feeding nematode families, taxa, and their feeding type in the root zone of pure Ammophila arenaria stands in
natural and introduced ranges.
Suborder: family Taxon Feeding type (after Yeates
et al. 1993)
Hoplolaimina: Heteroderidae Heterodera arenaria Cooper, 1955 Sedentary endoparasite
Heterodera sp. Sedentary endoparasite
Heteroderidae (juveniles) Sedentary endoparasite
Hoplolaimina: Meloidogynidae Meloidogyne maritima Jepson, 1987 Sedentary endoparasite
Meloidogyne sp. Sedentary endoparasite
Hoplolaimina: Pratylenchidae Pratylenchus brzeskii Karssen, Waeyenberge & Moens, 2000 Migratory endoparasite
Pratylenchus penetrans (Cobb, 1917) Migratory endoparasite
Pratylenchus scribneri Steiner, 1943 Migratory endoparasite
Pratylenchus sp. Migratory endoparasite
Tylenchina: Anguinidae Ditylenchus sp. Ectoparasite
Hoplolaimina: Hoplolaimidae Rotylenchus sp. Ectoparasite
Helicotylenchus pseudorobustus (Steiner, 1914) Ectoparasite
Helicotylenchus depressus Yeates, 1967 Ectoparasite
Helicotylenchus multicinctus (Cobb, 1893) Ectoparasite
Helicotylenchus sp. a Ectoparasite
Helicotylenchus sp. b Ectoparasite
Hoplolaimina: Tylenchorhynchidae Telotylenchus ventralis (Loof, 1963) Ectoparasite
Telotylenchus sp. Ectoparasite
Neodolichorhynchus microphasmis (Loof, 1960) Ectoparasite
Neodolichorhynchus dubius (Butschli, 1873) Ectoparasite
Neodolichorhynchus sp. Ectoparasite
Merlinius sp. Ectoparasite

Hoplolaimina: Belonolaimidae Morulaimus geniculatus Sauer, 1966 Ectoparasite

(syn. Scutellonema magnum Yeates, 1967
Hoplolaimina: Dolichodoridae Neodolichodorus arenarius (Clark, 1963) Ectoparasite
Criconematina: Criconematidae Criconema sp. Ectoparasite
Criconematina: Paratylenchidae Paratylenchus microdorus Andrassy, 1959 Ectoparasite
Paratylenchus sp. Ectoparasite
Criconematina: Hemicycliophoridae Hemicycliophora halophila Yeates, 1967 Ectoparasite
Hemicycliophora obesa Thorne, 1955 Ectoparasite
Hemicycliophora sp. Ectoparasite

Dorylaimina: Longidoridae Longidorus kuiperi Brinkman, Loof & Barbez, 1987 Ectoparasite
Longidorus sp. Ectoparasite
Diphtherophorina: Trichodoridae Paratrichodorus lobatus (Colbran, 1965) Ectoparasite
(syn. Trichodorus clarki Yeates, 1967)
Trichodorus sp. Ectoparasite

nematode taxa than adjacent Poaceae (t4 ¼ 1.84, The native species hypothesis (3) was only
P ¼ 0.9298; Figure 2a). Therefore, when consid- marginally rejected when testing for feeding-
ering all nematode taxa, the native species specialist (sedentary endoparasitic) nematodes
hypothesis (3) is rejected. Further analysis may (t8 ¼ )1.84, P ¼ 0.0519; Figure 2b). In this
elucidate to which extent A. arenaria has escaped case, feeding specialist taxa were not found on
from most its native species of root-feeding nem- A. arenaria in most introduced regions, except in
atodes, but it obviously is not exposed to fewer the USA. In Tasmania, there were cyst nema-
nematode taxa in the introduced range. todes detected in the root zone of the native
 739

8
Number of nematode taxa on
Ammophila arenaria 7 a a
6
5 ab
4
ab
3 b
2
1
0
Europe USA South Africa New Zealand Tasmania
Region

Figure 1. Average number of plant-feeding nematode taxa (±standard error) in the root zone of pure Ammophila arenaria stands
in natural and introduced ranges in five regions.

grass Spinifex sp., but not on A. arenaria. In the sampling period (Figure 3b). At peak plant
natural range, A. arenaria did not have fewer growth (August) there were 3 (for A. arenaria
feeding specialist nematode taxa than native and E. farctus) or 4 (C. epigejos) fewer taxa
plant species adjacent in the successional gradient recorded than the total number of taxa found
(t4 ¼ 2.14, P ¼ 0.9503; Figure 2b). over the entire growing season (Figure 3a).
The native species hypothesis (3) was rejected Therefore, in general, for these three plant spe-
when comparing number of generalist feeding cies a single sampling event, as has carried out
nematodes on A. arenaria and native plants in the for most sampling locations in the introduced
introduced range (t8 ¼ 0.78, P ¼ 0.772). There- regions, would have yielded only 60–77% (100

fore, A. arenaria did not have fewer generalist [number of nematode species in August divided
feeding (migratory endoparasite and ectoparasite) by the total number of nematode species
nematode taxa than similar native plant species in observed in 1 year]) of the actual nematode
the introduced range. In the natural range, taxon diversity.
A. arenaria also did not have fewer generalist For Himatangi Beach, New Zealand, one root-
feeding nematode taxa than adjacent Poaceae in feeding nematode species (Morulaimus geniculatus)
the successional gradient in coastal foredunes was present every month, while one (Hemicyclio-
(t4 ¼ 1.41, P ¼ 0.8849; Figure 2c). phora halophila) was present 4 out of 12 months
and the other two (Dolichodorus arenarius
The issue of sampling intensity around grasses in and Criconematidae juveniles) were present occa-
natural and introduced locations sionally and at very low densities. Moreover,
H. halophila was only found below 30-cm depth
We determined the cumulative number of plant- and would have been missed even given repeated
feeding nematode taxa found in the roots or in sampling of the topsoil. Thus a single sampling
the root zone sand of A. arenaria and two grasses would have the chance of missing 50–75% of all
in a natural sampling location in The Netherlands plant feeding nematode taxa present, but the
(Figure 3a). While the cumulative number of same error was applied to locations in which we
nematode taxa of the inner dune species C. epige- could control sampling intensity, by using data
jos was fairly constant over time, the cumulative from peak growth season only. The exceptions to
number of nematode taxa on the beach grass Ely- this sampling error are the literature records from
mus farctus and on A. arenaria steadily increased Scotland and Poland, that might have produced
with increased sampling (Figure 3a). an under- or overestimation owing to different
The total number of nematode taxa isolated sampling techniques, intensities, or sampling
each month shows some fluctuation during the dates.
Table 2. Plant-feeding nematode taxa on Ammophila arenaria (presence is indicated by +) in the natural range (Europe) and in four introduced regions (USA, South

740
Africa, New Zealand, and Tasmania).
Taxon Natural range Introduced range
Europe USA South Africa New Zealand Tasmania
PL S N Pn Ps Fa Fb SR SB SC HB D Bb Sf Kl Ko Kf Ks TV P C H S T B P D
1 2 3 4 5 6 7 1 2 3 4 1 2 3 4 5 6 7 8 1 2 3 4 5 6 1 2
Heterodera arenaria + +
Heterodera sp. a + + + + +
Heteroderidae (juveniles) + + +
Meloidogyne maritima + +
Meloidogyne sp. a + +
Pratylenchus brzeskii +
Pratylenchus penetrans +
Pratylenchus scribneri +
Pratylenchus sp. a + + + + + + + + + + + + + +
Ditylenchus sp. a + + + + + + + + + + + + + + + +
Rotylenchus sp. a + + + + + + + + + +
Helicotylenchus pseudorobustus +
Helicotylenchus depressus + +
Helicotylenchus multicinctus +
Helicotylenchus sp. a + +
Helicotylenchus sp. b + +
Telotylenchus ventralis + +
Telotylenchus sp. a + + + + + +
Neodolichorhynchus microphasmis + + +
Neodolichorhynchus dubius +
Neodolichorhynchus sp. a + + + +
Merlinius sp. a +
Morulaimus geniculatus + + + +
Neodolichodorus arenarius + +
Criconema sp. a + + + + + + + + +
Paratylenchus microdorus +
Paratylenchus sp. a + + + + + +
Hemicycliophora halophila +
Hemicycliophora obesa +
Hemicycliophora sp. a + + + +
Longidorus kuiperi +
Longidorus sp. a +
Paratrichodorus lobatus +
Trichodorus sp. a +
Abbreviations of the sampling locations are explained in the Materials and methods section (under ‘Geography’). Sedentary endoparasites (Heterodera spp. and Meloido-
gyne spp.) are considered as feeding specialists, whereas all other species are feeding generalists.
 741

10

All taxa Ammophila

8

0
(a) 0 2 4 6 8 10
Specialist taxa Ammophila

0
(b) 0 1 2
Generalist taxa Ammophila

0
0 2 4 6 8
(c) No. generalist nematode taxa on native species

Figure 2. Number of plant-feeding nematode taxa on native plant species plotted against the number of plant-feeding nematode
taxa on Ammophila arenaria. Closed symbols: natural range, open symbols: introduced range. (a) All plant-feeding nematodes, (b)
specialist feeding types (sedentary endoparasitic Heterodera spp. and Meloidogyne spp. only); note that some data were offset to
make individual data points visible and (c) generalist feeding types (ectoparasites) only.

Discussion found fewer specialist shoot-eating insects feeders

on introduced than on native plants. In the light
Mitchell and Power (2003) showed that invasive of this, we discuss the different hypotheses,
plants have fewer aboveground plant pathogen whether invasive A. arenaria may have escaped
and virus species in their introduced than in nat- its natural root-feeding nematodes, and how this
ural ranges, and that invasive plants also had may relate to the ecology of A. arenaria in natu-
fewer of these pathogen species than did similar ral and introduced regions.
species native to the introduced ranges. Our On average, A. arenaria did not have more
results for root-feeding nematodes, however, did root-feeding nematode taxa in its natural range
not fit into the general pattern observed for than in three out of four introduced regions
aboveground plant pathogens and viruses. Never- worldwide. This leaves little support for our first
theless, we found that specialist feeders (seden- hypothesis that an invasive plant may have fewer
tary endoparasites) of A. arenaria were absent in root-feeding nematode taxa in its introduced than
the introduced regions in the southern hemi- in its natural range. The only marked reduction
sphere. This agrees with observations by Jobin of the number of root-feeding nematodes was
et al. (1996) and Memmott et al. (2000), who observed in New Zealand, where some classes, for
742

Cumulative number of nematode taxa

12
(a)
10

0
Number of nematode taxa in each month

9
(b)
8
7
6
5
4
3
2
1
0
A M J J A S O N
Month of the year

Figure 3. Number of plant-feeding nematode taxa on Ammophila arenaria (r), an earlier successional grass Elymus farctus (j) and
a later successional grass Calammagrostis epigejos (N), in dunes of the Netherlands. (a) Cumulative number of taxa since the start
of monthly sampling; (b) average number of taxa at each sampling date.

example migratory endoparasitic nematodes, do of Yeates et al. (1993) and considered sedentary
not occur. The dominant native plant species in endoparasitic nematodes feeding specialists
New Zealand foredunes was taxonomically quite because of their capacity to initiate feeding struc-
distinct from the introduced species (in New Zea- tures within plant roots. In South Africa, Tasma-
land, the dominant native species belonged to the nia, and New Zealand, A. arenaria had no
Cyperaceae, while the introduced species belonged specialist feeders (the sedentary, endoparasitic
to the Poaceae). This conclusion confirms that of cyst and root knot nematodes), while in the USA
Knevel et al. (2004) who found less similarity we found cyst nematodes. Thus, in introduced
between nematodes in the root zone of A. arenaria regions in the southern hemisphere A. arenaria
and South African foredune dicotyledonous appeared to have escaped from both feeding spe-
plants than between A. arenaria and South Afri- cialist taxa, while in the USA plants had escaped
can foredune grasses (Poaceae). from one feeding specialist nematode taxon.
The second hypothesis was that in its intro- Therefore, our hypothesis on fewer feeding spe-
duced range A. arenaria would have fewer feed- cialists on the invasive plant species is supported
ing specialists than in its natural range, similar to in the introduced regions in the southern hemi-
the situation for aboveground insects on some sphere, but not in the introduced region in the
alien plant species (Jobin et al. 1996; Memmott northern hemisphere.
et al. 2000). For our feeding-specialist taxa, we In general, A. arenaria did not have fewer
used the classification of nematode feeding types nematode taxa than similar native plant species
 743

in the introduced range, which is not in support absent in the northern extreme, and in parts of
of our third hypothesis. However, the compari- the southern extreme of the natural range of
son of feeding specialist taxa on A. arenaria and A. arenaria. Therefore, the number of root-feed-
similar native species in the introduced range ing nematode taxa increased with geographic
resulted into a close to significant effect, so that range, while the number of (feeding-specialist)
the native species hypothesis may be supported taxa per site decreased towards the margins of
when considering feeding specialists, but not the natural range of the host plant.
when considering feeding generalists. In our comparison, we have averaged the num-
In our field survey along the west coast of the ber of taxa within sampling locations and consid-
USA, we also collected soil samples from Ammo- ered sampling locations within geographic
phila breviligulata, the American cogener of regions (Europe, USA, South Africa, Tasmania,
A. arenaria, natural to the east coast of North and New Zealand) as replicates. However, if the
America/USA (Maun 1998). At Sunset Beach number of feeding specialists is lower at the mar-
Road, Oregon, there was only one feeding spe- gins than in the centre of the geographic range,
cialist (Heterodera sp.) and one feeding generalist as it appears to be in Europe, then averaging the
(Tylenchorhynchus; currently Neodolichorhynchus number of taxa within a geographic range may
sp.), while at two locations in Delaware on the reduce the power of the analysis. Therefore,
east coast of the USA, where A. breviligulata is future analyses might also focus on taxon varia-
natural, seven genera (of which at least one had tion within and between natural and introduced
two species) had been found; one sedentary, regions or, alternatively, compare between natu-
endoparasitic feeding specialist and six migratory, ral and introduced regions with similar environ-
feeding generalists (Seliskar and Huettel 1993). mental conditions, for example climate and
This comparison might indicate an escape of rainfall.
A. breviligulata from its natural root-feeding A feeding specialist may not necessarily be a
nematode community, while A. breviligulata has host specialist. In north-western Europe, the feed-
not been reported to be exposed to Heterodera ing specialist Heterodera arenaria occurs on both
and Tylenchorhynchus spp. in its natural range, A. arenaria in the mobile (yellow) dunes (Cooper
except in lacustrine dunes along the Great Lakes and Harrison 1973; Cook 1982; Robinson et al.
in Canada (Little and Maun 1996). However, the 1996; Clapp et al. 2000) and on the beach grass
study by Seliskar and Huettel (1993) was more E. farctus (C.D. van der Stoel, W.H. van der Put-
intensive than ours, and our repeated sampling ten and H. Duyts, unpublished results). However,
throughout the year shows that one-time surveys the feeding specialist Meloidogyne maritima
may underestimate the actual diversity of root- occurs only on A. arenaria (Karssen et al. 1998a),
feeding nematodes in foredunes by 50–75%. Nev- while M. duytsi occurs only on E. farctus (Kars-
ertheless, the spectrum of nematode taxa to sen et al. 1998b). Cross-inoculation studies show
which A. breviligulata is exposed in its introduced that H. arenaria multiplies on both plants (van
range seems quite different from that in its natu- der Stoel 2001), but that both M. maritima and
ral range. M. duytsi seem more host specific (W.H. van der
In a survey on pathogens on grasses, Clay Putten and H. Duyts, unpublished results). While
(1995) concluded that the number of pathogen H. arenaria is not aggressively reducing the
species increased with grass geographic range. growth of its host plant (van der Stoel 2001), M.
This conclusion may also hold for root-feeding maritima does so, however, not in the presence of
nematodes on A. arenaria, but patterns for feed- H. arenaria and P. penetrans (E.P. Brinkman
ing generalists and specialists appear to be in et al., submitted for publication).
contrast. For example the ectoparasite (feeding It is not only the number of pathogenic species
generalist) genus Helicotylenchus. was represented left behind in the natural territory that is impor-
by different species in the nortern, central, and tant, but that there should also be information
southern locations of its natural range. The sed- on their role in the ecology of the invasive plants
entary endoparasites (feeding specialists) Hetero- (Mitchell and Power 2003). Feeding specialist
dera and Meloidogyne, on the other hand, were dune nematodes and feeding generalists show
744

interspecific competition, but this did not appear be more similar to the pattern observed for
to be strong enough to control the abundance of aboveground herbivorous insects (Memmott
the nematodes (Brinkman et al. 2004). However, et al. 2000). Our study shows that comparisons
interspecific control among feeding specialist between natural enemies on invasive plants in
nematodes of A. arenaria points at non-linear natural and introduced ranges requires full cover-
effects of diversity of feeding specialists. There- age of the entire natural and non-natural ranges,
fore, due to non-linear effects of nematode diver- as the number of feeding-specialist nematode
sity on growth of A. arenaria (E.P. Brinkman taxa seemed lower in the extremes than in the
et al., submitted for publication) the ecological central part of the natural range. Future studies
consequences of reduced diversity of feeding spe- may need to include comparisons between, as
cialist nematodes in the introduced regions can- well as within natural and introduced ranges.
not be directly interpreted as enemy release. On Data on numbers of nematode taxa on invasive
the other hand, abundance of most feeding types plants in their natural and introduced regions are
of nematodes in the root zone of A. arenaria an indication of enemy escape, but the ultimate
(excluding H. arenaria) relates positively to test requires that the ecological consequences of
growth reduction, so that reduced nematode the reduced root-feeders load needs to be quanti-
abundance may indicate at enhanced plant fied. This does not only apply to root feeders,
invasiveness. but also to aboveground feeders.
In natural dunes root-feeding nematodes are
not the only cause of growth reduction of A. are-
naria. Growth experiments with A. arenaria in Acknowledgements
South Africa (Knevel et al. 2004) and the USA
(Beckstead and Parker 2003) have shown that the We thank Roy Lubke and Brad Ripley for intro-
soils from these two introduced regions also pro- ducing WvdP to South African sand dunes and
duce growth-reducing potential even in the for collecting the soil samples in the USA, Kate
absence of feeding-specialist nematodes. These Lessells for collecting the soil samples in the
results are in agreement with studies in the natu- Camarque and Emma Watt for collecting the
ral range (van der Stoel et al. 2002), while field samples in Tasmania. This study has been
densities of feeding generalists were found to have performed as part of the EU-INVASS project
little impact on plant biomass production (de and the EU-EcoTrain project (contracts
Rooij-van der Goes 1995). Therefore, further IC18CT970145 and HPRN-CT-2002-00210 with
understanding of A. arenaria invasiveness will the European Commission). An ISAT grant from
require information on escape from other soil the Royal Society of New Zealand facilitated
enemies as well, such as pathogenic soil fungi. writing of this paper.
Our study shows that plants may escape from
their feeding-specialist natural root-feeding nema-
todes, but that they have a high chance of References
becoming exposed to almost as many feeding
generalist nematode taxa in their introduced Beckstead J and Parker IM (2003) Invasiveness of Ammophila
range. In the analysis made by Mitchell and arenaria: release from soil-borne pathogens? Ecology 84:
Power (2003), A. arenaria would fit in the cate- 2824–2831
Bever JD (1994) Feedback between plants and their soil com-
gory of ‘intensively used plants’. These plant spe-
munities in an old field community. Ecology 75: 1965–1977
cies had more aboveground pathogens than Bever JD (2003) Soil community feedback and the coexistence
plants that were not intensively used. Mitchell of competitors: conceptual frameworks and empirical tests.
and Power (2003) suggested that their study on New Phytologist 157: 465–473
aboveground pathogens and viruses may stand as Blossey B and Hunt-Joshi TR (2003) Belowground herbivory
by insects: influence on plants and aboveground herbi-
a model for belowground pathogens, but that
vores. Annual Review of Entomology 48: 521–547
suggestion needs further verification. However, Bongers T (1988) De nematoden van Nederland. Stichting
our analysis suggests that the escape pattern of Uitgeverij Koninklijke Nederlandse Natuurhistorische
invasive plants from root-feeding nematodes may Vereniging, Utrecht, Nederland
 745

Brinkman EP, van Veen JA and van der Putten WH (2004) Hertling UM and Lubke RA (2000) Assessing the potential
Plant recruitment of endoparasitic nematodes may for biological invasion – the case of Ammophila arenaria in
influence, but not regulate, ectoparasitic nematodes. South Africa. South African Journal of Science 96: 520–527
Applied Soil Ecology 27: 65–75 Heyligers PC (1985) The impact of introduced plants on
Brown VK and Gange AC (1992) Secondary plant succession – foredune formation in south-east Australia. Proceedings of
how is it modified by insect herbivory. Vegetatio 101: 3–13 the Ecological Society of Australia 14: 23–41
Brzeski MW (1998) Nematodes of Tylenchina in Poland and Huiskes AHL (1979) Biological flora of the British Isles: Am-
temperate Europe. Muzeum i Instytut Zoologii Polska Ak- mophila arenaria (L.) Link (Psamma arenaria (L.) Roem. et
ademia Nauk, Warszawa, Poland, 397 pp Schult.: Calamagrostis arenaria (L.) Roth). Journal of Ecol-
Buell AC, Pickart AJ and Stuart JD (1995) Introduction ogy 67: 363–382
history and invasion patterns of Ammophila arenaria on Jobin A, Schaffner U and Nentwig W (1996) The structure of
the north coast of California. Conservation Biology 9: the phytophagous insect fauna on the introduced weed Sol-
1587–1593 idago altissima in Switzerland. Entomologia Experimentalis
Callaway RM, DeLuca TH and Belliveau WM (1999) Bio- et Applicata 79: 33–42
logical-control herbivores may increase competitive ability Karssen G, van Aelst A and Cook R (1998a) Redescription
of the noxious weed Centaurea maculosa. Ecology 80: of the root-knot nematode Meloidogyne maritima Jepson,
1196–1201 1987 (Nematoda: Heteroderidae), a parasite of Ammophila
Clapp JP, van der Stoel CD and van der Putten WH (2000) arenaria (L.) Link. Nematologica 44: 241–253
Rapid identification of cyst (Heterodera spp., Globodera Karssen G, van Aelst A and van der Putten WH (1998b)
spp.) and root-knot (Meloidogyne spp.) nematodes on the Description of Meloidogyne duytsi n. sp. (Nematoda: Hetero-
basis of ITS2 sequence variation detected by PCR-SSCP deridae), a root-knot nematode from Dutch coastal foredun-
(PCR-Single-Strand Conformational Polymorphism) in cul- es. Fundamental and Applied Nematology 21: 299–306
tures and field samples. Molecular Ecology 9: 1223–1232 Keane RM and Crawley MJ (2002) Exotic plant invasions
Clark SE, van Driesche RG, Sturdevant N and Kegly S and the enemy release hypothesis. Trends in Ecology and
(2001b) Effect of root feeding insects on spotted knapweed Evolution 17: 164–170
(Centaurea maculosa) stand density. Southwestern Ento- Kisiel M (1970) On the systematic and geographical distribu-
mologist 26: 129–135 tion of the nematodes inhabiting sand dunes in an Ammo-
Clark SE, van Driesche RG, Sturdevant N, Elkinton J and phila arenaria plant community. Zeszyty Naukowe Wyszej
Buonaccorsi JP (2001a) Effects of site characteristics and Szkoly Rolniczej w Szczecinie 34: 151–193
release history on establishment of Agapeta zoegana (Lepi- Kliromonos JN (2002) Feedback with soil biota contributes
doptera: Cochylidae) and Cyphocleonus achates (Coleop- to plant rarity and invasiveness in communities. Nature
tera: Curculionidae), root-feeding herbivores of spotted 417: 67–70
knapweed, Centaurea maculosa. Biological Control 22: Knevel IC, Lans T, van der Putten WH, Menting FBJ and
122–130 Hertling UM (2004) The role of the enemy release hypothesis
Clay K (1995) Correlates of pathogen species richness in the and the biotic resistance hypothesis in the establishment of
grass family. Canadian Journal of Botany 73 (Suppl. 1A– the alien Ammophila arenaria in South Africa. Oecologia 141:
D): S42–S49 502–510
Cook R (1982) Cereal and grass hosts of some gramineous Levine JM, Vila M, D’Antonio CM, Dukes JS, Grigulis K
cyst nematodes. EPPO Bulletin 12: 99–411 and Lavorel S (2003) Mechanisms underlying the impacts
Cooper BA (1955) A preliminary key to British species of of exotic plant invasions. Proceedings of the Royal Society
Heterodera for use in soil examinations. In: McE Kevan of London, Series B, Biological Sciences 270: 775–781
DK (ed) Soil Zoology, pp 269–280. Butterworths Scientific Little LR and Maun MA (1996) The ‘Ammophila problem’
Publications, London revisited: a role for mycorrhizal fungi. Journal of Ecology
Cooper JI and Harrison BD (1973) The role of weed hosts 84: 1–7
and the distribution and activity of vector nematodes in Liebhold AM, MacDonald WL, Bergdahl D and Mastro VC
the ecology of tobacco rattle virus. Annals of Applied Bio- (1995) Invasion by Exotic Forest Pests: A Threat to Forest
logy 73: 53–66 Ecosystems. Forest Science Monograph (No. 30): ii + 49
De Deyn GB, Raaijmakers CE, Zoomer HR, Berg MP, De pp. Society of American Foresters, Bethesda, MD
Ruiter PC, Verhoef HA, Bezemer TM and van der Putten Maun MA (1998) Adaptations of plants to burial in coastal
WH (2003) Soil invertebrate fauna enhances grassland suc- sand dunes. Canadian Journal of Botany 76: 713–738
cession and diversity. Nature 422: 711–713 Memmott J, Fowler SV, Paynter Q, Sheppard, AW and Sy-
De Rooij-van der Goes PCEM (1995) The role of plant-para- rett P (2000) The invertebrate fauna on broom, Cytisus
sitic nematodes and soil-borne fungi in the decline of scoparius, in two native and two exotic habitats. Acta
Ammophila arenaria (L.) Link. New Phytologist 129: 611– Oecologia 21: 213–222
669 Mitchell CE and Power AG (2003) Release of invasive plants
Hertling UM and Lubke RA (1999) Use of Ammophila arena- from fungal and viral pathogens. Nature 421: 625–627
ria for dune stabilisation in South Africa and its current Owen SJ (1996) Ecological Weeds on Conservation Land in
distribution – perceptions and problems. Environtal New Zealand: A Database. Department of Conservation,
Management 24: 467–482 Wellington, New Zealand, 118 pp
746

Packer A and Clay K (2000) Soil pathogens and spatial pat- dera arenaria in Coastal Foredunes. PhD thesis
terns of seedling mortality in a temperate tree. Nature 404: Wageningen University, The Netherlands
278–281 van der Stoel CD, van der Putten WH and Duyts H (2002)
Packer A and Clay K (2002) Soil pathogens and Prunus sero- Development of a negative plant-soil feedback in the
tina seedlings and sapling growth near conspecific trees. expansion zone of the clonal grass Ammophila arenaria fol-
Ecology 84: 108–119 lowing root formation and nematode colonisation. Journal
Pimentel D (2002) Biological Invasions: Economic and of Ecology 90: 978–988
Environmental Costs of Alien Plant, Animal, and Microbe Verschoor BC, De Goede RGM, De Vries FW and Brussaard
Species. CRC Press, Boca Raton, FL, 369 pp L (2001) Changes in the composition of the plant-feeding
Reinhart KO, Packer A, van der Putten WH and Clay K nematode community in grasslands after cessation of fertil-
(2003) Plant-soil biota interactions and spatial distribution iser application. Applied Soil Ecology 17: 1–17
of black cherry in its native and invasive ranges. Ecology Virginia RA, Jarell WM, Whitford WG and Freckman DW
Letters 6: 1046–1050 (1992) Soil biota and soil properties in the surface rooting
Rejmanek M and Richardson DM (1996) What attributes zone of mesquite (Prosopis glandulosa) in historical and
make some plants more invasive? Ecology 77: 1665–1661 recently desertified Chihuahuan Desert habitats. Biology
Robinson AF, Orr CC and Abernathy JR (1978) Distribution and Fertility of Soils 14: 90–98
of Nothanguina phyllobia and its potential as a biological Wall JW, Skene KR and Neilson R (2002) Nematode com-
control agent for silver-leaf nightshade. Journal of Nema- munity and trophic structure along a sand dune succession.
tology 10: 362–366 Biology and Fertility of Soils 4: 293–301
Robinson AJ, Stone AR, Hooper DJ and Rowe JA (1996) A Wardle DA (2002) Communities and Ecosystems: Linking the
redescription of Heterodera arenaria Cooper 1955, a cyst Aboveground and Belowground Components. Princeton
nematode from marram grass. Fundamental and Applied University Press, Princeton, NJ
Nematology 19: 109–117 Wiedemann AM and Pickart AJ (1996) The Ammophila prob-
Seliskar DM and Huettel RN (1993) Nematode involvement lem on the Northwest Coast of North America. Landscape
in die-out of Ammophila breviligulata (Poaceae) on the and Urban Planning 34: 287–299
Mid-Atlantic coastal dunes. Journal of Coastal Research 9: Williamson M (1996) Biological Invasions. Chapman & Hall,
97–103 London
Stanton NL (1988) The underground in grasslands. Annual Yeates GW (1967) Studies on nematodes from dune sands. 9.
Review of Ecology and Systematics 19: 573–589 Quantitative comparison of the nematode faunas of six
Torchin ME, Lafferty KD, Dobson AP, McKenzie VJ and localities. New Zealand Journal of Science 10: 927–948
Kuris AM (2003) Introduced species and their missing par- Yeates GW (1968) An analysis of annual variation of the
asites. Nature 421: 628–630 nematode fauna in dune sand, at Himatangi Beach, New
van der Putten WH (2003) Plant defense below ground and Zealand. Pedobiologia 8: 173–207
spatio-temporal processes in natural vegetation. Ecology Yeates GW (1999) Effects of plants on nematode community
84: 2269–2280 structure. Annual Review of Phytopathology 37: 127–149
van der Putten WH and van der Stoel CD (1998) Plant para- Yeates GW and Williams PA (2001) Influence of three inva-
sitic nematodes and spatio-temporal variation in natural sive weeds and site factors on soil microfauna in New Zea-
vegetation. Applied Soil Ecology 10: 253–262 land. Pedobiologia 45: 367–383
van der Putten WH, van Dijk C and Peters BAM (1993) Yeates GW, Bongers T, De Goede RGM, Freckman DW and
Plant-specific soil-borne diseases contribute to succession Georgieva SS (1993) Feeding habits in soil nematode fami-
in foredune vegetation. Nature 362: 53–56 lies and genera – an outline for soil ecologists. Journal of
van der Stoel CD (2001) Specificity, Pathogenicity and Popu- Nematology 25: 315–331
lation Dynamics of the Endoparasitic Nematode Hedero-