Adaptive rather than non-adaptive evolution of Mimulus guttatus in its invasive range
Mark van Kleunena,b,c,* Markus Fischerb,c ,
"Centre for Invasion Biology, School of Biological and Conservation Sciences, University of KwaZulu Natal, P. Bag XOI Scottsville, Pietermaritzburg 3209, South Afi'ica
blnstitutefor Biochemistry and Biology, University of Potsdam, Maulbeeral/ee I, D 14469 Potsdam, Germany Clnstitute of Plant Sciences, University of Bern, Altenbergrain 21, CH 3013 Bern, Switzerland
Abstract
Adaptive and non-adaptive evolutionary processes are likely to play important roles in biological invasions but their relative importance has hardly ever been quantified. Moreover, although genetic differences between populations in their native versus invasive ranges may simply reflect different positions along a genetic latitudinal cline, this has rarely been controlled for. To study non-adaptive evolutionary processes in invasion of Mimulus guttatus, we used allozyme analyses on offspring of seven native populations from western North America, and three and four invasive populations from Scotland and New Zealand, respectively. To study quantitative genetic differentiation, we grew 2474 plants representing 17 native populations and the seven invasive populations in a common greenhouse environment under temporarily and permanently wet soil conditions. The absence of allozyme differentiation between the invasive and native range indicates that mUltiple genotypes had been introduced to Scotland and New Zealand, and suggests that founder effects and genetic drift played small, if any, roles in shaping genetic structure of invasive M. guttatus populations. Plants from the invasive and native range did not differ in phenology, floral traits and sexual and vegetative reproduction, and also not in plastic responses to the watering treatments. However, plants from the invasive range produced twice as many flower-bearing upright side branches than the ones from the native populations. Further, with increasing latitude of collection, vegetative reproduction of our experimental plants increased while sexual reproduction decreased. Plants from the invasive and native range shared these latitudinal clines. Because allozymes showed that the relatedness between native and invasive populations did not depend on latitude, this suggests that plants in the invasive regions have adapted to the local latitude. Overall, our study indicates that quantitative genetic variation of M. guttatus in its two invasive regions is shaped by adaptive evolutionary processes rather than by n,0n-adaptive ones.
Zusammenf assung
Obwohl sowohl adaptive als auch nichtadaptive evolutionare Prozesse eine wichtige Rolle fUr biologische Invasionen spielen konnen, wurde ihre relative Bedeutung bisher kaum quantifiziert. Zudem wurde bisher kaum dafUr
·Corresponding author. Institute of Plant Sciences, University of Bern, Altenbergrain 21, CH 3013 Bern, Switzerland. Tel.: + 41316314923;
fax: +41316314942.
Email address:vkleunen@ips.unibe.ch (M. van Kleunen).
First publ. in: Basic and Applied Ecology 9 (2008), 3, pp. 213-223
Konstanzer Online-Publikations-System (KOPS) URL: http://nbn-resolving.de/urn:nbn:de:bsz:352-133467
214
kontrolliert, ob genetische Unterschiede zwischen Populationen in fremden und einheimischen Gebieten nicht einfach durch unterschiedliche Positionen entlang eines durch geographische Breite verursachten genetischen Gradienten bedingt sein k6nnen. Um nichtadaptive evolutionare Prozesse bei der Invasion von Mimulus gutta/us zu untersuchen, benutzten wir Allozymanalysen von Nachkommen von sieben einheimischen Populationen aus Nordamerika und von drei und vier invasiven Populationen aus Schottland und Neuseeland. Um quantitativgenetische Differenzierung zu untersuchen, zogen wir 2474 Pflanzen aus 17 einheimischen und sieben invasiven Populationen in einem Experimentiergarten unter permanenter oder nur zeitweiser Bewasserung auf. Die fehlende Allozymdifferenzierung zwischen den einheimischen und invasiven Gebieten zeigt an, dass mehrere Genotypen nach Schottland und Neuseeland eingefUhrt wurden und dass Griindereffekte und genetische Drift wenn iiberhaupt nur eine kleine Rolle fUr die genetische Struktur invasiver M. gutta/us Populationen spielen. Pflanzen aus den einheimischen und invasiven Gebieten unterschieden sich weder phanologisch noch in sexueller oder vegetativer Reproduktion oder in ihren plastischen Reaktionen auf die Bewasserungsbehandlungen. Allerdings produzierten Pflanzen aus dem invasiven Gebiet doppelt so viele bliitentragende aufrechte Seitentriebe. AuI3erdem nahm mit zunehmendem Breitengrad sowohl fUr einheimische als auch fUr invasive Herkunftspopulationen die vegetative Reproduktion der experimentellen Pflanzen zu, wahrend die sexuelle Reproduktion abnahm. Da die Allozyme anzeigten, dass die Verwandtschaft zwischen einheimischen und invasiven Populationen nicht vom Breitengrad abhing, deutet das darauf hin, dass sich die Pflanzen in invasiven Gebieten an den lokalen Breitengrad angepasst haben. Insgesamt zeigt unsere Studie, dass die quantitativgenetische Variation von M. gutta/us in seinen beiden invasiven Regionen durch adaptive, und nicht durch nicht-adaptive, evolutiona.re Prozesse gepragt ist.
KeywOI'ds: Alien plants; Baker's law; EICA hypothesis; Founder effects; Genetic drift; Isoenzymes; Neobiota; Non indigenous plants; Selection
Introduction
Many species have been introduced from their native range into new regions where some of them have become invasive. Although an invasive alien species may have been pre-adapted to its new region, it is likely that post-introduction adaptive evolution increases invasiveness (Facon et aI., 2006; Lee, 2002; Miiller- Scharer, Schaffner, & Steinger, 2004; Strayer, Eviner, Jeschke, & Pace, 2006), and could result in genetic differentiation between the native and invasive range.
Research on this topic is dominated by the evolutionary increased competitive ability (EICA) hypothesis which suggests that due to a lack of natural enemies in the introduced range, invasive plants may have evolved a higher competitive ability at the costs of resistance to herbivores and pathogens (Blossey & N6tzold, 1995).
Post-introduction evolution of traits other than growth and resistance, however, has received hardly any attention (but see Blair & Wolfe, 2004).
In addition to the potential release from natural enemies, several other selective forces are likely to differ between the introduced and native range, and may have resulted in adaptive evolution. First, it is likely that adaptive phenotypic plasticity is selected for during invasions because it increases environmental tolerance
(e.g., Baker, 1974; Richards, Bossdorf, Muth, Gure-
vitch, & Pigliucci, 2006; Williams, Mack, & Black, 1995). Therefore, plants in the invasive range may have evolved higher levels of adaptive phenotypic plasticity
(Kaufman & Smouse, 2001). Second, in the invasive range, particularly shortly after introduction when populations are still small, plants might suffer from mate and pollinator limitation (cr., van Kleunen &
Johnson, 2005). Therefore, plants in the invasive range may have evolved increased vegetative reproduc- tion and floral characteristics that promote self- fertilization (Baker, 1955, 1974; Barrett & Husband, 1990), such as small size and short anther-stigma separation.
Plants may also be introduced at latitudes different from the one in their native range. Here, they may experience different climatic conditions such as tem- peratures, length of the growing season and day lengths that may affect life-history traits. Phenology and growth of plants often show latitudinal clines that have a genetic component both in the native (Neuffer & Hurka, 1986;
Olsson & Agren, 2002) and invasive (Kollmann &
Banuelos, 2004; Weber & Schmid, 1998) range. This implies that apparent genetic differences between invasive and native populations might simply reflect their different positions along the latitudinal cline. So far, however, latitude of collection has rarely been considered in comparative studies between plants from the native and invasive ranges (but see Maron, Vila, Bommarco, Elmendorf, & Beardsley, 2004). To get a better understanding of the importance of adaptive evolution for plant invasions, studies comparing plants from invasive and native populations should not be restricted to testing the EICA hypothesis but also test
for differences in other traits such as phenology, reproduction and floral characteristics, as well as phenotypic plasticity and latitudinal clines in these traits.
Quantitative genetic differences between plants from the invasive and native ranges are not necessarily the result of adaptive evolution in response to new selection pressures but could also be the result of non-adaptive evolutionary processes such as founder effects and genetic drift (e.g., Amsellem, Noyer, Le Bourgeois, &
Hossaert-McKey, 2000). Consistency of genetic differ- entiation between populations from the native range and the ones from several independently invaded regions would present strong evidence that founder effects and genetic drift have played minor roles in genetic differentiation compared to adaptation. Moreover, the importance of adaptive evolution relative to founder effects and genetic drift can be tested by comparison of quantitative genetic differentiation with neutral marker differentiation (Lande, 1992; Lynch, 1994). These approaches have only rarely been used in comparisons between native and invasive populations (but see Joshi
& Vrieling, 2005; Maron et aI., 2004), and never together in the same study.
We tested for quantitative genetic and neutral marker differentiation between invasive and native populations, and adaptation to latitude of the invasive region for the herbaceous plant Mimulus guttatus. This species is suitable to address these questions because it is invasive in more than one region (New Zealand and parts of Europe) and its native area (western North America) comprises a large latitudinal range (Mexico-Alaska).
Moreover, M. guttatus exhibits high levels of quantita- tive genetic variation in its native range (e.g., van Kleunen & Ritland, 2004) suggesting that rapid evolu- tion would be possible after introduction elsewhere. The species grows in temporarily and permanently wet habitats, which implies that plastic responses to water- ing conditions may be important.
We compared phenology, growth, floral traits and reproduction, and plasticity therein between invasive and native populations by growing 2474 plants repre- senting 17 populations from the whole latitudinal gradient in the native range, and three and four invasive populations from Scotland and New Zealand, respec- tively, in a common greenhouse environment under temporarily and permanently wet conditions. To test for neutral marker differentiation between invasive and native populations of M. guttatus, we carried out allozyme analysis on a total of 800 offspring represent- ing the seven invasive populations and seven of the 17 native populations. We use comparisons between plants from the two invasive regions and between quantitative and molecular genetic differentiation to distinguish between adaptive and non-adaptive evolutionary pro- cesses in the invasive range of M. guttatus.
215
Materials and methods
Study speciesThe yellow monkey flower M. guttatus Fisch. ex DC.
(Phrymaceae) is native to western North America and is invasive in New Zealand and parts of Europe. The species occurs in temporarily and permanently wet habitats such as streams, ditches and wet grasslands (Grant, 1924).
Shoots of M. guttatus have 0.1-1 m long branched stems. The upper side branches grow upright and produce flowers (further referred to as upright side branches), whereas most lower side branches are creeping and contribute to vegetative reproduction (further referred to as stolons).
M. guttatus has funnel-shaped, zygomorphic, yellow flowers that are 1--4cm in length. Each flower has a pair of short stamens and a pair of long stamens, which are usually exceeded by the single pistil. The self-compatible species is diploid (2n
=
28), and outcrossing rates are usually positively correlated with flower size and anther-stigma separation (Fenster & Ritland, 1994;van Kleunen & Ritland, 2004; but see Ivey & Carr, 2005).
Plant material
In 2002 and 2003, volunteers (see Acknowledgements) and the first author collected seeds from 17 populations of M. guttatus in its native range in western North America and from three and four populations in its invasive regions in Scotland and New Zealand, respec- tively (see Appendix A, Supplemantary Table 1).
Because the native origin of the plants that founded the invasive populations is unknown, it is more important to have a large representation of the native range than of the invasive regions. For each population, one full seed capsule was collected from each of 19 plants (further referred to as seed families) that were at least 1 m apart. We received fewer seed families for two populations from the native range (Loss Creek and Sandcut Creek, B.C.: n = 5 each), one population from Scotland (Tarland 2: n
=
14) and one from New Zealand (Otatara: n=
16) (see Appendix A, Supple- mantary Table I). In total, we used 420 seed families.Allozyme analysis
To estimate selectively neutral marker variation among and within invasive and native regions, we carried out allozyme analysis on offspring of the three populations from Scotland, the four populations from New Zealand, and seven populations from
216
western North America (see Appendix A, Suppleman- tary Table I).
On 21 June 2005, we sowed seeds of each of 10 seed families per population separately into trays (19 x 14 x 5 cm) filled with a 1:2 mixture of sand and commercial potting compost in a greenhouse. Because there were no or only few seedlings for IS of the seed families, we sowed seeds of an additional IS seed families on 6 July 2005. We collected fresh corollas from one to seven offspring per seed family for allozyme electrophoresis. Corollas were ground the same day in an extraction buffer containing 0.05 M Na2HP04 at pH 7.0, one drop of Tween-80 and 4.14mg BSA per IOml of extraction buffer, 0.023 M DIECA, 0.013 M DTT, 4.57 mM EDTA, 0.82 M sucrose and 5.25 mM PVP-40 (Leclerc-Potvin & Ritland, 1994). These extracts were kept at -80°C until use.
We used an electrode buffer of 0.04 M citric acid monohydrate adjusted to pH 6.1 with N-(3-aminopro- pyl)-morpholine, and 11% starch/5% sucrose gels in electrode buffer diluted I :20 with water (Ritland &
Ganders, 1987) to assay the following three poly- morphic enzyme systems: alcohol dehydrogenase (ADH) and phosphoglucoisomerase (2 loci; PGI-I and PGI-2) and 6-phophogluconic dehydrogenase (2 loci;
6PGD-I and 6PGD-2).
Common environment experiment Experimental set-up
We filled 72 mUlti-pot trays, each with 35 cells (diameter
=
6.0 cm, depth=
6.5 cm), with a 3: I mixture of commercial potting compost and sand. We placed the trays onto drip trays, and arranged them in three equally sized groups (blocks) in a greenhouse in Potsdam, Germany (latitude 52°24'N, longitude 13°0I'E). To test for phenotypic plasticity, we assigned half of the trays within each block to a temporarily wet treatment and the other half to a permanently wet treatment that simulated the different watering conditions in natural habitats of M. guttatus. Between 2 and 4 June 2004, for each of the six treatment-by-block combinations, we sowed ca. 10 seeds of each of the 420 seed families, representing the 17 native and seven invasive popula- tions, in one randomly chosen cell.During the first 3 weeks, we watered all trays from the bottom by pouring water into the drip trays to prevent seeds from floating into neighboring cells. Thereafter, when seeds had germinated in the majority (65%) of cells, we removed drip trays under the plants in the temporarily wet treatment to allow drainage. From then on, plants were top-watered at least every other day.
Thus, plants in the permanently wet treatment with drip trays experienced continuously wet to water- logged conditions and the ones in the temporarily wet
treatment experienced both wet and dry conditions.
Once a week, if necessary, we thinned seedlings to one per cell, thereby keeping the oldest seedling. Because some seeds did not germinate and a few plants did not survive, the experiment included 2474 instead of 2520 plants.
Measurements
To determine time from germination to anthesis, we recorded the presence of seedlings and subsequent flowers in each cell at least every second day. Between 2 and 6 August 2004, as measures of growth, we measured height and counted the number of upright side branches of all plants. As measures of vegetative reproduction, we counted the number of stolons and measured the length of the longest stolon. As an estimate of overall vegetative reproduction, we calcu- lated total stolon length by mUltiplying the number of stolons by the length of the longest stolon branch. As estimate of sexual reproduction, we counted the number of flowers (including flower buds and seed capsules). As estimates of floral size and anther-stigma separation, we measured on the most recently fully opened flower on each plant, the width and length of the corolla and the distance between the stigma tip and the upper pair of anthers. As a measure of corolla shape, we calculated the corolla length-width ratio.
Analyses F-statistics
We assayed 800 offspring representing 144 seed families from the seven native and seven invasive populations for the five allozyme loci. We used the computer program ML TR (Ritland, 2002) to recon- struct allozyme phenotypes of maternal plants from their progeny arrays. Then, we used the computer program TFPGA version 1.3 (Miller, 1997) to calculate hierarchical F-statistics among the three regions (FRT) and among populations within the regions (FST)' and Nei's pairwise genetic distances between populations from the allozyme phenotypes of maternal plants. To test whether variation among populations within regions differed between the three regions, we also calculated FST for each region separately. We tested for an association between Nei's pairwise genetic distances and differences in latitude between populations with the Mantel test using the computer program Mantel (Lied 1- off, 1999).
Common environment experiment
Because the number of populations was unbalanced among the regions of collection, we analyzed the data with restricted maximum likelihood analysis of variance implemented in the statistical software GenStat (Lawes
Agricultural Trust, Rothamsted, UK; Payne et aI., 2005). The fixed model included the factors 'watering treatment' (permanently wet and temporarily wet), 'range of collection' (invasive and native) and 'region of collection' (United States, New Zealand and Scot- land), the covariate 'latitude', and their interactions. We tested the fixed factors, the covariate and their interac- tions with the Wald-test statistic (type I). We used absolute instead of signed values of latitude because we were interested in the effect of distance to the equator and not in differences between the northern and southern hemisphere. We fitted 'latitude' before 'range of collection' and 'region of collection' to correct the latter two for potential latitudinal clines. Moreover, because we fitted 'range of collection' before 'region of collection', the latter only refers to differences between the two invasive regions. The random model included the factors 'block', 'tray', 'population' and 'seed family'.
'Tray' was nested within 'block' and 'watering treat- ment', 'population' was nested within 'region of collec- tion', and 'seed family' was nested within 'population'.
We tested the significance of random factors and their interactions with 'watering treatment' from the change in deviance after removing these terms from the model.
Both the Wald-test statistic (Dobson, 1990) and the change in deviance (Littell, Milliken, Stroup, & Wolfin- ger, 1996) are approximately chi-squared distributed. To achieve normality and homoscedasticity, time to an- thesis was 10gI0-transformed, and number of upright side branches, total stolon length and number of flowers were square-root transformed prior to analyses.
Results
Allozyme variation among invasive and native regions and among populations
All five allozyme loci were polymorphic (see Appen- dix A, Supplemantary Table 2; allele frequencies, ADH: 0.036,0.818,0.136,0.011; PGI-I: 0.021, 0.736, 0.243;
PGI-2: 0.050, 0.404, 0.165, 0.362, 0.019; 6PGD-I: 0.119, 0.874, 0.007; 6PGD-2: 0.007, 0.710, 0.283). One allele of ADH, one of PGI-2 and one of 6PGD-2 were only found in the native range, however, with low frequencies
«0.030). One allele of 6PGD-1 was only found in Scotland, also with low frequency (0.039). This suggests that overall there is little allozyme differentiation between the invasive and native regions of M. guttatus. Indeed, hierarchical F-statistics showed that there was no significant allozyme differentiation among western North America, New Zealand and Scotland (FRT
=
0.020,95% CI=
-0.019 to 0.051).Nei's pairwise genetic distances between popUlations were not significantly correlated with pairwise differ-
217
ences in latitude (r = 0.059, Mantel Z = 249.84, P = 0.301), implying that invasive populations were not more related to native popUlations of similar latitude than to the ones of other latitudes.
There was significant differentiation among popUla- tions within regions (FST
=
0.228, 95%cr =
0.085-0.353) but this was mainly due to significant differentia- tion among populations in the native range (FST
=
0.275, 95% CI=
0.108-0.390), while differentiation among populations within the invasive regions was not significant (Scotland: FST = 0.054, 95% CI = -0.007- 0.135; New Zealand: FST=
0.124, 95% CI = -0.024- 0.230).Common environment experiment
Phenology
Time to anthesis did not differ significantly between plants from populations in the invasive and native range and was not affected by latitude of collection (Fig. lA, Tables I and 2). However, time to anthesis differed significantly among populations within regions of collection and seed families within populations (Table I).
Time to anthesis was slightly, though significantly, delayed by on average 0.8 days in the permanently wet treatment when compared to the temporarily wet treatment (Table I). Plasticity of seed families in response to watering treatment in time to anthesis, however, did not differ among regions of collection, was not affected by latitude of collection, and did not differ among populations within regions, and seed families within popUlations (Table I).
Plant size and reproduction
Plant height, total stolon length and number of flowers did not differ significantly between the invasive and native ranges (Tables I and 2). The number of upright side branches, however, was significantly higher
(+ 81.8%) for plants from the invasive range than from
the native range (Tables I and 2), indicating an increased size in the invasive range. Moreover, among the invasive plants, the ones from Scotland produced significantly more upright side branches and grew non- significantly (P = 0.092) higher than the ones from New Zealand (Tables I and 2).
Plant height and the number of upright side branches were not affected by the latitude of collection (Figs. I B and C, Table I). Total stolon length, however, increased significantly with latitude of collection (Fig. 1 D, Table I) while the number of flowers decreased significantly with latitude (Fig. IE, Table 1). This indicates that vegetative reproduction is more pro- nounced in plants from high latitudes while sexual
218
A
(i) 'iii
1.8
Q) 1.7
£; c:
ro .9 1.6
Q)
~ E 1.5
.9
o 1.0
~ 9 o o o
• •
•
o1.4 I---~---~
c
1.6 (i)
-5 Q) 1.2 c: !!!
.0 '0 0.8
~ o 0.4 0- W
o
o
o lor!)
o o
• •
o
0.0 I---~---~
E
~ 5.0
~ 4.0
c;:::
'0 3.0 Q;
~ 2.0 ::>
t c: 1.0 0-
W A o o
0.0 I---~-~---~
G
o 1.6
~ 1.5
~ '§E 1.4 ..C:
0, 1.3 c:
~ 1.2 .!!l
~ 1.1 o
• • o
U 1.0 ./-_~-~-_ _ _ _ -~
B
40.0
:[ 30.0
1: .~ 20.0
.r:.
C ~ 10.0
•
•
o o
0.0 +---~-~-~~
D
12.0
£; g> 10.0
~ c: o
~
8.0
I
6.0 4.05-
2.0 Wo o
0.0 ./--~-~---~
F
35.0
E
oS 30.0£;
~ 25.0 .!!l
2 20.0
8
o o AI. 1.0
Q) A &
o o
o 0'0
q)
I
o15.0 .I---~-~-~~
H
E
oSo c:
~ II Q)
'"
ro E
~
a;
6.0 5.0 o
4.0 3.0 2.0
o
• •
o
1.0 ./--~-~---~
35 40 45 50 55 60
E
65
«
35 40 45 50 55 60 65Latitude rl Latitude rJ
Fig. 1. Mean (A) time to anthesis, (B) plant height, (C) number of upright side branches, (D) total stolon length, (E) number of flowers, (F) corolla width, (G) corolla length width ratio and (H) anther stigma separation of populations of Mimulus guttautus from different latitudes (absolute values) in the native range in North America (open circles), and the invasive ranges in New Zealand (filled triangles) and Scotland (filled squares). Regression lines, based on all 24 populations, are shown for traits that are significantly affected by the latitude of collection.
reproduction is more pronounced in plants from low latitudes.
The effect of latitude on vegetative and sexual reproduction did not differ significantly between the invasive and native range (Figs. 10 and E, Table I), indicating that plants from Scotland and New Zealand do not differ from the ones at similar latitudes in the native range. All four traits related to size and
reproduction differed significantly among populations within regions of collection and seed families within populations (Table I).
Plants in the permanently wet treatment produced significantly longer stems (+ 14.8%) and stolons (+ 88.1 %), and more upright side branches (+ 29.0%), though not significantly so (P
=
0.069; Table I), than plants in the temporarily wet treatment.219 Table I. Summary of restricted maximum likelihood analyses of variance of effects of watering treatment, latitude, range of collection (native versus invasive) and region of collection (of invasive populations: Scotland versus New Zealand) on phenology, growth, reproduction and floral traits of Mimulus guttatus
Time to Plant No. of upright Total stolon No. of Corolla Corolla Anther stigma anthesis height branches length flowers width length width ratio separation Fixed model
Watering 11.69*** 9.28** 3.14 31.30*** 0.03 14.44*** 3.52 1.69
treatment
Latitude 0.16 1.23 1.40 6.84** 4.24* 0.10 2.80 0.03
Range of 0.10 0.01 4.90* 2.33 0.18 2.19 0.42 0.19
collection
Region of 1.26 2.83 4.80* 0.12 0.52 2.74 0.05 0.87
collection (Ra)
Lx Ra 2.57 0.65 1.65 1.96 3.33 0.11 0.14 0.05
WxL 1.03 1.17 2.38 1.49 6.92** 0.08 0.17 0.15
Wx Ra 0.02 0.28 0.96 0.10 0.25 2.00 0.54 0.83
Wx Re (Ra) 0.27 0.68 0.42 0.09 1.65 0.61 2.66 0.97
Wx Lx Ra 2.94 3.65 0.82 1.26 1.99 1.79 7.90** 0.46
Random model
Block 6.50* 10.06** 3.77 12.52*** 8.31 ** 0.26 0.04 0.72
Tray 3.33 117.83*** 245.64*** 66.55*** 31.86*** 0.29 4.75* 7.91 **
Population (Re, 65.05*** 27.71*** 15.60*** 33.16*** 50.41 *** 46.10*** 23.10*** 30.05***
Ra)
Seed family (P, 83.14*** 55.32*** 15.26*** 12.53*** 96.36*** 5.08* 6.55* 25.93***
Re, Ra)
Wx P (Re, Ra) 0.00 24.56*** 9.43** 27.22*** 0.53 0.00 0.00 0.32
Wx S (P, Re, 0.00 0.09 0.57 1.00 1.22 0.00 0.02 0.00
Ra)
* P<0.05; ** P<O.OI; ·**P<O.OOI.
Fixed effects were tested with Wald tests, and random effects with the change in deviance after removing the effect from the full model. Both the Wald test and the change in deviance are chi squared distributed with I df. Time to anthesis was log 10 transformed, and number of upright side branches, total stolon length and number of flowers were square root transformed prior to analyses.
Table 2. The effects of range of collection on phenology, growth, reproduction and floral traits of Mimulus guttatus Trait
Time to anthesis (days) Plant height (cm)
No. of upright side branches Total stolon length (cm) No. of flowers
Corolla width (mm) Corolla length width ratio Anther stigma separation (mm)
Native range
Western North America 40.42 ± 1.29/1.25 21.72± 1.81
O.55±O.14/0.13 23.88 ± 6.93/6.05 6.36 ± 1.48/1.33 25.17±0.76
1.28±0.02 3.05±0.20
Invasive range Combined 41.31 ± 2.17/2.06 21.13±2.37
1.00 ± 0.25/0.22 16.13 ± 6.04/5.08 6.07 ± 2.26/ I. 90 27.16±0.72
1.27±0.02 2.99±0.17
Scotland 43.84 ± 1.83/1.75 23.28±2.62
I.22±0.18/0.16 28.77±3.19/3.03 4.77± 1.87/1.56 25.23±O.35
1.31 ±0.04 3.32±0.12
New Zealand 39.51 ± 3.39/3.12 19.52±3.77
0.84 ± 0.40/0.33 9.03 ± 6.36/4.68 7.16±4.34/3.32 28.61 ±O.33
1.24±0.00 2.75±0.22 Data are means± SE for non transformed data and ± upper SE/lower SE for transformed data after back transformation. Significant differences between the native range and combined invasive range, and between the two invasive regions are indicated in bold.
Plants from the invasive and native range did not differ in plasticity in response to the watering treatment. However, the negative effect of latitude of collection on the number of flowers was stronger in the temporarily than in the permanently wet treatment (significant watering treatment-by-Iatitude
interaction in Table I), indicating that the expression of quantitative genetic variation depends on the environment. Plasticity in stem height, the number of upright side branches and total stolon length differed significantly among populations within regions of collection (Table I).
220
Floral traits
Corolla width, corolla length-width ratio and an- ther-stigma separation did not differ significantly between plants from the invasive and native range (Tables I and 2). Among invasive plants, the ones from New Zealand had larger corollas than the ones from Scotland (+ 13.4%, Table 2) but this difference was not significant (P = 0.083, Table I). Corolla length-width ratio increased with latitude of collection (Fig. I G) but this effect was not significant (P = 0.072, Table I). All three floral traits differed significantly among popula- tions within regions of collection and among seed families within populations (Table I).
Plants in the permanently wet treatment had sig- nificantly larger corollas (+ 2.8%) and had slightly less elongated corollas, though not significantly so (P = 0.06 I; Table 2), than the ones in the temporarily wet treatment. Plasticities in floral traits did not differ between the invasive and native range, were not affected by latitude of collection and did not differ among populations within regions and seed families within populations (Table I).
Discussion
Founder effects and genetic drift
Founder effects and bottlenecks are likely to play important roles during invasion, especially shortly after introduction of a species (Barrett & Husband, 1990).
Because M. guttatus is frequently used in horticulture, it is likely that by now it has been introduced several times in its invasive regions. Indeed, the low level of allozyme differentiation among the three regions suggests that already most of the genetic variation in the native range of M. guttatus has been introduced to Scotland and New Zealand by now. Also, a recent review on studies assessing molecular genetic variation in invasive and native populations of I I species by Bossdorf et al. (2005) concludes that founder effects do not seem to playa major role for most invasive species.
Quantitative genetic differentiation between the invasive and native regions
Plants of M. guttatus from the invasive and native regions did not differ in phenology, height, sexual and vegetative reproduction and floral traits. However, plants from the invasive range produced twice as many flower-bearing upright side branches as the ones from the native range. Significant differentiation between the two invasive regions in this trait indicates that the higher number of upright side branches for plants of the invasive regions were mainly accounted for by the
populations from Scotland. This suggests that selective forces differed between the two invasive regions or that the differences are mainly a consequence of founder effects or genetic drift. The latter explanation is not likely for two reasons. First, the high amount of genetic variation for both traits within the invasive regions suggests that founder effects have played no or only a minor role. Second, our allozyme analysis did not reveal genetic differentiation among the invasive and native regions, which suggests no or only a minor role for genetic drift. Alternatively, maternal environmental carry-over effects (e.g., Galloway, 1995) may be responsible for the observed difference in branching frequency. For M. guttatus, however, there is no significant correlation between the volume of a seed and branching frequency (M. van Kleunen, unpublished data). This indicates that the higher branching fre- quency of invasive plants is not due to higher maternal seed provisioning in the invasive range. Therefore, we conclude that the higher number of upright side branches of plants from the invasive regions are a consequence of selection although this may not be consistent among invasive regions.
The higher number of flower-bearing upright side branches for plants from the invasive range is in line with the EICA hypothesis (Blossey & N6tzold, 1995).
As an alternative to adaptive evolution, branching frequency could also have evolved in response to selection by horticulturalists that most likely introduced the species. This may also be valid for many of the other species that have been used to test the EICA hypothesis. While some studies found results in favor of the EICA hypothesis (e.g., Blossey & N6tzold, 1995; Joshi &
Vrieling, 2005), others did not (e.g., Maron et aI., 2004) or even found opposing results (e.g., van Kleunen &
Schmid, 2003). A limitation of most other studies is that they cannot distinguish between the effects of selection, either natural or by man, and founder effects or genetic drift. As a consequence, it still remains unclear whether the EICA hypothesis holds for some species.
Baker (1955) suggested that plants capable of autono- mous self-fertilization may be favored during establish- ment in a new region where suitable mates and pollinators may be rare. M. guttatus is capable of autonomous self- fertilization, which is likely to have assisted the species during its invasion process. However, we did not find evidence that there has been evolution of floral traits that are associated with its mating system. This suggests that M. guttatus receives effective pollinator visits in its invasive range and has not suffered from or not responded to potential pollen limitation in the initially small populations. Indeed, outcrossing rates in invasive popula- tions of M. guttatus do not differ from the ones in native populations (M. van Kleunen, unpublished data). The main pollinators in the native range are honey bees, which have been introduced there, bumble bees and syrphid flies
(lvey & Carr, 2005; Kiang, 1972; Martin, 2004), and these insects also occur in Europe and New Zealand. Interest- ingly, both honey bees and bumble bees have been introduced in New Zealand suggesting that this has facilitated invasion by M. guttatus in the latter. As a consequence of the similar pollinator fauna, selection on floral traits of M. guttatus may not have been sufficiently different between the invasive and native regions to result in genetic differentiation in floral traits.
Latitudinal clines
Latitude has only rarely been considered in compar- isons between populations from invasive and native regions. A notable exception is a study by Maron et al.
(2004), who found that plants of Hypericum per/oratum from northern latitudes both in its native range Europe and its invasive range North America had higher fecundity than the ones from lower latitudes when grown at high latitudes while the reverse was true when grown at low latitudes. Unless invasive populations have been founded by offspring from similar latitude in the native range, this indicates that post-introduction adaptive evolution to geographical gradients may be important in invasive plants, and as pronounced as in native populations.
In our study, plants from the invasive ranges in Scotland and New Zealand and from the native range in North America shared a common latitudinal cline for sexual and vegetative reproductive allocation. The absence of a correlation between genetic marker distance and differences in latitude among populations indicates that invasive populations were not more related to populations from similar latitude in the native range. Therefore, it is unlikely that the invasive populations in Scotland and New Zealand have been founded from individuals from similar latitude in North America. This implies that M. guttatus has adapted to the climatic conditions associated with the latitudes in the invasive ranges. Therefore, we conclude that adaptive evolution may increase invasiveness even if it does not result in genetic differentiation between the invasive and native range.
Phenotypic plasticity
Environmental tolerance through high levels of phenotypic plasticity has often been suggested as a plant characteristic favoring invasiveness (Baker, 1974;
Williams et aI., 1995). Although several studies have included experimental treatments such as competition (Bossdorf et aI., 2005), herbivory (Buschmann, Edwards, & Dietz, 2005) and leaf area removal (van Kleunen & Schmid, 2003) in common garden experi- ments comparing invasive and native populations, only
221
few studies have explicitly focused on whether pheno- typic plasticity is higher in plants from the introduced range than in the ones from the native range (DeWalt, Denslow, & Hamrick, 2004; Kaufman & Smouse, 2001). We found significant plastic increases of time to anthesis, plant height, total stolon length and corolla width in response to permanency of the wet soil condition. Similarly, plants from permanently wet habitats had a longer time to anthesis, grew more and longer stolons and had larger flowers than the ones from temporarily wet habitats (van Kleunen, 2007), which suggests that most of the observed plastic responses are adaptive. However, there were no differences in plastic responses between plants from the native and intro- duced regions. This indicates that there has either been no selection or no response to selection for extremely plastic individuals in the introduced regions. Similarly, DeWalt et al. (2004) did not find differences in plasticity in growth rate, morphology and photosynthesis in response to shading for Clidemia hirta. On the other hand, Kaufman and Smouse (2001) found higher plasticity in leaf size and shape, and growth of Melaleuca quinquenervia in response to watering condi- tions and pH for populations from the invasive range in Florida than the ones from the native range in Australia.
However, more studies are required before general conclusions can be drawn on selection for plasticity during invasion (Richards et aI., 2006).
Conclusions
The number of studies comparing plants from the native and introduced ranges of invasive species has steadily accumulated over the last 10 years (Bossdorf et aI., 2005). Nevertheless, there are still many questions with regard to the importance of evolutionary processes in invasive plants (Facon et aI., 2006; Strayer et aI., 2006), as well as in native species exposed to invasive ones (Lau, 2006; Strayer et aI., 2006). Most studies have been restricted to tests of the EICA hypothesis, while tests of evolution of other life-history traits have been neglected. Moreover, most of these studies cannot distinguish between the role of adaptive and non- adaptive evolutionary processes. We therefore advocate approaches that enable distinction between adaptive and non-adaptive evolutionary processes in studies compar- ing invasive and native populations of a species. Overall, our results suggest that in M. guttatus adaptive evolution took place in the invasive range, while genetic drift and founder effects played a small, if any, role.
Acknowledgements
We thank Allison Butlen, Annie Truscot, Elizabeth Parnis, Jessica Ruvinsky, Linda Jennings, Marilyn
222
Barker, Brian Rance, Chris Woolmore, Lawrence Jane- way, Lowel Ahart, Nick Page, Nishanta Rajakaruna, Tom Belton and Tony Labanca for collecting seeds, and Anna Wojciechowska, Dorit Raudnitschka, Ines Schneider, Lena Blischke, Vanessa Pasqualetto and Stefan Dietrich for practical assistance. M.v.K. was supported by the Swiss National Science Foundation.
Appendix A . . Supplementary material
Supplementary data associated with this article can be found in the online version at doi:lO.1016/j.baae.
2007.03.006.
References
Amsellem, L., Noyer, J. L., Le Bourgeois, T., & Hossaert McKey, M. (2000). Comparison of genetic diversity of the invasive weed Rubus alceifolius Poir. (Rosaceae) in its native range and in areas of introduction, using amplified fragment length polymorphism (AFLP) markers. Molecular Ecology, 9, 443 455.
Baker, H. G. (1955). Self compatibility and establishment after 'long distance' dispersal. Evolution, 9, 347 349.
Baker, H. G. (1974). The evolution of weeds. Annual Review of Ecology and Systematics, 7, I 24.
Barrett, S. C. H., & Husband, B. C. (1990). The genetics of plant migration and colonization. In A. H. D. Brown, M.
T. Clegg, A. L. Kahler, & B. S. Weir (Eds.), Plant population genetics, breeding, and genetic resources (pp.
254 277). Sunderland: Sinauer Associates Inc.
Blair, A. c., & Wolfe, L. M. (2004). The evolution of an invasive plant: An experimental study with Silene latifolia.
Ecology, 85, 3035 3042.
Blossey, B., & Notzold, R. (1995). Evolution of increased competitive ability in invasive nonindigenous plants: A hypothesis. Journal of Ecology, 83, 887 889.
Bossdorf, 0., Auge, H., Lafuma, L., Rogers, W. E., Siemann,
E., & Prati, D. (2005). Phenotypic and genetic differentia
tion between native and introduced plant populations.
Oecologia, 144, I II.
Buschmann, H., Edwards, P. J., & Dietz, H. (2005). Variation in growth pattern and response to slug damage among native and invasive provenances of four perennial Brassi caceae species. Joumal of Ecology, 93, 322 334.
DeWalt, S. J., Denslow, J. S., & Hamrick, J. L. (2004).
Biomass allocation, growth, and photosynthesis of geno types from native and introduced ranges of the tropical shrub Clidermia hirta. Oecologia, 138, 521 531.
Dobson, A. J. (1990). An introduction to generalized lineal' models. London: Chapman & Hall.
Facon, B., Genton, B. J., Shykoff, J., Jarne, P., Estoup, A., &
David, P. (2006). A general eco evolutionary framework for understanding bioinvasions. Trends in Ecology and Evolution, 21, 130 135.
Fenster, C. B., & Ritland, K. (1994). Evidence for natural selection on mating system in Mimulus (Scrophulariaceae).
International Journal of Plant Science, 155, 588 596.
Galloway, L. F. (1995). Response to natural environmental heterogeneity: Maternal effects and selection on life history characters and plasticities in Mimulus guttatus. Evolution, 49, 1095 1107.
Grant, A. L. (1924). A monograph of the genus Mimulus.
Annals of the Missouri Botanical Gardens, 11, 99 193.
Ivey, C. T., & Carr, D. E. (2005). Effects of herbivory and inbreeding on the pollinators and mating system of Mimulus guttatus (Phrymaceae). American Journal of Botany, 92,1641 1649.
Joshi, J., & Vrieling, K. (2005). The enemy release and EICA hypothesis revisited: Incorporating the fundamental differ ence between specialist and generalist herbivores. Ecology Letters, 8, 704 714.
Kaufman, S. R., & Smouse, P. E. (2001). Comparing indigenous and introduced populations of Melaleuca quinquenervia (Cav.) Blake: Response of seedlings to water and pH levels. Oec%gia, 127, 487 494.
Kiang, Y. T. (1972). Pollination study in a natural population of Mimulus guttatus. Evolution, 26, 308 310.
Kollmann, J., & Banuelos, M. J. (2004). Latitudinal trends in growth and phenology of the invasive alien plant Impatiens glandulifera (Balsaminaceae). Diversity and Distributions, 10, 377 385.
Lande, R. (1992). Neutral theory of quantitative genetic variance in an island model with local extinction and colonization. Evolution, 46, 381 389.
Lau, J. A. (2006). Evolutionary responses of native plants to novel community members. Evolution, 60, 56 63.
Leclerc Potvin, c., & Ritland, K. (1994). Modes of self fertilization in Mimulus guttatus (Scrophulariaceae): A field experiment. American Journal of Botany, 81, 199 205.
Lee, C. E. (2002). Evolutionary genetics of invasive species.
Trends in Ecology and Evolution, 17, 386 391.
Liedloff, A. (1999). Mantel nonparametric test calculator, version 2.00. Brisbane: Queensland University of Technology.
Littell, R. c., Milliken, G. A., Stroup, W. W., & Wolfinger, R.
D. (1996). SAS ~ystem for mixed models. Cary: SAS Institute.
Lynch, M. (1994). Neutral models of phenotypic evolution. In L. A. Real (Ed.), Ecological genetics (pp. 86 108).
Princeton, NJ: Princeton University Press.
Maron, J. L., Vila, M., Bommarco, R., Elmendorf, S., &
Beardsley, P. (2004). Rapid evolution of an invasive plant.
Ecological Monographs, 74,261 280.
Martin, N. H. (2004). Flower size preferences of the honeybee (Apis mellifera) foraging on Mimulus guttatus (Scrophular iaceae). Evolutionary Ecology Research, 6, 777 782.
Miller, M. P. (1997). Tools for population genetic analyses (TFPGA) 1.3: A Windows program for the analysis of allozyme and molecular population genetic data. Computer software distributed by author.
MUlier Scharer, H., Schaffner, U., & Steinger, T. (2004).
Evolution in invasive plants: Implications for biological control. Trends in Ecology and Evolution, 19, 277 296.
Neuffer, B., & Hurka, H. (1986). Variation of development time until flowering in natural populations of Capse/la bursa pastoris (Cruciferae). Plant Systematics and Evolu tion, 152, 277 296.
Olsson, K., & Agren, J. (2002). Latitudinal population differentiation in phenology, life history and flower morphology in the perennial herb Lythrum salicaria.
lournal of Evolutionary Bioloyy, 15, 983 996.
Payne, R. W., Harding, S. A., Murray, D. A., Soutar, D. M., Baird, D. B., Welham, S. J., et a!. (2005). The yuide to GenStat release 8. Part 2: Statistics. Oxford: VSN Interna tiona!.
Richards, C. L., Bossdorf, 0., Muth, N. Z., Gurevitch, J., &
Pigliucci, M. (2006). Jack of all trades, master of some? On the role of phenotypic plasticity in plant invasions. Ecoloyy Letters, 9, 981 993.
Ritland, K. (2002). Extensions of models for the estimation of mating systems using n independent loci. Heredity, 88, 221 228.
Ritland, K., & Ganders, F. R. (1987). Covariation of selfing rates with parental gene fixation indices within populations of Mimulus yultatus. Evolution, 41, 760 771.
Strayer, D. L., Eviner, V. T., Jeschke, J. M., & Pace, M. L.
(2006). Understanding the long term effects of
223
species invasions. Trends in Ecoloyy and Evolution, 21, 645 651.
van Kleunen, M. (2007). Adaptive genetic differentiation in life history traits between populations of Mimulus yultatus with annual and perennial life cycles. Evolutionary Eco!oyy, doi:10.1007/sI0682 006 00197.
van Kleunen, M., & Johnson, S. D. (2005). Testing for ecological and genetic Allee effects in the invasive shrub Senna didymobotrya (Fabaceae). American Journa! of Botany, 92, 1124 1130.
van Kleunen, M., & Ritland, K. (2004). Predicting evolution of floral traits associated with mating system in a natural plant population. lournal of Evolutionary Bio!oyy, 17, 1389 1399.
van Kleunen, M., & Schmid, B. (2003). No evidence for an evolutionary increased competitive ability in an invasive plant. Ecoloyy, 84, 2816 2823.
Weber, E., & Schmid, B. (1998). Latitudinal population differentiation in two species of Solidayo (Asteraceae) introduced into Europe. American Journal of Botany, 85,
1110 1121. .
Williams, D. G., Mack, R. N., & Black, R. A. (1995).
Ecophysiology of introduced Pennisetum setaceum on Hawaii: The role of phenotypic plasticity. Eco!oyy, 76, 1569 1580.
