Macro-environmental modelling of the current distribution of Undaria pinnatifida (Laminariales, Ochrophyta) in northern Iberia

Macro-environmental modelling of the current distribution of Undaria pinnatifida

(Laminariales, Ochrophyta) in northern Iberia

--------------------------------

José C. Báez1, Jesús Olivero2, César Peteiro3, Francisco Ferri-Yáñez4, Carlos García-Soto3 &

Raimundo Real2 1Instituto Español de Oceanografía (IEO), Centro Oceanográfico de Málaga, 29640

Fuengirola, Malaga, Spain 2Universidad de Málaga (UMA), Facultad de Ciencias, 29071 Malaga, Spain 3Instituto Español de Oceanografía (IEO), Centro Oceanográfico de Santander, 39004

Santander, Spain 4International Institute for Geo-information Science and Earth Observation (ITC), 7500 AA

Enschede, The Netherlands

Manuscript published

Biological Invasions (2010), 12(7): 2131–2139. DOI: 10.1007/s10530-009-9614-1

SUMMARY

The Asian kelp Undaria pinnatifida was first reported in north-western Spain in 1988.

Since then, it has spread along the Galician coasts and towards the western Cantrabrian coast,

probably supported by new introductions related to aquaculture and private yachts. Despite

the high potential of this seaweed to invade new areas, it has not fully established itself in

specific areas along the eastern Cantabrian coast where it has been farmed. We attempted to

identify the macro-environmental determinants for the successful establishment of U.

pinnatifida in the northern Iberian coast. Thus, we built a model based on the significant

relationships between this kelp and several environmental factors using the Favourability

Function, a method based on binary logistic regression. We obtained a statistically significant

Favourability model where the chlorophyll concentration in summer was the most influential

variable and the photosynthetically active radiation in spring was a secondary variable that

best characterized the presence of U. pinnatifida. The ecological implications are discussed.

INTRODUCTION

There is an increasing worldwide interest in the study of foreign invasive species, given

their effects on native ecosystem structures and the way they modify global biodiversity

(López-Darías et al. 2007). Knowledge concerning ecological traits and interactions between

native and invasive species is therefore crucial in predicting and preventing biological

invasions (e.g. see Parker et al. 2006 and references therein). There is also interest in

predicting which regions or habitats are the most susceptible to be colonised by alien species,

as well as which factors favour increased invasiveness (e.g. see Heger &Trelp 2003). For this

purpose, analyses of foreign species distributions in their new ranges are useful, as they aid in

understanding the adaptive capability of these species in the occupied areas.

Biogeographic modelling studies are becoming fundamental tools for the rational

management of fauna and flora and for environmental conservation. Spatial models help

making inferences on the expected distribution of species as a function of the environmental,

biotic, spatial and human characteristics of the territory under study (Burnham &Anderson

2002). Peterson & Vieglais (2001) recommended modelling the ecological niche of an

invasive species to predict its future distribution. However, this aspect of biogeography has

scarcely been studied in relation to alien algae.

Undaria pinnatifida (Harvey) Suringar (Laminariales, Ochrophyta) is native to eastern

Asia where it is extensively cultivated for human consumption. This seaweed has also

undergone anthropogenic global spread due to aquaculture and maritime traffic (e.g. Floc’h et

al. 1991). As a result of this, U. pinnatifida has successfully invaded coastal areas around the

world, and is considered one of the three most invasive seaweed species in the European

Atlantic coasts (ICES 2007).

This alga was first found on the Atlantic coast of Spain in 1988 in a mussel farm in Ría de

Arousa (Galicia, N. W. Spain) (Santiago-Caamaño et al.1990). The presence of U. pinnatifida

in Galicia has been recently reviewed by Cremades et al. (2006). At present, this alien

seaweed is established along the west coast of Galicia, aquaculture appearing to be the most

likely vector for its introduction. This species was also reported in 1995 in Gijon harbour

(Asturias, N. Spain), the most likely vectors being private yachts (Salinas et al. 1996). Since

then, only one new additional citation has been noted in the Cantabrian Sea, in Cudillero

harbour, Asturias (Pérez-Ruzafa et al. 2002), even though U. pinnatifida has been cultivated

in Santander (Cantabrian Sea) since the early 2000s (Peteiro 2008).

Undaria pinnatifida has undergone selective dispersal since its arrival, such that it is now

occurring in sites where it was not introduced by human activity, but is absent from sites

where it has been cultivated in the past. The prediction of the future course of invasions has

been a major aim of ecological research ever since biological invasions have come into focus

(Heger & Trelp 2003). For this reason, it could be interesting modelling the current

distribution of this seaweed in northern Spain, about 20 years since its introduction, to

determine favourable areas for the species and to forecast its future course according to its

potential range. Although it is still in dispute whether it is feasible or not to produce accurate

predictions (Daehler & Carino 2000) for alien species, as they could never be in equilibrium

with its new environment, some authors have previously modelled the potential expansion

range of invasive species with a more than reasonable output. For example, Muñoz &Real

(2006) determined favourable areas for the monk parakeet Myiopsitta monachus, in Spanish

Peninsula to account for its current distribution and forecast its future course according to its

potential range.

For a foreign seaweed with a slow natural expansion rate, as it is the case of U. pinnatifida

in Iberia (Cremades et al. 2006), the optimization of its reproductive capacity would be

decisive in becoming established in new areas. In this context, we examined the most

important macro-environmental factors that determine the successful establishment of U.

pinnatifida, and tried to explain why there is no spreading of this alien species along the

eastern coast of the Cantabrian Sea. The initial assumptions are that, although the arrival of U.

pinnatifida seems to have been due to human factors (aquaculture and/or maritime traffic), the

ability of this seaweed to spread after its introduction does not only depend on human factors,

but also on the attributes of the algae and on the characteristics of the habitat being invaded

(Larson et al. 2001).

MATERIALS AND METHODS

Undaria pinnatifida, like all kelps, has a biphasic life cycle, with microscopic

gametophytes and macroscopic sporophytes. This study is based on all records currently

available for U. pinnatifida in northern Iberia, based on observations of individuals in the

sporophytic stage. Sporophytes are usually longer than 1 m in the northern Iberian coast, and

are easily distinguishable from any other native laminariales species in the study area.

The current coastal distribution of U. pinnatifida in northern Iberian Peninsula is shown in

Figure 1. The records were compiled from the work of Santiago-Caamaño et al. (1990),

Salinas et al. (1996), Pérez-Ruzafa et al. (2002), Cremades et al. (2006) and Peteiro (2008).

For the rest of the study area, U. pinnatifida was considered reliably absent according to

exhaustive phycological prospecting by Bárbara et al. (2006), Cremades et al. (2006) and

Martínez-Gil et al. (2007) in the western Cantabrian coast, and by Gorostiaga et al. (2004)

and our team in the eastern Cantabrian coast. We searched for U. pinnatifida in a non-

prospected coastal stretch between Gijon (where the last record was taken in 2002) and the

Basque region, with 26 winter visits in January, February, and March 2007–2008. Twelve of

those visits were made around Santander, where U. pinnatifida was cultivated in the early

2000s. We did not find any trace of sporophyte presences in the intertidal (3.5 m) and upper

subtidal (5 m) zones, where the most characteristic habitats for this species are found

(Cremades et al. 2006, ICES 2007).

Availability of energy, climatic stress, chemical potential for primary productivity, and

anthropogenic factors were considered to build a distribution model, each of which were

represented by a set of macroenvironmental variables (see Table 1):

Availability of energy. Some species and/or communities may need high values of

environmental energy to satisfy their physiological requirements and to maintain their

competitive capacity (Hutchinson 1959). Indicators of the influence of this factor include

positive associations between the presence of U. pinnatifida and mean annual solar radiation,

Photosynthetically Active Radiation (PAR) (average for each annual season) and seawater

surface temperatures in January and August (key months for the development of new

gametophytes and sporophytes, respectively, in the study area; see Cremades et al. 2006,

ICES 2007).

Climatic stress. A warm climate may cause physiological stress to some species that may

limit its development and reproduction (Lünning 1990). The influence of this factor could be

deduced by a negative relationship between the presence of U. pinnatifida and seawater

surface temperatures average for each annual season.

Chemical potential for primary productivity. Coastal primary productivity levels are closely

correlated with nitrate and phosphate concentrations in seawater (Lobban & Harrison 2000),

which, according to Mann & Lazier (1991) is correlated with water run-off. For this reason,

we searched for positive associations between the presence of U. pinnatifida and mean annual

run-off, the annual precipitation variation coefficient in the adjacent river basins, Chlorophyll

concentration, and Carbon based productivity. The two latter were computed for each annual

season.

Human factor. According to Larson et al. (2001) and Keeley et al. (2005), anthropogenic

disturbance favours the spread of alien species. Human concentrations next to the coast

usually imply effluents that may cause water eutrophication, and on the other hand, port cities

may also be related to alien species introductions. The distribution of harbours in the northern

Iberian Peninsula is roughly homogeneous, but commercial ports and the most important

yachting harbours are mainly located next the largest cities. As an indicator of the influence of

this factor, we used a negative relationship between the presence of U. pinnatifida and

distance to the nearest cities with more than 100,000 inhabitants and to those with more than

500,000 inhabitants.

To analyze the geographical distribution of U. pinnatifida in northern Iberia, the coastline

was divided into the coastal stretches of the 55 Galician-Cantabrian river basins. The presence

or absence of U. pinnatifida thus referred to these geographical units in the distribution

model. River basins strongly affect the exchange of water, sediments, energy and nutrients

among wetlands (Real et al. 1993), and this has a great influence on nutrient enrichment in

coastal waters near river mouths (Mann & Lazier 1991), thus affecting marine species

distribution along or near the coast. The values considered for the above variables, except for

seawater surface temperatures, are their average values at a 1 9 1 km resolution in every river

basin.

Marine variables (Carbon based productivity, Chlorophyll concentration, PAR, seawater

surface temperatures average for each annual season) were obtained from Moderate

Resolution Imaging Spectroradiometer (MODIS), provided by NASA. All marine variables

are average of 7 years (2001–2007 depending on data availability).

We used the Favourability Function (Real et al. 2006) to model the distribution of U.

pinnatifida according to the macro-environmental variables. The Favourability Function has

the form:

F = exp (y') / [1 + exp (y')],

where y' = α + β1x1 + β2x2 +• • •+βnxn – ln(n1 / n0), α is a constant, β1, β2, . . . , βn are the

coefficients of the n predictor variables x1, x2, . . . , xn, and n1 and n0 are, respectively, the

number of presences and absences of the species in the geographical unit. The difference

between Favourability (F) and probability (P) derived from the logistic binary regression is

the inclusion of the monomial –ln(n1/n0) in the former, by which geographical predictions for

the species are made independently of the initial presence/absence rate. Favourability was

computed by logistic binary regression, as it can be related to probability as follows:

F = [P / (1 – P)] / [(n1 / n0) + (P / [1 – P])].

The Favourability Function has been successfully used to model invasive species

distributions (e.g. Muñoz & Real 2006, Real et al. 2008) and logistic binary regressions have

proven useful to characterize the distributions of marine algae distributions in environmental

terms (e.g. Báez et al. 2004, 2005). Forward–backward stepwise logistic regression with the

set of predictor variables was performed. As a result, only a selection of these variables

appears in the final model, which is conditioned by the degree to which the inclusion or

deletion of a single variable significantly improves the model in every step. To control for the

increase in type I errors due to multiple tests (Benjamini & Hochberg 1995, García 2003), we

only accepted those variables that were significant under a False Discovery Rate (FDR) of q<

0.05, using the Benjamini & Hochberg (1995) procedure. We then applied the favourability

function on the subset of significant predictive variables. The relative importance of each

variable within the model was quantified using the Wald parameter, which follows a Chi-

square distribution (Zuur et al. 2007). We assessed the model's goodness-of-fit using the

Hosmer & Lemeshow test (Hosmer & Lemeshow 2000), and the discrimination capacity of

the model was evaluated with the receiving operating characteristic (ROC) curve. The area

under the ROC curve (AUC, see Fielding & Bell 1997, Lobo et al. 2008) provides a single-

number discrimination measure across all possible thresholds, thus avoiding subjectivity in

the threshold selection process, an AUC value higher than 0.9 is considered outstanding

(Hosmer & Lemeshow 2000, p. 162).

The favourability value predicted by the resulting model was geographically represented

by downscaling the equation to a 1 × 1 km grid. Values of the environmental variables

included in the model at a 1 × 1 km resolution were used for this.

Pair-wise Pearson correlation between the most important variables for Undaria

pinnatifida presence based on Akaike Information Criteria (AIC) of bivariate Logistic

regression models was made to visualize and evaluate the relation between them. AIC is a

measure of goodness of fit and can be calculated for each possible combination of explanatory

variables (Akaike 1973, Zuur et al. 2007). AIC is defined by:

AIC = n log(SS residual) + 2(p + 1) – n log(n)

RESULTS

Undaria pinnatifida was reported in 23 of the 55 Galician-Cantabrian river basins studied.

Chlorophyll concentration, PAR and average sea surface temperature (mostly for spring

and summer) have a close correlation between them (Table 2).

We obtained a statistically significant Favourability model where the chlorophyll

concentration in summer was the most influential variable (Wald = 5.362) and the

photosynthetically active radiation in spring was a secondary variable (Wald = 3.336). The

model goodness-of-fit was significant (Hosmer and Lemeshow's chi-square = 1.37, P =

0.986), and its discrimination capacity was outstanding (AUC = 0.987). The geographic

favourability, downscaled to 1 9 1 km resolution, is shown in Figure 2.

DISCUSSION

According to the favourability map presented in Figure 2, only the western coasts facing

the Atlantic Ocean, and those around Gijon facing the Cantabrian Sea, are highly favourable

areas (> 0.8) for U. pinnatifida. These extend to northern Portugal, which agrees with a recent

published record from Portugal by Araujo et al. (2009). Distance from the introduction sites

still has some capacity to explain the current distribution of this seaweed, as there is a

significant negative correlation between the number of records for this species in Galicia and

the distance to Ría de Arousa (Galicia, N.W. Spain), where this seaweed was first observed in

1988.

Chemical potential for primary productivity associated with chlorophyll concentration in

summer and with energy availability are involved in the characterization of the current

distribution of U. pinnatifida in northern Iberian coasts. We observed a longitudinal gradient

in which chlorophyll concentration in summer decreases westward along the coast where

most observations of this seaweed have been recorded. PAR decreases latitudinally towards

the north. Thus, this variable could explain the differences found along the Galician Coast.

Nevertheless, it is difficult to relate PAR directly with the biology of U. pinnatifida as this

seaweed presents low saturated light requirements, and low Irradiance-saturated

photosynthesis (Campbell et al. 1999). We interpret these results in a macro-ecologic context;

in agreement with many authors (Hutchinson 1959, Wright 1983, Davis et al. 2000), the

number of species in a given territory is positively correlated with the availability of energy

and resources. There are more large seaweeds, such as kelps, in ecosystems, where energy

availability is higher. These communities could create suitable habitats for U. pinnatifida, for

instance, by shadowing the habitat. Moreover, these results could explain why this species has

not colonised the highly oligotrophic Mediterranean sea from Thau lagoon, about 20 km

southwest of Montpelier, France, where it occurs since 1971; and also why U. pinnatifida is

found growing close to where urban waste water is discharged and thriving in nutrient-rich

polluted water in the Atlantic coasts (Curiel et al. 1998, Castric-Fey et al. 1999a, b, Cecere et

al. 2000). In the eastern coasts of the study area, where macro-ecological conditions seem to

be highly unfavourable for U. pinnatifida (see Figure 2), the model representation suggests

that human factors could be affecting its current distribution and, probably, also its future

spread.

Three isolated areas show intermediate environmental favourability values, that is, between

0.2 and 0.8, surrounded by a large low favourability territory. These are three river mouths

that are close to provincial capitals: Oviedo, Santander and Bilbao. This is in agreement with

previous cases as in Tasmania, where high rate of human eutrophization could facilitate the

establishment of U. pinnatifida (Valentine & Johnson 2003, Valentine et al. 2007). This kelp

has been found in waters surrounding Oviedo (in Gijón, Lastres and Cudillero, see Figure 1),

but the absence of U. pinnatifida in Santander, despite it has been cultivated in that city,

suggests that the macro-ecological conditions there are not suitable for the kelp (the

environmental favourability is only between 0.2 and 0.4). On the contrary, U. pinnatifida

could settle successfully next to Bilbao, where predicted favourability values are between 0.6

and 0.8. Despite this, the current absence of U. pinnatifida in Santander could be related to

Allee effects, which are decreases in the population growth rate caused by decreases in

population abundance (Courchamp et al. 1999). The probability of new species becoming

established generally appears to be related to the size of the initial populations, and founder

populations, which are typically small, are at great risk of extinction due to stochasticity and

Allee effects (Tobin et al. 2007). For benthic organisms with low mobility, the concurrence of

depensatory mortality and the Allee effects has been suggested (Ramírez-Félix & Manzo-

Monroy 2004). Thus, the lack of observations of U. pinnatifida settlements around Santander

may be a result of a low introduction rate combined with an low favourable environment.

Some authors have proposed the existence of a qualitative and quantitative biotic boundary

for seaweeds somewhere along the Cantabrian coast (Bay of Biscay), and differences in

seawater surface temperature have been postulated as one of the main causes of this

(Sauvageau 1897, Feldmann & Lami 1941, van den Hoek & Donze 1966, Gorostiaga et al.

2004, Bárbara et al. 2005). The warming of waters during summer in the Bay of Biscay is

considered responsible for the absence or very rare occurrence of cold temperature species,

such as large fucoids and kelps, which are common in Galicia and Britain (van den Hoek &

Donze 1966, Gorostiaga et al. 2004). The species composition is more adapted to warmer

waters in the east, which is agreement with Cheney's ratio (index for comparing seaweed

flora, expressed as (R + C)/O, where R: Rhodophyta, C: Chlorophyta and O: Ochorophyta;

see Cheney 1977), whose values are higher in the east (4.09) than in Galicia (2.94).

Temperature has been considered the most important factor in the broad scale distribution

of seaweeds (Lünning 1990). However, U. pinnatifida is able to survive in a wide range of

temperatures. Summer temperature is closely correlated with chlorophyll concentration,

because in summer there is an upwelling of cold eutrophic waters in the Biscay Bay which

decreases in importance towards the East along the coasts of Galicia and the Cantabrian Sea.

There is a close agreement between upwelling and chlorophyll concentration (Gil 2008).

Thus, differential east-west seaweed composition, and the current distribution of U.

pinnatifida in northern Iberia could be related to the chemical potential for primary

productivity and the energy availability, which are factors that are probably more involved in

the marine boundary than the surface seawater temperature.

In the northern coast of Spain, there are two sporophyte recruitment periods: late summer

and early winter. Macroscopic recruits appear in the winter months, mature during the spring,

and senesce during summer as water temperature increases (see review Cremades et al. 2006,

ICES 2007). Mature sporophytes release motile zoospores, which settle and develop into

microscopic gametophytes. The gametophyte stage of this species persists over the summer,

and can survive long periods while waiting for optimal conditions in order to mature.

Gametophyte development occurs once water temperatures drop below 20 ºC, with an

optimum range between 10 and 15 ºC (Morita et al. 2003a). Mature gametophytes undergo

sexual reproduction to produce embryonic sporophytes that subsequently develop into

macroscopic individuals. It has been reported that microscopic gametophytes can survive a

temperature ranging from -1 to 30 ºC (Saito 1975). The optimum growing temperatures for

the sporophyte fall from 10–15 to 17–20 ºC (Akiyama 1965, Saito 1975) and the total

acceptable range for sporophyte development is between 4–5 and 25 ºC (Akiyama & Kurogi

1982, Morita et al. 2003b). In August, the sea surface temperature gradient along the northern

Iberian coast falls from about 17 ºC in the west to 22 ºC in the east. These temperatures are

within the acceptable range for U. pinnatifida, but it is also true that the eastern half of the

study area shows temperatures outside the optimum range. Therefore, although temperatures

cannot, in principle, and by itself explain the absence of colonisation in the east of the

Cantabrian coast, we still consider that it is an important factor contributing to its presence. In

fact, Castric-Fey et al. (1999b) described the temperature range for possible sporophyte

recruitment in France as 5–20 ºC, recruitment peaks occurring at 13–17 ºC, which is totally

outside the sea surface temperature range in the eastern Cantabrian coast.

ACKNOWLEDGEMENTS

We are grateful to Ignacio Bárbara and Óscar Freire for providing us with access to the

Undaria pinnatifida data stored in the SANT-algae Herbarium (Index Herbariorum code).

This work has been partially financed by project CGL2006-09567/BOS (Ministerio de

Educación y Ciencia, Spain, and FEDER funds). We also thank to Dr. Daniel Simberloff and

one anonymous referee their comments.

REFERENCES

Akaike H. (1973). Information theory and an extension of the maximum likelihood principle.

In: Proceedings of the second international symposium on information theory. (Petrov

B.N. & Csaki F., eds.) pp. 267–281. Akademiai Kiado, Budapest, Hungary.

Akiyama K. (1965). Studies of ecology and culture of Undaria pinnatifida (Harvey) Sur. II.

Environmental factors affecting the growth and maduration of gametophyte. Bull. Tohoku

Reg. Fish. Res. Lab. (25): 143–170.

Akiyama K. & Kurogi M. (1982). Cultivation of Undaria pinnatifida (Harvey) Suringar, the

decrease in crops from natural plants following crops increase from cultivation. Bull.

Tohoku Reg. Fish. Res. Lab. (44): 91–100.

Álvarez C., Coma J.E., Lucena C., Porras J., San José M.A. & Quang N. (1979). Mapa

hidrogeológico Nacional. Explicación de los mapas de lluvia útil, de reconocimiento

hidrogeológico y de síntesis de los sistemas acuíferos. Instituto Geológico y Minero de

España, Madrid, Spain.

Araujo R., Barbara I., Tibaldo M., Berecibar E., Tapia P.D., Pereira R., Santos R. &

Pinto I.S. (2009). Checklist of benthic marine algae and cyanobacteria of northern

Portugal. Bot. Mar. 52(1): 24–46.

Báez J.C., Real R., Vargas J.M. & Flores-Moya A. (2004). A biogeographical analysis of

the genera Audouinella (Rhodophyta), Cystoseira (Phaeophyceae), and Cladophora

(Chlorophyta) in the western Mediterranean Sea and Adriatic Sea. Phycologia 43(4): 404–

415.

Báez J.C., Real R., Vargas J.M. & Flores-Moya A. (2005). Chorotypes of seaweeds from

the western Mediterranean Sea and the Adriatic Sea: An analysis based on the genera

Audouinella (Rhodophyta), Cystoseira (Phaeophyceae) and Cladophora (Chlorophyta).

Phycol. Res. 53: 255–265.

Bárbara I., Cremades J., Calvo S., López-Rodríguez M.C. & Dosil J. (2005). Checklist of

the benthic marine and brackish Galician algae (NW Spain). Anales Jard. Bot. Madrid

62(1): 69–100.

Bárbara I., Díaz P., Araújo R., Peña V., Berecibar E., Cremades J., Freire Ó.,

Baamonde S., Novo T., Calvo S., López Rodríguez M.C., Afonso-Carrillo J., De

Clerck O., Santos R., Sousa-Pinto I., Tibaldo M., Lagos V., López C., Sencilla A.,

Santolaria A., Díez I. & Veiga A.J. (2006). Adiciones corologicas y correcciones a la

flora bentónica marina del norte de la Península Ibérica. N.A.C.C. Biol. 15: 77–88.

Benjamini Y. & Hochberg Y. (1995). Controlling the false discovery rate: a practical and

powerful approach to multiple testing. J. R. Stat. Soc. Ser. B (Methodol.) 57(1): 289–300.

Burnham K.P. & Anderson D.R. (2002). Model selection and multimodel inference: a

practical information-theoretic approach. Springer-Verlag, New York, United Stated.

Campbell S.J., Bité J.S. & Burridge T.R. (1999). Seasonal patterns in the photosynthetic

capacity, tissue pigment and nutrient content of different developmental stages of Undaria

pinnatifida (Phaeophyta: Laminariales) in Port Phillip Bay, South-Eastern Australia. Bot.

Mar. 42(3): 231–243.

Castric-Fey A., Beaupoil C., Bouchain J., Pradier E. & L'Hardy-Halos M.T. (1999a).

The introduced alga Undaria pinnatifida (Laminariales, Alariaceae) in the rocky shore

ecosystem of the St. Malo area: morphology and growth of the sporophyte. Bot. Mar. 42:

71–83.

Castric-Fey A., Beaupoil C., Bouchain J., Pradier E. & L`Hardy-Halos M.T. (1999c).

The introduced alga Undaria pinnatifida (Laminariales, Alariaceae) in the rocky shore

ecosystem of the St. Malo area: growth rate and longevity of the sporophyte. Bot. Mar. 42:

83–96.

Cecere E., Petrocelli A. & Saracino O.D. (2000). Undaria pinnatifida (Fucophyceae,

Laminariales) spread in the central Mediterraneae: its occurrence in the Mar Piccolo of

Taranto (Ionian Sea, southern Italy). Cryptogamie, Algol. 21(3): 305–309.

Cheney D.F. (1977). R+C/P, a new and improved ratio for comparing seaweed floras. J.

Phycol. 13(Suppl.): 1–12.

Courchamp F., Clutton-Brock T. & Grenfell B. (1999). Inverse density dependence and

the Allee effect. Trends Ecol. Evol. 14(10): 405–410.

Cremades J., Freire Ó. & Peteiro C. (2006). Biología, distribución e integración del alga

alóctona Undaria pinnatifida (Laminariales, Phaeophyta) en las comunidades bentónicas

de las costas de Galicia (NW de la Península Ibérica). Anales Jard. Bot. Madrid 63(2):

169–187.

Curiel D., Bellemo G., Marzocchi M., Scattolin M. & Parisi G. (1998). Distribution of

introduced Japanese macroalgae Undaria pinnatifida, Sargassum muticum (Phaeophyta)

and Antithamnion pectinatum (Rhodophyta) in the Lagoon of Venice. Hydrobiologia 385:

17–22.

Daehler C. & Carino D. (2000). Predicting invasive plants: prospects for a general screening

system based on current regional models. Biol. Invasions 2(2): 93–102.

Davis M.A., Grime J.P. & Thompson K. (2000). Fluctuating resources in plant

communities: a general theory of invasibility. J. Ecol. 88: 528–534.

Feldmann J. & Lami R. (1941). Flore et végétation marines de la côte basque française.

Bull. Soc. Bot. Fr. 88: 123–142.

Fielding A.H. & Bell J.F. (1997). A review of methods for the assessment of prediction

errors in conservation presence/absence models. Environ. Conserv. 24(1): 38–49.

Floc´h J.Y., Pajot R. & Wallentinus I. (1991). The japanese brown alga Undaria pinnatifida

on the coast of France and its possible establishment in European waters. J. Cons. Int.

Explor. Mer 47: 379–390.

Font I. (2000). Climatología de España y Portugal. Universidad de Salamanca, Salamanca,

Spain.

García L.V. (2003). Controlling the false discovery rate in ecological research. Trends Ecol.

Evol. 18(11): 553–554.

Gil J. (2008). Macro and mesoscale physical patterns in the Bay of Biscay. J. Mar. Biol.

Assoc. U.K. 88(2): 217–225.

Gorostiaga J.M., Santolaria A., Secilla A., Casares C. & Díez I. (2004). Check-list of the

Basque coast benthic algae (North of Spain). Anales Jard. Bot. Madrid 61(2): 155–180.

Heger T. & Trepl L. (2003). Predicting biological invasions. Biol. Invasions 5(4): 313–321.

Hosmer D.W. & Lemeshow S. (2000). Applied logistic regression. John Wiley & Sons, New

York, United Stated.

Hutchinson G.E. (1959). Homage to Santa Rosalía, or why are there so may kinds of

animals? . Am. Nat. 93: 145–159.

ICES (2007). Alien species alert: Undaria pinnatifida (wakame or Japanese kelp). ICES

Coop. Res. Rep. No. 283: 1–36.

IGN (1999). Mapa de carreteras. Península Ibérica, Baleares y Canarias. Instituto

Geográfico Nacional, Ministerio de Fomento, Madrid, Spain.

Keeley J.E., Baer-Keeley M. & Fotheringham C.J. (2005). Alien plant dynamics following

fire in Mediterranean-climate California shrublands. Ecol. Appl. 15(6): 2109–2125.

Larson D.L., Anderson P.J. & Newton W. (2001). Alien plant invasion in mixed-grass

prairie: Effects of vegetation type and anthropogenic disturbance. Ecol. Appl. 11(1): 128–

141.

Lobban C.S. & Harrison P.J. (1997). Seaweed ecology and physiology. Cambridge

University Press, Cambridge.

Lobo J.M., Jimenez-Valverde A. & Real R. (2008). AUC: a misleading measure of the

performance of predictive distribution models. Global Ecol. Biogeogr. 17(2): 145–151.

López-Darías M., Lobo J.M. & Gouat P. (2008). Predicting potential distributions of

invasive species: the exotic Barbary ground squirrel in the Canarian archipelago and the

west Mediterranean region. Biol. Invasions 10(7): 1027–1040.

Lüning K. (1990). Seaweeds: their environment, biogeography and ecophysiology. John

Wiley & Sons, New York; USA.

Mann K.H. & Lazier J.R.N. (1991). Dynamics of marine ecosystems: biological-physical

interactions in the oceans. Blackwell Scientific Publications, New York, United Stated.

Martínez-Gil M., Gallardo T., Díaz P. & Bárbara I. (2007). Aportaciones al conocimiento

de las algas marinas bentónicas del litoral comprendido entre el estuario del río Quejo y

Punta de la Mesa, Noja, Cantabria, España. Bot. Complutensis 31: 41–53.

Morita T., Kurashima A. & Maegawa M. (2003a). Temperature requirements for the

growth and maturation of the gametophytes of Undaria pinnatifida and Undaria

undarioides (Laminariales, Phaeophyceae). Phycol. Res. 51: 154–160.

Morita T., Kurashima A. & Maegawa M. (2003b). Temperature requirements for the

growth of young sporophytes of Undaria pinnatifida and Undaria undarioides

(Laminariales, Phaeophyceae). Phycol. Res. 51: 266–270.

Muñoz A.R. & Real R. (2006). Assessing the potential range expansion of the exotic monk

parakeet in Spain. Diversity Distrib. 12(6): 656–665.

Parker J.D., Burkepile D.E. & Hay M.E. (2006). Opposing effects of native and exotic

herbivores on plant invasions. Science 311(5766): 1459–1461.

Pérez-Ruzafa I., Menéndez J. & Salinas J.M. (2002). Mapas de distribución de algas

marinas de la Península Ibérica e Islas baleares. XV. Undaria pinnatifida (Harvey)

Suringar (Laminariales, Phaeophyceae). Bot. Complutensis 26: 147–151.

Peteiro C. (2008). A new record of the introduced seaweed Undaria pinnatifida

(Laminariales, Phaeophyceae) from the Cantabrian Sea (northern Spain) with comments on

its establishment. Aquat. Invasions 3(4): 413–415.

Peterson A.T. & Vieglais D.A. (2001). Predicting species invasions using ecological niche

modeling: New approaches from bioinformatics attack a pressing problem. BioScience

51(5): 363–371.

Ramirez-Félix E. & Manzo-Monroy H.G. (2004). Modelling the use of two fishery access

rights, concessions and licences, in the red sea urchin, Strongylocentrotus franciscanus

(Agassiz), fishery at Santo Tomas, Baja California, Mexico. Cienc. Mar. 30(4): 547–560.

Real R., Barbosa A.M. & Vargas J.M. (2006). Obtaining environmental favourability

functions from logistic regression. Environmental and Ecological Statistics 13(2): 237–

245.

Real R., Marquez A.L., Estrada A., Munoz A.R. & Vargas J.M. (2008). Modelling

chorotypes of invasive vertebrates in mainland Spain. Diversity Distrib. 14(2): 364–373.

Real R., Vargas J.M. & Antúnez A. (1993). Environmental influences on local amphibian

diversity: the role of floods on river basins. Biodiversity Conserv. 2(4): 376–399.

Saito Y. (1975). Undaria. In: Advance in phycology in Japan. (Tokida J. & Hirose H., eds.)

pp. 304–320. Dr. W. Junk, Hague.

Salinas J.M., Llera E.M. & Fuertes C. (1996). Nota sobre la presencia de Undaria

pinnatifida (Harvey) Suringar (Laminariales, Phaeophyta) en Asturias (mar Cantábrico).

Bol. Inst. Esp. Oceanogr. 12(1): 77–79.

Santiago Caamaño J., Durán Neira C. & Acuña Castroviejo R. (1990). Aparición de

Undaria pinnatifida en las costas de Galicia (España). Un nuevo caso en la problemática

de introducción de especies foráneas. Informes Técnicos C.I.S. No. 3: 43 pp.

Sauvageau C. (1897). Note préliminaire sur les algues marines du Golfe de Gascogne. J. Bot.

11: 1–64.

Tobin P.C., Whitmire S.L., Johnson D.M., Bjornstad O.N. & Liebhold A.M. (2007).

Invasion speed is affected by geographical variation in the strength of Allee effects. Ecol.

Lett. 10(1): 36–43.

Valentine J.P. & Johnson C.R. (2003). Establishment of introduced of the introduced kelp

Undaria pinnatifida in Tasmania depends on disturbance to nativa algal assemblages. J.

Exp. Mar. Biol. Ecol. 295: 63–90.

Valentine J.P., Magierowski R.H. & Johnson C.R. (2007). Mechanisms of invasion:

establishment, spread and persistence of introduced seaweed populations. Bot. Mar.

50(5/6): 351–360.

van den Hoek C. & Donze M. (1966). The algal vegetation of the rocky côte basque (SW

France). Bull. Centr. Etud. Rech. Sci., Biarritz 6: 289–319.

Wright D.H. (1983). Species-energy theory: an extension of species-area theory. Oikos: 496–

506.

Zuur A.K., Ieno E.N. & Smith G.M. (2007). Analysing ecological data. Springer, New

York, United Stated.

Table 1. Variables used to model the distribution of Undaria pinnatifida in northern

peninsular Spain

Variable Units

Annual precipitation variation coefficient a

Mean carbon based productivity for each annual season b gC m−2 year−1

Mean chlorophyll concentration for each annual season b mg m−3

Distance to the nearest city with more than 100,000 inhabitants c Km

Distance to the nearest city with more than 500,000 inhabitants c Km

Mean annual run-off d mm

Mean annual solar radiation a kwh m−2 day−1

Mean photosynthetically active radiation (PAR) for each annual season b μmol photons m−2 day−1

Mean seawater surface temperatures for the coldest month (January) b ºC

Mean Seawater surface temperatures for the warmest month (August) b ºC

Mean seawater surface temperatures for each annual season b ºC

Sources: a Font (2000); b MODIS, provided by NASA, data set is available from the website: http://modis.gsfc.nasa.gov/; c IGN (1999); d Álvarez et al. (1979)

Table 2. Pair-wise Pearson correlation between the most important variables for Undaria pinnatifida presence based on AIC of bivariate Logistic

regression models

cb_sum chl_spr chl_sum par_spr par_sum par_aut rad_win mst_sum mst_jan mst_aug AIC

cb_sum 1 0.893** 0.857** 0.822** 0.903** 0.744** 0.866** –0.513** –0.704** –0.606** 102.84

chl_spr 0.893** 1 0.827** 0.729** 0.740** 0.702** 0.821** –0.248 –0.516** –0.340* 177.07

chl_sum 0.857** 0.827** 1 0.744** 0.860** 0.752** 0.635** –0.568** –0.730** –0.620** 93.12

par_spr 0.822** 0.729** 0.744** 1 0.794** 0.816** 0.691** –0.631** –0.683** –0.703** 103.96

par_sum 0.903** 0.740** 0.860** 0.794** 1 0.663** 0.757** –0.644** –0.736** –0.706** 91.42

par_aut 0.744** 0.702** 0.752** 0.816** 0.663** 1 0.675** –0.539** –0.527** –0.639** 173.93

rad_inv 0.866** 0.821** 0.635** 0.691** 0.757** 0.675** 1 –0.265* –0.450** –0.370** 212.15

mst_sum –0.513** –0.248 –0.568** –0.631** –0.644** –0.539** –0.265* 1 0.732** 0.967** 158.91

mst_jan –0.704** –0.516** –0.730** –0.683** –0.736** –0.527** –0.450** 0.732** 1 0.727** 109.10

mst_aug –0.606** –0.340* –0.620** –0.703** –0.706** –0.639** –0.370** 0.967** 0.727** 1 172.45

Key: cb_sum, mean Carbon based productivity for Summer; chl_spr, mean Chlorophyll concentration for Spring; chl_sum, mean Chlorophyll concentration for Summer; par_sum, mean PAR

for Summer; par_aut, mean PAR for Autumn; rad_win, mean annual solar radiation for Winter; mst_sum; Mean Seawater surface temperatures for Summer; mst_jan, Mean Seawater surface

temperatures for the coldest month (January); mast_aug, Mean Seawater surface temperatures for the warmest month (August)

* Correlation significant at 0.05 threshold

** Correlation significant at 0.01 threshold

Figure 1. Distribution of Undaria pinnatifida along the north Spanish coast. Key: Filled

circles U. pinnatifida recorded

Figure 2. Favourability of Undaria pinnatifida in Northern Spain according to the

favourability function (Real et al. 2006) using chlorophyll concentration in summer and

photosynthetically active radiation in spring as predictor variables. Points indicate records of

U. pinnatifida

Figure 1.

Figure 2.