Advertisement
Research Article

Invasion of the Red Seaweed Heterosiphonia japonica Spans Biogeographic Provinces in the Western North Atlantic Ocean

  • Christine Newton mail,

    newton.c@husky.neu.edu

    Affiliation: Marine Science Center, Northeastern University, Nahant, Massachusetts, United States of America

    X
  • Matthew E. S. Bracken,

    Affiliation: Marine Science Center, Northeastern University, Nahant, Massachusetts, United States of America

    X
  • Megan McConville,

    Affiliation: Department of Biological Sciences, University of Rhode Island, Kingston, Rhode Island, United States of America

    X
  • Katherine Rodrigue,

    Affiliation: Department of Biological Sciences, University of Rhode Island, Kingston, Rhode Island, United States of America

    X
  • Carol S. Thornber

    Affiliation: Department of Biological Sciences, University of Rhode Island, Kingston, Rhode Island, United States of America

    X
  • Published: April 24, 2013
  • DOI: 10.1371/journal.pone.0062261

Abstract

The recent invasion of the red alga Heterosiphonia japonica in the western North Atlantic Ocean has provided a unique opportunity to study invasion dynamics across a biogeographical barrier. Native to the western North Pacific Ocean, initial collections in 2007 and 2009 restricted the western North Atlantic range of this invader to Rhode Island, USA. However, through subtidal community surveys, we document the presence of Heterosiphonia in coastal waters from Maine to New York, USA, a distance of more than 700 km. This geographical distribution spans a well-known biogeographical barrier at Cape Cod, Massachusetts. Despite significant differences in subtidal community structure north and south of Cape Cod, Heterosiphonia was found at all but two sites surveyed in both biogeographic provinces, suggesting that this invader is capable of rapid expansion over broad geographic ranges. Across all sites surveyed, Heterosiphonia comprised 14% of the subtidal benthic community. However, average abundances of nearly 80% were found at some locations. As a drifting macrophyte, Heterosiphonia was found as intertidal wrack in abundances of up to 65% of the biomass washed up along beaches surveyed. Our surveys suggest that the high abundance of Heterosiphonia has already led to marked changes in subtidal community structure; we found significantly lower species richness in recipient communities with higher Heterosiphona abundances. Based on temperature and salinity tolerances of the European populations, we believe Heterosiphonia has the potential to invade and alter subtidal communities from Florida to Newfoundland in the western North Atlantic.

Introduction

Non-native species invasions have become a primary focus of research on global change in the past decade [1], [2]. The importance of marine invasions is highlighted by the significant economic and ecological impacts often associated with these species following a successful invasion. While only a small fraction of all introduced species can successfully thrive in a new habitat, their impacts can be dramatic [3], [4]. Economically, invasive marine species have been responsible for the collapse of fisheries and losses in aquaculture, tourism, and marine infrastructure [5]. Invaders may also have substantial ecological impacts by modifying the habitat in which they invade, displacing native species, and altering food webs and community structure [6]. Additionally, marine invasive species have been identified as a major threat to biodiversity [1], [5], [7].

The invasive red seaweed Heterosiphonia japonica Yendo (hereafter Heterosiphonia), recently discovered in the western North Atlantic Ocean, poses a threat to native biodiversity and ecosystem functioning. First reported in Rhode Island waters in 2007, this species is morphologically and genetically identical to invasive populations of Heterosiphonia in the eastern North Atlantic [8], [9]. Heterosiphonia was first recorded in France in 1984 and has since become widespread along European coastlines [10]. While the exact vector of introduction to the western North Atlantic is unknown, this species was likely introduced from Europe via ballast water early in the new century [8].

In contrast to its European invaded range [10], [11], [12], Heterosiphonia is not particularly abundant in its native range in the western North Pacific Ocean, comprising less than 1% of the macroalgal biomass and only occurring sporadically throughout the year [13], [14]. In both its native and invaded ranges, Heterosiphonia occupies shallow, subtidal habitats and is present either on rocky substrata or epiphytic on other macroalgal species, although it has also been found in sandy and soft sediment habitats ([11], [13], C. Newton, pers. obs.).

Initial reports limited the western Atlantic distribution of this species to Rhode Island [8]. However, reports from the invasion of Heterosiphonia in Europe suggest that the species is capable of rapid dispersal associated with broad thermal and salinity tolerances and high fecundity due to vegetative propagation of fragmented pseudolaterals [15], [16]. This has led to a wide geographic distribution in Europe, with reports of the invasive alga from Norway to Italy [10]. Based on thermal tolerances across its European range, Heterosiphonia has the potential to invade western Atlantic waters from Newfoundland through Florida [16].

Based on these predictions from the eastern Atlantic invasion of Heterosiphonia, we sought to: (1) ascertain the present invaded range and relative abundance of Heterosiphonia in western Atlantic coastal waters; (2) determine the biological attributes that are associated with recipient communities; (3) establish whether more diverse communities have higher resistance to Heterosiphonia invasion; and (4) assess the proportion of drift Heterosiphonia in wrack mats washed ashore. We addressed these goals by surveying shallow subtidal communities and adjacent beaches along the northwestern Atlantic shoreline from New York to Maine, USA.

Methods

We conducted subtidal community surveys at 19 sites between Cape Elizabeth, Maine (43°37′N, 70°12′W) and Waterford, Connecticut (41°17′N, 72°09′W), including Southold, New York (41°18′N, 71°55′W), during the summer of 2012 (Fig. 1). This geographic range covers over 700 km of coastline while spanning a well-known biogeographic barrier in the western Atlantic Ocean; Cape Cod, Massachusetts, separates the Acadian biogeographic province from the more southerly Virginian province. These two provinces are characterized by marked differences in water temperature and community structure [17], [18]. Cape Cod is a well-known southern limit for many cold water marine species, as the southward flowing Labrador Current brings down cooler waters, before swinging east along the arm of Cape Cod, and finally returning in a northeasterly direction. However, waters south of Cape Cod are more influenced by the warmer Gulf Stream, particularly during the summer months [17].

thumbnail

Figure 1. Presence of Heterosiphonia japonica in the Western Atlantic Ocean.

Numbers correspond to locations listed in Table 1 where Heterosiphonia was found in SCUBA, shallow subtidal, and intertidal surveys. Circles indicate locations of in situ SCUBA surveys, with filled circles corresponding to locations where Heterosiphonia was found and open circles indicating locations where Heterosiphonia was absent. Squares correspond to locations where Heterosiphoina was found drifting in the shallow subtidal or as intertidal wrack.

doi:10.1371/journal.pone.0062261.g001

Surveys were conducted in both biogeographic provinces between 0 and 6.25 m depth using SCUBA, as Heterosiphonia is most commonly found within this depth range. Sites were chosen to include a variable range of exposure, from locations exposed to ocean swells to protected bays.

We also conducted weekly surveys of intertidal wrack mats at five sites in the northern range of our surveys from 28 June 2012 through 02 August 2012 (Table 1). These sites were chosen to encompass a variety of exposures and local topographies. Surveys were conducted at low tide each week. Similar to our subtidal surveys, a 20 m transect was laid parallel to the mean-low water line at each site and a 0.0625 m2 quadrat was used to sample every 2 m along the transect. The contents of each quadrat were collected into individual bags and immediately returned to the laboratory where any Heterosiphonia present in the quadrat was sorted out. The wet weight of all Heterosiphonia and other remaining macrophytes present was recorded after being spun in a salad spinner 15X to remove excess moisture [6].

thumbnail

Table 1. Locations of Heterosiphonia japonica surveys.

doi:10.1371/journal.pone.0062261.t001

Statistical Analyses

Multivariate data were analyzed using Primer v. 6.0 (Primer-E Ltd., Plymouth, UK) to compare differences in subtidal community structure. Bray-Curtis similarity matrices were constructed on square-root transformed percent-cover data. We then ran a PERMANOVA (Permutational Multivariate Analysis of Variance; [19]) to determine if community compositions differed north and south of Cape Cod, Massachusetts. Univariate data were analyzed using JMP v. 9.0 (SAS Institute, Inc., Cary, North Carolina, USA). Regression was used to determine the relationship between species richness and abundance of Heterosiphonia. We conducted analyses of variance (ANOVAs) to assess spatial and temporal differences for intertidal drift surveys of Heterosiphonia. To assess temporal variability in our subtidal surveys we were only able to repeat subtidal surveys at four locations, two north (sites 14 & 15, Table 1) and two south (sites 21 & 22, Table 1). We were unable to satisfy the homogeneity of variances assumption for analyses of variances via transformation for these data [20]. Therefore, we used a generalized linear model (proc GENMOD in SAS v. 9.2, SAS Institute, Inc., Cary, North Carolina, USA) with a Poisson distribution and log link to assess temporal differences of Heterosiphonia abundance between early summer and late summer.

Ethics Statement

No specific permits were required for the field surveys, as they were conducted from public access points or Northeastern University property (Marine Science Center, Nahant, Massachusetts, USA). The study did not involve any endangered or protected species or any protected locations.

Results

Subtidal community composition differed significantly between the Acadian province and the Virginian province (p<0.001; Fig. 2). Northern subtidal communities were primarily composed of Chondrus crispus, Heterosiphonia japonica, crustose coralline algal species, and Corallina officinalis, which collectively comprised over 60% of the sessile species cover (means of 20%, 17%, 12% and 11%, respectively). While Chondrus and Heterosiphonia remained the most abundant species in southern subtidal communities (13% and 7% of the cover, respectively), Phyllophora pseudoceranoides also made up a substantial portion of the subtidal community (9%). There was also a greater diversity in community structure at southern sites, with 22 different species comprising 90% of the community. By contrast, only 13 species comprised 90% of the shallow subtidal community in the north (Fig. 2). Results were similar when Heterosiphonia was removed from the analysis (p<0.001).

thumbnail

Figure 2. Relative abundances of sessile species (seaweeds and sessile invertebrates) in subtidal communities.

North and south refer to the biogeographical barrier at Cape Cod, Massachusetts. Species listed comprised 80% of the overall community, and data are means ±1 At each site, a 20 m transect was haphazardly placed in the subtidal zone, at approximately the mid point of the species’ typical depth range (mean depth = 2.0±0.11 m). We placed a 0.0625 m2 quadrat every 2 m along the transect. Within each quadrat, the percent cover of each macroalgal and sessile invertebrate species was recorded. When present, a subsample of Heterosiphonia individuals was collected from each site. Upon collection, specimens were returned to the laboratory, where field identifications were confirmed under a compound microscope (100X) using characteristics from Schneider [8]. All individuals collected were either vegetative or tetrasporic. No fertile gametophytes were found during our sampling. At an additional 12 sites within our subtidal survey range, we examined the intertidal and/or shallow subtidal zone for evidence of drifting Heterosiphonia individuals. At each of these sites, individuals were collected and identified under a field microscope (40X) or laboratory microscope (100X). Due to logistical constraints, we clustered all crustose coralline algal species together, as in situ identifications to the species level proved impossible for this group. When possible, a subsample of Heterosiphonia individuals from most locations was pressed and deposited in the University of Rhode Island (KIRI) or Northeastern University (HNUB) herbarium collections.

doi:10.1371/journal.pone.0062261.g002

Despite these differences in community compositions, Heterosiphonia was present at all but two of the subtidal sites surveyed (it was absent at Nubble Light, York, Maine and Town Neck Beach, Sandwich, Massachusetts) in both the Acadian province and the Virginian province. Relative abundances of Heterosiphonia ranged from 0.0 to 100 percent coverage to the north of Cape Cod (mean = 17.34±2.86%), while relative abundances were slightly lower south of Cape Cod (0.0% to 52.63%, mean = 7.41±1.19%; one-way ANOVA, F1,183 = 11.35, p<0.001). The average abundance of Heterosiphonia was lower in subtidal communities where species richness was higher (R2 = 0.30, p = 0.02; Fig. 3). During our surveys of intertidal wrack mats, Heterosiphonia comprised an average of 20% of biomass; however, abundances of up to 65% were seen on multiple occasions.

thumbnail

Figure 3. Relationship between sessile species richness and Heterosiphonia japonica abundance.

Data presented are per plot, across all sites (regression; R2 = 0.30, p = 0.02).

doi:10.1371/journal.pone.0062261.g003

Heterosiphonia abundances also exhibited temporal and spatial variability, both subtidally and in intertidal wrack mats. Subtidal Heterosiphonia abundances were two orders of magnitude higher at sites surveyed during the early summer (May/June) than at the end of the summer (X2 = 1676.52, p<0.001; Table 2). While intertidal wrack mat surveys were not conducted during the month of May, the abundance of Heterosiphonia was 40% higher during June than any of the other months surveyed (F2,370 = 9.34, p<0.001; Table 2). As expected, we found differences among sites surveyed (F4,370 = 15.95, p<0.001). There was also a significant site*week interaction for our intertidal drift surveys (p<0.001).

thumbnail

Table 2. Seasonality of Heterosiphonia japonica abundances.

doi:10.1371/journal.pone.0062261.t002

Discussion

While initial reports restricted the distribution of Heterosiphonia to Rhode Island [8], [9], this invader now occurs across a much larger biogeographic range. From this study alone, we have determined Heterosiphonia has become established in subtidal communities along >700 km of the western Atlantic coastline from Cape Elizabeth, Maine through Waterford, Connecticut. Furthermore, recent reports have documented the presence of Heterosiphonia in Atlantic Canada [9], and Heterosiphonia has been reported as far west as the mouth of the Connecticut River (J. Foertch, pers. comm.). This extensive range, with a continuous distribution spanning at least 700 km, became evident within only five years of the initial report of Heterosiphonia in Rhode Island in 2007 [9]. While we can not confirm how long Heterosiphonia may have been present (but unnoticed) in New England, if Heterosiphonia was indeed initially limited to sites in Rhode Island, as suggested by Schneider [8], then this incredibly rapid rate of expansion is comparable to the expansion throughout its invaded range along European coastlines; within 5 years of initial reports, the invader had spread >830 km along the Norwegian coast [10], [11]. Although the possibility of multiple introductions via shipping vectors throughout its western Atlantic range exists, it is also possible that Heterosiphonia successfully expanded its invaded range through rapid growth and nutrient uptake rates, high reproductive potential due to fragmentation, and release from natural herbivory (A. Drouin & N. Low, pers. comm.).

Based on the wide range of temperature and salinity tolerances of the European populations of Heterosiphonia, we believe this invader will continue to rapidly expand its geographic range along the western Atlantic coast, ultimately achieving a temperate to subtropical distribution and potentially invading locations from Florida to Newfoundland [16]. We are currently working to determine the lethal temperature and salinity ranges for the western Atlantic Ocean populations of Heterosiphonia. However, these populations are an exact genetic match to European populations [8], for which the thermal and salinity tolerances are known (0°C to 30°C, 10 to >30 psu; [16]). Thus, whereas this invader’s eventual range will likely be impacted by changes in temperature and salinity associated with climate change, a more pressing current concern is the likelihood that it will continue to expand its range rapidly, both northward and southward, to fill its thermal niche (i.e., its temperature and salinity tolerances allow it to grow well beyond its current geographic range).

Heterosiphonia was able to invade subtidal communities both north and south of Cape Cod, Massachusetts, a well-known biogeographical barrier. Despite the historical differences in both abiotic conditions and biological community structure between the two biogeographic provinces [17], [18], Heterosiphonia has become one of the most abundant macrophytes, on average, in these communities. However, the abundance of Heterosiphonia also appears to be spatially variable, with Heterosiphonia comprising up to 79% of total macrophyte cover at some locations, whereas in other communities where Heterosiphonia is present, it occupies <1% of the shallow subtidal community (Table 1). These data may be the result of our survey design, which was intended to rapidly assess the invader’s geographic range by maximizing the number of sites visited over the course of four months. As a result, we were not able to visit all sites on a regular basis.

However, we were able to capture some of the temporal variability in Heterosiphonia abundances in subtidal communities. During this study, four sites (two north and two south of Cape Cod, Massachusetts) were surveyed both at the beginning of the summer (May/June) and at the end of the summer (August). Heterosiphonia abundances were two orders of magnitude higher during May/June than during the end of the summer (Table 2). However, populations of Heterosiphonia appeared to be recovering following the end of this survey (September and October). Additionally, large populations of Heterosiphonia were present during the previous fall of 2011. Therefore, despite this decrease in abundance towards late summer, we postulate that Heterosiphonia populations may be experiencing a seasonal growth cycle, with extremely high abundances during the late spring (Table 2) and fall (late September – October). However, towards the middle of August, the abundance of Heterosiphonia was significantly reduced, even becoming undetectable in some locations where it was formerly abundant (Table 2). The seasonal pattern observed during 2012 is consistent with reproductive observations from European populations, where necrosis in cells of Heterosiphonia pseudolaterals was most prominently observed in late summer and fall [15], suggesting individuals were shedding small fragments. As fragmented pseudolaterals can function as vegetative propagules [15], the abscission of these structures during the late summer may lead to a secondary peak in the abundance of Heterosiphonia upon settlement and regrowth.

The temporal patterns observed in subtidal Heterosiphonia abundance paralleled patterns seen in intertidal wrack mats. Peak abundances of attached, subtidal Heterosiphonia were seen during May and began to decline in June (Table 2). In contrast, maximum intertidal wrack abundances did not begin to decline until July. This suggests that Heterosiphonia individuals became detached from the substratum during the late spring and early summer, existing primarily as drifting individuals. Indeed, many drifting specimens were observed during early summer months at various sites during our SCUBA surveys. The loss of these individuals from subtidal populations could further explain the decline in Heterosiphonia abundances seen during our surveys in late summer and provide further support for the likely importance of excised pseudolaterals as a means of reproduction for this invasive macrophyte, particularly in contributing to a second peak in abundance during early fall. However, it is currently unknown whether Heterosiphonia populations will maintain their high abundances through winter conditions (e.g. lower temperature, higher storm frequency and intensity).

Our surveys suggest that Heterosiphonia may have already altered subtidal community structure in areas it has invaded, as we observed lower seaweed species richness in communities characterized by greater Heterosiphonia abundance. These patterns are consistent with local extinction of native macroalgae due to competition with Heterosiphonia. It is also possible that communities with greater native macrophyte diversity are more resistant to invasion by Heterosiphonia [21], and we are currently conducting experiments to evaluate these possibilities.

Very little is currently known about the impacts of invasive seaweed species; ecological impacts have only been studied for ~6% of the 277 known invasive seaweed species. Of these, only 6 studies have examined ecological impacts in situ [22], [23]. Collectively, our observations highlight community characteristics and population fluctuations across the current known range of a newly-discovered invasive macrophyte. The invasion and expansion of Heterosiphonia across multiple biogeographic provinces in the western North Atlantic Ocean provides an opportunity to understand the spread, impacts, and mechanisms associated with a marine invasion, providing critical information for management and amelioration of the impacts of this species and other marine invaders.

Acknowledgments

The authors would like to thank V. Perini, I. Rosenthal, and N. Low for their diving assistance and A. Saco for help with beach wrack surveys. We thank the diving safety officers and their respective universities for diving reciprocity: T. Lyman (Northeastern University), A. Watson (University of Rhode Island), and J. Godfrey (University of Connecticut). This is contribution number 297 from the Marine Science Center of Northeastern University.

Author Contributions

Conceived and designed the experiments: CN MESB CST. Performed the experiments: CN MM KR. Analyzed the data: CN MESB MM KR CST. Contributed reagents/materials/analysis tools: CN MESB CST. Wrote the paper: CN MESB CST.

References

  1. 1. Stachowicz JJ, Terwin JR, Whitlatch RB, Osman RW (2002) Linking climate change and biological invasions: ocean warming facilitates nonindigenous species invasions. Proceedings of the National Academy of Science USA 99: 15497–15500. doi: 10.1073/pnas.242437499
  2. 2. Rahel FJ, Olden J (2008) Assessing the effect of climate change on aquatic invasive species. Conservation Biology 22: 521–533. doi: 10.1111/j.1523-1739.2008.00950.x
  3. 3. Mack RN, Simberloff D, Lonsdale WM, Evans H, Colout M, et al. (2000) Biotic invasions: causes, epidemiology, global consequence and control. Ecological Applications 10: 689–710. doi: 10.1890/1051-0761(2000)010[0689:bicegc]2.0.co;2
  4. 4. Molnar JL, Gamboa RL, Revenga C, Spalding MD (2008) Assessing the global threat of invasive species to marine biodiversity. Frontiers of Ecology and the Environment 6: 485–492. doi: 10.1890/070064
  5. 5. Bax N, Williamson A, Aguero M, Gonzalez E, Geeves W (2003) Marine invasive alien species: a threat to global biodiversity. Marine Policy 27: 313–323. doi: 10.1016/s0308-597x(03)00041-1
  6. 6. Jones E, Thornber CS (2010) Effects of habitat-modifying invasive macroalgae on epiphytic algal communities. Marine Ecology Progress Series 400: 87–100. doi: 10.3354/meps08391
  7. 7. Parker IM, Simberloff D, Lonsdale WM, Goodell K, Wonham M, et al. (1999) Impact: toward a framework for understanding the ecological effects of invaders. Biological Invasions 1: 3–19. doi: 10.1023/a:1010034312781
  8. 8. Schneider CW (2010) Report of a new invasive alga in the Atlantic United States: “Heterosiphonia” japonica in Rhode Island. Journal of Phycology 46: 653–657. doi: 10.1111/j.1529-8817.2010.00866.x
  9. 9. Savoie AM, Saunders GW (2013) First record of the invasive red alga Heterosiphonia japonica (Ceramiales, Rhodophyta) in Canada. BioInvasions Records 2: 27–32. doi: 10.3391/bir.2013.2.1.04
  10. 10. Sjøtun K, Husa V, Pena V (2008) Present distribution and possible vectors of introductions of the alga Heterosiphonia japonica (Ceramiales, Rhodophyta) in Europe. Aquatic Invasions 3: 377–394. doi: 10.3391/ai.2008.3.4.3
  11. 11. Husa V, Sjøtun K, Lein TE (2004) The newly introduced species Heterosiphonia japonica Yendo (Dasyaceae, Rhodophyta): geographical distribution and abundance at the Norwegian southwest coast. Sarsia 89: 211–217. doi: 10.1080/00364820410006600
  12. 12. Husa V, Sjøtun K, Brattenborg N, Lein TE (2008) Changes of macroalgal biodiversity in sublittoral sites in southwest Norway: impact of an introduced species or higher temperature? Marine Biology Research 4: 414–428. doi: 10.1080/17451000802232874
  13. 13. Choi CG, Oh SJ, Kang IJ (2009) Subtidal marine algal community of Jisepo in Geoge, Korea. Journal of the Faculty of Agriculture, Kyushu University 55: 39–45.
  14. 14. Kang C-K, Choy EJ, Son Y, Lee J-Y, Kim JK, et al. (2008) Food web structure of a restored macroalgal bed in the eastern Korean peninsula determined by C and N stable isotope analyses. Marine Biology 153: 1181–1198. doi: 10.1007/s00227-007-0890-y
  15. 15. Husa V, Sjøtun K (2006) Vegetative reproduction in “Heterosiphonia japonica” (Dasyaceae, Ceramiales, Rhodophyta), an introduced red alga on European coasts. Botanica Marina 49: 191–199. doi: 10.1515/bot.2006.024
  16. 16. Bjaerke MR, Rueness J (2004) Effects of temperature and salinity on growth, reproduction and survival in the introduced red alga Heterosiphonia japonica (Ceramiales, Rhodophyta). Botanica Marina 47: 373–380. doi: 10.1515/bot.2004.055
  17. 17. Humm HJ (1969) Distribution of marine algae along the Atlantic coast of North America. Phycologia 7: 43–53. doi: 10.2216/i0031-8884-7-1-43.1
  18. 18. Wethey DS (2002) Biogeography, competition, and microclimate: the barnacle Chthamalus fragilis in New England. Integrative and Comparative Biology 42: 872–880. doi: 10.1093/icb/42.4.872
  19. 19. Anderson MJ (2001) A new method for non-parametric multivariate analysis of variance. Austral Ecology 26: 32–46. doi: 10.1111/j.1442-9993.2001.01070.pp.x
  20. 20. Underwood AJ (1997) Experiments in Ecology. New York: Cambridge University Press. 504 p.
  21. 21. Stachowicz JJ, Fried H, Osman RW, Whitlatch RB (2002) Biodiversity, invasion resistance, and marine ecosystem function: reconciling pattern and process. Ecology 83: 2575–2590. doi: 10.1890/0012-9658(2002)083[2575:birame]2.0.co;2
  22. 22. Thomsen MS, Wernberg T, Tuya F, Silliman BR (2009) Evidence for impacts of nonindigenous macroalgae: a meta-analysis of experimental field studies. Journal of Phycology 45: 812–918. doi: 10.1111/j.1529-8817.2009.00709.x
  23. 23. Williams SL, Smith JE (2007) A global review of the distribution, taxonomy, and impacts of introduced seaweeds. Annual Review of Ecology, Evolution, and Systematics 38: 327–359. doi: 10.1146/annurev.ecolsys.38.091206.095543