Next Article in Journal
Spatio-Temporal Distribution of Limnomysis benedeni Czerniavsky, 1882 and Other Invaders in the Danube Floodplain Kopački Rit Nature Park, Croatia
Previous Article in Journal
Background Mortality of Wildlife on Renewable Energy Projects
Previous Article in Special Issue
First Insights into the Mitochondrial DNA Diversity of the Italian Sea-Slater Across the Strait of Sicily
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Diversity
Volume 17
Issue 9
10.3390/d17090629
diversity-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Evidence in Support of the Kelp Conveyor Hypothesis

by
Cody M. Brooks
* and
Gary W. Saunders
Centre for Environmental and Molecular Algal Research, Department of Biology, University of New Brunswick, 10 Bailey Dr., Fredericton, NB E3B 5A3, Canada
*
Author to whom correspondence should be addressed.
Diversity 2025, 17(9), 629; https://doi.org/10.3390/d17090629
Submission received: 24 July 2025 / Revised: 28 August 2025 / Accepted: 3 September 2025 / Published: 7 September 2025
(This article belongs to the Special Issue Marine Nearshore Biodiversity—2nd Edition)
Browse Figures
Versions Notes

Abstract

The flora of Haida Gwaii includes many macroalgal species, initially assumed endemic, which are largely absent from southern British Columbia but which were subsequently collected in California. One explanation for these disjunct distributions is the kelp conveyor hypothesis, which proposes non-buoyant macroalgae hitchhike on kelp rafts from central California to Haida Gwaii. Using mitochondrial COI-5P data, we adopt a weight-of-evidence approach and summarize broad patterns of allelic distribution and genetic differentiation across 11 species of red algae collected in California, Haida Gwaii and southern Vancouver Isl., British Columbia, to assess the impact of the kelp conveyor hypothesis. Although this hypothesis was based on species with disjunct distributions, we emphasize species with continuous distributions. In support of this hypothesis, we demonstrate low genetic differentiation between Haida Gwaii and California in 9 of 11 species consistent with significantly higher gene flow than from California to Vancouver Isl., and 13% of all alleles observed on Haida Gwaii were disjunct. These observations are consistent with predictions of the kelp conveyor hypothesis. Results here shed light on a previously cryptic source of gene flow which has impacted a considerable proportion of the red algal flora of Haida Gwaii.

1. Introduction

1.1. Kelp Rafting Violates Isolation by Distance

Kelp rafts can be important drivers of oceanographic connectivity [1,2] and shift marine community compositions by introducing alien species [3] or distributing colonising or epiphytic organisms which themselves are poor dispersers. Kelp rafting is common in the southern oceans where species of Durvillaea Bory are responsible for up to 94% of rafts [4]. These rafts may be capable of drifting tens of thousands of kilometres [1] and carry sediment [5] and hundreds of passenger species [6,7,8]. In the context of marine biogeography, identifying marine communities impacted by kelp rafting—and the source of those rafts—can be critical to understanding species distributions.
Wright [9] originally proposed his concept of isolation-by-distance (IBD)—that genetic and geographic distance are positively correlated—and IBD generally holds for macroalgae [10]. Non-buoyant seaweed typically disperse through short-lived spores (4–11 days) [11] subject to local currents and upwelling [12,13] that drift at best several kilometers before settling in new habitat [14]. This short-range dispersal aligns well with IBD; in a metastudy of 49 biogeography studies, IBD held in all 49 cases [10]. For non-buoyant passengers on kelp rafts, this long-range and relatively infrequent dispersal would violate IBD and produce high genetic similarity between geographically distant populations and may present disjunct distributions of species or alleles [15,16].

1.2. The Setting of Haida Gwaii

The remote archipelago of Haida Gwaii, British Columbia, Canada, is home to a high number of reportedly endemic species, both terrestrial and marine [17,18]. These high rates of endemism have been historically attributed to extensive glacial refugia on Haida Gwaii that isolated local populations and led to strong genetic differentiation over time [19]. In marine species present on Haida Gwaii and the surrounding mainland, recent studies have demonstrated that contemporary barriers to gene flow exist across the Queen Charlotte Sound (between Haida Gwaii and mainland British Columbia) [20]. This combination of genetic isolation by ancient glacial refugia and modern barriers to gene flow in the surrounding area have resulted in a Haida Gwaii flora that is decidedly distinct from even nearby regions. During ongoing surveys of the aquatic flora, Saunders [21] observed 33 macroalgae that were hypothesized as being endemic to Haida Gwaii. However, collections completed in central California ~1500 km to the south of Haida Gwaii uncovered all 33 of these species, which were largely absent from the southern end of Vancouver Isl., BC (~350 km southeast) leading to the proposal of the kelp conveyor hypothesis to explain these disjunct distributions [21].

1.3. The Kelp Conveyor Hypothesis

The kelp conveyor hypothesis posits that kelp rafts dislodged in central California drift northward on the Davidson current and carry non-buoyant passenger species to Haida Gwaii. Drift bottles [22] and long-term hydrographic survey data [23,24] show transport on the Davidson current begins near central California and extends northward along the Pacific coast, through Queen Charlotte Sound and reaches as far north as Unimak, Alaska. At a rate of 20–30 cm/s [22,23], kelp dislodged in California could reach Haida Gwaii in ~100 days. The most likely rafting species in California are Macrocystis pyrifera (Linnaeus) C.Agardh and Nereocystis luetkeana (K.Mertens) Postels & Ruprecht and while longevity of kelp rafts varies by temperature and species, those of M. pyrifera can remain afloat for up to 109 days in water temperatures exceeding 20 °C, and likely longer in the cooler winter temperatures of the north Pacific [8] when kelp dislodgement is at its peak [25,26,27]. The combination of slow raft degradation, the near-shore and northbound Davidson current and high dislodgement rates are predicted to synthesize into a long-range pathway of kelp rafting between California and Haida Gwaii.
The presence of long-range gene flow between California and Haida Gwaii would result in these two populations being more similar than the populations of California are to those in S. Vancouver Isl., after correcting for geographic distance. Whereas Saunders [21] developed the kelp conveyor hypothesis based on disjunct species, the present study tests for population-level effects of kelp rafting on the Haida Gwaii flora by leveraging extensive collections made during routine barcoding surveys of Haida Gwaii, S. Vancouver Isl. and California, i.e., for species with continuous ranges along the NE Pacific coast. These barcoding surveys targeted the 5′ end of the mitochondrial cytochrome c oxidase (COI-5P) gene which has a relatively low mutation rate for population-level studies, and consequently we adopt a weight-of-evidence approach by summarizing these relatively weak genetic patterns across 11 macroalgal species to assess if there are broad trends in genetic connectivity between Haida Gwaii and California.

2. Materials and Methods

2.1. Species Selection

This study compiles data from 11 species previously collected as part of routine barcoding surveys spanning 1996–2017. Only species with a minimum of five collections within a 10 km radius and found in all three geographical regions were retained (Figure 1). The 11 species meeting these criteria and used in this study are: Bossiella frondifera (Manza) P.W.Gabrielson, K.A.Miller, Martone & K.R.Hind; Calliarthron tuberculosum (Postels & Ruprecht) E.Y.Dawson; Callophyllis edentata Kylin; Cryptopleura ruprechtiana (J.Agardh) Kylin; Erythrophyllum delesserioides J.Agardh; Mastocarpus intermedius S.C.Lindstrom, Hughey & Martone; Mazzaella rosea (Kylin) Fredericq; Neoporphyra perforata (J. Agardh) L.-E.Yang & J.Brodie; Osmundea spectabilis (Postels & Ruprecht) K.W.Nam; Polyneura latissima (Harvey) Kylin; and Prionitis sternbergii (C.Agardh) J.Agardh. Details for the specimens used in this study are provided in Table S1.

2.2. Sample Collection

The complete dataset used for assessing the vertical distribution of target species consists of 1376 collections across 11 species from California, Oregon, Washington and British Columbia (including S. Vancouver Isl. and Haida Gwaii) (Table S1). A reduced dataset was used for haplotype and differentiation analyses and consisted of 808 collections for which COI-5P data were available from California, S. Vancouver Isl. and Haida Gwaii (Table 1). Collections were made in the intertidal or subtidal using SCUBA. From each collection, ~1 cm2 of material was preserved in silica and returned to the Connell Memorial Herbarium (UNB) at the University of New Brunswick. DNA was extracted with a QIAGEN TissueLyser II and QIAxtractor prior to amplification [28]. Amplification targeted the 5′ end of the cytochrome c oxidase subunit I gene (COI-5P) following protocols outlined in [29] prior to editing by eye in Geneious v8.1.9 [30]. Edited sequences were deposited into the Barcode of Life Data Systems (BOLD dataset DS-KELPCON2) and GenBank (actual primers used for amplification and sequencing recorded with each GenBank accession, Table S1).

2.3. Data Analyses

All analyses were carried out using R v4.5.0 [31]. Observed and expected allele richness was calculated using the non-parametric Chao-1 richness estimator [32] using 9999 bootstrap replicates in the rareNMtests package [33]. Patterns of allele distribution were summarized by categorizing each allele across all species as follows: (1) ubiquitous alleles, present in all three regions; (2) northern alleles, present only in S. Vancouver Isl. and Haida Gwaii; (3) disjunct alleles, present in Haida Gwaii and California but absent from S. Vancouver Isl; (4) private alleles, present only in one region.
Population differentiation was assessed between regions as Jost’s differentiation D calculated using mmod v1.3.3 [34]. To account for effects of isolation-by-distance, pairwise Jost D values were standardized according to geographic distance (in km) as in [35] using the most frequently sampled site per region as a geographic reference point (Figure 1) to calculate geographic distance between regions. To test for significant differentiation between regions, pairwise D values were permuted 999 times and compared to the observed values and exact p-values were calculated. To determine if differentiation between California-Haida Gwaii was lower than California-S. Vancouver Isl., a Wilcoxon test was performed on the standardized D estimates of both population pairs considering all species simultaneously.

3. Results

Across 11 species using all collections (expanded dataset, Table S1) three species occurred predominantly in the mid to upper intertidal (M. intermedius, N. perforata and P. sternbergii), eight species occurred predominantly in the low intertidal to subtidal (B. frondifera, C. edentata, C. ruprechtiana, E. delesserioides, M. rosea, O. spectabilis and P. latissima) and only C. tuberculosum was found throughout the high intertidal to subtidal (Figure 2). Only M. intermedius was confined to the intertidal among the species studied here. Owing to the typically subtidal attachment of the kelp raft forming species, the kelp conveyor hypothesis should largely impact species that occur in the subtidal.
Amplicon lengths varied by species from 482–662 bp after trimming and are reported in Table S1. Overall, observed allele counts were relatively low—between 3–4 alleles per region on average and generally met the lower bound estimates of the Chao-1 estimator (Table 2), however 95% confidence intervals of expected allele counts are indicative of unseen diversity (Table S2), particularly in California where diversity is highest (Figure S1). Disjunct alleles found in California and Haida Gwaii but not in S. Vancouver Isl. were observed in the predominantly subtidal species C. edentata, E. delesserioides, and M. rosea and Chao-1 estimates predict no overlooked diversity in S. Vancouver Isl. for these three species (Table 2 and Table S2). For E. delesserioides, disjunct alleles represented 90% of all Haida Gwaii collections. Broad patterns of allele distribution across all 11 species are summarized in Figure 3. Overall, disjunct alleles represented ~13% of all allelic diversity on Haida Gwaii (Figure 3). Private alleles were generally rare in Haida Gwaii, accounting for only 3.7% of all Haida Gwaii collections; private alleles were more common in S. Vancouver Isl. (13.7%) and abundant in California (41.8%). A subset of alleles presented a northern-only distribution, occurring in S. Vancouver Isl. and Haida Gwaii but not in California (Figure 3).
Pairwise population differentiation for all species is reported in Table 3 as Jost’s differentiation (D) and standardized by geographic distance. Overall, differentiation between regions was low but often significant; significant differentiation was observed between Haida Gwaii and S. Vancouver Isl. in 9 of 11 species, between S. Vancouver Isl. and California in 9 of 11 species and between Haida Gwaii and California populations in only four of 11 species (Table 3).
In considering Haida Gwaii-California directly against S. Vancouver Isl.-California, estimates of D suggest higher gene flow from California to Haida Gwaii than California to S. Vancouver Isl. in 9 of 11 species (Figure 4), and the Wilcoxon test found that the overall mean differentiation was significantly lower for Haida Gwaii-California than for S. Vancouver Isl.-California (p = 0.015, Cohen’s d = 1.93). Results of the Wilcoxon test found no significant correlation between subtidal species and kelp rafting effects (p = 0.067, Cohen’s d = 1.68) rather both intertidal and subtidal species appear to be frequently affected.

4. Discussion

In support of the kelp conveyor hypothesis, we present evidence for elevated gene flow from California to Haida Gwaii and patterns of allele sharing between these two distant populations consistent with kelp rafting in red algae. Saunders [21], based on observed disjunct species, suggested the kelp conveyor may affect as much as 9% of the Haida Gwaii flora. However, kelp rafting patterns at the population level for species with continuous distributions were apparent in 9 of 11 species, suggesting kelp rafting to Haida Gwaii may be much more common than previously supposed. While evidence for kelp rafting is strong, the frequency of private alleles and highly differentiated populations additionally support previous conclusions of glacial refugia on Haida Gwaii, resulting in a complex demographic landscape for this remote archipelago (Figure 3).

4.1. Gene Flow Across Barriers

It is evident that there exists some connectivity in macroalgal populations between Haida Gwaii and California, and this would typically require continuous, genetic connectivity between these two populations. Permuted estimates of D in this study revealed significant differentiation between Haida Gwaii and S. Vancouver Isl. in nine species and support previous conclusions of a genetic break across the Queen Charlotte Sound (QCS) (Table 3). Another significant genetic break exists at Cape Blanco, Oregon, USA, resulting from a combination of physical structure and coastal upwelling [36,37]. Despite at least two significant barriers to gene flow, these distant populations are nevertheless connected. Kelp-rafted passengers have been shown to “jump the fence” over such barriers in other studies [38,39] and would result in long-range connectivity and short-range dissimilarity, consistent with the patterns of disjunct distribution observed in this study (e.g., Erythrophyllum delesserioides, Figure 4).

4.2. Community-Level Impacts

In published studies kelp rafting is a known driver of dispersal and connectivity between island populations of the kelp rafts themselves [40] and a variety of marine fauna [1,8]. Rafts of Macrocystis pyrifera have been associated with more than 800 species of passenger fauna [6,7,8] in the northeast Pacific, and the presence of kelp rafting effects in 9 of 11 species in this study suggest dispersal by kelp rafts is likewise common to many species of macroalgae (Table 3). Evidence of kelp-rafted invertebrates in this region are already evident in the literature; Kelly and Palumbi [41] investigated 50 marine macroinvertebrates and uncovered higher genetic connectivity among subtidal species than intertidal species, and red abalone shells (Haliotis rufescens Swainson) have been found attached to Nereocystis luetkeana holdfasts in southern Haida Gwaii, more than 800 km north of their range [42,43,44].
Across the macroalgal community, kelp rafting is likely both uneven and highly stochastic. The intertidal species Mastocarpus intermedius (specimens with exclusively intertidal habitat data, n = 63) in this study was strongly differentiated across all pairwise comparisons and may not be capable of overcoming the genetic break between Haida Gwaii and S. Vancouver Isl. (Queen Charlotte Sound) or the break between S. Vancouver Isl. and California (Cape Blanco, Oregon). Conversely, Haida Gwaii populations of the low intertidal to subtidal E. delesserioides (88 of 92 collections low intertidal or subtidal; Figure 2) appears to experience considerable gene flow from California such that, of 32 collections on Haida Gwaii, a disjunct Haida Gwaii-California allele was observed in 29 E. delesserioides collections. Standardized estimates of D for E. delesserioides were two orders of magnitude higher for Haida Gwaii-S. Vancouver Isl. (0.000431) than for Haida Gwaii-California (0.000009), so this long-range connectivity is not likely to be a symptom of low differentiation overall in this species.
This study examined seven species which occurred primarily in the low intertidal to subtidal zones (B. frondifera, C. edentata, C. ruprechtiana, E. delesserioides, M. rosea, O. spectabilis and P. latissima) (Figure 2) and kelp conveyor effects were present in all seven species. Additionally, kelp conveyor effects were detected in two predominantly intertidal species (N. perforata and P. sternbergii) (Table 3); it should be said that while N. perforata was predominantly in the mid and high intertidal zones (145 of our 169 collections), three gametophytes were collected subtidally to depths as much as 13 m (Table S1) and members of this group (Bangiales) have a microscopic sporophyte stage (conchocelis) which eludes collection and is known to occur subtidally [45] within stones, shells and coralline algae [46,47], all of which themselves are frequent passengers of kelp rafting [1,2]. The intertidal species P. sternbergii was also, occasionally, found at depths of 6–10 m (Table S1), and could certainly co-occur with the holdfasts of the rafting species M. pyrifera or N. luetkeana. Only M. intermedius occurred strictly in the intertidal zone and lacked any signs of kelp rafting between California and Haida Gwaii (Table 3), which is consistent with the kelp conveyor hypothesis. The final species, C. tuberculosum was abundant throughout the high to subtidal zones (Figure 2) but COI-5P variation in this species was weak—only six alleles observed across 131 collections (Table 2)—resulting in a panmictic population and low Jost D values across all population pairs (Table 3).

4.3. Uneven Sampling Sizes & Low Diversity

Most collections used in this study were haphazard and, consequently, sample sizes were uneven across species and sites. While the species accumulation curves indicated low sample coverage (Figure S1), the Chao1 estimator is most effective when populations are dominated by one to three alleles (as was the case in this study) [32] and the Chao1 results generally estimated a lower bound of 0–2 overlooked alleles (Table 1). The three species with disjunct alleles were considered well-sampled in S. Vancouver Isl. by the Chao-1 estimator, indicating these disjunct alleles are not sampling artefacts (Table S2). Unobserved alleles in California were common owing to relatively small sample sizes in California, however additional California samples would produce one of four outcomes: (1) alleles which are also observed in S. Vancouver Isl. and Haida Gwaii; (2) private alleles present only in California; (3) alleles present in S. Vancouver Isl. but not Haida Gwaii; or (4) alleles observed in Haida Gwaii but not S. Vancouver Isl., indicative of kelp rafting effects. In the former three cases, these new samples would not lessen the existing evidence of kelp rafting between California and Haida Gwaii and in the latter case, additional kelp rafting effects would further support the kelp conveyor hypothesis.
The COI marker has relatively low resolution for detecting population-level patterns and may have led to weaker signals than a highly variable marker (e.g., cox2–3) or microsatellites, however most sequence data used in this study was derived from previous barcoding surveys for which COI was well-suited. Nevertheless, COI has been used to assess demographic patterns in species of Chondrus Stackhouse in the northwest Pacific [48], and COI diversity was adequate for detecting a kelp rafting signal in this study (Table 3). Low resolution is of particular concern at small geographic scales where populations are least differentiated [9], however Haida Gwaii and S. Vancouver Isl. are nearly 800 km distant by coastal shortest-path distance and the relatively conserved COI proved sufficient to detect differentiation and 5–6 alleles per species. No significant trend of habitat (intertidal/subtidal) was detected, and this may reflect a uniform impact of kelp rafting on subtidal/intertidal species or be an artefact of the low resolution in COI. However, as discussed above, only Mastocarpus intermedius was strictly intertidal of the species studied here, and it lacked signal consistent with the kelp conveyor hypothesis, which was also a prediction of the kelp conveyor hypothesis [21].

5. Conclusions

Studies of biogeography depend upon strong foundational knowledge regarding sources of biodiversity and factors affecting speciation. Here, we present patterns of differentiation and disjunct allele distributions which are consistent with the kelp conveyor hypothesis across nine species of red macroalgae. Rather than limited to rare events, the kelp conveyor appears to be a relatively continuous and important process in shaping population genetics of the Haida Gwaii flora. We emphasize that the species used in this study are all macroalgae belonging to the division Rhodophyta and so the conclusions here pertain to red macroalgae. The proportion of kelp-rafted diversity on Haida Gwaii remains unknown and likely extends beyond red algae (indeed of the 33 species in the original publication of the kelp conveyor hypothesis three each were green and brown algae [21]) yet the recent appearance of at least one species of Scinaia Bivona-Bernardi on Haida Gwaii [49] raises the possibility that new arrivals are occurring on ecologically relevant timescales.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/d17090629/s1, Table S1 provides the metadata for the expanded dataset (n = 1376) used in this study, including exact site names, site coordinates, habitat (high intertidal to subtidal 16 m+) and date of collection. Table S2 provides expected allele counts estimated by the Chao-1 estimator including 95% confidence intervals per species and region. Figure S1 provides rarefaction curves of allele richness per species and region. Barcode of Life (BOLD) accession numbers and GenBank numbers for every sample are also provided.

Author Contributions

Conceptualization, G.W.S.; Methodology, C.M.B. and G.W.S.; Software, C.M.B.; Validation, C.M.B. and G.W.S.; Formal Analysis, C.M.B.; Investigation, C.M.B. and G.W.S.; Resources, G.W.S.; Data Curation, G.W.S.; Writing—Original Draft Preparation, C.M.B.; Writing—Review and Editing, C.M.B. and G.W.S.; Visualization, C.M.B.; Supervision, G.W.S.; Project Administration, G.W.S.; Funding Acquisition, G.W.S. All authors have read and agreed to the published version of the manuscript.

Funding

This research was supported by the Natural Sciences and Engineering Research Council of Canada, the Canada Foundation for Innovation and the New Brunswick Innovation Foundation.

Acknowledgments

We thank the myriad of lab colleagues who supported this work, most notably Tanya Moore, Amanda Savoie, Gina Filloramo, Trevor Bringloe, Josh Evans, Marie Dankworth. The cooperative Gwaii Haanas Archipelago Management Board—Council of the Haida Nation, Fisheries and Oceans Canada, and Parks Canada—and Parks Canada’s Conservation and Restoration Program are sincerely thanked for supporting this research, as are the staff of Gwaii Haanas National Park Reserve, National Marine Conservation Area Reserve, and Haida Heritage Site, for extensive support of our field studies in this amazing region. Our thanks also to Lynn Lee and Leandre Vigneault for their field support in Haida Gwaii.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Fraser, C.; Nikula, R.; Waters, J. Oceanic rafting by a coastal community. Proc. R. Soc. 2011, 278, 649–655. [Google Scholar] [CrossRef] [PubMed]
  2. Wichmann, C.-S. Floating kelps in Patagonian Fjords: An important vehicle for rafting invertebrates and its relevance for biogeography. Mar. Biol. 2012, 159, 2035–2049. [Google Scholar] [CrossRef]
  3. Avila, C.; Angulo-Preckler, C.; Martín-Martín, R.P.; Figuerola, B.; Griffiths, H.J.; Waller, C.L. Invasive marine species discovered on non–native kelp rafts in the warmest Antarctic island. Sci. Rep. 2020, 10, 1639. [Google Scholar] [CrossRef]
  4. Smith, S.D.A. Kelp rafts in the Southern Ocean. Glob. Ecol. Biogeogr. 2002, 11, 67–69. [Google Scholar] [CrossRef]
  5. Craw, D.; Waters, J. Long distance kelp-rafting of rocks around southern New Zealand. J. Geol. Geophys. 2018, 61, 428–443. [Google Scholar] [CrossRef]
  6. Foster, M.; Schiel, D. The Ecology of Giant Kelp Forests in California: A Community Profile. In US Fish & Wildlife Service Biological Report; The U.S. Fish and Wildlife Service (FWS): Falls Church, VA, USA, 1985; Volume 85, p. 152. [Google Scholar]
  7. Bushing, W.W. Biogeographic and ecological implications of kelp rafting as a dispersal vector for marine invertebrates. In The Fourth California Islands Symposium: Update on the Status of Resources; Halvorson, W.L., Maender, G.J., Eds.; Santa Barbara Museum of Natural History: Santa Barbara, CA, USA, 1994; pp. 103–110. [Google Scholar]
  8. Hobday, A.J. Persistence and transport of fauna on drifting kelp (Macrocystis pyrifera (L.) C. Agardh) rafts in the Southern California Bight. J. Exp. Mar. Biol. Ecol. 2000, 253, 75–96. [Google Scholar] [CrossRef]
  9. Wright, S. Isolation by distance. Genetics 1943, 28, 114–138. [Google Scholar] [CrossRef]
  10. Durrant, H.M.S.; Burridge, C.P.; Kelaher, B.P.; Barrett, N.S.; Edgar, G.J.; Coleman, M.A. Implications of macroalgal isolation by distance for networks of marine protected areas. Conserv. Biol. 2013, 28, 438–445. [Google Scholar] [CrossRef]
  11. Santelices, B. Patterns of reproduction, dispersal and recruitment in seaweeds. In Oceanography and Marine Biology Annual Review; Barnes, M., Ed.; Aberdeen University Press: Aberdeen, Scotland, 1990; Volume 28, pp. 177–276. [Google Scholar]
  12. Kumagai, N.H.; Molinos, J.G.; Yamano, H.; Takao, S.; Fujii, M.; Yamanaka, Y. Ocean currents and herbivory drive macroalgae-to-coral community shift under climate warming. Proc. Natl. Acad. Sci. USA 2018, 115, 8990–8995. [Google Scholar] [CrossRef]
  13. De Jode, A.; David, R.; Haguenauer, A.; Cahill, A.; Erga, Z.; Guillemain, D.; Sartoretto, S.; Rocher, C.; Marjorie, S.; Le Gall, L.; et al. From seascape ecology to population genomics and back. Spatial and ecological differentiation among cryptic species of the red algae Lithophyllum stictiforme/L. cabiochiae, main bioconstructors of coralligenous habitats. Mol. Phylogenet. Evol. 2019, 137, 104–113. [Google Scholar] [CrossRef]
  14. Gaylord, B.; Reed, D.; Raimondi, P.; Washburn, L.; McLean, S. A physically based model of macroalgal spore dispersal in the wave and current-dominated nearshore. Ecology 2002, 83, 1239–1251. [Google Scholar] [CrossRef]
  15. Van den Hoek, C. The possible significance of long-range dispersal for the biogeography of seaweeds. Helgol. Mar. Res. 1987, 41, 261–272. [Google Scholar] [CrossRef]
  16. Macaya, E.C.; Lopez, B.; Tala, F.; Tellier, F.; Thiel, M. Float and raft: Role of buoyant seaweeds in the phylogeography and genetic structure of non-buoyant associated flora. In Seaweed Phylogeography: Adaptation and Evolution of Seaweeds Under Environmental Change; Hu, Z.M., Fraser, C., Eds.; Springer: Dordrecht, The Netherlands, 2016; pp. 97–130. [Google Scholar]
  17. Calder, J.A.; Taylor, R.L. Flora of the Queen Charlotte Islands, Part I, Systematics of Vascular Plants; Research Branch, Canada, Department of Agriculture Monograph No. 4 Queen’s Printer: Ottawa, ON, Canada, 1968. [Google Scholar]
  18. Warner, B.C.; Mathewes, R.W.; Clague, J.J. Ice-free conditions on the Queen Charlotte Islands, British Columbia, at the height of the late Wisconsin glaciation. Science 1982, 218, 675–677. [Google Scholar] [CrossRef]
  19. Shafer, A.B.A.; Cullingham, C.I.; Côté, S.D.; Coltman, D.W. Of glaciers and refugia: A decade of study sheds new light on the phylogeography of northwestern North America. Mol. Ecol. 2010, 19, 4589–4621. [Google Scholar] [CrossRef]
  20. Keever, C.C.; Sunday, J.; Puritz, J.B.; Addison, J.A.; Toonen, R.J.; Grosberg, R.K.; Hart, M.W. Discordant distribution of populations and genetic gariation in a sea star with high dispersal potential. Evolution 2009, 63, 3214–3227. [Google Scholar] [CrossRef]
  21. Saunders, G.W. Long distance kelp rafting impacts seaweed biogeography in the northeast Pacific: The kelp conveyor hypothesis. J. Phycol. 2014, 50, 968–974. [Google Scholar] [CrossRef] [PubMed]
  22. Schwartzlose, R.A.; Reid, J.L. Near-shore circulation in the California Current. Calif. Mar. Res. Comm. CalCOFI Rept. 1972, 16, 57–65. [Google Scholar]
  23. Penrose, L.W. Analysis of the inshore California Current System off Central California using naval oceanographic office survey data from 1997 to 2002. Master’s Thesis, Naval Postgraduate School, Monterey CA, USA, September 2012. [Google Scholar]
  24. Thomson, R.E.; Krassovski, M.V. Poleward reach of the California undercurrent extension. J. Geophys. Res. 2010, 15, C09027. [Google Scholar] [CrossRef]
  25. Dayton, P.K.; Tegner, M.J. Catastrophic storms, El Niño, and patch stability in a Southern California kelp community. Science 1984, 224, 283–285. [Google Scholar] [CrossRef] [PubMed]
  26. Rothäusler, E.; Gómez, I.; Hinojosa, I.A.; Karsten, U.; Tala, F.; Thiel, M. Effect of temperature and grazing on growth and reproduction of floating Macrocystis spp. (Phaeophyceae) along a latitudinal gradient. J. Phycol. 2009, 45, 547–549. [Google Scholar] [CrossRef]
  27. Hobday, A.J. Abundance and dispersal of drifting kelp Macrocystis pyrifera rafts in the Southern California Bight. Mar. Ecol. Prog. Ser. 2000, 195, 101–116. [Google Scholar] [CrossRef]
  28. Saunders, G.W.; McDevit, D.C. Methods for DNA barcoding photosynthetic protists emphasizing the macroalgae and diatoms. In DNA Barcodes: Methods and Protocols; Kress, W.J., Erickson, D.L., Eds.; Springer: Dordrecht, The Netherlands, 2012; pp. 207–222. [Google Scholar]
  29. Saunders, G.W.; Moore, T.E. Refinements for the amplification and sequencing of red algal DNA barcode and Red ToL phylogenetic markers: A summary of current primers, profiles and strategies. Algae 2013, 28, 31–43. [Google Scholar] [CrossRef]
  30. Kearse, M.; Moir, R.; Wilson, A.; Stones-Havas, S.; Cheung, M.; Sturrock, S.; Drummond, A. Geneious Basic: An integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics 2012, 28, 1647–1649. [Google Scholar] [CrossRef] [PubMed]
  31. R Core Team. R-Project. Available online: https://www.R-project.org (accessed on 6 May 2025).
  32. Chao, A. Non-parametric estimation of the number of classes in a population. Scand. J. Stat. 1984, 11, 265–270. [Google Scholar]
  33. Cayuela, L.; Gotelli, N.J.; Colwell, R.K. Ecological and biogeographic null hypotheses for comparing rarefaction curves. Ecol. Monogr. 2015, 85, 437–455. [Google Scholar] [CrossRef]
  34. Winter, D. Mmod: An R library for the calculation of population differentiation statistics. Mol. Ecol. Res. 2012, 12, 1158–1160. [Google Scholar] [CrossRef]
  35. Galarza, J.A.; Carreras-Carbonell, J.; Macpherson, E.; Pascual, M.; Roques, S.; Turner, G.F.; Rico, C. The influence of oceanographic fronts and early-life-history traits on connectivity among littoral fish species. Proc. Natl. Acad. Sci. USA 2009, 106, 1473–1478. [Google Scholar] [CrossRef]
  36. Magnell, B.A.; Bray, N.A.; Winant, C.D.; Greengrove, C.L.; Largier, J.; Borchardt, J.F.; Bernstein, R.L.; Dorman, C.E. Convergent shelf flow at Cape Mendocino. Oceanography 1990, 4, 4–11. [Google Scholar] [CrossRef]
  37. Barth, J.A.; Pierce, S.D.; Smith, R.L. A separating coastal upwelling jet at Cape Blanco, Oregon and its connection to the California Current System. Deep Sea Res. Part II 2000, 47, 783–810. [Google Scholar] [CrossRef]
  38. Nikula, R.; Spencer, H.G.; Waters, J.M. Passive rafting is a powerful driver of transoceanic gene flow. Biol. Lett. 2013, 9, 20120821. [Google Scholar] [CrossRef]
  39. Leese, F.; Agrawal, S.; Held, C. Long-distance island hopping without dispersal stages: Transportation across major zoographic barriers in a Southern Ocean isopod. Sci. Nat. 2010, 97, 583–594. [Google Scholar] [CrossRef]
  40. Guillemin, M.; Valero, M.; Tellier, F.; Macaya, E.; Destombe, C.; Faugeron, S. Phylogeography of seaweeds in the South East Pacific: Complex Evolutionary Processes Along a Latitudinal Gradient. In Seaweed Phylogeography; Hu, Z.-M., Fraser, C., Eds.; Springer: Dordrecht, The Netherlands, 2016; pp. 251–277. [Google Scholar]
  41. Kelly, R.P.; Palumbi, S.R. Genetic structure among 50 species of the northeastern Pacific rocky intertidal community. PLoS ONE 2010, 5, e8594. [Google Scholar] [CrossRef]
  42. Merilees, B. Records of the Red Abalone in British Columbia. Vic. Nat. 2008, 63, 11. [Google Scholar]
  43. Sloan, N.A.; McDevit, D.C.; Saunders, G.W. Further to the occurrence of red abalone, Haliotis rufescens, in British Columbia. Can. Field Nat. 2010, 124, 238. [Google Scholar] [CrossRef]
  44. Hobday, A.J.; Tegner, M.J. The Warm and the Cold: Influence of Temperature and Fishing on Local Population Dynamics of Red Abalone. Rep. Calif. Coop. Ocean. Fish. Investig. 2002, 43, 74–96. [Google Scholar]
  45. Saunders, G.W.; Brooks, C.M. Metabarcoding extends the distribution of Porphyra corallicola (Bangiales) into the Arctic while revealing novel species and patterns for Conchocelis stages in the Canadian flora. Diversity 2023, 15, 677. [Google Scholar] [CrossRef]
  46. Drew, K. Conchocelis-phase in the life-history of Porphyra umbilicalis (L.) Kütz. Nature 1949, 164, 748–749. [Google Scholar] [CrossRef]
  47. Martinez, E. The conchocelis-phase of porphyra (Rhodophyta) in the intertidal of San-Juan island, Washington, USA. Phycologia 1990, 29, 391–395. [Google Scholar] [CrossRef]
  48. Yang, M.Y.; Kim, M.S. Phylogeography of economic seaweeds Chondrus (Gigartinales, Rhodophyta) in the Northwest Pacific based on rbcL and COI-5P genes. Algae 2022, 37, 135–147. [Google Scholar] [CrossRef]
  49. Brooks, C.M.; Saunders, G.W. First record of Scinaia cf. johnstoniae (Nemaliales, Rhodophyta) in Gwaii Haanas, British Columbia. BioInvasions Rec. 2021, 10, 270–276. [Google Scholar] [CrossRef]
Diversity 17 00629 g001
Figure 1. Map of sites in this study. Each point represents a site (radius 10 km) assigned to one of three regions: (1) Haida Gwaii (green), (2) S. Vancouver Isl. (blue), or (3) California (orange). The arrow shows the predicted path of kelp rafting along the Davidson current. Black circles indicate additional records lacking genetic data used to develop vertical distribution profiles for each species (Table S1). Triangles indicate the most frequently sampled site per region and were used for calculating coastal shortest-path distance between regions.
Figure 1. Map of sites in this study. Each point represents a site (radius 10 km) assigned to one of three regions: (1) Haida Gwaii (green), (2) S. Vancouver Isl. (blue), or (3) California (orange). The arrow shows the predicted path of kelp rafting along the Davidson current. Black circles indicate additional records lacking genetic data used to develop vertical distribution profiles for each species (Table S1). Triangles indicate the most frequently sampled site per region and were used for calculating coastal shortest-path distance between regions.
Diversity 17 00629 g001
Diversity 17 00629 g002
Figure 2. Vertical distribution of all collections from the expanded dataset. Collections used for differentiation and haplotype analyses are indicated in blue and collections used only for vertical analyses (lacking COI-5P data) are in grey.
Figure 2. Vertical distribution of all collections from the expanded dataset. Collections used for differentiation and haplotype analyses are indicated in blue and collections used only for vertical analyses (lacking COI-5P data) are in grey.
Diversity 17 00629 g002
Diversity 17 00629 g003
Figure 3. A summary of allele distributions observed in (A) Haida Gwaii, (B) S. Vancouver Isl., and (C) California across all species in this study. Disjunct alleles (those present in California and Haida Gwaii only; orange) represent putative kelp rafting effects and account for 13.3% of all alleles observed in Haida Gwaii populations. Numbers in parentheses indicate sample size across all species for each region.
Figure 3. A summary of allele distributions observed in (A) Haida Gwaii, (B) S. Vancouver Isl., and (C) California across all species in this study. Disjunct alleles (those present in California and Haida Gwaii only; orange) represent putative kelp rafting effects and account for 13.3% of all alleles observed in Haida Gwaii populations. Numbers in parentheses indicate sample size across all species for each region.
Diversity 17 00629 g003
Diversity 17 00629 g004
Figure 4. Jost D (standardized per km) between California-Haida Gwaii (orange) and California-S. Vancouver Isl. (blue) for all species. Vertical lines are the mean D values across species. The difference in means is significant (p = 0.015).
Figure 4. Jost D (standardized per km) between California-Haida Gwaii (orange) and California-S. Vancouver Isl. (blue) for all species. Vertical lines are the mean D values across species. The difference in means is significant (p = 0.015).
Diversity 17 00629 g004
Table 1. Sample sizes used for differentiation and haplotype analyses (reduced dataset, n = 808) for all species across regions used in this study (Table S1).
Table 1. Sample sizes used for differentiation and haplotype analyses (reduced dataset, n = 808) for all species across regions used in this study (Table S1).
SpeciesCaliforniaHaida GwaiiS. Vancouver Isl.
Bossiella frondifera12166
Calliarthron tuberculosum284954
Callophyllis edentata157048
Cryptopleura ruprechtiana308420
Erythrophyllum delesserioides83212
Mastocarpus intermedius93017
Mazzaella rosea1188
Neoporphyra perforata121644
Osmundea spectabilis13199
Polyneura latissima124021
Prionitis sternbergii9937
Table 2. Allele richness reported as alleles observed/expected. Expected allele counts are estimates derived from Chao-1 allele estimator. Detailed 95% confidence intervals of expected allele counts are provided in Table S2.
Table 2. Allele richness reported as alleles observed/expected. Expected allele counts are estimates derived from Chao-1 allele estimator. Detailed 95% confidence intervals of expected allele counts are provided in Table S2.
SpeciesCaliforniaHaida GwaiiS. Vancouver Isl.
Bossiella frondifera4/53/42/2
Calliarthron tuberculosum2/24/45/7
Callophyllis edentata5/53/31/1
Cryptopleura ruprechtiana5/52/23/3
Erythrophyllum delesserioides6/83/32/2
Mastocarpus intermedius2/22/24/4
Mazzaella rosea5/73/41/1
Neoporphyra perforata4/62/24/4
Osmundea spectabilis5/62/24/5
Polyneura latissima3/37/116/9
Prionitis sternbergii5/51/11/1
Table 3. Jost D standardized per kilometer of geographic distance. Kelp rafting effects (bold type) are considered when HG-Cal is less than SVI-Cal. Differentiation marked with * is significant as assessed by permutation testing.
Table 3. Jost D standardized per kilometer of geographic distance. Kelp rafting effects (bold type) are considered when HG-Cal is less than SVI-Cal. Differentiation marked with * is significant as assessed by permutation testing.
SpeciesHaida Gwaii–CaliforniaS. Vancouver Isl.–CaliforniaHaida Gwaii–S. Vancouver Isl.
Bossiella frondifera0.0000143 *0.0000299 *0.0000254 *
Calliarthron tuberculosum0.00000540.00000330.0000554 *
Callophyllis edentata0.0000210 *0.0000908 *0.0000530 *
Cryptopleura ruprechtiana0.0000449 *0.0000901 *0.0000290 *
Erythrophyllum delesserioides0.0000066 *0.0002481 *0.0005711 *
Mastocarpus intermedius0.0000624 *0.0000453 *0.0001304 *
Mazzaella rosea0.00000580.0000249 *0.0000103 *
Neoporphyra perforata0.0000374 *0.0000450 *0.0000472 *
Osmundea spectabilis0.00002000.0000679<0.00001
Polyneura latissima0.0000451 *0.0000666 *0.0000071
Prionitis sternbergii0.0000440 *0.0000663 *<0.00001 *
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Brooks, C.M.; Saunders, G.W. Evidence in Support of the Kelp Conveyor Hypothesis. Diversity 2025, 17, 629. https://doi.org/10.3390/d17090629

AMA Style

Brooks CM, Saunders GW. Evidence in Support of the Kelp Conveyor Hypothesis. Diversity. 2025; 17(9):629. https://doi.org/10.3390/d17090629

Chicago/Turabian Style

Brooks, Cody M., and Gary W. Saunders. 2025. "Evidence in Support of the Kelp Conveyor Hypothesis" Diversity 17, no. 9: 629. https://doi.org/10.3390/d17090629

APA Style

Brooks, C. M., & Saunders, G. W. (2025). Evidence in Support of the Kelp Conveyor Hypothesis. Diversity, 17(9), 629. https://doi.org/10.3390/d17090629

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop