Abstract
Augmentative biocontrol aims to control established pest populations through enhancement of their indigenous enemies. To our knowledge, this approach has not been applied at an operational scale in natural marine habitats, in part because of the perceived risk of adverse non-target effects on native ecosystems. In this paper, we focus on the persistence, spread and non-target effects of the sea urchin Evechinus chloroticus when used as biocontrol agent to eradicate an invasive kelp from Fiordland, New Zealand. Rocky reef macrobenthic assemblages were monitored over 17 months in areas where the indigenous algal canopy was either removed or left intact prior to the translocation of a large number of urchins (>50 ind.·m−2). Urchin densities in treated areas significantly declined ∼9 months after transplant, and began spreading to adjacent sites. At the end of the 17-month study, densities had declined to ∼5 ind.·m−2. Compared to controls, treatment sites showed persistent shifts from kelp forest to urchin barrens, which were accompanied by significant reductions in taxa richness. Although these non-target effects were pronounced, they were considered to be localised and reversible, and arguably outweigh the irreversible and more profound ecological impacts associated with the establishment of an invasive species in a region of high conservation value. Augmentative biocontrol, used in conjunction with traditional control methods, represents a promising tool for the integrated management of marine pests.
Introduction
Biological invasions can have profound impacts on ecosystem functioning, by altering community structure, native species richness and ecological processes [1], [2]. The magnitude and extent of these impacts vary across temporal and spatial scales [3], but they can be extensive and irreversible [4]. Historically, a range of tools has been used to mitigate impacts associated with marine pests, including physical [5], [6], chemical [7], [8] and biological [9], [10] treatments. Compared to terrestrial and freshwater systems, control or eradication tools used in the marine environment are often simplistic, labour intensive and implementation at a large-scale is generally not feasible [6]. In this context, there is growing interest in developing integrated, cost-effective and environmentally-sound marine pest management tools. Biological control (biocontrol), specifically augmentative biocontrol with indigenous agents, stands out as a promising approach, but is one that has not yet been used in natural marine habitats at an operational scale [9], [11]. Augmentative biocontrol can be considered in the context of two recognised strategies [12], [13]. The first is the inoculation approach, in which biocontrol agents are released with the expectation that they will multiply and reduce pest populations for an extended period [12], [14]. The other is an inundation strategy which relies exclusively on shorter term effects of the release of a large number of agents [12].
Critical to the success of augmentative biocontrol are considerations of the ecology, population dynamics and behaviour of specific control agents before their release in natural habitats [15], [16]. This includes assessment of organisms in terms of their persistence, and spread to new habitats or areas [17]. Simultaneously, it is necessary to consider their direct or indirect non-target effects; the release of indigenous natural enemies into novel environments may lead to collateral impacts such as reduced biodiversity and altered ecosystem functioning [18].
Urchins have long been recognised as a strong structuring force of benthic communities throughout coastal areas worldwide [19], [20], [21]. Previously, urchin species have been trialled as biocontrol agents only in small-scale experiments both on artificial structures [22], [23] and in natural habitats [24]. The New Zealand sea urchin, Evechinus chloroticus (hereafter Evechinus), has been used during trials on suspended artificial structures, and has been found to control a wide range of algae and invertebrate species, including several high profile marine pests [25].
More recently, Evechinus was used as part of a multi-agency response in a New Zealand fiord to eradicate an incursion of the non-indigenous Asian kelp Undaria pinnatifida [26], which provided a unique opportunity to evaluate urchin biocontrol at an operational scale. An overview of the eradication programme and its outcomes will be described elsewhere by the agencies involved. In this paper, we describe the persistence and spread of Evechinus that were transplanted using an inundation approach during the eradication effort, and quantify the associated changes in subtidal benthic assemblages at transplant sites. Such knowledge provides information on both the non-target effects of Evechinus as a biocontrol agent, and also its potential efficacy as a control agent for other species. Finally, we consider some of the benefits and limitations of using urchins for managing marine pests in natural habitats.
Methods
Ethic statement
All required permits for this study were obtained from the Department of Conservation and Ministry for Primary Industries. No locations were privately-owned and the studies did not involve endangered or protected species.
Study sites
The study was conducted in Sunday Cove (45°36′S, 166°44′E), a small, shallow embayment of approximately 2.1 ha surface area in Breaksea Sound, Fiordland (Fig. 1). Fiordland is subjected to high rainfall (5300–6300 mm yr−1) that creates a permanent low-salinity surface layer (<30 ‰) of variable depth in the fiords, but is typically 1–2 m deep in outer fiord areas such as Sunday Cove [27]. This low-salinity layer results in stratified habitats, with the surface layer dominated by euryhaline organisms. Deeper areas are characterised by dense ‘kelp forest’ habitats dominated by large canopy forming brown macroalgae, such as Ecklonia radiata and Carpophyllum flexuosum (hereafter Ecklonia and Carpophyllum, respectively), and diverse understory assemblages [28], [29]. At mid-depths sea urchins may remove large areas of kelp forest and form habitats termed ‘urchin barrens’, dominated by encrusting algal forms.
Figure 1. Map of the study sites.
Filled circles denote ‘Control’, empty circles ‘Adjacent’, empty squares ‘No Canopy’ and filled squares ‘Canopy’ sites
During the eradication programme, a total of 30,000–35,000 Evechinus individuals (mean test diameter = 11.3 cm, ±1.5 SE) were collected [26] by divers from sites in the outer reaches of Breaksea Sound. They were transported by boat in ∼70 L self-draining plastic containers within ∼1 h to four treatment sites: two ‘Canopy’ and two ‘No Canopy’ sites. These sites were in established kelp forest habitats at depths of 5–7 m within Sunday Cove (Fig. 1). The plastic containers were then submerged by divers and Evechinus were manually placed within each of the treatment sites. Sampling sites were then established in a circular formation at each of the treatment sites, with sampling occurring within a 5 m radius of a weighted central marker. At two of the sites (No Canopy sites), the algal canopy was removed before adding Evechinus, while at the other two sites (Canopy sites) the algal canopy was left intact. The treatment sites (i.e. both Canopy and No Canopy) were chosen based on previous locations of mature Undaria pinnatifida individuals found during the eradication programme [26]. Two ‘Adjacent’ sites within Sunday Cove (<50 m from the treatment sites) and two ‘Control’ sites outside of Sunday Cove (>500 m from the treatment sites) were also selected (Fig. 1). Adjacent sites were established to assess the effects on areas contiguous to the treatment sites, as a result of the anticipated urchin spread. Sites were chosen so that they were geographically interspersed and, when possible, chosen randomly from a larger pool of candidate sites. Evechinus were transplanted to the No Canopy sites before the first sampling occasion, resulting in a mean density of 52.5 ind.·m−2 (±2.7 SE). Whereas, at the two Canopy sites, Evechinus was added in August 2011 (after the first sampling occasion), resulting in a mean density of 51.9 ind.·m−2 (±2.8). Natural mean densities of Evechinus at the Adjacent and Control sites were 0.1 (±0.1) and 0.3 (±0.2) ind.·m−2, respectively.
Sampling
Sites were sampled on four occasions: in August 2011 immediately before Evechinus were transplanted to the two Canopy sites, then again in October 2011, April 2012 and December 2012. Counts of stipes of Ecklonia and Carpophyllum, as well as Evechinus individuals were obtained from ten random 1 m2 quadrats at each of the eight sites. To sample the understory community, eight haphazardly placed 0.24 m2 photo–quadrat images were taken with a digital underwater camera. Before taking each photograph, blades of large brown algae and individual Evechinus were moved outside each quadrat so that the understory was visible. At the two Canopy sites, the spread of Evechinus from the initial seeding location was recorded by measuring their densities within 1 m2 quadrats (n = 5) at three locations: the site perimeter (edge) and two distances beyond (5 and 10 m).
The photo-quadrat images were analysed using the Coral Point Count software [30], with 50 stratified random points overlaid on each image. All sessile taxa >1 mm in size were identified to the lowest possible taxonomic level and their percentage cover estimated. Taxa that could not be identified at species or genus level were grouped by morphological criteria (Table 1).
Table 1. List of taxa (not including Evechinus) recorded on the photo quadrats at the study sites. Column ‘Group’ refers to variable names used in Fig. 5.
Taxon | Category | Group |
Echnidermata | ||
Patiriella regularis | Echinoderm | Asteroidea |
Australostichopus mollis | Echinoderm | Holothuroidea |
Holothurian unid. | Echinoderm | Holothuroidea |
Brachiopoda | ||
Terebratella sanguinea | Terebratella sanguinea | Brachiopoda |
Porifera | ||
Cymbastela tricalyciformis | Sponge | Demospongiae |
Crella incrustans | Sponge | Demospongiae |
Dysidea sp. | Sponge | Demospongiae |
Haliclona sp. | Sponge | Demospongiae |
Latrunculia fiordensis | Sponge | Demospongiae |
Encrusting sponge unid. | Sponge | Demospongiae |
Erect sponge unid. | Sponge | Demospongiae |
Strongylacidon conulosum | Sponge | Demospongiae |
Tethya sp. | Sponge | Demospongiae |
Cnidaria | ||
Hydroid unid. | Hydroid | Hydrozoa |
Anthothoe albocincta | Anthozoan | Anthozoa |
Corynactis australis | Anthozoan | Anthozoa |
Chlorophyta | ||
Caulerpa brownii | Green algae | Bryopsidales |
Codium gracile | Green algae | Bryopsidales |
Green filamentous | Green algae | Ulvaceae |
Ulva spp. | Green algae | Ulvaceae |
Phaeophycea | ||
Carpophyllum flexuosum | Brown canopy | Fucales |
Ecklonia radiata | Brown canopy | Laminariales |
Brown filamentous algae | Brown filamentous | Phaeophyceae |
Carpomitra costata | Brown herbaceous | Sporochnales |
Dictyota spp. | Brown herbaceous | Dictyotales |
Halopteris spp. | Brown herbaceous | Leptothecata |
Zonaria turneriana | Brown herbaceous | Littorinimorpha |
Microzonia velutina | Brown herbaceous | Cutleriales |
Ralfsia verrucosa | Encrusting algae | Ralfsiales |
Rhodophyta | ||
Coralline crustose algae | Coralline crustose algae | Corallinales |
Coralline turfing algae | Coralline turfing algae | Corallinales |
Hildenbrandia spp. | Encrusting algae | Hildenbrandiales |
Red complex foliose algae | Red complex foliose algae | Rhodophyta |
Red filamentous algae | Red filamentous algae | Rhodophyta |
Red laminar foliose algae | Red laminar foliose algae | Rhodophyta |
Bryozoa | ||
Beania sp. | Bryozoan | Bryozoa |
Erect bryozoan | Bryozoan | Bryozoa |
Catenicellid bryozoan | Bryozoan | Bryozoa |
Margaretta barbata | Bryozoan | Bryozoa |
Membranipora membranacea | Bryozoan | Bryozoa |
Tunicata | ||
Cnemidocarpa bicornuta | Cnemidocarpa bicornuta | Tunicata |
Aplidium sp. | Colonial ascidian | Tunicata |
Botrylloides sp. | Colonial ascidian | Tunicata |
Botryllus sp. | Colonial ascidian | Tunicata |
Colonial ascidian | Colonial ascidian | Tunicata |
Didemnidea | Colonial ascidian | Tunicata |
Diplosoma spp. | Colonial ascidian | Tunicata |
Mollusca | ||
Cellana radians | Gastropod | Mollusca |
Gastropod unid. | Gastropod | Mollusca |
Mytilus galloprovincialis | Bivalve | Mollusca |
Polychaeta | ||
Galeolaria hystrix | Polychaeta | Polychaeta |
Spirobis spp. | Polychaeta | Polychaeta |
Other | ||
Bare rock | Inorganic | Other |
Sand | Inorganic | Other |
Shell hash | Inorganic | Other |
Biofilm | Organic | Other |
Detritus | Organic | Other |
Other organic | Organic | Other |
Data analyses
To investigate treatment effects, data were analysed with Population-Averaged Generalized Estimating Equations (PA-GEEs), an extension of Generalized Linear Models for correlated data [31]. This procedure was used because sites were sampled repeatedly through time and, although randomly positioned quadrats were sampled at each time, data could not be treated as independent in conventional analyses. PA-GEEs provide estimates of model parameters and standard errors that take into account correlated observations to avoid spurious correlations. Additionally, PA-GEEs estimate a scale parameter, which is an adjustment for overdispresion. A first order autoregressive AR(1) working correlation matrix was used in all of the analyses, assuming observations from a given site separated by one sampling occasion were likely to be more similar than those separated by longer time periods. Models were fitted for data on densities of Evechinus, Ecklonia and Carpophyllum using a Poisson error distribution and a log-link, whereas Gaussian errors were used for the number of taxa and Shannon diversity. The number of taxa and Shannon's diversity index were calculated from the data obtained in the photo-quadrats, having first removed organic (i.e. detritus and other organic) and inorganic (i.e. sand, bare rock and shell hash) cover data (Table 1). The experimental design comprised three factors: ‘Treatment’ (fixed, four levels: Canopy, No Canopy, Adjacent, Control), ‘Time’ (fixed, four sampling dates: August 2011, October 2011, April 2012 and December 2012) and ‘Site’ (random, two levels, nested in factor Treatment). Comparisons were based on the Control treatment level and the first sampling occasion (i.e. August 2011) as reference points. For the spread experiment data from the two Canopy sites, the effect of time and distance on Evechinus density was also examined by fitting a PA-GEE model as described above. Time (months since addition) and distance (meters from the edge of each site) were included as covariates in the model, while site was included as a categorical factor. All PA-GEE models were selected using the Quasi Likelihood under the Independence Model criterion (QIC), a modification of the Akaike's Information Criterion for PA-GEEs models [32], and validated by inspecting the deviance residuals. PA- GEEs models were run using the library geepack [33] in the software R [34].
Differences in assemblage structure between experimental treatments were tested using PERMANOVA [35] based on Bray–Curtis similarities of log(x+1) transformed data. The same experimental design was used as described above for the univariate analyses, with factors ‘Treatment’ (fixed, four levels), ‘Time’ (fixed, four levels) and ‘Site’ (random, two levels, nested in factor Treatment). Significant terms were then investigated using a posteriori pair-wise comparisons with the PERMANOVA t statistic and 999 permutations. SIMPER analysis [36] was used to identify the percentage contribution of each species (or taxon) to the observed differences between assemblages in the different treatments. Taxa that consistently discriminated between treatments (ratio of the average dissimilarity and standard deviation ≥3) were displayed as vectors in principal coordinate ordination plots (PCO).
Results
Changes in Evechinus densities
Field observations indicated that Evechinus individuals quickly relocated to suitable areas after being transplanted, moving from sandy, biogenic and vertical (or steeply sloping) habitats into relatively flat rocky areas. Over time, significant reductions in Evechinus densities were observed at both Canopy and No Canopy sites in relation to the initial mean densities of 51.9 (±2.8 SE) and 52.5 (±2.7) ind.·m−2, respectively (Fig. 2A, Table 2). At these sites, densities rapidly declined between October 2011 and April 2012, but in December 2012 had stabilised at mean values of approximately 5 ind.·m−2. Evechinus densities at the Adjacent sites increased from mean values of 0.15 (±0.1) and 0.45 (±0.3) ind.·m−2 in August and October 2011, to 4.95 (±1.16) and 6.0 (±1.26) ind.·m−2 in April and December 2012, respectively (Fig. 2A, Table 2). Urchin densities at the Control sites remained stable throughout the study at <5 ind.·m−2 (Fig. 2A). Overall, there was no significant spatial variation among sites within each treatment (Table 2).
Figure 2. Changes in sea urchin density through time for each A) treatment and B) distance.
Data were pooled across sites (n = 20).
Table 2. Results of Population-Averaged Generalized Estimating Equations (PA-GEEs) examining the effects of time and treatment on the density of Evechinus, Ecklonia and Carpophyllum.
Evechinus | Ecklonia | Carpophyllum | Number of taxa | Shannon diversity | |||||||||||
Estimate | SE | P | Estimate | SE | P | Estimate | SE | P | Estimate | SE | P | Estimate | SE | P | |
Site | 0.11 | 0.07 | 0.62 | 0.19 | ** | −0.04 | 0.13 | 1.07 | 0.50 | * | 0.05 | 0.09 | |||
Oct-11 x Adjacent | 2.27 | 1.10 | −0.10 | 0.31 | 0.25 | 0.22 | 2.54 | 1.22 | * | 0.52 | 0.18 | ** | |||
Apr-12 x Adjacent | 2.13 | 1.17 | * | −0.86 | 0.60 | −0.21 | 0.75 | 0.44 | 1.58 | 0.36 | 0.16 | * | |||
Dec-12 x Adjacent | 2.91 | 0.94 | ** | −1.81 | 0.52 | *** | −1.98 | 0.77 | * | −0.11 | 2.02 | 0.47 | 0.39 | ||
Oct-11 x Canopy | 1.16 | 0.92 | −45.89 | 0.82 | *** | −44.51 | 0.84 | *** | −5.63 | 1.59 | *** | −0.66 | 0.19 | *** | |
Apr-12 x Canopy | −3.64 | 0.94 | *** | −45.64 | 0.88 | *** | −44.28 | 0.84 | *** | −6.09 | 1.66 | *** | −1.00 | 0.20 | *** |
Dec-12 x Canopy | −3.21 | 0.95 | *** | −45.64 | 0.83 | *** | −43.81 | 0.86 | *** | −3.91 | 2.27 | * | −0.10 | 0.39 | |
Oct-11 x No Canopy | 0.73 | 0.94 | −0.04 | 1.00 | −0.35 | 1.02 | 2.84 | 1.93 | 0.25 | 0.46 | |||||
Apr-12 x No Canopy | −3.17 | 0.93 | *** | 0.21 | 1.05 | −0.12 | 1.02 | 3.13 | 2.27 | 0.72 | 0.44 | ||||
Dec-12 x No Canopy | −2.91 | 0.94 | ** | 0.22 | 1.01 | 0.35 | 1.04 | 3.12 | 2.52 | 0.80 | 0.56 | ||||
Scale Parameter | 3.09 | 0.48 | 0.91 | 0.17 | 0.96 | 0.26 | 6.00 | 0.72 | 0.17 | 0.02 | |||||
Correlation Parameter | 0.20 | 0.14 | 0.43 | 0.12 | 0.07 | 0.07 | 0.44 | 0.06 | 0.55 | 0.05 |
Main effects for Treatment and Date are not shown, as all interactions were significant. Treatment effect comparisons are in reference to the controls and the first sampling occasion in August 2011. *P<0.05, **P<0.01 and ***P<0.001.
Reductions in Evechinus densities at the two Canopy transplant sites were reflected in the spread experiment results. The patterns of spread evident from density changes were similar between the two Canopy sites, and more strongly related to time since Evechinus transplant (P<0.001) than distance from the sites (Fig. 2B, Table 3). In August and October 2011, densities at all distances from the Canopy sites were <1.5 ind.·m−2. However, concomitant with the reduced densities at Canopy sites in April 2012 (see Fig. 2A), there was an outward movement of Evechinus from the point of initial transplant, reflected by increased densities at the spread sites (e.g. ∼20 ind.·m−2 at the 10 m site) (Fig. 2A). By December 2012, densities had decreased to mean values between 4.6 (±0.7) and 7.7 (±1.2) ind.·m−2 for all distances (Fig. 2B), which were comparable to the residual densities at the Canopy sites (Fig. 2A).
Table 3. Population-Averaged Generalized Estimating Equations (PA-GEEs) results examining the effects of time, distance and site on Evechinus density. Log-link and Poisson errors. **P<0.01.
Estimate | SE | P | |
Distance | −0.08 | 0.07 | |
Time | 0.07 | 0.01 | ** |
Site | −0.28 | 0.16 | |
Distance x Time | 0 | 0.01 | |
Scale Parameter | 11.7 | 1.57 | |
Correlation Parameter | 0.64 | 0.05 |
Effects of Evechinus on canopy forming algae
Mean densities of Ecklonia at the beginning of the experiment (August 2011) were approximately 5 ind.·m−2 for all treatments, with the exception of the No Canopy sites where the canopy had been manually removed by divers prior to Evechinus transplant (Fig. 3A). After Evechinus transplant to the Canopy sites (i.e. August 2011), there were significant reductions in Ecklonia densities at the Canopy and Adjacent sites over time (Fig. 3A, Table 2). Canopy sites were completely denuded of Ecklonia by the first resampling (i.e. October 2011). However, at the Adjacent sites there was a gradual decline in Ecklonia over time to 0.7 ind.·m−2 in December 2012 (Fig. 3A), simultaneous with the gradual increase in Evechinus densities at those sites (Fig. 2A). Additionally, there was a significant variation in densities of Ecklonia between sites within given treatments (Table 2).
Figure 3. Treatment effect through time on density of A) Ecklonia radiata and B) Carpophyllum flexuosum.
Data were pooled across sites (n = 20).
The spatio-temporal pattern of density change for Carpophyllum was comparable to that described for Ecklonia. Mean densities of Carpophyllum at the beginning of the study ranged from 0.5 to 1.8 ind.·m−2 for all treatments except the No Canopy sites where the algae were manually removed (Fig. 2B). Densities at the Control sites remained relatively stable throughout the study, whereas no individuals were recorded at the Canopy sites from October 2011 onwards (Fig. 3B, Table 2). At the Adjacent sites there was a small increase in Carpophyllum densities in October 2011, after which densities gradually declined to zero by December 2012.
Changes in the number of taxa, diversity and structure of understory assemblages
Significant changes in the mean number of taxa through time were observed in response to the urchin treatments, with Canopy and No Canopy sites having approximately half the number of taxa recorded in the Control and Adjacent sites from October 2011 onwards (Fig. 4A, Table 2). At the start of the experiment in August 2011, the mean number of taxa at the Control, Adjacent and Canopy sites was 11.9, 12.0 and 13.1 (±0.6, 0.8 and 0.5, respectively). In contrast, at the No Canopy sites where Evechinus had already been transplanted, the mean number of taxa was 4.6 taxa (±0.4). In October 2011, the mean number of taxa at the Canopy sites was significantly reduced to 4.0 (±0.5), while no significant changes were observed in the other treatments (Table 2). Patterns observed both in April and December 2012 were similar to that described for October 2012, with significantly fewer taxa at both Canopy and No Canopy sites compared to the Control and Adjacent sites (Fig. 4A, Table 2). In December 2012, both of the urchin-treated sites displayed a slight increase in their mean number of taxa compared to April 2012, but this trend was not significant. This pattern was confirmed by a similar trend observed for the Shannon's diversity index (Fig. 4B, Table 2); both Canopy and No Canopy sites had lower diversity in October 2011 and April 2012 compared to the controls, which then increased in December 2012 (Fig. 4B).
Figure 4. Treatment effect through time on A) mean number of taxa and B) Shannon diversity.
Data were pooled across sites (n = 16).
Significant shifts in understory community structure were observed in response to the treatments over time, as indicated by a significant Treatment x Time interaction (P<0.05, Table 4), despite significant spatial variability at the site-scale within treatments (P<0.001, Table 4). The PCO plots illustrate the overall pattern among treatments for each of the sampling dates (Fig. 5). At the beginning of the study in August 2011, the understory assemblages at the Control, Adjacent and Canopy sites (before Evechinus addition) did not differ significantly (pair-wise comparison P>0.05). Assemblages were characterised by a mixture of large canopy forming brown algae, turfing coralline algae, red foliose algae, small brown herbaceous algae (both filamentous and foliose forms), green algae, colonial ascidians and bryozoans. These taxa collectively accounted for >75% of observed similarities among these sites. In contrast, the No Canopy sites were significantly different to all other treatments (pair-wise comparison P<0.05). These sites resembled an urchin barren, being largely dominated by crustose algae, echinoderms (other than Evechinus; see Table 1), organic and inorganic matter and the solitary ascidian Cnemidocarpa bicornuta.
Table 4. PERMANOVA of differences in benthic assemblage structure based on Bray-Curtis dissimilarities of log transformed data (using 9999 permutations).
Source | df | MS | F | P |
Treatment | 3 | 52999 | 6.92 | ** |
Time | 3 | 18060 | 6.25 | *** |
Site (Treatment) | 4 | 7653.5 | 14.75 | *** |
Treatment x Time | 9 | 7885.2 | 2.73 | *** |
Time x Site (Treatment) | 12 | 2888.3 | 5.56 | *** |
Residual | 224 | 519.02 |
Figure 5. Principal coordinates ordination (PCO) biplots illustrating treatment effects on the understory macrobenthic assemblages.
Plots are based on Bray-Curtis similarities of log(x+1) transformed data
During the second sampling occasion (October 2012) both Canopy and No Canopy sites exhibited urchin barren characteristics. These sites were dominated by crustose algae, echinoderms (e.g. sea stars), the brachiopod Terebratella sanguinea and both organic and inorganic matter, which together accounted for >90% of their similarity (Fig. 5). In contrast, Adjacent and Control sites remained characterised by kelp forest assemblages and were significantly different to the treated sites (pair-wise comparison P<0.05). A large proportion of the similarity of the Adjacent and Control sites was accounted for by red foliose algae, large brown fucoid algae, crustose algae, small brown filamentous and foliose algae, coralline turf, and suspension feeding invertebrates such as sponges and bryozoans.
By the third sampling occasion (April 2012), the Control sites remained significantly distinct from the urchin treated sites (pair-wise comparison P<0.05), but assemblages at the Adjacent sites (where Evechinus densities had increased) were transitional between the control and treated sites. The substratum cover at Adjacent sites had similarities with an urchin barren, and was characterised by bare rock, organic detritus, crustose algae, coralline turf, Cnemidocarpa bicornuta and encrusting sponges (Fig. 5). Nonetheless, the Adjacent sites still had a diverse complement of species that was comparable to the controls, highlighting that the patterns in Fig. 5 reflect trends in taxon structure rather than composition. Assemblages in the Canopy and No Canopy sites continued to be largely dominated by crustose algae, inorganic and organic matter and echinoderms, such as the sea cucumber Australostichopus mollis and the seastar Patiriella regularis. A similar pattern was observed during the last sampling occasion (December 2012), with only the Control sites differing to the treated sites (Fig. 5, pair-wise comparison P<0.05), but assemblages at the Adjacent sites were marginally distinct to the treated sites (pair-wise comparison P<0.1). The Control sites remained dominated by kelp forest assemblages, while the Canopy and No Canopy sites were characterised by urchin barren assemblages (crustose algae, organic and inorganic matter and Cnemidocarpa bicornuta). However, a relatively larger cover of red and green ephemeral algae was recorded in these sites compared to previous sampling occasions.
Discussion
Changes in Evechinus densities
Over the course of the study Evechinus densities declined by one order of magnitude at the treated sites, from ∼50 to 5 ind.·m−2. Adjacent sites had similar urchin densities to the treated sites after nine months. Natural densities of Evechinus are variable, although in urchin barrens these are typically around 5 ind.·m−2 [37], [38], [39]. Nonetheless, densities of over 50 ind.·m−2 have been documented in natural systems [40], which is comparable to the enhanced transplant densities in the present study. Evechinus can adapt to situations of high conspecific densities and associated low food availability, through adjustment in growth and reproductive output [41]. As such, they can persist at high densities (e.g. 5 ind.·m−2) for many years and be replenished by recruitment of new individuals. Recruitment of new Evechinus was not observed during the course of this study (authors' personal observation). Additionally, increased recruitment in response to locally enhanced adult densities would not be expected, as recruitment dynamics in this species are largely influenced by larval advection and other abiotic factors that operate at the whole-fiord scale [42].
The temporal reduction in densities within the treated sites is consistent with the spread experiment, which confirmed that Evechinus moved >10 m away from the edge of the transplant areas in a relatively short period of time (<9 months). Based on tagged individuals, Dix [40] concluded that Evechinus individuals dispersed over maximum distances of 5 m in six months, and hypothesised that their movement was related to food availability. The relatively higher spread observed in the present study is possibly related to the scarcity of food at the treated sites, where very high Evechinus densities quickly created barren areas. Biocontrol agents with low spread rates may necessitate the redistribution of individuals or a high density of release for effective control to be exerted. On the other hand, high rates of spread may lead to a reduction in densities in target areas to a point where control is ineffective. Given such issues, it has been argued that the use of species with intermediate rates of spread is likely to maximize the probability of successful augmentative biocontrol [17].
Effects of Evechinus on benthic assemblages
At the treated sites, Evechinus rapidly (in <3 months) grazed the canopy forming algae (Ecklonia and Carpophyllum) and most of the understory assemblages, including turfing, foliose and filamentous algae, bryozoans, colonial ascidians and sponges. Evechinus preferentially graze on laminarian algae, specifically Ecklonia [43], but can consume a wide range of other algal and invertebrate species at relatively high rates [37], [43].
The low-richness urchin barren assemblage that resisted Evechinus grazing pressure included encrusting algal forms, and to some extent the brachiopod Terebratella sanguinea and the solitary ascidian Cnemidocarpa bicornuta. These findings are consistent with the well-recognised role of Evechinus as a primary driver of community structure of shallow subtidal rocky reef assemblages in New Zealand [37], [39], [41], [43], [44]. Elsewhere, increased abundances of Evechinus are strongly related to reductions in canopy forming algae, understory algae and encrusting invertebrates [37]. In this respect, it is of interest that Evechinus densities at Adjacent sites were similar to treated sites at the end of the study (almost certainly a reflection of spread from the treated sites), yet the range of taxa remained comparable to controls. This contrast likely reflects the different site histories; despite Adjacent and treated sites reaching comparable Evechinus densities, the Adjacent sites never experienced the intense grazing pressure that would have resulted from the initial high density urchin transplant at treated sites.
Broader perspective on benthic effects
The dramatic changes to benthic assemblage structure highlights the potential for substantial non-target effects associated with generalist biocontrol agents. This issue is particularly relevant in areas of high conservation value such as in the present study. Some trepidation around the use of biocontrol has resulted from early experiences, particularly classical biocontrol in terrestrial systems, where procedures were less regulated and non-target effects were less considered [16]. It is common for marine pest management programmes (e.g. physical or chemical treatments) to be implemented with a lack of proven and cost-effective control tools, on the basis that prompt response will give the greatest chance of success [5], [6], [45]. In such instances, ecological risks are likely to be relatively unimportant; for example, perhaps at worst leading to local-scale effects that are reversible. By contrast, the use of marine biocontrol agents that have not been exhaustively evaluated in terms of their efficacy and non-target effects is probably not advisable.
Crucial to the assessment of non-target effects is the magnitude and spatio-temporal scale of the impacts. The choice of Evechinus as a biocontrol agent was based largely on previous ecological knowledge of the species (e.g. high consumption rates, generalist feeding habits and wide distribution). Relevant to the issue of non-target effects in the present study is that Evechinus were transplanted from urchin barrens within the same fiord as the transplant sites. In this sense, the regional population remained largely unchanged; it was simply redistributed. Although the non-target effects of transplant were dramatic, and occurred outside of the initial transplant sites (e.g. at Adjacent sites), they could be reversed by the removal or destruction of Evechinus from treated areas. However, based on Evechinus longevity (10–15 years) and growth rates (1–2 cm yr−1) [46], it can be expected that the transplanted population will likely decline over the subsequent years. Additionally, the continued spread of Evechinus will likely reduce densities to levels that enable large brown algae and understory assemblages to re-establish in the longer term. Existing studies suggest that the threshold density for recovery is <2 ind.·m−2 [47]. Based on these considerations, it is expected that the biocontrol effects of Evechinus will be localised and reversible. In natural systems, shifts from urchin barren to kelp forest have also been related to the recovery of sea urchin predators [48], mass sea urchin mortality due to storms [49] and diseases [50], no-take marine reserves protection [39], [44] and harvesting [47].
As well as the spatial and temporal scales of non-target effects, distinction also needs to be made in terms of the species that are affected by biocontrol [18]. Impacts on species should not be considered equally; special consideration should be given to endangered species, species that provide crucial ecosystem services as well as species of commercial value [51]. The taxonomic resolution described in the present study does not enable such issues to be fully addressed. At a broad level, the assemblages found at the study sites were comparable to those inhabiting similar habitats (i.e. kelp forest) in the wider area [28], [29], [52]. However, it is clear that ecologically important species were adversely affected as a result of the Evechinus transplants. For example, Ecklonia is an ecosystem engineer that provides habitat, nurseries and food (both directly and through detrital additions to the food chain) to many other species [53], [54]. However, in the context of marine pest management, the significance of localised non-target effects from biocontrol must be weighed up against the potentially irreversible and regional-scale effects resulting from the establishment of an invasive species.
Broader consideration of the utility of urchins in biocontrol
Despite the potential for non-target effects, the present study illustrates the potential efficacy of augmentative biocontrol. Such approaches, when used in conjunction with more traditional management measures [55] may provide a useful way forward for future invasive species responses. For example, diver-based pest response programmes would benefit from the use of generalist biocontrol agents that denude canopy, by aiding visual searches. This benefit may prove crucial to the success of marine pest eradication attempts, as one of the primary problems with manual diver removal is the failure to detect individuals before they successfully reproduce [56], [57].
It is conceivable that a similar biocontrol approach could be applied in other regions, using the same or different urchin species, or in fact other generalist grazers or predators. The use of such taxa is not only restricted to the control of algae, but can clearly be extended to a range of other pests, both invertebrates and algae. For example, Evechinus efficiently removed colonial ascidians; although this was a non-target effect in the present context, it is relevant in a wider context that ascidians are an invertebrate group that can show invasive traits [58].
For Evechinus specifically, applications are restricted to mainly hard bottom substrata, because the species has limited grazing activity in soft bottom and biogenic habitats [40]. These habitats may constitute a refuge for the settlement of invasive algae. Similarly, high substrate rugosity or biogenic complexity may create important refuges, where invasive species may escape urchin grazing pressure. This situation was evident in the present study, and may have contributed to persistence of some species that would otherwise have been vulnerable. In the context of the study region itself, vertical walls were common in the shallowest zones of the treated sites, and the wider area was subject to variable or reduced salinity in surface layers that likely created conditions that urchins could not tolerate [59].
More broadly, augmentative biocontrol approaches are likely to have greatest efficacy and practicality in local-scale applications that target spatially restricted areas. It would not be practical to use this strategy for pests that were geographically dispersed, certainly because of logistical constraints, and perhaps because of unacceptable large-scale non-target effects. Additionally, augmentative biocontrol is perhaps more suited to the purpose of population control rather than eradication, except where used alongside a broader suite of tools [60]. A final consideration is that, because the settlement and dispersal of invasive species can be facilitated by disturbance [61], urchin barrens have been suggested as favourable habitats for being colonised by marine pests [62]. This could result in a counteractive effect in which the use of biocontrol enhanced, rather than controlled, the spread of a target pest. For example, in Tasmania the invasive alga Undaria pinnatifida can successfully recruit and persist in urchin barren areas [63]. In this context, it is crucial that monitoring is continued in treated areas after the cessation of an eradication programme, or the potential removal of the control agents is considered.
Acknowledgments
Thanks to Richard Kinsley and Kath Blakemore (Department of Conservation), Jennie Brunton and Tim Riding (Ministry for Primary Industries) and Environment Southland for providing logistic support for this project. Thanks to the Southern Winds' crew. Emma Newcombe helped with image analyses. For comments on early version of the manuscript we thank Lauren Fletcher and Emma Newcombe.
Funding Statement
This research was funded by National Institute of Water and Atmospheric Research (NIWA) under Coasts and Oceans Research Programme 4 – Marine Biosecurity (2012/13 SCI). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
References
- 1. Ruiz GM, Fofonoff P, Hines AH, Grosholz ED (1999) Non-indigenous species as stressors in estuarine and marine communities: Assessing invasion impacts and interactions. Limnol Oceanogr 44: 950–972. [Google Scholar]
- 2. Molnar JL, Gamboa RL, Revenga C, Spalding MD (2008) Assessing the global threat of invasive species to marine biodiversity. Front Ecol Environ 6: 485–492. [Google Scholar]
- 3. Strayer DL, Eviner VT, Jeschke JM, Pace ML (2006) Understanding the long-term effects of species invasions. Trends Ecol Evol 21: 645–651. [DOI] [PubMed] [Google Scholar]
- 4. Edelist D, Rilov G, Golani D, Carlton JT, Spanier E (2013) Restructuring the Sea: profound shifts in the world's most invaded marine ecosystem. Divers Distrib 19: 69–77. [Google Scholar]
- 5. Coutts ADM, Forrest BM (2007) Development and application of tools for incursion response: lessons learned from the management of the fouling pest Didemnum vexillum . J Exp Mar Biol Ecol 342: 154–162. [Google Scholar]
- 6. Hopkins GA, Forrest BM, Jiang W, Gardner JPA (2011) Successful eradication of a non-indigenous marine bivalve from a subtidal soft-sediment environment. J Appl Ecol 48: 424–431. [Google Scholar]
- 7. Carver CE, Chisholm A, Mallet AL (2003) Strategies to mitigate the impact of Ciona intestinalis (L.) biofouling on shellfish production. J Shellfish Res 22: 621–631. [Google Scholar]
- 8. Piola RF, Dunmore RA, Forrest BM (2010) Assessing the efficacy of spray-delivered "eco-friendly" chemicals for the control and eradication of marine fouling pests. Biofouling 26: 187–203. [DOI] [PubMed] [Google Scholar]
- 9. Lafferty KD, Kuris AM (1996) Biological control of marine pests. Ecology v77: p1989 (1912).. [Google Scholar]
- 10. Coquillard P, Thibaut T, Hill DRC, Gueugnot J, Mazel C, et al. (2000) Simulation of the mollusc Ascoglossa Elysia subornata population dynamics: application to the potential biocontrol of Caulerpa taxifolia growth in the Mediterranean Sea. Ecol Model 135: 1–16. [Google Scholar]
- 11. Anderson LWJ (2007) Control of invasive seaweeds. Bot Mar 50: 418–437. [Google Scholar]
- 12. Eilenberg J, Hajek A, Lomer C (2001) Suggestions for unifying the terminology in biological control. Biocontrol 46: 387–400. [Google Scholar]
- 13.Hajek AE (2004) Natural Enemies: An Introduction to Biological Control: Cambridge University Press.
- 14. Secord D (2003) Biological control of marine invasive species: cautionary tales and land-based lessons. Biol Invasions 5: 117–131. [Google Scholar]
- 15. Murdoch WW, Briggs CJ (1996) Theory for biological control: Recent developments. Ecology 77: 2001–2013. [Google Scholar]
- 16. Thomas MB, Willis AJ (1998) Biocontrol - risky but necessary? Trends Ecol Evol 13: 325–329. [DOI] [PubMed] [Google Scholar]
- 17. Heimpel G, Asplen M (2011) A ‘Goldilocks’ hypothesis for dispersal of biological control agents. Biocontrol 56: 441–450. [Google Scholar]
- 18. Simberloff D, Stiling P (1996) How risky is biological control? Ecology 77: 1965–1974. [Google Scholar]
- 19. Lawrence JM (1975) On the relationships between marine plants and sea urchins. Oceanogr Mar Biol Annu Rev 13: 213–286. [Google Scholar]
- 20. Steneck RS, Graham MH, Bourque BJ, Corbett D, Erlandson JM, et al. (2002) Kelp forest ecosystems: biodiversity, stability, resilience and future. Environ Conserv 29: 436–459. [Google Scholar]
- 21. Poore AGB, Campbell AH, Coleman RA, Edgar GJ, Jormalainen V, et al. (2012) Global patterns in the impact of marine herbivores on benthic primary producers. Ecol Lett 15: 912–922. [DOI] [PubMed] [Google Scholar]
- 22. Lodeiros C, García N (2004) The use of sea urchins to control fouling during suspended culture of bivalves. Aquaculture 231: 293–298. [Google Scholar]
- 23. Switzer SE, Therriault TW, Dunham A, Pearce CM (2011) Assessing potential control options for the invasive tunicate Didemnum vexillum in shellfish aquaculture. Aquaculture 318: 145–153. [Google Scholar]
- 24. Conklin EJ, Smith JE (2005) Abundance and spread of the invasive red algae, Kappaphycus spp., in Kane'ohe Bay, Hawai'i and an experimental assessment of management options. Biol Invasions 7: 1029–1039. [Google Scholar]
- 25.Atalah J, Forrest BM (2011) Developing Augmentative Biocontrol Tools for Marine Fouling Assemblages. 7th International Conference on Marine Bioinvasions, Barcelona. Available from: http://wwwicmbinfo/. Accessed 2011 Dec 5.
- 26.Fiordland Marine Guardians website (2011) Annual Report for the year ended 30 June 2011. Available from http://www.fmg.org.nz/sites/default/files/fmg-annual-report-2011.pdf. Accessed 2013 Mar 3.
- 27. Gibbs MT, Bowman MJ, Dietrich DE (2000) Maintenance of near-surface stratification in Doubtful Sound, a New Zealand fjord. Estuar Coast Shelf Sci 51: 683–704. [Google Scholar]
- 28.Nelson WA, Villouta E, Neill KF, Williams GC, Adams NM, et al.. (2002) Marine macroalgae of Fiordland, New Zealand. Tuhinga: 117–152.
- 29.Wing SR, Jack L (2010) Biological Monitoring of the Fiordland (Te Moana o Atawhenua) Marine Area and Fiordland's Marine Reserves. Department of Conservation. pp. 94.
- 30. Kohler KE, Gill SM (2006) Coral Point Count with Excel extensions (CPCe): A Visual Basic program for the determination of coral and substrate coverage using random point count methodology. Comput Geosci 32: 1259–1269. [Google Scholar]
- 31. Liang KY, Zeger SL (1986) Longitiudinal data analysis using generalized linear models. Biometrika 73: 13–22. [Google Scholar]
- 32. Pan W (2001) Akaike's information criterion in generalized estimating equations. Biometrics 57: 120–125. [DOI] [PubMed] [Google Scholar]
- 33. Halekoh U, Hojsgaard S, Yan J (2006) The R Package geepack for Generalized Estimating Equations. J Stat Softw 15: 1–11. [Google Scholar]
- 34.R Development Core Team (2013) R: A language and environment for statistical computing. Vienna, Austria. R Foundation for Statistical Computing. Available from: http://www.R-project.org/. Accessed 2012 Jun 3.
- 35. Anderson MJ (2001) A new method for non-parametric multivariate analysis of variance. Austral Ecol 26: 32–46. [Google Scholar]
- 36. Clarke KR (1993) Non-parametric multivariate analyses of changes in community structure. Aust J Ecol 18: 117–143. [Google Scholar]
- 37. Ayling AM (1981) The role of biological disturbance in temperate subtidal encrusting communities. Ecology 62: 830–847. [Google Scholar]
- 38. Choat JH, Schiel DR (1982) Patterns of distribution and abundance of large brown algae and invertebrate herbivores in subtidal regions of northern New Zealand. J Exp Mar Biol Ecol 60: 129–162. [Google Scholar]
- 39. Shears NT, Babcock RC (2003) Continuing trophic cascade effects after 25 years of no-take marine reserve protection. Mar Ecol Prog Ser 246: 1–16. [Google Scholar]
- 40. Dix TG (1970) Biology of Evechinus chloroticus (Echinoidea: Echinometridae) from different localities. N Z J Mar Freshw Res 4: 267–277. [Google Scholar]
- 41. Andrew NL (1988) Ecological aspects of the common sea urchin, Evechinus chloroticus, in northern New Zealand: a review. N Z J Mar Freshw Res 22: 415–426. [Google Scholar]
- 42. Lamare MD (1998) Origin and transport of larvae of the sea urchin Evechinus chloroticus (Echinodermata: Echinoidea) in a New Zealand fiord. Mar Ecol Prog Ser 174: 107–121. [Google Scholar]
- 43. Schiel DR (1982) Selective feeding by the echinoid, Evechinus chloroticus, and the removal of plants from subtidal algal stands in northern New Zealand. Oecologia 54: 379–388. [DOI] [PubMed] [Google Scholar]
- 44. Babcock RC, S Kelly, Shears NT, Walker JM, Willis TJ (1999) Changes in community structure in marine reserves. Mar Ecol Prog Ser 189: 125–134. [Google Scholar]
- 45. Bax N, Carlton JT, Mathews-Amos A, Haedrich RL, Howarth FG, et al. (2001) The control of biological invasions in the world's oceans. Conserv Biol 15: 1234–1246. [Google Scholar]
- 46. Dix T (1972) Biology of Evechinus chloroticus (Echinoidia: Echinometridae) from different localities. N Z J Mar Freshw Res 6: 48–68. [Google Scholar]
- 47.Villouta E (2000) Potential ecological impacts of harvesting kina (Evechinus chloroticus) in Fiordland. Wellington: Conservation Advisory Science Notes No. 286. Department of Conservation.
- 48. Estes JA, Duggins DO (1995) Sea otters and kelp forests in Alaska - generality and variation in a community ecological paradigm. Ecol Monogr 65: 75–100. [Google Scholar]
- 49. Ebeling AW, Laur DR (1988) Fish populations in kelp forests without sea otters: effects of severe storm damage and destructive sea urchin grazing. Ecol Stud 65: 169–191. [Google Scholar]
- 50. Scheibling RE, Hennigar AW (1997) Recurrent outbreaks of disease in sea urchins Strongylocentrotus droebachiensis in Nova Scotia: Evidence for a link with large-scale meteorologic and oceanographic events. Mar Ecol Prog Ser 152: 155–165. [Google Scholar]
- 51.Hopper KR (2001) Research Needs Concerning Non-target Impacts of Biological Control Introductions. Publications from USDAARS/UNL Faculty Paper 478. Available from: http://digitalcommonsunledu/usdaarsfacpub/478. Accessed 2012 Oct 5.
- 52. Shears N, Babcock R (2007) Quantitative description of mainland New Zealand's shallow subtidal reef communities. Science for Conservation 280: 1–128. [Google Scholar]
- 53.Tait LW, Schiel DR (2011) Legacy Effects of Canopy Disturbance on Ecosystem Functioning in Macroalgal Assemblages. PLoS ONE 6. [DOI] [PMC free article] [PubMed]
- 54. Wernberg T, Thomsen MS, Tuya F, Kendrick GA (2011) Biogenic habitat structure of seaweeds change along a latitudinal gradient in ocean temperature. J Exp Mar Biol Ecol 400: 264–271. [Google Scholar]
- 55.Piola RF, Denny CM, Forrest BM, Taylor MD (2009) Marine biosecurity: management options and response tools. In: Williams PA, Clout MN, editors. Invasive Species Management: A Handbook of Principles and Techniques. United Kingdom: Oxford University Press. pp. 205–231. [Google Scholar]
- 56. Hewitt CL, Campbell ML, McEnnulty F, Moore KM, Murfet NB, et al. (2005) Efficacy of physical removal of a marine pest: the introduced kelp Undaria pinnatifida in a Tasmanian Marine Reserve. Biol Invasions 7: 251–263. [Google Scholar]
- 57.Hunt L, Chadderton L, Stuart M, Cooper S, Carruthers M (2009) Results of an attempt to control and eradicate Undaria pinnatifida in Southland, New Zealand, April 1997 - November 2004. Department of Conservation, Invercargill, New Zealand. 48p.
- 58. Dijkstra J, Sherman H, Harris LG (2007) The role of colonial ascidians in altering biodiversity in marine fouling communities. J Exp Mar Biol Ecol 342: 169–171. [Google Scholar]
- 59. Lawrence JM (1975) The effect of temperature-salinity combinations on the functional well-being of adult Lytechinus variegatus (Lamarck) (Echinodermata, Echinoidea). J Exp Mar Biol Ecol 18: 271–275. [Google Scholar]
- 60. Anderson L (2005) California's Reaction to Caulerpa taxifolia: A Model for Invasive Species Rapid Response. Biol Invasions 7: 1003–1016. [Google Scholar]
- 61. Bulleri F, Alestra T, Ceccherelli G, Tamburello L, Pinna S, et al. (2011) Determinants of Caulerpa racemosa distribution in the north-western Mediterranean. Mar Ecol Prog Ser 431: 55–67. [Google Scholar]
- 62. Coma R, Serrano E, Linares C, Ribes M, Díaz D, et al. (2011) Sea Urchins Predation Facilitates Coral Invasion in a Marine Reserve. PLoS ONE 6: e22017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 63. Valentine JP, Johnson CR (2005) Persistence of the exotic kelp Undaria pinnatifida does not depend on sea urchin grazing. Mar Ecol Prog Ser 285: 43–55. [Google Scholar]