SEARCH

SEARCH BY CITATION

Keywords:

  • habitat loss;
  • inhibition;
  • recruitment;
  • regime-shift;
  • restoration;
  • state;
  • turf

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

1.  Policy initiatives that seek to recover lost habitats require the capacity to anticipate and suppress the mechanisms that drive loss. The replacement of forested landscapes by simple landscapes comprising of opportunistic or ‘weedy’ species represents an increasingly common phenomenon across human-dominated systems. The failure of subtidal forests to recover from natural and human disturbance and their ultimate replacement by degraded habitats is recognized globally. The current lack of knowledge on whether such shifts can be reversed jeopardizes considerations of restoration policy within increasingly human-dominated landscapes.

2.  We critically assessed the model that recovery of canopies within remnant kelp forests in degraded landscapes (i.e. turf-forming algae that carpet space) is slower than in adjacent forested landscapes, but may be increased by removing turfs.

3.  After generating experimental disturbance, canopies recovered to their former state within forested landscapes, but not in remnant forests in degraded landscapes. Removal of turfs from spaces between remnant forests, however, enabled canopies to recruit and subsequently develop covers that matched those in remnant forests.

4.  Whilst the supply of canopy-forming propagules to degraded landscapes is likely to decline with gap expansion, we show that improvements to forest resilience and restoration are possible via policies that result in a reduction of turf covers. These results also support the model that regime-shifts need not be a product of synchronized loss, but can occur as a result of reduced rates of canopy-recruitment over broad areas and many years. Indeed, patterns of canopy-loss over several decades redouble attention to the human-mediated conditions that enable turfs to retain space (i.e. elevated nutrient and sediment loads via coastal runoff).

5.Synthesis and applications. We demonstrate that future restoration is a possible outcome of polices that promote ecosystem recovery. In doing so, we reduce uncertainty about policy initiatives that aim to upgrade the recycling potential of wastewater treatment plants (e.g. nearly 45% of South Australia’s metropolitan wastewater) to improve the quality of water needed to restore subtidal forests. Uncertainty about resilience-building and restoration management are redressed by demonstrating that the feedbacks maintaining regime-shifted landscapes are not necessarily permanent.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

The failure of ecosystems to recover from human disturbance has galvanized ecologists around theories relating to restoration and conservation (Hobbs & Norton 1996; Dobson, Bradshaw & Baker 1997). The restoration of previous states or conservation of current conditions requires knowledge of the mechanisms that hasten or resist recovery, especially those that follow disturbance of sufficient size that allows alternate habitats to colonize. The failure of systems to recover can lead to regime-shifts whereby early successional and fast-growing assemblages may become permanent (Duarte et al. 2009). Knowledge of the interactions between human-induced stressors and natural processes are needed for the development of models that more accurately predict conditions whereby humans can transform acute disturbances into chronic stress (Nystrom, Folke & Moberg 2000) and identify the circumstances in which regime-shifts can be reversed.

On temperate coasts, there is concern about the permanent replacement of perennial canopy-forming algae (i.e. structurally complex and highly productive habitats), with opportunistic taxa such as filamentous turf-forming algae (i.e. comparatively simple habitats, see review: Krause-Jensen et al. 2008). The replacement of canopy-forming ‘forests’ with ‘turfed’ landscapes, relates to the physiology of algal opportunists that enable them to persist in human-dominated landscapes (e.g. elevated nutrients; Worm et al. 1999). This type of regime-shift has become of global concern, particularly where historically low-nutrient coastal waters have been enriched by subsidies of nutrients from activities that relate to intensified land use (North Sea, Eriksson, Johansson & Snoeijs 2002; Baltic Sea, Österblom et al. 2007; Australia, Connell et al. 2008; Mediterranean Sea, Mangialajo, Chiantore & Cattaneo-Vietti 2008). Whilst the drivers and consequences of anthropogenic forcing have been widely documented (e.g. eutrophication and overfishing; previous citations), models that account for the mechanisms that hasten or resist change are notoriously difficult to test (Scheffer et al. 2001).

Models that account for changes in the ability of aquatic systems to respond to disturbance (e.g. ‘adaptive capacity’; Gunderson 2000) and its consequences for habitat degradation (e.g. loss of forest canopies), generally recognize modified water quality (i.e. increasing nutrient and sediment loads) and biotic interactions (i.e. increased competition) as key drivers of change (reviews: Steneck, Vavrinec & Leland 2004; Connell 2007; Krause-Jensen et al. 2008). These models also acknowledge the mediating influence of biogeography on the relative extent to which eutrophication and overfishing can alter the abundance of species that facilitate or inhibit canopy recruitment (Connell & Irving 2008). Nevertheless, a quite general model in subtidal ecology centres on the important influence of water quality on the maintenance of interactions that facilitate forest recovery.

Models of kelp forest dynamics recognize these communities as dynamic systems that are constantly reshaped by disturbance-recovery cycles (Dayton et al. 1992). The phenomenon whereby forests fail to recover along increasingly human-dominated coastlines relates to coastal activities that modify environmental conditions that facilitate the persistence of turfs which inhibit canopy recruitment (review: Airoldi 2003). In refining this model, current research reveals that the natural seasonal contraction of turf covers (i.e. during winter months) does not occur under conditions (i.e. experimental) and locations (i.e. urban) with sustained levels of elevated nutrients (S.D. Connell, B.D. Russell, D. Gorman & L. Airoldi unpublished). Whilst the ability of turfs and their associated sediments to inhibit the recruitment of canopy formers has been demonstrated (Kennelly 1987a), there remains no quantitative evidence that the removal of turfs results in increased rates of canopy recovery. In the absence of this evidence, models that seek to understand the processes of canopy facilitation and inhibition, as mediated by water quality, tend to be elusive, and the potential for future restoration efforts remain uncertain. The need for this information is particularly urgent for those managing landscapes in which propagule supply is likely to decline with continuing declines in canopy cover; and thereby further reduce forest recruitment.

The research presented in this study redresses this information gap, and proceeded in two steps. First, we tested the hypothesis that rates of canopy recovery are indeed substantially less in turf-dominated habitats (hereafter ‘turfed landscapes’) relative to canopy-dominated landscapes (hereafter ‘forested landscapes’). Secondly, we tested the hypothesis that canopy recovery would be enhanced within turfed landscapes if turfs were removed from the spaces between remnant forests.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Location and timing of study

Forested landscapes were characterized by uninterrupted stands of canopy-forming algae reaching ∼40 m diameter and commonly punctuated by small gaps of 1–3 m diameter. Turfed landscapes were characterized by carpets of uninterrupted turf-forming algae reaching ∼20 m diameter and commonly punctuated by small remnant forests of 1–12 m diameter (Fig. 3; Gorman, Russell & Connell 2009). At a continental scale, forested landscapes span thousands of kilometres across Australian coast (Connell & Irving 2008) and can be punctuated by turfed landscapes spanning several hundreds of metres on urbanized coast (e.g. Connell et al. 2008).

image

Figure 3.  Correlation between covers of turf and accumulated sediment post-canopy loss. Open circles = forested landscapes, closed circles = turfed landscapes.

Download figure to PowerPoint

This study was carried out in the Gulf St Vincent of South Australia (see Fig. S1, Supporting information) for which Connell et al. (2008) provide a historical description of urbanization and associated loss of forests. Briefly, this coast was colonized by Europeans in 1836, and as coastal urbanization intensified from the early 1980s, water quality declined along with a reduction in forest canopy covers (up to 70%). Canopy covers were originally similar to non-urbanized coast across the broader region (i.e. ∼100 km) and the broader biogeographic province (i.e. Flindersia; ∼2000 km). Whilst much of the original areal covers of forest have been replaced by turfed habitat, small remnants of forests are common. Turfs are made up of taxonomically diverse, but morphologically similar groups of filamentous algae (e.g. Feldmannia spp., see Russell & Connell 2007; Roberts & Connell 2008). Our experiments were performed on subtidal rock (4–10 m depth) at three forested and three degraded sites selected at random along an 18 km stretch of metropolitan coast based on previous habitat mapping (see Connell et al. 2008). Experiments spanned 12 months (2007–2008) and their initiation was timed to coincide with the onset of storm-driven canopy disturbance (Kennelly 1987b) and winter sporulation (May) of the most abundant canopy-forming algal species (i.e. Cystoseiraceae, Hotchkiss 1999; Ecklonia radiata, Novaczek 1984; most Sargassaceae, Womersley 1987).

Landscape-scale variation in canopy recovery

If the novel environmental conditions that characterize turfed landscapes (i.e. increased covers of turf and accumulated sediments) can indeed reduce rates of recovery, we would expect that the re-establishment of canopies would be slower in these habitats than forest landscapes where turf covers are naturally sparse. To evaluate this hypothesis, we tested whether canopy recovery in the wake of experimental disturbance would be greater within forested than turfed landscapes. ‘Recovery’ was calculated as the difference in percentage cover of canopy-forming algae between pre-clearance covers (i.e. before) and upon termination of the experiment (i.e. after). The experiment was terminated once the percentage cover of canopy-forming algae within all manipulated plots (= 18) did not differ from all adjacent un-manipulated plots (n = 18) across all forested landscapes (see Results). This rule removed the potential influence of seasonal variation in canopy cover on estimates of recovery between landscapes (i.e. ∼8 months for this first component of study).

Experimental disturbance was produced by clearing canopy-forming algae from replicate circular areas of 1·5 m2; a size within the range of natural storm-driven canopy-loss (Kennelly 1987b) and which is sufficiently large to allow turfs to colonize free from canopy scour (Melville & Connell 2001; Wernberg & Connell 2008). Clearances were created in May 2007 and involved cutting algae (i.e. kelp and fucoids >5 cm high) above their holdfast as typically occurs through storm driven losses of canopies (Goodsell & Connell 2005). Replicate experimental plots (= 6) were created at each of three replicate sites nested within forested and turfed landscapes (see Fig. S1, Supporting information). Within turfed landscapes, clearings were carried out within remnant patches of forest (∼5–12 m diameter) that had similar canopy covers to canopy-dominated landscapes (anova: F1,8 = 1·39; = 0·303). Percentage covers of canopy-forming adults that had recruited (i.e. Cystophora spp., Ecklonia radiata, and Sargassum spp.) were quantified within experimental plots after 4 and 8 months using a 1 × 1 m quadrate comprising of 25 random intersects (see Drummond & Connell 2005). The most abundant canopy-former (i.e. Ecklonia radiata) reaches maturity in canopy gaps within 8 months (Kirkman 1981).

A characteristic of filamentous turfs is their ability to accumulate large amounts of sediment (Airoldi 2003), a physical characteristic that inhibits canopy recruitment (Devinny & Volse 1978). We augmented our interpretation of the effects of landscape on canopy recovery by quantifying the temporal evolution of turf covers and sediment accumulation among treatments (i.e. at 0, 4 and 8 months). Turf covers within experimental plots were estimated using (= 3) replicate 25 × 25 cm quadrates (25 random points method; Drummond & Connell 2005) that provided an average measure of percentage cover within each plot. Net sediment accumulation was estimated by collecting loose sediment within a 10 × 10 cm quadrate using a vacuum. In the laboratory the collected sediment was filtered though Whatman™ (Whatman Ltd., Maidstone, UK) (27 cm ø) grade 1 qualitative filter papers using distilled water. The filters and retained filtrate were dried to a constant weight (60 °C for ∼48 h) and reported as g dry wt cm2.

The aim of this experiment was to test for variation in the recovery of canopies among forests and turfed landscapes. Recovery was initially tested using analysis of covariance (ancova), where the main treatment was landscape; replicate sites were nested within landscape type and turf covers were treated as a covariate. ancova initially tested for differences in the slope of the relationship between recovery and turf covers for the two landscapes types. As there was no difference in the slopes describing the relationship between recovery and turfs among landscape types, we were then able to test for variation in recovery rates between landscape types, after accounting for the effect of turf cover.

Can turf removal facilitate canopy recovery?

We tested the prediction that the removal of turfs and associated sediment from turfed landscapes would facilitate the recovery of forests. At each of the three turfed landscapes (Fig. S1, Supporting information) we removed turf and associated sediment from 12 replicate 1 × 1 m plots within turf-covered gaps (∼10 m diameter) using paint scrapers and wire brushes during May 2007. We recognize that this procedure may alter the natural rate of forest recruitment; including the unintended removal of encrusting corallines whose presence may facilitate recruitment (Camus 1994). Our intention in this study was not so much to quantify absolute rates of recruitment among treatments, but to use a procedure that would quantify relative rates of recovery (as per hypothesis), and allow us to assess the applicability of such removal for restoration efforts. If the forthcoming reductions in nutrient discharge (i.e. water recycling initiatives) are to reduce turfs, there is a further need to improve propagule supply by assisting the expansion of remnant forests within degraded turfed landscapes (i.e. removal of turfs for propagule settlement). These experimental manipulations, as well as untouched controls (= 12), were interspersed at distances <1 m from adjacent stands of canopy within turf-dominated gaps at each site. Percentage covers of canopy-forming adults that recruited to the plots (i.e. Cystophora spp., Ecklonia radiata, and Sargassum spp.) were quantified after 12 months using a 1 × 1 m quadrate comprising 25 random intersects (Drummond & Connell 2005).

Recovery was estimated as the difference in percentage cover of canopy-forming algae between pre-clearance covers sampled in remnant forests (i.e. before) and the termination of the experiment (i.e. after). The experiment was terminated after c. 12 months (April 2008); once the percentage cover of canopy-forming algae within all 12 plots of turf removal did not differ from all 12 adjacent forest (∼5–8 m diameter); again to control for any potential influence of seasonal variation. anova was used to compare canopy recovery among orthogonal treatments (i.e. turf removal plots cf. controls) at each replicate-turfed landscape site (Fig. S1, Supporting information).

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Landscape-scale variation in canopy recovery

Patterns of canopy recovery differed among landscape types (Fig. 1). ancova detected no interaction between landscape and turf covers for both sampling periods (ancova; 4 months: F1,24 = 3·58; = 0·070 and 8 months: F1,24 = 0·00; = 0·986), thereby enabling tests of the main effects (Table 1). Negligible variation was evident in the recovery of canopies among turfed landscapes and forests after 4 months, but clear differences emerged after 8 months. Turf covers were generally less (∼30%) and canopy recovery significantly greater within forests compared with turfed landscapes (∼660 %). After 8 months, canopy covers within forests did not differ significantly from temporal controls at the same site, whereas those at turfed sites were less than the equivalent controls (anova: F1,30 = 17·07; = 0·002; Student-Newman-Keuls (SNK) tests on interaction term; turfed: clearances< controls, forests: clearances = controls). We recognize the conservative nature of this comparison (cf. stands in remnants opposed to forested landscapes) and therefore, comparison powerfully demonstrates the validity of the hypothesis (i.e. recovery is substantially slower in regime shifted landscapes).

image

Figure 1.  Recovery of canopy-forming algae within experimental plots of canopy clearance (= 6 clearances per site) within turfed and forested landscapes (n = 3 sites per landscape type) after 8 months. Open circles = forested landscapes, closed circles = turfed landscapes.

Download figure to PowerPoint

Table 1. ancova testing for differences in canopy recovery between landscape types (i.e. three turfed vs. three forested sites) after 4 and 8 months (= 6 plots per site)
Treatmentd.f.MSF-value
  1. Data were square root (+ 1) transformed to meet the assumptions of heterogeneity of variance. Analysis treated ‘landscape’ as fixed, ‘site’ as random and nested within ‘landscape’; and ‘turf covers’ as a continuous covariate. Significance: NS, not significant; > 0·05; **< 0·01.

Canopy recovery after 4 months
 Landscape10·010·000NS
 Site (landscape)433·921·70NS
 Turf covers119·890·996NS
 Residual29281·13 
Canopy recovery after 8 months
 Landscape14415·4517·08**
 Site (landscape)4431·401·82NS
 Turf covers10·270·001NS
 Residual29237·02 

In forests, fucoids comprised a smaller proportion of the total canopy than Ecklonia radiata: untouched treatments (i.e. 98%E. radiata, 2% fucoids) and recovered treatments (62%E. radiata, 38% fucoids). The relative proportions of these taxa varied considerably among untouched canopies in turfed landscapes (i.e. ranging 3–95%E. radiata, and fucoids the remainder) and were similar in the unrecovered treatments.

Understanding variation in the capacity of forests to recover among landscape types may be improved by observations of turf development and sediment accumulation after canopy loss. The development of turfs and associated accumulation of sediments within treatments of canopy clearance revealed considerable variation in the nature of this association between landscapes (Fig. 2a,b). Turf covers were initially sparse (mean: <8%; across all sites) and were relatively uniform under canopies within both landscape types (anova: F1,30 = 2·37; = 0·198). After 8 months turf covers within turfed landscapes were ∼3 times greater than those in forested landscapes (Fig. 2a; anova: F1,30 = 28·43; = 0·006). Similarly, while the amount of sediment under canopies (dry weight) did not vary between landscape types before clearance (anova: F1,30 = 0·49; = 0·523), eight months after clearance, the weight of accumulated of sediment was ∼10 times greater among turfed than forested sites (Fig. 2b; anova: F1,30 = 22·12; = 0·009). These patterns (Fig. 2a,b) were opposite to the rate of canopy development (Fig. 2c). The relationship between turfs and sediments differed among landscape types (ancova: F1,60 = 4·37; = 0·014), with turfed landscapes characterized by more extensive covers of turfs and greater sediment accumulation (Fig. 3). There were significant positive relationships between turfs and sediments for both landscape types, albeit the correlation for turfed landscapes (Pearson’s correlation coefficient, = 0·757, P = 0·001) was considerably stronger than that for forested landscapes (= 0·624, P = 0·001).

image

Figure 2.  The influence of landscape-type (i.e. turfed cf. forested landscapes) on (a) the development of turfs, (b) the accumulation of sediment and (c) canopy recovery within plots of canopy clearance. Each landscape is represented by a mean ± SEM for each of three turfed and three forested sites (= 6 replicate plots per site).

Download figure to PowerPoint

Can turf removal increase rates of canopy recovery?

The removal of turfs and associated sediment from turfed landscapes resulted in greater rates of canopy recovery than controls plots where turfs were not removed (Fig. 4). Twelve months after initial clearance, recovery within removal plots was ∼250% greater than it was within control plots. Turf removal resulted in recovery that was relatively consistent across sites (mean ± SE: 89·3 ± 2·5%; across all sites) in contrast to control plots that displayed significant variation in recovery among sites, with means ranging from 11·9% to 61·4% (Fig. 4; Table 2, ‘Treatment × Site’ interaction). At the end of the experiment (∼12 months), canopy covers within turf removal plots did not differ significantly from temporal controls, and both were significantly greater than non-manipulated plots (anova: F2,75 = 13·79; = 0·016; SNK-test; turf removals = adjacent canopy > non-removal controls). The taxonomic composition of canopies varied among the three sites; a characteristic of temperate Australia (i.e. Flindersia: Goodsell et al. 2004; Connell & Irving 2008). One site was dominated by fucoids, and recovery was dominated by fucoids (e.g. 96% in both untouched-canopies and turf removal-treatments), composition at another (31%E. radiata, 68% fucoids) was also reflected in recovering canopies (35%E.radiata, 65% fucoids), but the pattern at a third site (dominated by E. radiata, 95%) showed the reverse pattern in recovering canopies (fucoids, 74%).

image

Figure 4.  A comparison of canopy recovery between treatments from which turfs were removed (turf removal) and left untouched (non-removal) 12 months previously. Each treatment is represented by a mean ± SEM for each of three turfed sites (= 12 replicate per treatment per site).

Download figure to PowerPoint

Table 2. anova testing for the effect of turf on canopy recovery (turf removal vs. non-removal-treatments; = 12 per treatment) among three replicate-turfed landscapes
Treatmentd.f.MSF-value
  1. Data were ArcSin (%) transformed to meet the assumptions of heterogeneity of variance. ‘Treatment’ was treated as fixed and ‘sites’ as random (n = 12 replicate plots per site). SNK (Student-Newman-Keuls) tests on significant interaction ‘Treatment × Site’ term showed, for turf removals, site 1 = site 2 = site 3; and for controls, site 1 < site 3 < site 2. Significance: NS, not significant; > 0·05; ***< 0·001.

Turf removal150705·2014·27NS
Site23862·711·09NS
Turf removal × site23552·2911·46***
Residual66309·98 

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Recovery of canopy-forming algae was substantially less in turfed landscapes than it was in forested landscapes. These data suggest that the natural disturbances involved in gap creation (i.e. small, storm driven gaps of <10 m along Australian coast; Connell 2007) need to be separated from the human-derived stressors that facilitate the expansion of gaps along human-dominated coasts (i.e. increased sediment and nutrient loads; Duarte 1995; Gorgula & Connell 2004; Airoldi, Balata & Beck 2008). Our results support the idea that loss of forests need not be a synchronized event, but instead may be the consequence of declining water quality that may over many years reduce rates of canopy recovery and thereby enable gaps to expand over broad areas. We propose that gap persistence, therefore, is an important component of gap expansion (i.e. individual gaps coalescing). Our demonstration of the inhibitory effects of turfs on canopy recovery not only provides a mechanism that underpins regime-shifts along human-dominated coasts, but also provides a powerful demonstration of the potential for reversing the human-mediated conditions that enable turfs to retain space (i.e. restoration).

Natural communities are often founded on spatially heterogeneous mosaics of habitat-forming species mediated by disturbance-recovery cycles (Pickett & White 1985; Diaz-Pulido & McCook 2002). Whilst turfs are natural to subtidal landscapes even on undeveloped coastlines (e.g. Irving & Connell 2006), human activities provide the environmental conditions needed to enable them to retain space throughout periods of natural senescence (winter loss of cover and biomass; S.D. Connell, B.D. Russell, D. Gorman & L. Airoldi, unpublished) which, coincides with the period of canopy recruitment (Novaczek 1984; Hotchkiss 1999). This timing is important because the positive feedbacks between canopy and understorey (i.e. kelp – crust associations: Irving, Connell & Gillanders 2004) are disrupted by turfs that overgrow understorey taxa (i.e. crusts) upon which canopies recruit. While canopies inhibit turfs (Connell 2003, 2005), turfs inhibit the recruitment of canopies (i.e. this study) and create their own positive feedbacks needed to retain space (i.e. turf-sediment associations; Airoldi 2003). In essence, therefore, each habitat (canopy cf. turf) is maintained by different feedback mechanisms that can switch to produce contrasting habitats (e.g. forested to turfed landscapes).

It is often simpler to identify the general causes of regime-shifts (e.g. pollution and nutrient overloading; Duarte 1995; Rabalais 2002) than to identify the factors that maintain these shifts (i.e. the mechanisms that prevent a return to a pre-existing state). This is particularly true of forest dynamics (marine or terrestrial; Eriksson et al. 2002; Worrall, Lee & Harrington 2005) where the agents involved in gap expansion do not occur in a consistent and uniform manner that is easily attributed to a single factor or event. Humans can reduce the capacity of habitats to recover when their activities transform an acute disturbance (such as storm events; Wernberg & Connell 2008) into a chronic stress that makes recovery slow or apparently irreversible (see Gorgula & Connell 2004). We propose that coastal eutrophication can work in concert with natural disturbance to facilitate the persistence of turfs in gaps, thereby creating a chronic stress that results in long-term forest decline over broader areas. Such links between declining water quality and the structure of subtidal communities have been evidenced throughout the globe (Duarte 1995; Cloern 2001; Deegan et al. 2002) and in the context of our study we refer to the 70% reduction of forest cover spanning 30 years (Connell et al. 2008). Spatial variation in canopy loss has been linked to variation in nutrient release among alternate catchment-types (i.e. urban, agricultural and natural) and its incorporation into opportunistic algal tissue (i.e. isotopic indicators), and is believed to account for contemporary sizes and proportions of forests relative to turfs (Gorman et al. 2009).

We recognize that gap persistence reflects an effect of inhibition by turfs as well as a reduction in spore supply that may be exacerbated with increasing gap size. Canopy-formers are short dispersers (e.g. Kendrick 1994) and efforts to reduce the dominance of turfs will also need to enhance spore supply. Whilst the rate of recovery in forested landscapes is rapid, recovery in large gaps would seem slow or less possible without manipulation to increase spore supply. Mangialajo et al. (2008) propose active transplantation of canopy-formers as a means of aiding the restoration of forested landscapes. This approach may have merit as it is likely to increase propagule supply (Worm et al. 1999) and extend the area over which canopies can maintain free substratum for canopy recruitment (Irving & Connell 2006). Additionally, canopy transplantation is likely to facilitate the recovery of large gaps more rapidly than lateral expansion of remnant forests (Emmerson & Collings 1998). In this regard, restoration research will need to follow active efforts to increase spore supply and reduce water pollution.

This study joins with the idea that restoration ecology provides a powerful tool for speeding the recovery of damaged landscapes (Dobson et al. 1997). Future restoration efforts are possible if the link between elevated coastal water nutrient concentrations and persistent covers of turfs can be interrupted. Yet, marine managers have little or no jurisdiction over land-based activities that elevate nutrient and sediment loads (e.g. catchment management). Current developments in coastal policy (i.e. marine-protected areas) act primarily on extraction (what is removed from the sea) rather than addition (what is added to the sea). Initiatives that reduce coastal nutrient loading and consequently the persistence and areal extent of turfed landscapes, are likely to enhance forest resilience and restoration by interrupting the positive feedback between turfs and sediments (i.e. turf-sediment associations; Airoldi 2003) that inhibit the recruitment and prevent subsequent recovery of canopy-formers after disturbance. These efforts are realistic when viewed in light of new policy initiatives that aim to increase wastewater recycling for residential and industrial use (i.e. nearly 45% of Adelaide’s wastewater; ‘Water Proofing Adelaide’ policy of the South Australian Government) and conservation organizations keen to assist with reseeding (i.e. ‘Reef Watch’ of South Australia).

In conclusion, natural communities are often founded on spatially heterogeneous mosaics of habitat-forming species mediated by disturbance-recovery cycles (Pickett & White 1985); whereby ephemeral components (e.g. turfs) function as temporary or successional units (Diaz-Pulido & McCook 2002). In human-dominated landscapes, however, the altered environmental conditions favour early successional and fast-growing assemblages that respond rapidly to increased resources to retain space in these novel environments. The increasing prevalence of human-altered environments can induce permanent shifts to fundamentally different landscapes with altered productivity and structure (e.g. lower rates of productivity and accumulation of carbon and nitrogen: Copertino, Connell & Cheshire 2005) and function (e.g. habitats to fewer species of animals: Goodsell & Connell 2002). While ecologists are concerned that such regime-shifts may be permanent (reviews: Scheffer & Carpenter 2003; Folke et al. 2004) we show that it may be possible to reverse canopy-to-turf shifts along temperate rocky coastlines, particularly if done in conjunction with resilience-building policies (e.g. improving coastal water quality).

Humans are an integral part of the environment, and have a long and continuing history of manipulating environmental conditions that encourage unwanted regime-shifts (Jackson et al. 2001). With an increasing appreciation for restoring habitats, resilience-based management provides a framework for evaluating the links between habitats and humans; linking ecological resilience to their governance (Hughes et al. 2005). Restoration and resilience-building policies are necessarily based on the ecological capacity to anticipate or reverse the trend of unwanted regime-shifts; a particularly relevant but vexing challenge for ecologists. Not all components of resistance will be as fully understood as ecologists may hope for (as recognized by Gunderson & Holling 2002), and therefore resilience-building management may need to accommodate flexible and open learning (Folke et al. 2004) to provide the adaptability and capacity needed to enhance resilience in the face of uncertainty. We take significant steps to redress uncertainty about restoration policy by demonstrating that the feedbacks which maintain regime-shifts need not be permanent, thereby opening opportunities for improving the resilience and restoration of subtidal systems on rocky coasts.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

The authors thank the farsighted recognition of P. von Baumgarten (Principle Marine Policy Adviser) and her department (The Department of Environment and Heritage) for commissioning this research to guide development of new policy on coastal management. The Australian Research Council (ARC Linkage) provided leverage funds to assist DEH achieve our work in an applied context. These research outcomes have been used by the South Australian Department of Premier and Cabinet as part of its coastal planning and water recycling initiatives. Field assistance was provided by J. Brooks, J.P. Livore and J.R. Naumann. DG was supported by a Wildlife Conservation Foundation Research Grant (0254), and SDC by a QEII Fellowship.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • Airoldi, L. (2003) The effects of sedimentation on rocky coast assemblages. Oceanography and Marine Biology: An Annual Review, 41, 161236.
  • Airoldi, L., Balata, D. & Beck, M. (2008) The Gray Zone: relationships between habitat loss and marine diversity and their applications in conservation. Journal of Experimental Marine Biology and Ecology, 366, 815.
  • Camus, P.A. (1994) Recruitment of the intertidal kelp Lessonia nigrescens Bory in northern Chile: successional constraints and opportunities. Journal of Experimental Marine Biology & Ecology, 184, 171181.
  • Cloern, J.E. (2001) Our evolving conceptual model of the coastal eutrophication problem. Marine Ecology Progress Series, 210, 223253.
  • Connell, S.D. (2003) The monopolization of understorey habitat by subtidal encrusting coralline algae: a test of the combined effects of canopy-mediated light and sedimentation. Marine Biology, 142, 10651071.
  • Connell, S.D. (2005) Assembly and maintenance of subtidal habitat heterogeneity: synergistic effects of light penetration and sedimentation. Marine Ecology Progress Series, 289, 5361.
  • Connell, S.D. (2007) Water quality and the loss of coral reefs and kelp forests: alternative states and the influence of fishing. Marine Ecology (eds S.D.Connell & B.M.Gillanders), pp. 556568. Oxford University Press, Melbourne.
  • Connell, S.D. & Irving, A.D. (2008) Integrating ecology with biogeography using landscape characteristics: a case study of subtidal habitat across continental Australia. Journal of Biogeography, 35, 16081621.
  • Connell, S.D., Russell, B.D., Turner, D.J., Shepherd, S.A., Kildea, T., Miller, D.J., Airoldi, L. & Cheshire, A. (2008) Recovering a lost baseline: missing kelp forests on a metropolitan coast. Marine Ecology Progress Series, 360, 6372.
  • Copertino, M., Connell, S.D. & Cheshire, A. (2005) The prevalence and productivity of turf-forming algae on a temperate subtidal coast. Phycologia, 44, 241248.
  • Dayton, P.K., Tegner, M.J., Parnell, P.E. & Edwards, P.B. (1992) Temporal and spatial patterns of disturbance and recovery in a kelp forest community. Ecological Monographs, 62, 421445.
  • Deegan, L.A., Wright, A., Ayvazian, S.G., Finn, J.T., Golden, H., Merson, R.R. & Harrison, J. (2002) Nitrogen loading alters seagrass ecosystem structure and support of higher trophic levels. Aquatic Conservation: Marine and Freshwater Ecosystems, 12, 193212.
  • Devinny, J.S. & Volse, L.A. (1978) Effects of sediments on the development of Macrocystis pyrifera gametophytes. Marine Biology, 48, 343348.
  • Diaz-Pulido, G. & McCook, L.J. (2002) The fate of bleached corals: patterns and dynamics of algal recruitment. Marine Ecology Progress Series, 232, 115128.
  • Dobson, A.P., Bradshaw, A.D. & Baker, A.J.M. (1997) Hopes for the future: restoration ecology and conservation biology. Science, 277, 515522.
  • Drummond, S.P. & Connell, S.D. (2005) Quantifying percentage cover on subtidal organisms on rocky coasts: a comparison of the costs and benefits of standard methods. Marine & Freshwater Research, 56, 865876.
  • Duarte, C.M. (1995) Submerged aquatic vegetation in relation to different nutrient regimes. Ophelia, 41, 87112.
  • Duarte, C.M., Conley, D.J., Carstensen, J. & Sanchez-Camacho, M. (2009) Return to Neverland: shifting baselines affect eutrophication restoration targets. Estuaries and Coasts, 32, 2936.
  • Emmerson, L.M. & Collings, G.J. (1998) Macroalgal recruitment in artificially disturbed areas: interactive effects of temporal and spatial scale. Marine and Freshwater Research, 49, 541546.
  • Eriksson, B.K., Johansson, G. & Snoeijs, P. (2002) Long-term changes in the macroalgal vegetation of the inner Gullmar Fjord, Swedish Skagerrak coast. Journal of Phycology, 38, 284296.
  • Folke, C., Carpenter, S., Walker, B., Scheffer, M., Elmqvist, T., Gunderson, L. & Holling, C.S. (2004) Regime shifts, resilience, and biodiversity in ecosystem management. Annual Review of Ecology Evolution and Systematics, 35, 557581.
  • Goodsell, P.J. & Connell, S.D. (2002) Can habitat loss be treated independently of habitat configuration? Implications for rare and common taxa in fragmented landscapes. Marine Ecology Progress Series, 239, 3744.
  • Goodsell, P.J. & Connell, S.D. (2005) Disturbance initiates diversity in recruitment of canopy-forming algae: interactive effects of canopy-thinning and substratum availability. Phycologia, 44, 632639.
  • Goodsell, P.J., Fowler-Walker, M.J., Gillanders, B.M. & Connell, S.D. (2004) Variations in the configuration of algae in subtidal forests: implications for invertebrate assemblages. Austral Ecology, 29, 350357.
  • Gorgula, S.K. & Connell, S.D. (2004) Expansive covers of turf-forming algae on human-dominated coast: the relative effects of increasing nutrient and sediment loads. Marine Biology, 145, 613619.
  • Gorman, D., Russell, B.D. & Connell, S.D. (2009) Land-sea connectivity: linking human modification of subsidies to altered subtidal habitats. Ecological Applications, 19, 11141126.
  • Gunderson, L.H. (2000) Ecological resilience – in theory and application. Annual Review of Ecology and Systematics, 31, 425439.
  • Gunderson, L.H. & Holling, C.S. (2002) Panarchy: Understanding Transformations in Human and Natural Systems. Island Press, Washington.
  • Hobbs, R.J. & Norton, D.A. (1996) Towards a conceptual framework for restoration ecology. Restoration Ecology, 4, 93110.
  • Hotchkiss, S. (1999) Life history strategies of three species of Cystophora (Phaeophyta, fucales) from a shallow subtidal community in South Australia. Doctor of Philosophy, University of Adelaide, Adelaide.
  • Hughes, T.P., Bellwood, D.R., Folke, C., Steneck, R.S. & Wilson, J. (2005) New paradigms for supporting the resilience of marine ecosystems. Trends in Ecology & Evolution, 20, 380386.
  • Irving, A.D. & Connell, S.D. (2006) Predicting understorey structure from the presence and composition of canopies: an assembly rule for marine algae. Oecologia, 148, 491502.
  • Irving, A.D., Connell, S.D. & Gillanders, B.M. (2004) Local complexity in patterns of canopy-benthos associations produces regional patterns across temperate Australasia. Marine Biology, 144, 361368.
  • Jackson, J.B.C., Kirby, M.X., Berger, W.H., Bjorndal, K.A., Botsford, L.W., Bourque, B.J., Bradbury, R.H., Cooke, R., Erlandson, J., Estes, J.A., Hughes, T.P., Kidwell, S., Lange, C.B., Lenihan, H.S., Pandolfi, J.M., Peterson, C.H., Steneck, R.S., Tegner, M.J. & Warner, R.R. (2001) Historical overfishing and the recent collapse of coastal ecosystems. Science, 293, 629638.
  • Kendrick, G.A. (1994) Effects of propagule settlement density and adult canopy on survival of recruits of Sargassum spp. (Sargassaceae, Phaeophyta). Marine Ecology Progress Series, 103, 129140.
  • Kennelly, S.J. (1987a) Inhibition of kelp recruitment by turfing algae and consequences for an Australian kelp community. Journal of Experimental Marine Biology and Ecology, 112, 4960.
  • Kennelly, S.J. (1987b) Physical disturbances in an Australian kelp community. I. Temporal effects. Marine Ecology Progress Series, 40, 145153.
  • Kirkman, H. (1981) The first year in the life history and the survival of the juvenile marine macrophyte, Ecklonia radiata (Turn.) J. Agardh. Journal of Experimental Marine Biology & Ecology, 55, 243254.
  • Krause-Jensen, D., Sagert, S., Schubert, H. & Bostrom, C. (2008) Empirical relationships linking distribution and abundance of marine vegetation to eutrophication. Ecological Indicators, 8, 515529.
  • Mangialajo, L., Chiantore, M. & Cattaneo-Vietti, R. (2008) Loss of fucoid algae along a gradient of urbanisation, and structure of benthic assemblages. Marine Ecology Progress Series, 358, 6374.
  • Melville, A.J. & Connell, S.D. (2001) Experimental effects of kelp canopies on subtidal coralline algae. Austral Ecology, 26, 102108.
  • Novaczek, I. (1984) Development and phenology of Ecklonia radiata at two depths in Goat Island Bay, New Zealand. Marine Biology, 81, 189197.
  • Nystrom, M., Folke, C. & Moberg, F. (2000) Coral reef disturbance and resilience in a human-dominated environment. Trends in Ecology & Evolution, 15, 413417.
  • Österblom, H., Hansson, S., Larsson, U., Hjerne, O., Wulff, F., Elmgren, R. & Folke, C. (2007) Human-induced trophic cascades and ecological regime shifts in the Baltic sea. Ecosystems, 10, 877889.
  • Pickett, S.T.A. & White, P.S. (1985) The Ecology of Natural Disturbances and Patch Dynamics. Academic Press, New York.
  • Rabalais, N.N. (2002) Nitrogen in aquatic ecosystems. Ambio, 31, 102112.
  • Roberts, B.K. & Connell, S.D. (2008) Detecting benthic responses to human-induced change: effectiveness of alternate taxonomic classification and indices. Marine Ecology Progress Series, 358, 7584.
  • Russell, B.D. & Connell, S.D. (2007) Response of grazers to sudden nutrient pulses in oligotrophic versus eutrophic conditions. Marine Ecology Progress Series, 349, 7380.
  • Scheffer, M. & Carpenter, S.R. (2003) Catastrophic regime shifts in ecosystems: linking theory to observation. Trends in Ecology & Evolution, 18, 648656.
  • Scheffer, M., Carpenter, S., Foley, J.A., Folke, C. & Walker, B. (2001) Catastrophic shifts in ecosystems. Nature, 413, 591596.
  • Steneck, R.S., Vavrinec, J. & Leland, A.V. (2004) Accelerating trophic-level dysfunction in kelp forest ecosystems of the western North Atlantic. Ecosystems, 7, 323332.
  • Wernberg, T. & Connell, S.D. (2008) Physical disturbance and subtidal habitat structure on open rocky coasts: effects of wave exposure, extent and intensity. Journal of Sea Research, 59, 237248.
  • Womersley, H.B.S. (1987) The Marine Benthic Flora of Southern Australia, Part II. South Australian Government Printing Division, Adelaide.
  • Worm, B., Lotze, H.K., Bostroem, C., Engkvist, R., Labanauskas, V. & Sommer, U. (1999) Marine diversity shift linked to interactions among grazers, nutrients and propagule banks. Marine Ecology Progress Series, 185, 309314.
  • Worrall, J.J., Lee, T.D. & Harrington, T.C. (2005) Forest dynamics and agents that initiate and expand canopy gaps in Picea-Abies forests of Crawford Notch, New Hampshire, USA. Journal of Ecology, 93, 178190.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Fig. S1. Map showing the interspersion of sampling of alternate landscape types.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
JPE_1711_sm_EconnellSF1.doc3848KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.