- Top of page
- Material and Methods
Marine algae are known to provide habitats for a wide range of marine organisms. Populations of marine epiphytal invertebrates are generalists and are less adapted to live in only one macroalga species. However, there are some examples of local adaptation and, in particular, amphipods have shown strong host specificity. Amphitholina cuniculus, an amphipod with an alga-burrowing habit, has been mainly observed in Bifurcaria bifurcata mats on the intertidal rocky shores of the southern region of the Galician coast (NW Spain; 42º11.27′ N, 8º48.25′ W). We designed a laboratory experiment conducted in June 2009 to analyse the association between the amphipod and the macroalga. In particular, we hypothesised that (i) this species would exhibit different behaviour during the day and at night, and (ii) adults of A. cuniculus would prefer B. bifurcata as habitat rather than Fucus vesiculosus or Sargassum muticum. Results supported the hypothesis that adults of A. cuniculus preferred B. bifurcata, although such preferences did not show the day/night variability predicted. This amphipod might be considered a specialist species, at least on rocky intertidal shores along the Galician coast.
- Top of page
- Material and Methods
The association of small invertebrates with larger sessile organisms such as plants or macroalgae that provide habitat and/or food has been a major topic in ecological and evolutionary studies, in both terrestrial and marine systems. However, patterns of habitat use in terrestrial and marine systems differ. For example, most species of terrestrial herbivorous insects show habitat specialisation (Fox & Morrow 1981), whereas most marine invertebrates are associated with multiple macroalgal families as habitats (Hay & Fenical 1988). Some marine invertebrates, however, are specialised and it has been argued that an understanding of these rare cases may provide insight into the ecology and evolution of feeding and habitat specialisation in general (Poore et al. 2000; Sotka 2007).
For small mobile epifauna, seaweeds are a refuge from stressful conditions associated with life on rocky intertidal shores. For example, seaweeds can provide cool and protective canopies and interstices and ameliorate the influence of rapid changes in temperature, desiccation and hydrodynamic forces on intertidal shores (Bates 2009). In addition, seaweeds may also offer shelter from predation and food for some species of invertebrates (Cronin et al. 1995; Bates 2009).
Patterns of distribution of mobile epifauna are frequently explained by the behaviour of the species under consideration, or of those with which they interact (Chapman 2000). Preference for different habitats, for example, may occur during site selection and settlement of larvae (Highsmith 1982), during exploratory behaviour of post-larvae and juveniles (Pardo et al. 2007), and during the period of adult life (Aikins & Kikuchi 2001). Changes in behaviour may also occur during a daily cycle due to changes in patterns of activity of epifauna. For example, small soft-bodied herbivores tend to move at night when predators are less active (Hay et al. 1987; Bell 1991; Brawley 1992). In contrast, large herbivores such as fishes, urchins and some gastropods exhibit high mobility with relatively little probability of being successfully attacked by their predators (Hay et al. 1987). In particular, some amphipods are more active at night than during the day. For example, the filter-feeding amphipod Ericthonius brasiliensis lives in domiciles it constructs by curling terminal segments of its host seaweed Halimeda tuna, but exclusively at night (Sotka et al. 1999). Criteria for epifauna host choice include intrinsic factors of the host such as cell wall components, defensive biochemistry, nutritive value and palatability, toughness and architectural complexity of seaweeds (Hawkins & Hartnoll 1983; Chemello & Milazzo 2002; Van Alstyne & Houser 2003). In addition, habitat selection may be determined by abundance, size and patchiness of habitat (Kumagai 2008).
When animals are associated with a particular habitat, it is often assumed that they have actively ‘selected’ that habitat. A common problem is the probable or actual misuse of the term ‘preference’. It is always necessary to test (and be able to provide evidence to support) the model that behaviour of the animals can indeed account for the observed patterns (see Olabarria et al. 2002 and references therein). To avoid confounding preference (an active behaviour) with other causes of greater numbers being found in some habitats, it is also necessary to determine the pattern of occupancy of each type of habitat when available alone (i.e. when there is no choice and therefore no preference). If organisms choose a certain habitat (i.e. show a preference) they should select more of that habitat when presented with a mixture than would be the case when each of the various choices is presented alone. Therefore, experiments to demonstrate preferences are quite complex.
On intertidal shores, amphipods are often among the most dominant macroepifauna on macroalgae and seagrasses (Hay et al. 1987; Bell 1991; Aikins & Kikuchi 2001; Bates 2009). Despite generalist habits of most species of this group, some organisms have shown a strong host specificity (Duffy & Hay 1991; Poore et al. 2000), and they have been suggested to play a similar role to insects in terrestrial habitats because they are locally abundant, small relative to the host that they use for both habitat and food, and can have large impacts on host assemblage structure (Duffy & Hay 2000). An interesting group of amphipods that has been used as a model in evolutionary approaches and in several studies on ecology and behaviour is the family Ampithoidae (Arrontes 1999; Poore & Steinberg 1999; Cruz-Rivera & Hay 2001; Sotka 2003; Poore & Hill 2006). Previous studies have revealed that ampithoid amphipods differ in host specificity and composition (Poore et al. 2008), suggesting that evolutionary history may affect the ability of this family to colonise diverse algal taxa. As an example, the genus Peramphithoe rarely uses available macrophytes in the order Dictyotales, and as a consequence, displays a more restricted host range than other genera. However, other species from the family are very often associated with the host genus Sargasssum (Poore et al. 2008).
Amphitholina cuniculus (Stebbing 1874) is a small species that belongs to the family Ampithoidae and exhibits an alga-burrowing habit (Myers 1974). This species has been found burrowing into the thalli of different macroalgae such as Alaria esculenta (Linnaeus) Greville, Fucus vesiculosus Linnaeus or Bifurcaria bifurcata R. Ross (Myers 1974; Viejo 1999; Gestoso et al. 2010). A previous study done on the Galician coast showed that this species was especially abundant in B. bifurcata, although it also appeared sporadically associated with Sargassum muticum Yendo (Fensholt) (Gestoso et al. 2010) and with F. vesiculosus (I. Gestoso, personal observation). Due to this differential pattern of distribution, the study suggested that A. cuniculus could exhibit a preference for certain habitats.
Based upon these observations we designed a laboratory experiment in which we analysed the association of A. cuniculus with the most abundant algae that form mixed stands on low intertidal rocky shores along the Galician coast (Northwest Spain). An important aspect of the experimental design used here was the unconfounding of preference (active behavioural choice) from any other reasons for different occupancies of different types of habitat (e.g. differences in accessibility). Many studies on selection of habitats by different invertebrates have not used an experimental design like that used here, i.e. also determining the pattern of occupancy of each habitat when there is no choice (Durante & Chia 1991; McDonald & Bingham 2010). Here, we tested the hypotheses that (i) this species would exhibit different behaviour during the daily cycle (day versus night), and (ii) adults of A. cuniculus would prefer B. bifurcata as habitat rather than F. vesiculosus or S. muticum.
- Top of page
- Material and Methods
A total of 864 individuals of Amphitholina cuniculus were placed into different habitats at the beginning of the experiments but only 763 individuals (88%) were recovered in any of the three habitats at the end of experiments (i.e. after 4 h).
The hypothesis that A. cuniculus exhibit different preference behaviour during the day than at night was not supported. Although proportions of amphipods varied significantly among treatments, this variation was consistent during the daily cycle (i.e. no significant main effect of Time of day, or significant interaction of Time of day with any other factor; Table 1). In addition, response was consistent in the two trials (i.e. no significant main effect of Trial, or interaction of Trial with any other factor; Table 1).
Table 1. ANOVA analysis of proportions of Amphitholina cuniculus in experimental treatments. Time of day (day versus night) and treatment (six levels) were fixed factors, and Trial (two levels) was a random factor (n = 6)
|Time of day||1||0.0508||2.19||0.3785|
|Treatment||5||0.4433||7.00|| 0.0261 |
|Trial × Time of day||1||0.0232||0.50||0.4804|
|Trial × Treatment||5||0.0633||1.37||0.2420|
|Time of day × Treatment||5||0.0939||4.06||0.0752|
|Trial × Time of day × Treatment||5||0.0231||0.50||0.7762|
|Residual||120||0.0463|| || |
|Total||143|| || || |
As Trial and Time of day did not have any effect on the proportions of amphipods found in each habitat, replicates of the different trials and different experiments (day and night) were pooled for analyses of preference (n = 24).
There were significant differences among the proportions of individuals in different habitats (Table 2). All analyses supported the hypotheses derived from the model in which adults of A. cuniculus prefer Bifurcaria bifurcata.
Table 2. ANOVA analyses of proportions of Amphitholina cuniculus in experimental treatments. Data from the two experiments (day and night) and two trials were pooled (n = 24)
|Hypothesis 1: nB/N1, nB/N4 > nS/N2, nF/N3 (more in preferred habitat)|
|Among treatments||3||0.7169||15.48|| 0.0000 |
|Residual||92||0.0463|| || |
|Hypothesis 2: nS/N5 > nS/N2 (fewer in non-preferred habitats where there is choice)|
|Among treatments||1||0.4020||6.39|| 0.0150 |
|Residual||46||0.0629|| || |
|Hypothesis 3: nF/N6 > nF/N3|
|Among treatments||1||0.4134||7.37|| 0.0093 |
|Residual||46||0.0561|| || |
To test the second hypothesis of preference (see 'Material and Methods'), a priori determined contrasts were used which demonstrated that a greater mean proportion of the individuals were recovered of those initially placed in B. bifurcata habitat (B in treatments 1 and 4; Fig. 3A) than in Sargassum muticum (S in treatment 2) or Fucus vesiculosus habitats (F in treatment 3) (contrast: nB/N1, nB/N4 versus nS/N2, nF/N3; F1,92 = 16.1769; P = 0.0001). This was entirely consistent with the hypothesis. There were also greater proportions of A. cuniculus in B. bifurcata when choices of preferred habitat were available (treatment 1) than when choices were not available (treatment 4) (contrast: nB/N1 versus nB/N4; F1,92 = 5.8315; P = 0.0177), probably due to the random movement of amphipods. Choice treatments of non-preferred habitats showed similar proportions (treatments 2 and 3) (contrast: nS/N2 versus nF/N3, F1,92 = 0.3732; P = 0.5428). [Correction added after online publication, 22 May 2013: ‘No’ deleted from start of preceding sentence.] In addition, there were smaller proportions of amphipods in non-preferred habitats when choices were available (treatments 2 and 3) than in those habitats when choices were not available (treatments 5 and 6) (Table 2, Fig. 3B,C). [Correction added after online publication, 22 May 2013: text ‘(treatments 5 and 6)’ and ‘(treatments 2 and 3)’ transposed in preceding sentence]. These results, therefore, supported hypotheses 2 and 3.
Figure 3. Mean (+ SE) proportion of Amphitholina cuniculus remaining in treatments in the habitats where they were initially placed: (A) proportion of individuals in multiple choice and no choice treatments; (B) proportion of individuals remaining in Sargassum muticum habitat in choice (T2) and no choice (T5) treatments; (C) proportion of individuals in Fucus vesiculosus habitat in choice (T3) and no choice (T6) treatments.
Download figure to PowerPoint
- Top of page
- Material and Methods
The hypothesis that this species exhibits different behaviour during the daily cycle was not supported. In contrast, results confirmed the hypothesis that adults of Amphitholina cuniculus show a preference for Bifurcaria bifurcata habitat.
Previous studies have found that marine invertebrates may exhibit different behaviour during a daily cycle (Hay et al. 1987). In particular, amphipods are more active at night than during the day (e.g. Edgar 1983; Hay et al. 1987; Sotka et al. 1999). These organisms may constrain their foraging movements to avoid predators because the risk of predation increases when moving between seaweeds. However, our results indicated that A. cuniculus did not change patterns of selection of habitat during the daily cycle. In general, individuals were quite active during the day and at night in both trials (I. Gestoso, personal observation). Nevertheless, we have to take into account that experiments under laboratory conditions have certain constraints. For example, the absence of large predators in the laboratory experiments might have an effect on the behaviour of this species because individuals may change behaviour in response to stimuli from predators (Chapman 2000).
In marine habitats, small herbivorous invertebrates tend to be generalists and very few species are host-plant specialists (Hay & Steinberg 1992; McDonald & Bingham 2010). In addition, most amphipods associated with seaweeds are habitat generalists (Duffy & Hay 1991). However, A. cuniculus might be considered a specialist species at least on rocky intertidal shores along the Galician coast (Gestoso et al. 2010). The active behaviour shown by this species in the experiments may explain in part the distribution of this amphipod. For example, after widespread recruitment, adults of A. cuniculus could choose the favoured habitats (Underwood & Denley 1984). Alternatively, populations could distribute randomly across available habitats, but differential mortality would lead to a decrease of individuals in unfavourable habitats (Russo 1987).
Several works have pointed out that distributions of amphipods are rarely determined by differential mortality due to predation, active selection of habitat being a more important factor in determining patterns of distribution (Duffy & Hay 1991; Aikins & Kikuchi 2001; Poore 2004). Many factors related to intrinsic characteristics of macroalgae such as morphology, toughness, longevity, nutritional value or content of polyphenols may influence the choices of species (Hawkins & Hartnoll 1983; Chemello & Milazzo 2002; Van Alstyne & Houser 2003; McDonald & Bingham 2010). More complex morphologies may allow amphipods to escape detection or attack by visual predators (Hay 1991) or prevent dislodgement by waves (Sotka 2007). In this particular case, morphological complexity did not seem to play an important role. Although not formally analysed, Fucus vesiculosus and Sargassum muticum have a more complex morphology than B. bifurcata, which was the preferred habitat. Fucus vesiculosus and S. muticum are more branched and present vesicles, whereas B. bifurcata is a less frondose and cylindrical macroalga. Furthermore, algal toughness and stability of habitat (influenced by longevity of algae) have been reported to be more important factors than morphology (McDonald & Bingham 2010). A greater stability of habitat might favour certain species of amphipods, especially in habitats where risk of predation is high or environmental conditions are stressful (Duffy & Hay 1991). The fact that B. bifurcata is a perennial alga might explain why the amphipod A. cuniculus prefers this species over the pseudoperennial S. muticum.
The chemical traits of the host may also affect the susceptibility of amphipods to predation, and predation on small invertebrates may be reduced due to their association with chemically rich hosts that are avoided by omnivorous fish (Poore et al. 2000). Bifurcaria bifurcata is well known for its capacity to synthesise a wide variety of acrylic diterpenes (Ortalo-Magné et al. 2005). Amphitholina cuniculus might be able to tolerate these compounds because it supposes a lower risk of being consumed than if it selects more palatable species (Taylor & Steinberg 2005). Nevertheless, within the family Ampithoidae, responses to non-polar metabolites in host algae are very variable, with species showing negative relationships (Peramphithoe, Biancolina), positive relationships (Exampithoe) or no relation (Ampithoe spp.) (Poore et al. 2008). In addition, the nutritional value of host seaweeds may influence the selection of host alga (Cruz-Rivera & Hay 2000).
In conclusion, it is very likely that the amphipod should get some benefit from selecting B. bifurcata against F. vesiculosus and S. muticum. For example, the amphipod might have greater fitness (i.e. maximising growth or mate encounter rates), get a good quality food and refuge from abiotic stress and predators, or avoid competition with other species, among other possibilities. Unfortunately, this experiment did not allow us to determine the underlying mechanisms of habitat choice and, therefore, the evolutionary forces that maintain this restricted host choice are unclear.
Preference experiments in the laboratory should be extrapolated to natural systems with caution (Chapman 2000). For example, the laboratory preference assays were conducted on small scales that may not represent the scales on which organisms discriminate among hosts in the field. Therefore, factors such as availability of preferred habitat or proximity of available hosts (Jonsen et al. 2001) might influence the distribution of amphipods in the field. Indeed, the spatial arrangement of plants affects strongly patterns of colonisation of marine invertebrates associated with vegetated substrates (Bell et al. 2001). Not only colonisation, but also post-colonisation processes (competition, predation, disturbance, immigration/emigration, etc.) may be important in determining patterns of spatial distribution of these organisms (Olabarria 2002, and references therein). Thus, it is essential that laboratory studies be accompanied by field tests of related hypotheses, so that differences between field-based behaviour and laboratory-based behaviour can be measured and the relevance of the laboratory studies sensibly evaluated. We are concerned about these difficulties but tried to make the conditions as realistic as possible, maintaining amphipods in the laboratory along with natural algae prior to the experiment. The experiments were also repeated twice to examine the consistency of patterns. In addition, these experiments were based on observations from a previous field study in which the distribution of mobile epifauna, including this species, was evaluated at different spatio-temporal scales. It would be desirable to carry out some parts of these experiments in the field if logistic problems could be overcome, using the information gained here as a guide. Nevertheless, results suggested that the behaviour of these amphipods may explain in part their distribution in the field.