|

Received: 21 February 2017    Accepted: 5 November 2017

DOI: 10.1111/jai.13553

ORIGINAL ARTICLE

Site fidelity of intertidal fish to rockpools

J. Roma1  | M. Dias1  | C. Vinagre1  | A. C. F. Silva2

1
MARE – Marine and Environmental
Sciences Centre, Faculdade de Ciências da Summary
Universidade de Lisboa, Lisboa, Portugal Gobius paganellus, Lipophrys pholis and Coryphoblennius galerita are wide-­spread inter-
2
CERIS, Instituto Superior
tidal fish that spend their earlier life stages in rock pools, and yet very little is known
Técnico, Universidade de Lisboa, Lisboa,
Portugal about their site fidelity behaviour. For these species, fidelity to rockpools may result in
increased fitness costs in a predicted scenario of warmer sea water, due to the low
Correspondence
Joana Roma, MARE – Marine and thermal inertia of these water bodies. In this context, it is relevant to investigate these
Environmental Sciences Centre, Faculdade de
species’ site fidelity. We made a mark-­recapture study to assess the mentioned spe-
Ciências da Universidade de Lisboa, Lisboa,
Portugal. cies’ movements within and between rockpools. We tagged a total of 530 individuals
Email: joana.roma.isa@gmail.com
of the aforementioned species with the Visible Implant Elastomer and tracked their
Funding information movement for a 7-­month period. We found that site fidelity and specific rockpools
Fundação para a Ciência e Tecnologia, Grant/
conditions are important factors in distribution of intertidal blennies and gobies. We
Award Number: UID/MAR/04292/2013
also examined the relations between rockpool volume, depth and site fidelity. We
found that G. paganellus tends to remain in its original marking pool, with an average
recapture rate of 20.5%, but showed no evidence of inter-­pool movement. Rockpool
depth, however, proved to be important in the blennies’ movements. Our findings are
among the first to prove that a mark-­recapture method can be successfully used to
track intertidal fish movements. In particular, we showed that G. paganellus presents
site fidelity in intertidal rockpools during its early ontogeny for a period of two to three
months.

1 |  INTRODUCTION little is currently known about how the rockpools’ inhabitants inter-
act with their physical setting, both in ordinary and extreme thermal
One of the greatest challenges that researchers currently face is under- conditions, such as in the event of heat waves due to climate change.
standing the impact of global warming over biodiversity (Schwenk, To correctly predict changes in the patterns of rockpool usage by its
Padilla, Bakken, & Full, 2009). It is predicted that extreme climatic inhabitants when faced with thermal variations, it is first necessary
events such as heat waves will increasingly become more frequent in to determine the still largely unknown standard behavioural patterns.
Europe (IPCC, 2014). The coastal habitat and in particular the inter- Hence, we aimed to investigate the habitat distribution behaviour of
tidal zone have been identified as some of the most vulnerable habi- the common intertidal gobies and blennies in the Portuguese intertidal
tats to this type of climatic pressure (Helmuth, Mieszkowska, Moore, rockpools.
& Hawkins, 2006). Temperature increases have already caused shifts Our targeted species—Gobius paganellus, Lipophrys pholis and
in the current distribution of intertidal species (Helmuth et al., 2002). Coryphoblennius galerita—are all commonly found at the intertidal
The rocky intertidal zone is often an extremely productive habitat rockpools in the Eastern Atlantic coasts (Bath, 1990; Miller, 1990;
(White, Hose, & Brown, 2015), subject daily to severe changes on its Zander, 1986), and are important prey for commercially import-
physical, biological and chemical features due to its water/land inter- ant predator fish species (Patzner, Van Tassell, Kovačić, & Kapoor,
face. Within this shifting zone rockpools’ temperatures vary widely 2011). However, much is unknown about these species’ habitat
due to their relative little thermal inertia, in turn, related to the small distribution including among others, whether or not they pres-
water volume and prolonged exposure to low-­tide conditions (Morris ent any form of site fidelity, choose to move around the habitat
& Taylor, 1983; Vinagre et al., 2013, 2016). Despite their ubiquity, and/or keep coming back to it with a certain frequency (homing

J Appl Ichthyol. 2017;1–7. © 2017 Blackwell Verlag GmbH |  1

wileyonlinelibrary.com/journal/jai  
 |
2       ROMA et al.
ROMA et al. |
      3

F I G U R E   1   Representation of the sampled rockpools’ (rectangles of varied sizes numbered P1 through P12) location in relation to each other.
Maximum length and width are represented within the rockpools in the sampling stations (a) Cabo Raso 1, (b) Cabo Raso 2 and (c) Raio Verde 1.

behaviour) or, if they abandon their current habitat altogether at

2.2 | Statistical analysis
each high tide. Available studies on the matter focused on the
distribution behaviour of other species and habitats, particularly The following independent variables and experimental design were
in coastal (Shima, McNaughtan, Geange, & Wilkinson, 2012) and used to test for significant differences in the response variables: rock-
estuarine areas (Jud & Layman, 2012). In those cases, the distri- pool volume (fixed, two levels: below 0.8 m3 and above 1.0 m3) and
bution behaviour was influenced by water depth, water body vol- rockpool depth (fixed, two levels: below 0.3 m and above 0.3 m). Three
ume, sediment type, light penetration and the presence of aquatic response variables were assessed: (i) average number of recaptures, (ii)
macrophytes (Connolly, 1994; Griffiths, 2001; Whitfield, 1999). average size of all marked individuals and, (iii) average size of all recap-
Supported by this information, and because rockpool volume and tured ­individuals. These were chosen to examine the influence of rock-
depth are also known to strongly influence the rockpool commu- pool characteristics on the movement patterns and population structure
nity structure (Martins, Hawkins, Thompson, & Jenkins, 2007), we associated with the movements. A PERMANOVA analysis was used to
focused on these two factors as potentially important environmen- assess the hypotheses stated in the introduction. Prior to analysis, the
tal conditions for fish distribution. data were fourth root transformed in order to obtain a more balanced
We tested the following hypothesis for each of the species sampled weight distribution between variables. The number of used permutations
in this study: (i) the species does not present rockpool fidelity and this is was 9,999 and the similarity measure used was the Euclidean distance.
independent of time; (ii) if present, the scale of the species’ movement Depending on the PERMANOVA results, we performed a
range between rockpools is small (<5 m) and; (iii) rockpool depth and vol- SIMPER similarity analysis to identify which response variable had
ume contribute to explain the observed movement patterns. the most contribution for the significant differences found in the
PERMANOVA analysis. We also made a multi-­dimensional scal-
ing (MDS) analysis to obtain a visual ordination of the differences
2 | MATERIALS AND METHODS
between samples.

2.1 | Sampling sites and techniques

Sampling was performed at three rocky beaches in the Portuguese 3 | RESULTS
West coast (38º42′N, 9º28′W and 38°42′N, 9°29′W). Twelve rock-
pools (P1–P12) were sampled monthly and bimonthly (depending on In total, 530 fish were tagged in this study, of which 83.2% were G.
rockpool access due to tidal variation), during a 7-­month (July 2012 to paganellus (441), 9.4% were L. pholis (50) and 7.4% were C. galerita
January 2013) period (Figure 1). (39).
The rockpool average area was 4.66 ± 1.28 m2 (range: 0.79– The total number of tagged animals for volume 1 pools (<0.8 m3)
16.90 m2) and medium depth was 0.26 ± 0.2 m (range: 0.18–0.40 m). was 153, and for volume 2 (>1.0 m3) was 377. The total number of
The pools were categorized accordingly to: A. volume: volume 1— tagged animals for depth 1 pools (<0.3 m) was 196, and for depth 2
below 0.8 m3 (7 pools: P5, P6, P8, P9, P10, P11 and P12) and volume (>0.3 m) was 334.
2—above 1.0 m3 (5 pools: P1, P2, P3, P4 and P7) and; B. depth: depth
1—below 0.3 m (5 pools: P7, P9, P10, P11 and P12) and depth 2—
3.1 | Gobius paganellus
above 0.3 m (7 pools: P1, P2, P3, P4, P5, P6 and P8).
We chose the mark-­recapture technique for this study as it is one The highest number of tagged individuals in a single rockpool was 107
of the most widely accepted sampling methodologies for the distribu- and the lowest was 12. With the exception of a single rockpool, G.
tion studies of small fishes (Imbert, Beaulaton, Rigaud, & Elie, 2007). paganellus was recaptured at least once in all 12 sampled rockpools.
The visible implant elastomer (VIE) was chosen as the internal tag as it The average recapture rate across all rockpools was 20.5%. The highest
causes little to no damage or disturbance to the living organism, espe- recapture number in a single event was 44 individuals in P2, although
cially to such small sample subjects (some were as small as 2 cm or less). the highest average percentage (of all recapture events of previously
This implant tag has been successfully used for rockpool fishes in Knope, marked individuals in a single pool) was 35.8% in P4 (Figure 2). No
Tice, and Rypkema (2017). The individuals were caught in rockpools with marked G. paganellus were recaptured after the 4th recapture event,
hand-­held nets, tagged with VIE by injecting a hypodermic syringe just 4 months after the tagging time. The average recapture rate in pool
above the left lateral line and their length, location and species registered, volume <0.8 m3 was 20.8% and 21.1% for pools >1.0 m3. The aver-
followed by immediate release into their original rockpool. The target ani- age recapture rate of pools with depth <0.3 m was 23.1% and 19.3%
mals were tagged for each pool, and subsequently the marked individuals for those with depth >0.3 m. Concerning the recaptures rates within
were examined for recaptures on a monthly or bi-­monthly basis. Each the same pool where individuals were initially tagged (henceforth,
pool was assigned a different colour. “home pool”), the average size of marked individuals (3.6 ± 1.0 cm,
 |
4       ROMA et al.

F I G U R E   2   Overview of Gobius paganellus recaptured tagged individuals, represented as the average percentage of recaptured individuals in
all recapture occasions, per sampled rockpool (P1 through P12).

range 2–9.5 cm) was smaller than that of individuals later recaptured recaptured individuals). However, the same parameters varied signifi-
(4.5 ± 1.4 cm, range 2.7–9.9 cm). cantly with rockpool depth (Table 1b and Figure 3). The SIMPER anal-
As for external recaptures, i.e., the number of marked individuals ysis showed that all three analysed parameters were greater at deeper
found outside their home pool, only one rockpool (P5) registered 8.9% rockpools, i.e., rockpools ≥0.3 m.
of its tagged individuals in other pools.

4 | DISCUSSION
3.2 | Lipophrys pholis
In total, 50 L. pholis individuals were tagged in all 12 pools, but only Our study was the first of its kind to use this mark-­recapture meth-
five individuals were recaptured afterwards in four of those pools. No ods applied to the distribution of intertidal fish species in rockpools.
tagged L. pholis were further recaptured after the 1st recapture event The main hypotheses of this study were successfully addressed by
(within the same month of tagging). The average size of individuals the methodology. However, it was not possible to validate our sec-
when they were first captured (5.1 ± 1.5 cm, range 1.9–11.3 cm) ond hypothesis (the scale of its movement range between rockpools
was larger than that of individuals recaptured in later occasions is smaller than 5 m) for all studied species since there were very few
(4.4 ± 2.2 cm, range 3.9–10.5 cm). Concerning external recaptures, recaptured individuals that were caught outside their home pools, and
only one rockpool registered one of its individuals in other pools, rep- hence no accurate conclusion could be derived from our data.
resenting 20% of its total marked individuals. Next, the specific results obtained for each species will be dis-
cussed, addressing the remaining hypothesis: (i) it does not have rock-
pool fidelity and this is independent of time; and (ii) rockpool depth
3.3 | Coryphoblennius galerita
and volume contribute to explain the movement patterns.
A total of 39 C. galerita individuals were tagged in all 12 pools, but only
six individuals were recaptured afterwards in five of those pools. No
4.1 | Gobius paganellus
marked individuals were further recaptured after the 1st recapture
event (within the same month of tagging). The average size of marked Gobius paganellus generally remained in the same rockpool, suggesting
individuals (5.0 ± 1.6 cm, range 3–11 cm) was smaller than individuals a site fidelity behaviour and avoidance to cross into nearby pools. This
later recaptured (6.8 ± 3.0 cm, range 4.2–10.5 cm). No external recap- fidelity behaviour is potentially associated with: (i) increased predator
tures were registered. refuge conditions, such as comparatively higher algae canopy cover
There were no significant differences between rockpool volumes or increased crevice availability (Baeza & Thiel, 2007; Baeza, Farías,
(Table 1a) for the tested parameters (average number of recaptured Luppi, & Spivak, 2010); (ii) territorial and food defence behaviours
individuals, average size of marked individuals and average size of (Baeza & Thiel, 2003, 2007; Emlen & Oring, 1977) and; (iii) preference
ROMA et al. |
      5

T A B L E   1   Results of PERMANOVA
Source df SS MS Pseudo-­F P(perm) Unique perms
analysis on Coryphoblennius galerita
according to the following experimental (a)
design: (a) rockpool volume (fixed, two Volume 1 3.8786 3.8786 1.3103 0.3404 512
levels: below 0.8 m3 and above 1.0 m3) and
Residuals 22 65.121 2.9601
(b) rockpool depth (fixed, two levels: below
0.3 m and above 0.3 m) Total 23 69
(b)
Depth 1 13.138 13.138 5.1741 0.0101 512
Residuals 22 55.862 2.5392
Total 23 69

F I G U R E   3   MDS analysis of the

analysed Coryphoblennius galerita variables
(average number of recaptured individuals,
average size of marked individuals and
average size of recaptured individuals) at
each rockpool and analysed by respective
depth.

for a habitat that individuals were previously familiar with (Carlson, and among the IPCC temperature variation prediction models, one
2005). already predicts a temperature increment as high as 4.8°C within this
Mostly, only juveniles of G. paganellus (i.e., fish <7 cm, the average century alone (IPCC, 2014). Considering that this species is known to
size at first maturity) were collected. During the early winter sampling exclusively inhabit the intertidal habitat, and if this becomes inhospi-
period almost no individuals were found in the sampling areas. This table due to high temperatures, the fish may migrate into habitats with
absence may be explained by high mortality rates in earlier ontogenic lower temperatures. This would ensure future survival in more resilient
stages, or by simply migrating outside the sampling area due to a bio- habitats, such as deeper rockpools lower in the intertidal zone or even
logical driver, such as reproductive behaviour. The reproductive sea- channels found in this type of habitat (Faria & Almada, 2001a, 2001b;
son occurs between January and June (Azevedo & Simas, 2000), hence Vinagre et al., 2013).
potentially explaining the absence of tagged individuals. Gibson (1972,
1999) also proposes that this species favours rockpools in the lower
4.2 | Lipophrys pholis and Coryphoblennius galerita
intertidal zone. However, no movements between different areas of
the intertidal zone or even to the subtidal have yet been documented. We did not acquire enough data (i.e., sampled individuals) of these
Site fidelity behaviour can lead to the prediction that when faced species to be able to ascertain if they display any site fidelity or inter-­
with global warming effects, G. paganellus living in rockpools may rockpool behaviours. Few individuals were caught and tagged and
face challenging if not adverse conditions, potentially affecting their even less were later recaptured. While L. pholis individuals were all
physiology and ecology. While during the sampling days of the study juveniles (adult size ~16 cm), we found adult C. galerita measuring up
rockpool water temperatures never surpassed 28°C in the shallowest to 9.5 cm.
rockpools. If these same temperatures rise as high as 33.1°C (previ- Around October and at 1 year of age, L. pholis is known to aban-
ously verified as this species’ thermal tolerance limit) (Vinagre et al., don intertidal rockpools in favour of near-­subtidal rock fissures and
2013), this can lead to thermal stress or even desiccation, compro- corridors (Faria & Almada, 2001a, 2001b; Faria, Almada, & Gonçalves,
mising the survival of this species (Brierley & Kingsford, 2009). In 1996; Gibson, 1972), potentially explaining its disappearance from
southern Europe where our study was located, days with tempera- our study area around the same time. We could find no evidence
tures above 35°C are expected to increase to more than 40 days per in literature that attested the same for C. galerita, however Nieder
year (Miranda, Coelho, Tomé, & Valente, 2002). Within this context (1993) found that C. galerita apparently exhibits both site fidelity
 |
6       ROMA et al.

and movement around the same rockpool for only 2 months, which hypothesis with the prawn Betaeus lilianae and description of its part-
might explain its apparently premature disappearance from the nership with the crab Platyxanthus crenulatus. Journal of Experimental
Marine Biology and Ecology, 389, 85–92.
study site. Almada, Dores, Pinheiro, Pinheiro, and Santos (1983) ver-
Baeza, J. A., & Thiel, M. (2003). Predicting territorial behavior in symbiotic
ified that both blennies possibly show site fidelity not to a single crabs using host characteristics: A comparative study and proposal of a
rockpool, but rather to a group of rockpools. Later, Jorge, Almada, model. Marine Biology, 142, 93–100.
Gonçalves, Duarte-­Coelho, and Almada (2012) in a laboratory set- Baeza, J. A., & Thiel, M. (2007). The mating system of symbiotic crusta-
ceans. A conceptual model based on optimality and ecological con-
ting, verified a homing behaviour in adult L. pholis individuals which
straints. In J. E. Duffy, & M. Thiel (Eds.), Reproductive and social behavior:
is a behaviour often associated with site fidelity, albeit not a con- Crustaceans as model systems (pp. 245–255). Oxford: Oxford University
firmation. Altogether, these studies indicate that some type of site Press.
fidelity could be shown by C. galerita. Bath, H. (1990). Blenniidae. In J. C. Quero, J. C. Hureau, C. Karrer, A.
Post, & L. Saldanha (Eds.), Check-list of the fishes of the eastern tropi-
However, we successfully proved that C. galerita seems to prefer
cal Atlantic (CLOFETA) (Vol. 2, pp. 905–915). Lisbon, Paris: JNICT/SEI/
deeper rockpools, a choice also verified by Nieder (1993) who observed UNESCO.
more blennies in deeper pools and a theory previously confirmed by Bellquist, L. F., Lowe, C. G., & Caselle, J. E. (2008). Fine-­scale movement pat-
Martins et al. (2007). Deeper pools may offer different prey availability terns, site fidelity, and habitat selection of ocean whitefish (Caulolatilus
princeps). Fisheries Research, 91, 325–335.
and additional predator refuge opportunity due to larger algae canopy
Brierley, A. S., & Kingsford, M. J. (2009). Impacts of climate change on ma-
cover, which are characteristics that fishes displaying homing and site
rine organisms and ecosystems. Current Biology, 19, R602–R614.
fidelity behaviours often seem to favour (e.g., Abecasis, Bentes, & Erzini, Carlson, P. (2005). Homing, habitat fidelity and kin recognition in Eurasian
2009; Bellquist, Lowe, & Caselle, 2008; Harasti, Lee, Gallen, Hughes, & perch (Perca fluviatilis) (Master thesis dissertation), Uppsala University,
Stewart, 2015; Mason & Lowe, 2010; March, Alós, Grau, & Palmer, 2011). Sweden.
Connolly, R. M. (1994). Removal of seagrass canopy: Effects on small fish
In conclusion, the study successfully showed that intertidal fish
and their prey. Journal of Experimental Marine Biology and Ecology, 184,
species can display site fidelity behaviours to rockpools, and that pool 99–110.
characteristics may strongly influence their movement choices. Future Emlen, S. T., & Oring, L. W. (1977). Ecology, sexual selection and the evolu-
studies using individual tagging procedures would enable a more com- tion of mating systems. Science, 197, 215–223.
Faria, C., & Almada, V. (2001a). Agonistic behaviour and control of ac-
prehensive assessment of intertidal fish movements.
cess to hiding places in two intertidal blennies, Lipophrys pholis and
Coryphoblennius galerita (Pisces: Blenniidae). Acta Ethologica, 4, 51–58.
Faria, C., & Almada, V. (2001b). Microhabitat segregation in three rocky
ACKNOWLE DG E MEN TS
intertidal fish species in Portugal: Does it reflect interspecific competi-
tion? Journal of Fish Biology, 58, 145–159.
We would like to thank all those involved in the fieldwork. This work
Faria, C., Almada, V. C., & Gonçalves, E. J. (1996). Juvenile recruitment,
was funded by the “Fundação para a Ciência e Tecnologia”, Portugal, growth and maturation of Lipophrys pholis (Pisces: Blenniidae), from the
through the researcher position awarded to C. Vinagre and the strate- west coast of Portugal. Journal of Fish Biology, 49, 727–730.
gic project UID/MAR/04292/2013 granted to MARE. Gibson, R. N. (1972). The vertical distribution and feeding relationships of
intertidal fish on the Atlantic cost of France. Journal of Animal Ecology,
41, 189–207.
O RCI D Gibson, R. N. (1999). Movement and homing in intertidal fishes. In M. H.
Horn, K. L. M. Martin, & M. A. Chotkowski (Eds.), Intertidal fishes: Life in
J. Roma  http://orcid.org/0000-0002-5884-4623 two worlds (pp. 97–125). San Diego, CA: Academic Press.
Griffiths, S. P. (2001). Factors influencing fish composition in an Australian
M. Dias  http://orcid.org/0000-0003-0447-6009 intermittently open estuary. Is stability salinity-­dependent? Estuarine,
C. Vinagre  http://orcid.org/0000-0003-2146-7948 Coastal and Shelf Science, 52, 739–751.
Harasti, D., Lee, K. A., Gallen, C., Hughes, J. M., & Stewart, J. (2015).
A. C. F. Silva  http://orcid.org/0000-0003-4734-6538 Movements, home range and site fidelity of snapper (Chrysophrys
auratus) within a temperate marine protected area. PLoS ONE, 10,
e0142454.
Helmuth, B., Harley, C. D. G., Halpin, P. M., O’Donnell, M., Hofmann, G. E.,
REFERENCES
& Blanchette, C. A. (2002). Climate change and latitudinal patterns of
Abecasis, D., Bentes, L., & Erzini, K. (2009). Home range, residency and intertidal thermal stress. Science, 298, 1015–1017.
movements of Diplodus sargus and Diplodus vulgaris in a coastal lagoon: Helmuth, B., Mieszkowska, N., Moore, P., & Hawkins, J. (2006). Living on
Connectivity between nursery and adult habitats. Estuarine, Coastal and the edge of two changing worlds: Forecasting the responses of rocky
Shelf Science, 85, 525–529. intertidal ecosystems to climate change. Annual Review of Ecology,
Almada, V. C., Dores, J., Pinheiro, A., Pinheiro, M., & Santos, R. S. (1983). Evolution, and Systematics, 37, 373–404.
Contribuição para o estudo do comportamento de Coryphoblennius Imbert, H., Beaulaton, L., Rigaud, C., & Elie, P. (2007). Evaluation of visible
galerita (L.) (Pisces, Blenniidae). Memórias do Museu do Mar: Série implant elastomer as a method for tagging small European eels. Journal
Zoológica, 2, 166. of Fish Biology, 71, 1546–1554.
Azevedo, J. M. N., & Simas, A. M. V. (2000). Age and growth, reproduction IPCC (2014). The regional impacts of climate change: An assessment of vulner-
and diet of sublittoral population of the rock goby Gobius paganellus ability. UK: Cambridge University Press.
(Teleostei: Gobiidae). Hydrobiologia, 440, 129–135. Jorge, P. E., Almada, F., Gonçalves, A. R., Duarte-Coelho, P., & Almada, V. C.
Baeza, J. A., Farías, N. E., Luppi, T. A., & Spivak, E. D. (2010). Refuge size, group (2012). Homing in rocky intertidal fish. Are Lipophrys pholis L. able to
living and symbiosis: Testing the “resource economic monopolization” perform true navigation? Animal Cognition, 15, 1173–1181.
ROMA et al. |
      7

Jud, Z., & Layman, C. A. (2012). Site fidelity and movement patterns of in- Patzner, R. A., Van Tassell, J. L., Kovačić, M., & Kapoor, B. G. (Eds.) (2011).
vasive lionfish, Pterois spp., in a Florida estuary. Journal of Experimental The biology of gobies (p. 685). Enfield, NH: Science Publishers.
Marine Biology and Ecology, 414–415, 69–74. Schwenk, K., Padilla, D. K., Bakken, G. S., & Full, R. J. (2009). Grand chal-
Knope, M. L., Tice, K. A., & Rypkema, D. C. (2017). Site fidelity and hom- lenges in organismal biology. Integrative and Comparative Biology, 49,
ing behaviour of intertidal sculpins revisited. Journal of Fish Biology, 90, 7–14.
341–355. Shima, J. S., McNaughtan, D., Geange, S. W., & Wilkinson, S. (2012).
March, D., Alós, J., Grau, A., & Palmer, M. (2011). Short-­term residence Ontogenetic variation in site fidelity and homing behaviour of a tem-
and movement patterns of the annular seabream Diplodus annularis perate reef fish. Journal of Experimental Marine Biology and Ecology,
in a temperate marine reserve. Estuarine, Coastal and Shelf Science, 92, 416–417, 162–167.
581–587. Vinagre, C., Dias, M., Roma, J., Silva, A., Madeira, D., & Diniz, M. S. (2013).
Martins, G. M., Hawkins, S. J., Thompson, R. C., & Jenkins, S. R. (2007). Critical thermal maxima of common rocky intertidal fish and shrimps –
Community structure and functioning in intertidal rock pools: Effects A preliminary assessment. Journal of Sea Research, 81, 10–12.
of pool size and shore height at different successional stages. Marine Vinagre, C., Leal, I., Mendonça, V., Madeira, D., Narciso, L., Diniz, M. S., &
Ecology Progress Series, 329, 43–55. Flores, A. A. V. (2016). Vulnerability to climate warming and acclima-
Mason, T. J., & Lowe, C. G. (2010). Home range, habitat use, and site fidelity tion capacity of tropical and temperate coastal organisms. Ecological
of barred sand bass within a southern California marine protected area. Indicators, 62, 317–327.
Fisheries Research, 106, 93–101. White, G. E., Hose, G. C., & Brown, C. (2015). Influence of rock-­pool charac-
Miller, P. J. (1990). Gobiidae. In J. C. Quero, J. C. Hureau, C. Karrer, A. teristics on the distribution and abundance of inter-­tidal fishes. Marine
Post, & L. Saldanha (Eds.), Check-list of the fishes of the eastern tropical Ecology, 36, 1332–1344.
Atlantic (CLOFETA) (Vol. 2, pp. 925–951). Lisbon, Paris: JNICT/SEI/ Whitfield, A. K. (1999). Ichthyofaunal assemblages in estuaries: A South
UNESCO. African case study. Reviews in Fish Biology and Fisheries, 9, 151–186.
Miranda, P. M. A., Coelho, F. E. S., Tomé, A. R., & Valente, M. A. (2002). 20th Zander, C. D. (1986). Blenniidae. In P. J. P. Whitehead, M.-L. Bauchot, J.-
century Portuguese climate and climate scenarios. Scenarios, impacts C. Hureau, J. Nielsen, & E. Tortonese (Eds.), Fishes of the north-eastern
and adaptation measures – SIAM project. In: F.D. Santos, K. Forbes & Atlantic and the Mediterranean (pp. 1096–1112). Paris: UNESCO.
Moita R (Eds.), Climate change in Portugal: Scenarios, impacts and adapta-
tion measures (SIAM Project). Lisbon, Portugal: Gradiva, 454 pp.
Morris, S., & Taylor, A. C. (1983). Diurnal and seasonal variation in physico-­
How to cite this article: Roma J, Dias M, Vinagre C, Silva ACF.
chemical conditions within intertidal rock pools. Estuarine, Coastal and
Shelf Science, 17, 339–355.
Site fidelity of intertidal fish to rockpools. J Appl Ichthyol.
Nieder, J. (1993). Distribution of juvenile blennies (Coryphoblennius galer- 2017;00:1–7. https://doi.org/10.1111/jai.13553
ita and Lipophrys pholis, Pisces, Blenniidae) in small tide-­pools on the
Algarve coast of southern Portugal: Result of low-­tide lottery or strate-
gic habitat selection? Bonner Zoologische Beiträge, 44, 133–140.