recruitment disturbances on two interacting barnacle species

Transcription

1 Ecology , Competing species in a changing climate: effects of Blackwell Publishing Ltd recruitment disturbances on two interacting barnacle species CARL JOHAN SVENSSON, EMMELI JOHANSSON and PER ÅBERG Department of Marine Ecology, Göteborg University, Box 461, SE Göteborg, Sweden Summary 1. The climate is changing and data-based simulation models can be a valuable tool for predicting population response to such changes and investigate the mechanisms of population change. In this study, a data-based two-species matrix model was constructed to explore the possible effects of elevated sea surface temperature (i.e. climate change) on the interaction between open populations of the south Atlantic barnacle species Chthamalus montagui and the boreal species Semibalanus balanoides in the north-east Atlantic. 2. First, the model was used to perform an elasticity analysis to determine the relative importance of recruitment and survival in the interaction. Further, three scenarios of changes in recruitment, related to climate change, were investigated with model simulations: (i) increased frequencies of low recruitment for S. balanoides; (ii) increased frequencies of high recruitment for C. montagui; (iii) a combination of (i) and (ii). 3. Model simulations showed that in present environmental conditions, S. balanoides occupied most of the space and dominated the interaction through high recruitment and survival. These results matched independent field observations, which validated the model for further analyses. 4. The elasticity analyses showed that although free space was available there was competition for space during recruitment intervals. It was also shown that both populations were sensitive to changes in recruitment. 5. Introducing the three scenarios of recruitment disturbances led to large changes in species abundance and free space. The most significant changes were found when scenario (i) and (ii) were combined, producing a shift in species dynamics towards C. montagui dominance. This demonstrates that recruitment can be an important mechanism in the interaction between populations and that the population response to changes in recruitment depends on the added response of interacting species. 6. In a more general context, this model shows that increased sea surface temperature could rapidly lead to increased competition from southern species at higher latitudes. This might accelerate the effects of climate change on the species distribution at these latitudes and eventually lead to changes in community dynamics on temperate and subarctic shores. Key-words: changes in recruitment, competition for space, increase of sea surface temperature, open populations, sensitivity. Ecology (2006) 75, doi: /j x Ecological Society Introduction The well-documented increase of atmospheric greenhouse gases is promoting a rise in global air and sea surface Correspondence; C. J. Svensson, Department of Marine Ecology, Göteborg University, Box 461, SE Göteborg, Sweden. Tel.: +46(0) carl-johan.svensson@marbot.gu.se temperature (e.g. Edwards et al. 2002). In the North Atlantic, this has led to an increase of the positive index of the North Atlantic Oscillation (NAO), adding variability to the weather (Southward 1991; Weaver et al. 2001) and affecting ecosystem functioning and ecological functions for a number of species (e.g. Fromentin & Planque 1996; Reid, Holliday & Smyth 2001; Beaugrand & Reid 2003; Root et al. 2003; Svensson et al. 2005).

2 766 C. J. Svensson, E. Johansson & P. Åberg However, as most climate variables are measured over broad scales and many ecological studies are carried out at small scales, it is often difficult to correlate the two types of data and make predictions for the future (Levin 1992). Also, in connecting climate change to ecological changes, the significance of results depends on the demographic detail of species data, as many factors other than climate are influencing population dynamics (Caley et al. 1996; Menge 2000; Brown et al. 2001; Bertness & Ewanchuk 2002). Models that are designed to predict effects of environmental disturbances on species must therefore operate at relevant scales and incorporate the natural variability in species demography and the interaction between species (Sanford 1999; Brown et al. 2001; Schwilk & Ackerly 2005). Many marine benthic invertebrates, such as barnacles, have life cycles comprising a sessile adult stage and freeswimming pelagic larvae (Thorson 1950). These larvae undergo morphological development before competent to settle and must typically spend a minimum amount of time in open water, resulting in long distance dispersal. Though this behaviour could be means of avoiding crowding, settling larvae have been shown to respond positively to chemical cues from adult conspecifics (e.g. Hills et al. 1998) and fully developed larvae often attempt settling in habitats that are populated by conspecifics and other species (Kent, Hawkins & Doncaster 2003). Thus, competition for space can be an important factor in determining settlement success (e.g. Thorson 1950; Hixon, Pacala & Sandin 2002): first, larvae compete with other larvae for suitable surface habitat (Roughgarden, Iwasa & Baxter 1985; Minchinton & Scheibling 1993); secondly, as attached cyprids and growing juveniles, density-dependent mortality is experienced through crowding, which can be expressed either by neighbours overgrowing or undercutting each other (e.g. Connell 1961; Bertness 1989) or, neighbours bulging in hummocks that are easily torn from the surface by wave action (Barnes & Powell 1950; Connell 1961). In the north-east Atlantic, the barnacle species Chthamalus montagui Southward, Chthamalus stellatus (Poli) and Semibalanus balanoides (L.) occupy the intertidal, where they compete for space and possibly food (Connell 1961; Southward 1991). These species are common across large scales and extensively studied (e.g. Connell 1961; Southward 1976, 1991; Range & Paula 2001; O Riordan et al. 2004). S. balanoides, which has high recruitment and fast growth, is the stronger competitor for space in the mid and lower part of the intertidal while C. montagui, having greater tolerance for desiccation, is more likely to be found higher up (Connell 1961; Jenkins et al. 2001; O Riordan et al. 2004). C. stellatus is commonly found at mid to high shore (O Riordan et al. 2004), being less abundant or absent at sheltered shores (Range & Paula 2001). Several studies have shown that latitudinal gradients in recruitment can regulate the relative abundances of interacting species (e.g. Carroll 1996; Connolly & Roughgarden 1999; Connolly, Menge & Roughgarden 2001; Bertness & Ewanchuk 2002). In relation to this, we know that for both C. montagui and S. balanoides, the climate and associated weather at higher latitudes are major factors determining recruitment success (e.g. Carroll 1996; O Riordan et al. 2004). For S. balanoides, which produce one brood per year, recruitment success depends on the matching of larvae release with spring diatom bloom (Thorson 1950; Crisp & Clegg 1960). During winter spring periods with higher than normal sea surface temperature, the spring bloom occurs later compared with the long-term seasonal cycle, accompanied by a reduction in diatom numbers (Edwards et al. 2002), leading to low recruitment for S. balanoides. C. montagui, on the other hand, produce numerous small broods in rapid succession throughout summer and larvae are small, depending on flagellates for food. At the northern borders of its range, where the average temperature is low, the recruitment season is reduced to late summer and early autumn (Kendall & Bedford 1987; O Riordan et al. 2004). This apparent sensitivity of recruitment to temperature may explain why Southward (1991) found that the relative abundance of the two species correlated with inshore winter sea surface temperatures 2 years earlier. In those observations, low mean inshore temperature was followed by an increase of adult S. balanoides and a decrease of adult C. montagui and the pattern was reversed following winters with high mean inshore temperatures (Southward 1991). As C. montagui is a more southern species and less abundant than S. balanoides at higher latitudes, this interaction presents a good model system for looking at possible effects of climate change on open populations. In this study, interactions in a simplified two-species system are investigated using model simulations of a two-species, stage-structured open population matrix model. The model describes the interaction between two barnacle species, C. montagui at the northern border of its range and S. balanoides well within its latitudinal distribution. Detailed demographic data for the construction of the model were obtained from studies in Ireland, Isle of Man and Portugal (Hyder et al. 2001; Jenkins et al. 2001). The model, however, is theoretical and not specific to any location and tries to mimic places where these two species compete. In this, we acknowledge that the situation could be more complex in places where also C. stellatus is present. Two main questions were investigated: (1) What is the relative importance of a disturbance in recruitment and adult survival in one of the species for the abundance of both species and free space? (2) What is the effect of changes in the recruitment in one or both species, related to elevated sea surface temperature, on the abundance and free space at a local scale and the distribution at a regional scale? The first question was investigated through proportional perturbation analyses (i.e. elasticity) and the second one with model simulations of three possible scenarios: (1) decreased recruitment for S. balanoides; (2) increased recruitment for C. montagui; (3) a combination of scenario (1) and scenario (2).

3 767 Modelling changes in species interaction Table 1. Size-classes (mm 2 ), mean projection matrices (± SD) and mean recruitment state functions (± SD) (R for recruitment, LR for low recruitment and HR for high recruitment) during summer to winter (period 1) and winter to summer (period 2) for the S. balanoides population (a) and C. montagui population (b) (a) S. balanoides population From To 1 (0 11 8) 2 ( ) 3 (> 20) Period ± ± ± ± ± 0 46 R = 16 5 ± 7 11 (cm 2 ) LR = 0 97 ± 0 79 (cm 2 ) Period ± ± ± ± ± ± 0 44 (b) C. montagui population From To 1 (0 5 3) 2 ( ) 3 ( ) 4 (> 14 0) R = 0 81 ± 0 46 (cm 2 ) HR = ± 8 57 (cm 2 ) Period ± ± ± ± ± ± ± 0 51 Period ± ± ± ± ± ± ± 0 48 Methods STUDY ORGANISMS C. montagui is a common intertidal barnacle found on rocky shores ranging from the north of Scotland to Mauritania and also in the Mediterranean (Crisp, Southward & Southward 1981). Timing of breeding and releasing of larvae vary with latitude (Kendall & Bedford 1987; Burrows, Hawkins & Southward 1992). Eggs are brooded internally and larvae are released only if the temperature is greater than 15 C (Burrows et al. 1992). In Ireland, the recruitment for C. montagui is mostly restricted to the period August until early October (Kendall & Bedford 1987). S. balanoides is a boreo-arctic species found in the intertidal on both sides of the north Atlantic. In Europe, it extends over a large latitudinal range but is found mainly from north-west Spain (Barnes 1958) to Spitzbergen (Barnes 1957). S. balanoides requires temperatures below 12 C for successful gonad and brood development and breeding occurs during the autumn or winter (Crisp & Clegg 1960). The release of larvae is induced by local environmental conditions (King et al. 1993) and settlement generally occurs between March and May. Successful recruitment depends on matching larvae release with the spring bloom (Crisp & Clegg 1960; Connell 1961). DATA AND MATRICES Matrices used in this study were extracted from Hyder et al. (2001) for C. montagui and Jenkins et al. (2001) for S. balanoides. The matrices were produced from data that were collected during 1 year and included eight spatially independent quadrats for each of two annual periods (see Hyder et al and Jenkins et al for details). Quadrats were randomly distributed in areas that were dominated by the studied species and included between 200 and 500 individuals of all ages and sizes. At the end of each period, survival and recruitment were calculated, growth measured for individuals that had survived and size determined for newly recruited individuals. C. montagui and S. balanoides individuals were then separated into four and three size-classes, respectively, which were used to produce 16 transition matrices (eight matrices for each of two annual periods) and eight linear recruitment functions, expressing the number of recruits per area of free space, for each of the two species. Here we present the mean matrices (± SD) and recruitment state functions (± SD) for each of the two species (Table 1) [for the individual matrices, see Hyder et al. (2001) for C. montagui and Svensson et al. (2004) for S. balanoides]. THE MATRIX MODEL The mathematical theory behind matrix population models is extensive (see Caswell 2001 for a review). First, a column vector representing the number of individuals in each size class is defined. In this model, this meant four classes for C. montagui (C 1,1, C 2,1, C 3,1 and C 4,1 ) and three classes for S. balanoides (S 5,1, S 6,1 and S 7,1 ), producing the following population vector (V) (see Hyder et al and Svensson et al for details):

4 768 C. J. Svensson, E. Johansson & P. Åberg C 1,1 C 2,1 C 3,1 V = C 4,1 S 5,1 S 6,1 S 7,1 By projecting a transition matrix on to this vector, the result is a new vector (of the same size) that displays the number of individuals for each species and in each size class one time-step later. In a deterministic model there is only one matrix describing the environment, eventually producing constant asymptotic estimates of population variables. By adding variability to the matrix or by using several matrices, the variability in model outcome increases. Further nonlinearity can be added by introducing dependent transition elements (i.e. replacing transition elements with functions). In this analysis, the transition matrix (A) consisted of two submatrices. The survivals transitions for C. montagui were represented in the upper left corner and the survival transitions for S. balanoides were represented in the lower right corner, producing a matrix with seven rows and seven columns. All other elements were set at 0: A = S 1, G 2,1 S 2, G 3,1 G 3,2 S 3, G 4,1 G 4,2 G 4,3 S 4, S 5, G 6,5 S 6, G 7,5 G 7,6 S 7,7 where S represents individuals that survive but stay within a size-class and G stands for individuals that survive and grow into another size-class. > As recruitment for both species was dependent on available free space, recruitment was not included in the transition matrix. Instead, the recruitment procedure was separated from the rest of the projection and, in model simulations, recruitment at time t (R t ) was calculated independently as: R t = R ft (A O Ct O St ) eqn 1 where R ft is the linear recruitment function at time t, A is total area (25 cm 2 ) and O Ct and O St, respectively, are the space occupied by individuals of C. montagui and S. balanoides at time t. Hence, free space was defined as space without barnacles and other potential occupiers of space were not included in the model. In nature, C. montagui settles and recruits during summer and autumn while S. balanoides settles and recruits during spring. Therefore, in model simulations, C. montagui recruited during period 1 (summer autumn) and S. balanoides during period 2 (winter spring). Based on the measured variability in survival and recruitment, represented by the transition matrices and recruitment functions, a periodic and stochastic twospecies matrix model was constructed. The model was constructed on certain biologically derived and tested assumptions (Hyder et al. 2001; Svensson et al. 2004). The assumptions were: 1. Individuals could either survive in their current size-class, increase in size to any of the other sizeclasses or die. 2. The population was considered open, with all recruitment into the first size-class. 3. Mortality was size-specific and calculated from data and independent of vital rates. 4. There was no spatial covariation between mortality and recruitment. 5. Recruitment was partly limited by space. In the model, the simulated time-step equalled 1 year and was divided into two periods. Measured variability was included in the model by, for each time-step and period, randomly drawing each transition element and recruitment factor for each species from a normal distribution based on their mean and standard deviation and introducing them into the transition matrix. As the survival of individuals cannot exceed 1, the sum of the elements in a column cannot be larger than 1. Therefore, if the stochastic process produced a column that summed to more than 1, each element in that column was replaced with its relative proportion. The formula for this was: 7 a = a a tij ij ij i = 0 eqn 2 where a tij is the transformed element, a ij is the original element and i is the row number. The process was repeated for all j columns. The initial population vector v(0) was set to zero and a stochastic sample path representing a total of years was produced. A sample path of this length is not used to predict population size after years, but as a means of estimating response variables and their variation at the stochastic steady state. In simulations, the generated sequence of transition matrices together with the recruitment functions was applied to v(0) to produce a sequence of population vectors (v(1), v(2),, v(10 000)). The stochastic growth rate (λ s ), that is the geometric mean of the growth rate, was then calculated numerically by averaging a number of one-step estimates of log λ s over time units and then calculating its natural exponent (see Caswell 2001 for details). To exclude transient effects and allow the population to reach stochastic steady state, the first 1000 years was disregarded from the analysis. For all simulations, the average population size (N), population growth rate (λ s ), population structure, recruitment and occupied space and the variation were calculated for each period and species. To compare the variation in population size (N) and recruitment between species with different means, the variation in population size (N) and recruitment were converted to coefficients of variation.

5 769 Modelling changes in species interaction PERTURBATION ANALYSIS A perturbation analysis was designed to investigate the relative importance of different vital rates (i.e. survival, growth and recruitment) to barnacle densities and free space. As the model includes nonlinear properties (i.e. stochastic variability and space-limited recruitment) classic growth rate elasticity was not applicable (Caswell 2001). An alternative perturbation analysis was constructed, which comprised making a small positive proportional change to each transition element and recruitment separately and then calculate the change in mean population size (N Cm and N Sb ) and available free space (F). This produced three different perturbation estimates (for N Cm N Sb and F) for each of the transitions in the life cycle. An identical Markov chain was used for all simulations, eluding differences in population variables resulting from variability in the stochastic sample path. Perturbation estimates (P ij ) were calculated accordingly: P ij = ((X nij X o )/X o )/( p 1) eqn 3 Table 2. Mean population size (N), population growth rate (λ s ) and recruitment (R), including variation (V) or the coefficient of variation (CV), for each species (C. montagui and S. balanoides) and, the mean amount of free space (F) per 25 cm 2, including variation, measured at the end of each period Model Period N CV (N) λ s V (λ s ) R CV (R) C. montagui S. balanoides F V where X nij is the value of the studied variable, resulting from a proportional change p to element e ij, X o is the original value of the same variable and (p 1) is a scaling factor correcting for the size of factor p. The analysis was performed with a set of different p-values ranging on a logarithmic scale from 1 1 to (1 1, 1 01,, ). Perturbation estimates converged at p = 1 01 and was determined at an accuracy of SIMULATIONS ON ALTERED RECRUITMENT Three scenarios were designed to simulate the effects of changes in recruitment on barnacle densities and free space. First, decreasing recruitment for S. balanoides was simulated (scenario 1) and, secondly, increasing recruitment for C. montagui was simulated (scenario 2). Thirdly, scenario 3 comprised simulating scenario 1 and 2 concurrently. Data on high recruitment for C. montagui were taken from Lisbon, Portugal (Hyder et al. 2001), and estimates of low recruitment for S. balanoides were taken from the Isle of Man (Svensson et al. 2005). For each species there were two normal distributions of recruitment, one around a low mean and one around a high mean (Table 1). In the original model, C. montagui recruitment values (eqn 1) were drawn from their low mean distribution and S. balanoides recruitment values (eqn 1) from their high mean distribution. For scenario 1, the low mean distribution for S. balanoides was included in the recruitment process at 100 different discrete frequencies, spanning in between 0 and 1 (0, 0 01, 0 02, 1). In scenario 2, the same procedure was repeated for C. montagui, by including the high mean distribution at 100 discrete frequencies. In scenario 3, scenario 1 and 2 were simulated simultaneously. An identical Markov chain was used in all simulations. Barnacle population sizes and amount of occupied space were recorded for each frequency in all three scenarios. Fig. 1. Population structure at the end of period 1 and period 2 for C. montagui (A) and S. balanoides (B). Results POPULATION VARIABLES S. balanoides had higher mean population size than C. montagui at the end of both periods (Table 2), though C. montagui had greater variability. Free space was always available, changing slightly from period 1 to period 2. S. balanoides had a linearly decreasing number of individuals in consecutive size categories (Fig. 1) and population structure was similar in both periods. C. montagui showed a similar pattern at the end of period 1 but at the end of period 2, size-class 1 and 2

6 770 C. J. Svensson, E. Johansson & P. Åberg Fig. 2. The elasticity of free space (F), at the end of period 1 and period 2, to a proportional change of 1 01 to the matrix elements in the matrices for period 1 and period 2. (a) Elasticity of F at the end of period 1 to changes to the matrix elements in the period 1 matrix; (b) elasticity of F at the end of period 1 to changes to the matrix elements in the period 2 matrix; (c) elasticity of F at the end of period 2 to changes to the matrix elements in the period 1 matrix; (d) elasticity of F at the end of period 2 to changes to the matrix elements in the period 2 matrix. *Marks the matrix elements (i.e. life cycle transitions) that are 0 and therefore lack elasticity. G = growth; S = survival; R = recruitment; i,j = 1 4 represent transition elements for C. montagui; i,j = 5 7 represent transition elements for S. balanoides. had similar numbers of individuals. Few C. montagui individuals reached size-class 3 and 4. PERTURBATION ANALYSIS As expected from the differences in density between the two species, free space was consistently most sensitive to changes in the vital rates of S. balanoides (Fig. 2a d). Free space was most sensitive to changes in the survival of size-class 1 and 2 individuals (S 5,5 and S 6,6 ) and to recruitment. However, free space was also sensitive to changes in the recruitment of C. montagui (Fig. 2a,c). C. montagui population size (N Cm ) was most sensitive to changes in the recruitment of C. montagui (positive effects) (Fig. 3a,c) and the recruitment and survival of S. balanoides (predominantly negative effects) (Fig. 3a d). Next to recruitment, changes to the matrix element S 6,6 for period 2 (survival of individuals in size-class 2) had the largest effects on N Cm (Fig. 3b,d). Though less pronounced, N Cm was also sensitive to changes in intraspecific survival and growth (positive effects). The population size of S. balanoides (N Sb ) was consistently most sensitive to changes in intraspecific survival (S 5,5 and S 6,6 ) (both positive and negative effects) and recruitment (positive effects) (Fig. 4a d). Changes to the matrix element S 5,5 for period 1 had a positive effect on N Sb at the end of period 1 but a negative effect at the end of period 2 (Fig. 4a,c). A summary and schematic picture of the results is presented in Fig. 5, which shows that changes in S. balanoides vital rates had significant effects on C. montagui and free space, while the effects of changes in vital rates of C. montagui were mainly intraspecific. SIMULATIONS ON ALTERED RECRUITMENT Increased frequencies of low recruitment for S. balanoides resulted in an exponential decrease in N Sb (Fig. 6a,b) accompanied by an exponential increase in the proportion of free space, reaching as high as 80% (Fig. 7a). N Cm only increased slightly, despite higher proportions of free space. With increased frequencies of high recruitment for C. montagui, N Cm increased to approximately 13 times its original value (Fig. 6c,d). The proportion of free space increased slightly (Fig. 7b). Simultaneously, N Sb decreased and, at high frequencies, N Cm was equal to or higher than N Sb. Simultaneous simulations of increased frequencies of low recruitment for S. balanoides and increased frequencies of high recruitment for C. montagui resulted in a rapid and linear decrease of N Sb accompanied by a simultaneous and equally rapid increase of N Cm (Fig. 6e,f ). From low to high frequencies, N Cm increased from 10 to approximately 300 individuals. The mean proportion of free space decreased slightly at the end of period 1 and increased slightly at the end of period 2 (Fig. 7c).

7 771 Modelling changes in species interaction Fig. 3. The elasticity of population size of C. montagui (N Cm ), at the end of period 1 and period 2, to a proportional change of 1 01 to the matrix elements for period 1 and period 2. (a) Elasticity of N Cm at the end of period 1 to changes to the matrix elements in the period 1 matrix; (b) elasticity of N Cm at the end of period 1 to changes to the matrix elements in the period 2 matrix; (c) elasticity of N Cm at the end of period 2 to changes to the matrix elements in the period 1 matrix; (d) elasticity of N Cm at the end of period 2 to changes to the matrix elements in the period 2 matrix. *Marks the matrix elements (i.e. life cycle transitions) that are 0 and therefore lack elasticity. G = growth; S = survival; R = recruitment; i,j = 1 4 represent transition elements for C. montagui; i,j = 5 7 represent transition elements for S. balanoides. Fig. 4. The elasticity of population size of S. balanoides (N Sb ), at the end of period 1 and period 2, to a proportional change of 1 01 to the matrix elements for period 1 and period 2. (a) Elasticity of N Sb at the end of period 1 to changes to the matrix elements in the period 1 matrix; (b) elasticity of N Sb at the end of period 1 to changes to the matrix elements in the period 2 matrix; (c) elasticity of N Sb at the end of period 2 to changes to the matrix elements in the period 1 matrix; (d) elasticity of N Sb at the end of period 2 to changes to the matrix elements in the period 2 matrix. *Marks the matrix elements (i.e. life cycle transitions) that are 0 and therefore lack elasticity G = growth; S = survival; R = recruitment; i,j = 1 4 represent transition elements for C. montagui; i,j = 5 7 represent transition elements for S. balanoides.

8 772 C. J. Svensson, E. Johansson & P. Åberg Discussion The model used in this study was developed to describe the interaction between two open populations of barnacles and demonstrate and how this interaction could change if sea surface temperature increases. Initially, simulation showed us that in a simple interaction such as this one, the species with the highest recruitment could control the interaction through high space occupancy. However, it was also shown that juvenile and adult survival could be important for maintaining high space occupancy after recruitment. As expected from theory (e.g. Caley et al. 1996), this demonstrates that local interactions between open populations can be sensitive to recruitment disturbances of either species. Changes to climate that affect recruitment are therefore likely to influence on species composition. However, results also Fig. 5. A schematic summary of how changes to the recruitment and the survival/growth transitions of C. montagui (C.m.) and S. balanoides (S.b.) affect the density dynamics of each species and the proportion of free space (F). The thickness of the arrows represents the relative strength of the effects. showed that species response to disturbances could depend on the added response of interacting species. This is one explanation to why it has been hard to predict effects of climate change even in relatively simple and well-studied systems (e.g. Brown et al. 2001; Bertness & Ewanchuk 2002). Adding complexity to this, open populations are usually affected by factors operating at large spatial and temporal scales (e.g. Levin 1992). None the less, through scenario simulations with a databased model, we believe this study incorporated some of that complexity and reduced the uncertainties in predicting the future. With that in mind, it is likely that increased sea surface temperatures will lead to that southern species of barnacles extend their northern range and expand as competitors for space at higher latitudes, adding to the direct effects of temperature change on boreal and arctic species (Svensson et al. 2005). Surely these results can be extended to other marine organisms that compete for space within large latitudinal ranges (for examples, see Sagarin et al. 1999). In drawing the above conclusions, it should be noted that simulations in this study were based on data from one year and two time-periods only. However, data that were used for model parameterization consisted of several thousands of individuals and simulation outputs were similar to earlier long-term investigations of C. montagui and S. balanoides (Southward 1991; Jenkins et al. 2001; O Riordan et al. 2004; Svensson et al. 2004). This reassures the accuracy of the model and adds certainty to model predictions. Though it can always be argued that a model is too simple and thus not realistic. For example, in the present model near shore water Fig. 6. The population density of C. montagui (----) and S. balanoides ( ) at the end of period 1 (a,c,e) and period 2 (b,d,f) from model simulations with: (a,b) increased frequencies (0, 0 01, ) of low recruitment for S. balanoides; (c,d) increased frequencies (0, 0 01, ) of high recruitment for C. montagui; (e,f) increased frequencies (0, 0 01, ) of low recruitment for S. balanoides coupled with increased frequencies of high recruitment for C. montagui.

9 773 Modelling changes in species interaction Fig. 7. Proportion of free space at the end of period 1 (----) and period 2 ( ) from model simulations with increased frequencies (0, 0 01, ) of altered recruitment: (a) increased frequencies of low recruitment for S. balanoides; (b) increased frequencies of high recruitment for C. montagui; (c) increased frequencies of low recruitment for S. balanoides coupled with increased frequencies of high recruitment for C. montagui. circulation, which could increase or decrease larval supply, was not included. This could have led to a greater variation in the modelled recruitment as a response to increased sea surface temperature. On the other hand, part of this effect was implicitly incorporated in data. Secondly, C. stellatus, the third barnacle species commonly found together with C. montagui and S. balanoides was not included in the model. C. stellatus is, however, similar in biology to C. montagui (O Riordan et al. 2004) and also less abundant or absent at many shores (Range & Paula 2001). Finally, changes in adult survival was not included in the model scenarios, though it is known that higher temperatures can have negative effects on the adult survival of S. balanoides (Barnes & Powell 1950; Bertness 1989; Southward 1991). While we did not find such correlations in our data, it is possible that the survival of S. balanoides was slightly overestimated in model simulations. However, we were looking at the effects of changes in recruitment and we chose to keep the model as simple as possible but proficient enough for answering our two main questions. Not surprisingly, S. balanoides was shown to have significant effects, not only on free space, but also on the abundance of C. montagui. This demonstrates that competition can occur when free space is still available (Hixon et al. 2002). Simulations even showed that the amount of free space could become greater when competition is increased as a result of decreased recruitment for the larger species due to lower proportions of space during recruitment intervals (i.e. scenario 2). One explanation to the constant presence of free space in model simulations is that although settling is often gregarious on larger scales (i.e. metres and more), on small scales (i.e. cm to mm), larvae have been shown to settle at a distance from conspecifics and other species to avoid crowding during growth (Hawkins 1983; Jenkins, Norton & Hawkins 1999; Kent et al. 2003). This behaviour could be hidden in the recruitment estimates, which were measured some time after settlement. However, it is well known that interacting barnacles compete also as adults (Connell 1961). For example, high recruitment may promote lower adult survival due to crowding at high densities, creating patches of free space (Barnes & Powell 1950; Connell 1985; Roughgarden et al. 1985). Arguably the model outputs could have changed if competition between adults (i.e. density dependence) was included. On the other hand, the survival factor was not detached from the model and the elasticity analyses explicitly showed that the survival of adult S. balanoides had large effects on free space and C. montagui. Furthermore, Connell (1961) showed that many adult C. montagui die from unknown densityindependent causes and others have shown that when free space is available, mortality between the juvenile and adult stages of S. balanoides may be density-independent (Connell 1985; Svensson et al. 2004). We believe that this favours a method where stochastic data-based mortality rather than density-dependent mortality is included in the model. Mixing stochasticity and density dependence would, of course, improve the accuracy of the model but it would also make the model much more complicated, and require better data. Forde & Raimondi (2004) proposed that changes in recruitment for interacting species often influence community composition initially but that such changes are short-lived, suggesting that many communities function at a stable equilibrium. This has been shown a valid proposition for some ecosystems (e.g. Corbin & D Antonio 2004), yet, in this study, simulations showed that long-term disturbances in recruitment could substantially and permanently change the relative abundance of species. Increased frequencies of high recruitment for C. montagui (scenario 2) produced C. montagui population sizes equal to densities found in warmer regions (Range & Paula 2001) and, as a consequence, S. balanoides densities were reduced. Naturally these effects were stronger when increased frequencies of low recruitment for S. balanoides (scenario 3) were incorporated. Hence, although C. montagui is less abundant at higher latitudes, they have a competitive potential that could be realized with increased recruitment. This might explain why C. montagui have been advancing into areas formerly dominated by S. balanoides (Southward & Crisp 1954; Kendall & Bedford 1987; Southward 1991). Similar northerly expansions of marine

10 774 C. J. Svensson, E. Johansson & P. Åberg southern species connected to a warming of the sea surface are well documented in other parts of the northern hemisphere (Barry et al. 1995; Sagarin et al. 1999) and it is not surprising to find that southern populations of barnacles are extending northwards. However, this study shows in more detail how recruitment could be an important mechanism controlling these extensions, and, the consequences for species composition in these regions. Subsequent effects on ecosystem functioning will then depend on the biology of C. montagui in comparison with S. balanoides. For example, could C. montagui substitute S. balanoides as food and shelter to other organisms and how would a diluted larval release over a longer period (C. montagui) compare with a concentrated larval release during a restricted period (S. balanoides)? These questions are interesting, though not targeted in this study. Svensson et al. (2005) showed that isolated populations of S. balanoides were relatively resistant to increased frequencies of low recruitment. However, by including a natural competitor into the system (this study), the effects of lowered recruitment were much greater. This shows that the buffering qualities of high adult survival depend also on the interaction with other species (Brown et al. 2001). In some interactions the effects can be the opposite and species become even more resistant to environmental disturbances by compensatory actions from interacting species (Bertness & Ewanchuk 2002; Jiang & Morin 2004). In this study, however, the increased presence of C. montagui had direct negative effects on S. balanoides density and the same pattern can be found in Southward s time series, in which C. montagui and S. balanoides abundance at Cellar beach, south Devon have been monitored for 40 years in relation to the height on shore, solar cycle and inshore sea surface temperature (Southward 1991). Therefore, given the simplicity of the system, the known sensitivity of relative abundance to changes in sea surface temperature and how easy it is to monitor barnacles, these two species appear as good indicators of climate change at temperate latitudes. As such, C. montagui is a representation of other southern species that are extending their distribution northward (for some examples, see Barry et al and Sagarin et al. 1999). In terrestrial ecosystems, similar scenarios of low land trees or plants expanding beyond their transition zones on mountain ridges have been used as indicators of climate change (e.g. Hattenschwiler & Korner 1995; Heegaard & Vandvik 2004). Continuing to study and monitor marine counterparts to these systems, such as the interaction between C. montagui and S. balanoides, would provide valuable information for understanding and predicting, not only the effects of climate change, but also other globally occurring events, such as the introduction and establishment of introduced species (Ruiz et al. 1997, 1999). Acknowledgements This study was performed as a part of the project EUMAR (EU EVK3-CT ) and partly funded by the Adlerbertska Foundation. Data were attained from EUROROCK (EU MAS3-CT ) and we would like to thank Pedro Range for analysing the photographs. We also would like to thank Dr Stuart Jenkins for valuable comments on the manuscript and Dr Michael Burrows for valuable discussions about barnacles and climate change. This study complies with the current laws of the country in which it was performed. References Barnes, H. (1957) The northern limits of Balanus balanoides (L.). Oikos, 8, Barnes, H. (1958) Regarding the southern limits of Balanus balanoides (L.). Oikos, 9, Barnes, H. & Powell, H.T. (1950) The development, general morphology and subsequent elimination of barnacle populations, Balanus crenatus and B. balanoides, after a heavy initial settlement. Ecology, 19, Barry, J.P., Baxter, C.H., Sagarin, R.D. & Gilman, S.E. (1995) Climate-related, long-term faunal changes in a California rocky intertidal community. Science, 267, Beaugrand, G. & Reid, P.C. (2003) Long-term changes in phytoplankton, zooplankton and salmon related to climate. Global Change Biology, 9, Bertness, M.D. (1989) Intraspecific competition and facilitation in a northern acorn barnacle population. Ecology, 70, Bertness, M.D. & Ewanchuk, P.J. (2002) Latitudinal and climatedriven variation in the strength and nature of biological interactions in New England saltmarshes. Oecologia, 132, Brown, J.H., Whitham, T.G., Ernest, S.K.M. & Gehring, C.A. (2001) Complex species interactions and the dynamics of ecological systems: long-term experiments. Science, 293(5530), Burrows, M.T., Hawkins, S.J. & Southward, A.J. (1992) A comparison of reproduction in co-occurring chthamalid barnacles, Chthamalus stellatus (Poli) and Chthamalus montagui Southward. Journal of Experimental Marine Biology and Ecology, 160, Caley, M.J., Carr, M.H., Hixon, M.A., Hughes, T.P., Jones, G.P. & Menge, B.A. (1996) Recruitment and the local dynamics of open marine populations. Annual Review of Ecology and Systematics, 27, Carroll, M.L. (1996) Barnacle population dynamics and recruitment regulation in southcentral Alaska. Journal of Experimental Marine Biology and Ecology, 199, Caswell, H. (2001) Matrix Population Models, 2nd edn. Sinauer Associates, Inc. Publishers, Sunderland, MA. Connell, J.H. (1961) The influence of interspecific competition and other factors on the distribution of the barnacle Chthamalus stellatus. Ecology, 42, Connell, J.H. (1985) The consequences of variation in initial settlement vs. post-settlement mortality in rocky intertidal communities. Journal of Experimental Marine Biology and Ecology, 93, Connolly, S.R. & Roughgarden, J. (1999) Theory of marine communities: competition, predation and recruitmentdependent interaction strength. Ecological Monographs, 69, Connolly, S.R., Menge, B.A. & Roughgarden, J. (2001) A latitudinal gradient in recruitment of intertidal invertebrates in the northeast Pacific Ocean. Ecology, 82, Corbin, J.D. & D Antonio, C.M. (2004) Competition between native perennial and exotic annual grasses: implications for an historical invasion. Ecology, 85(5),

11 775 Modelling changes in species interaction Crisp, D.J. & Clegg, D.J. (1960) The induction of the breeding condition in Balanus balanoides (L.). Oikos, 11, Crisp, D.J., Southward, A.J. & Southward, E.C. (1981) On the distribution of the intertidal barnacles Chthamalus stellatus, Chthamalus montagui and Euraphia depressa. Journal of the Marine Biological Association of the United Kingdom, 61, Edwards, M., Beaugrand, G., Reid, P.C., Ashley, A.R. & Jones, M.B. (2002) Ocean climate anomalies and the ecology of the North Sea. Marine Ecology Progress Series, 239, Forde, S.E. & Raimondi, P.T. (2004) An experimental test of the effects of variation in recruitment intensity on intertidal community composition. Journal of Experimental Marine Biology and Ecology, 301, Fromentin, J.-M. & Planque, B. (1996) Calanus and environment in the North Atlantic. II. Influence of the North Atlantic Oscillation on C. finmarchicus and C. helgolandicus. Marine Ecology Progress Series, 134, Hattenschwiler, S. & Korner, C. (1995) Responses to recent climate warming of Pinus sylvestris and Pinus cembra within their montane transition zone in the Swiss alps. Journal of Vegetation Science, 6(3), Hawkins, S.J. (1983) Interactions of patella and macroalgae with settling Semibalanus balanoides. Journal of Experimental Marine Biology and Ecology, 71, Heegaard, E. & Vandvik, V. (2004) Climate change affects the outcome of competitive interactions application of principal response curves. Oecologia, 139, Hills, J.M., Thomason, J.C., Milligan, J.L. & Richardson, M. (1998) Do barnacle larvae respond to multiple settlement cues over a range of spatial scales. Hydrobiologia, 375/376, Hixon, M.A., Pacala, S.W. & Sandin, S.A. (2002) Population regulation: historical context and contemporary challenges of open vs. closed systems. Ecology, 83, Hyder, K., Åberg, P., Johnson, M.P. & Hawkins, S.J. (2001) Models of open populations with space limited recruitment: extension of theory and application to the barnacle Chthamalus montagui. Ecology, 70, Jenkins, S.R., Norton, T.A. & Hawkins, S.J. (1999) Settlement and post-settlement interactions between Semibalanus balanoides (L.) (Crustacea: Cirripedia) and three species of fucoid canopy algae. Journal of Experimental Marine Biology and Ecology, 236, Jenkins, S.R., Åberg, P., Cervin, G., Coleman, R.A., Delany, J., Hawkins, S.J., Hyder, K., Myers, A.A., Paula, J., Power, A.M., Range, P. & Hartnoll, R.G. (2001) Population dynamics of the intertidal barnacle Semibalanus balanoides at three European locations: spatial scales of variability. Marine Ecology Progress Series, 217, Jiang, L. & Morin, P.J. (2004) Temperature-dependent interactions explain unexpected responses to environmental warming in communities of competitors. Ecology, 73(3), Kendall, M.A. & Bedford, M.L. (1987) Reproduction and recruitment of the barnacle Chthamalus montagui at Aberystwyth (mid-wales). Marine Ecology Progress Series, 38, Kent, A., Hawkins, S.J. & Doncaster, C.P. (2003) Population consequences of mutual attraction between settling and adult barnacles. Ecology, 72, King, P.A., McGrath, D., Morgan, R., Fitzgerald, O., Mullins, P. & Raleigh, J. (1993) Reproduction and settlement of the barnacle Semibalanus Balnoides (L.) in Galway Bay. Proceedings of the Royal Irish Academy, 93B, Levin, S.A. (1992) The problem of pattern and scale in ecology; The Robert H. MacArthur award lecture. Ecology, 73, Menge, B.A. (2000) Recruitment vs. postrecruitment processes as determinants of barnacle population abundance. Ecological Monographs, 70, Minchinton, T.E. & Scheibling, R.E. (1993) Free space availability and larval substratum selection as determinants of barnacle population structure in a developing rocky intertidal community. Marine Ecology Progress Series, 95, O Riordan, R.M., Arenas, F., Arrontes, J., Castro, J.J., Cruz, T., Delany, J., Martinez, B., Fernandez, C., Hawkins, S.J., McGrath, D., Myers, A.A., Oliveros, J., Pannacciulli, F.G., Power, A.M., Relini, G., Rico, J.M. & Silva, T. (2004) Spatial variation in the recruitment of Chthamalus montagui Southward and Chthamalus stellatus (Poli) (Crustacea: Cirripedia) over an European scale. Journal of Experimental Marine Biology and Ecology, 304, Range, P. & Paula, J. (2001) Distribution, abundance and recruitment of Chthamalus (Crustacea: Cirripedia) populations along the central coast of Portugal. Journal of the Marine Biological Association of the United Kingdom, 81, Reid, P.C., Holliday, N.P. & Smyth, T.J. (2001) Pulses in the eastern margin current and warmer water off the north west European shelf linked to North Sea ecosystem changes. Marine Ecology Progress Series, 215, Root, T.L., Price, J.T., Hall, K.R., Schneider, S.H., Rosenzweig, C. & Pounds, J.A. (2003) Fingerprints of global warming on wild animals and plants. Nature, 421, Roughgarden, J., Iwasa, Y. & Baxter, C. (1985) Demographic theory for an open marine population with space-limited recruitment. Ecology, 66, Ruiz, G.M., Carlton, J.T., Grosholz, E.D. & Hines, A.H. (1997) Global invasions of marine and estuarine habitats by non-indigenous species: mechanisms, extent and consequences. American Zoology, 37, Ruiz, G.M., Fofonoff, P., Hines, A.H. & Grosholz, E.D. (1999) Non-indigenous species as stressors in estuarine and marine communities: assessing invasion impacts and interactions. Limnology and Oceanography, 44, Sagarin, R.D., Barry, J.P., Gilman, S.E. & Baxter, C.H. (1999) Climate-related change in an intertidal community over short and long time scales. Ecological Monographs, 69, Sanford, E. (1999) Regulation of keystone predation by small changes in ocean temperature. Science, 283(5410), Schwilk, D.W. & Ackerly, D.D. (2005) Limiting similarity and functional diversity along environmental gradients. Ecology Letters, 8, Southward, A.J. (1976) On the taxonomic status and distribution of Chthamalus stellatus (Cirripedia) in the North-east Atlantic region with a key to the common intertidal barnacles of Britain. Journal of the Marine Biological Association of the UK, 56, Southward, A.J. (1991) Forty years of changes in species composition and population-density of barnacles on a rocky shore near Plymouth. Journal of the Marine Biological Association of the United Kingdom, 71, Southward, A.J. & Crisp, D.J. (1954) Recent changes in the distribution of the intertidal barnacles Chtalamus stellatus Poli and Balanus balanoides L. in the British Isles. Journal of Animal Ecology, 23, Svensson, C.J., Jenkins, S.R., Hawkins, S.J., Myers, A.A., Range, P., Paula, J., O Riordan, R.M. & Åberg, P. (2004) Models of open populations with density dependent recruitment in stochastic environments: the relative importance of recruitment and survival in populations of Semibalanus balanoides. Marine Ecology Progress Series, 275, Svensson, C.J., Jenkins, S.R., Hawkins, S.J. & Åberg, P. (2005)