Introduction

Somatic production of macrozoobenthos is an important parameter for the study of ecosystem dynamics and a quantitative measure of a population’s function in the ecosystem (Dolbeth et al. 2005). In natural populations, secondary production is mainly a function of individual growth, recruitment patterns, and mortality observed in nature (Sardá 1997). Therefore, it is directly related to the life-cycle of any given species and a key parameter in population ecology (Sardá et al. 2000). Studies of secondary production have been used to improve the understanding of important ecological issues such as energy transfer within communities, monitoring of benthic communities (Wilber and Clarke 1998), rational management of aquatic resources (Downing 1984), and food web analyses (Benke 1998). Moreover, measures of production are informative because they represent the rate at which organic matter is made available to higher trophic levels and reflect the relative importance of organisms as consumers and nutrient recyclers (Taylor 1998).

Population production is often expressed as the annual production to biomass ratio (P/B) in hopes of yielding standardized production values that can be compared among species or populations with differing biomasses (Plante and Downing 1989). The P/B ratio, also called turnover rate, is generally correlated with different biotic and environmental variables such as life span, body weight, and temperature (Robertson 1979; Banse and Mosher 1980; Schwinghamer et al. 1986; Sprung 1993; Cusson and Bourget 2005). However, few estimates of productivity of isopods are available, particularly when compared to marine amphipods (sensu Cusson and Bourget 2005). Isopods are among the most abundant species of the sandy beach macrofauna. They are generally scavengers (McLachlan and Brown 2006) and regulate the energy flow in the beach ecosystem (sensu Hayes 1974). Thus, estimates of production of species of this group are important to understand the energy flow on sandy beaches.

Cirolanid isopods are conspicuous members of the supralittoral and intertidal fringes of sandy beaches around the world (McLachlan and Jaramillo 1995) and might be one of the dominant groups in terms of abundance, biomass, and production (Glynn et al. 1975; Dexter 1977; Lercari and Defeo 2003; Veloso et al. 2003; Petracco 2008). The genus Excirolana is the most ubiquitous intertidal invertebrate in low-latitude temperate, subtropical, and tropical sandy beaches of the Americas (McLachlan and Brown 2006). Excilorana armata, a representative of this genus, occurs along the Atlantic coast of South America, from the state of Rio de Janeiro, Brazil, to Buenos Aires Province, Argentina (Castro and Brum 1969). This highly substratum-specific species shows a clear preference for fine sand (Defeo et al. 1997; Yannicelli et al. 2002; Lozoya et al. 2010) and is abundant on dissipative and intermediate morphodynamic beaches. On the other hand, the other representative of Excirolana in Brazil, E. braziliensis, occurs on beaches with a wider range of morphodynamic states (Defeo et al. 1992, 1997; Giménez and Yannicelli 1997; Defeo and Martínez 2003; Petracco et al. 2010).

The scavenger E. armata is a primary consumer, exposed to high predation pressure (Lercari et al. 2010), and is used as food by fishes of economic importance (Bergamino et al. 2011). On Una beach, E. armata is one of the dominant species in terms of abundance and biomass and showed stable isotope signature of nitrogen close to the filter feeders (Petracco 2008; Petracco et al. 2010). Thus, considering the scarcity of information on the contribution of E. armata and other cirolanid isopods to the energy flow in beach ecosystems, and that estimates of production of E. armata population are essential to determine the amount of food available to subsequent links in the food web, this study aims to estimate the somatic and gonad productions of E. armata on this beach.

Materials and methods

Study area

Una beach (24°27′S; 47°06′W) is an exposed beach (sensu McLachlan 1980) located in the Ecological Station Jureia-Itatins, São Paulo state, Brazil. The sampling area is situated in the northern segment of the beach, with a dissipative morphodynamic state (Souza and Souza 2004) and gentle profile (mean slope = 2.46%; SD ± 0.27), with sands ranging from fine to very fine (mean grain size = 0.13 mm; SD ± 0.01). The mean surf zone temperature and salinity were 23.00°C (SD ± 2.15) and 32.77 (SD ± 1.77), respectively. For more details of the study area, see Petracco et al. (2010).

Sampling and laboratory procedures

Sampling was carried out monthly during spring low tides from December 2003 to November 2005. Three fixed transects (placed 8 m apart) were sampled from the base of the foredunes to the waterline. Between December 2003 and April 2004, sampling units were obtained every 8 m along each transect with a metallic cylinder (20 cm in diameter) and to a depth of 20 cm. From May 2004 to the end of the study, the sampling units were obtained every 4 m. The samples were sieved using a 0.5 mm mesh and the isopods retained were fixed in 4% buffered formalin. For logistic reasons, the samples of E. armata in February 2004 and January 2005 were not collected. Individuals were measured from the tip of the cephalon to the end of the telson and grouped into 0.5 mm size classes. Each individual was sexed under a stereomicroscope, according to Dexter (1977). Individuals were classified as juveniles (length < 3 mm; Petracco et al. 2010) and adults. Adults were categorized as females, ovigerous females (with presence of eggs/embryos), and males.

Eggs/embryos incubated in the marsupium of females collected in the first year (December 2003 to November 2004) were removed, counted, and categorized into four developmental stages according to Jones (1970) and Martínez and Defeo (2006) (see also Petracco et al. 2010). In order to obtain the ash-free dry weight, the individuals were pooled into 0.25 mm length classes and dried at 70°C for 48 h to measure the dry weight (DW). Ash weight (AW) was obtained after burning the dried individuals in a muffle furnace for 4 h at 500°C. Ash-free dry weight (AFDW) was calculated by subtracting the ash weight from the dry weight.

Length–weight relationship and secondary production

The relationship between length and weight for each sex was calculated by linear regression analysis, with the data converted into logarithms using the equation: logW = loga + b · logL, where W is the ash-free dry weight per individual (g AFDW), L is the length of the size class (mm), and a and b are constants.

Annual somatic production (P s) was estimated by the weight-specific growth rate method (Crisp 1984; Brey 2001) from the length-frequency distribution obtained from all pooled samples, the length-weight relationship, and the von Bertalanffy growth function parameters. The annual production is given by the equation:

$$ P_{\text{s}} = \sum {\sum {f_{\text{i}} \cdot w_{\text{i}} \cdot G_{\text{i}} } } , $$

where f i is the annual mean number of individuals in length class i, w i is the mean individual weight in the length class i, calculated from the mean length in length class i, and G i is the weight-specific growth rate in length class i obtained through the equation:

$$ G_{\text{i}} = b \cdot K \cdot [(L_{\infty } /L_{\text{i}} ) - 1], $$

where b is the exponent of the length-weight relationship, K and L ∞ are von Bertalanffy growth function parameters estimated for E. armata (see Petracco et al. 2010), and L i is the mean length in length class i. For the estimates of somatic production, the juveniles were divided equally in males and females, since the sex ratio did not differ significantly from 1:1 (Petracco et al. 2010).

Production was expressed in running meter (g and mg m−1 year−1) and also in square meter (g m−2 year−1) to enable it to be compared with other studies of production. The P s/B ratio was calculated by the ratio between annual somatic production (P s) and annual mean biomass (B). The somatic production and the P s/B ratio were estimated for the two sampling years separately; December 2003 to November 2004 (2004: first year) and December 2004 to November 2005 (2005: second year). Monthly productions of males and females were calculated from the monthly length-frequency distributions (see Petracco et al. 2010) obtained from samples pooled from the three transects. The monthly P s/B ratio of females was estimated from the production in each month and the respective biomass.

Annual gonad production (P g) was estimated for the first year of study (December 2003 to November 2004), according to Shafir and Field (1980), as the product of number of eggs/embryos per female and the monthly abundance of ovigerous females with broods in late developmental stages. Since the incubation period estimated for E. armata is longer than 1 month (ca 2 months; Petracco et al. 2010), only the ovigerous females bearing embryos in the developmental stages three and four described by Martínez and Defeo (2006) were considered to estimate the abundance of eggs/embryos in each month. Annual gonad production (P g) is given by the equation:

$$ P_{\text{g}} = \sum {A_{\text{i}} \cdot {\text{Ne}}_{\text{i}} \cdot {\text{We}}} , $$

where A i is the monthly abundance of ovigerous females (ind. m−1) bearing embryos in the developmental stages three and four in length class i; Nei is the number of eggs/embryos female−1 in length class i, obtained from the length-fecundity relationship (Petracco et al. 2010), and We is the individual weight of the last marsupial developmental stage (manca stage). The individual weight of the last marsupial developmental stage (We = 0.0445 mg AFDW) was obtained by measuring 50 individuals (L = 1.39 mm ± 0.16 SD) and from the length-weight relationship. The gonad production was estimated for the period between June and November 2004 because the number of ovigerous females sampled between December 2003 and May 2004 was very low (≤3 individuals/month) to estimate the percentage of ovigerous females bearing embryos at stages three and four. However, the fact that ovigerous female abundances between June and November totaled 90% of the annual abundance of ovigerous females allowed a reliable estimate of gonad production to be obtained. Moreover, in 2 years of study, E. armata showed the same proportion of ovigerous females (ca 25%; Petracco et al. 2010).

Correlation analyses were used to assess the relationship between monthly P s/B ratio of females and ovigerous female abundance, and between monthly P s/B ratio of females and monthly mean length of females. Nested one-way ANOVA was used to test the null hypothesis that there was no significant difference in the abundance of ovigerous females between years and among months, with months nested in years. Nested one-way ANOVA was also employed to test the null hypothesis that there was no significant difference in the mean length of females between years and among months.

The annual weight-specific gonad production (GP) was estimated by the ratio between the annual gonad production (P g) and the population mean annual biomass (GP = P g /B; Brey 1995). On the population level, this measure of reproductive output is a measure equivalent to the P s/B ratio, generally used to describe somatic productivity in benthic invertebrate populations (Brey 1995).

Results

The observed relationships between length and weight obtained for females (logW (g AFDW) = −4.074 + 2.458 · logL (mm), n = 54, r 2 = 0.89, P < 0.0001) and males (logW (g AFDW) = −4.718 + 2.558 · logL (mm), n = 56, r 2 = 0.94, P < 0.0001) were used for biomass and production estimates. The annual somatic production estimated for males (8.10 and 7.87 g AFDW m−1 year−1) and females (7.47 and 9.38 g AFDW m−1 year−1) resulted in annual somatic productions of population of 15.57 and 17.25 g AFDW m−1 year−1 for 2004 and 2005, respectively. These annual somatic values of production of population were similar to those calculated from the sum of monthly productions (16.91 and 17.10 g AFDW m−1 year−1 for 2004 and 2005, respectively). The mean annual biomass of males (2.01 and 2.47 g AFDW m−1) and females (2.41 and 3.03 g AFDW m−1) totaled mean annual biomasses of 4.42 (2004) and 5.50 g AFDW m−1 (2005). The somatic production of males and females was concentrated between August 2004 and March 2005 with a high peak in November 2004 (Fig. 1a).

Fig. 1
figure 1

a Monthly variation in somatic production of males and females; b monthly variation in somatic turnover rate (P s/B; open circles) of females and in abundance (closed circles; mean ± SE) of ovigerous females from December 2003 to October 2005 on Una beach

Full size image

The P s/B ratios for males ranged between 4.02 and 3.19 year−1 for 2004 and 2005, while, for females, the P/B ratio was 3.10 for both years. The P s/B ratios of the population ranged between 3.55 year−1 (2004) and 3.16 year−1 (2005). The monthly P s/B ratios of females were higher from January to May and lower between June and November in both years of study, i.e., there was an inverse pattern of temporal distribution of mean length of females and of abundance of ovigerous females (n = 21, r = −0.88, P < 0.0001 and n = 21, r = −0.58, P < 0.01, respectively; Fig. 1b). The abundance of ovigerous females and the mean length of females differed among months (nested ANOVA F 20,44 = 6.89, P < 0.0001; nested ANOVA F 20,44 = 4.87, P < 0.0001, respectively), but not between years.

The distribution of annual production for the length classes of males showed the same pattern in both years. A high peak of production in 3.5 mm length class and a high production in other intermediary length classes (4.0–4.5 mm) were observed (Fig. 2a, b). For the females, a different distribution pattern of production occurred between the years. However, high productions also occurred in intermediary length classes (3.5 and 4.5 mm) with a decrease in production in the 4.0 mm class (Fig. 2c, d).

Fig. 2
figure 2

Annual distribution of production (open circles) and abundance (bars) of a, b males and c, d females for different length classes from December 2003 to November 2004 (first year) and December 2004 to November 2005 (second year) on Una beach

Full size image

The gonad production (P g: 1.07 g AFDW m−1 year−1) amounted to 15 and 6% of female production and total population production (somatic + gonad production), respectively. A higher gonad production occurred in 5.5 and 6.0 mm length classes. The proportion P g/P s increased with individual size (ca 90% in the 7.5 mm size class; Fig. 3). The annual weight-specific gonad production (P g/B ratio) was 0.24 year−1.

Fig. 3
figure 3

Annual distribution of somatic production (P s), gonad production (P g), and the P g/P s ratio for females of different length classes from December 2003 to November 2004

Full size image

Discussion

High productions of males and females of E. armata on Una beach between August 2004 and March 2005 can be explained by the high abundance in this period (Petracco et al. 2010), since these variables were strongly related for both males (n = 21, r = 0.88, P < 0.0001) and females (n = 21, r = 0.87, P < 0.0001). The higher P s/B ratios estimated for males in relation to females, mainly in 2005, can be attributed to the high abundance of individuals of the 3.5 mm length class. These small individuals grow fast and, consequently, have a high weight-specific growth rate contributing to 30% of the annual production of males which is twice the percentage observed for female individuals of this class.

The somatic production of E. armata on Una beach was higher than the value estimated for this species on a dissipative beach situated in southern Brazil (Souza 1998), despite the similarity in annual mean biomass (ca 0.18 g AFDW m−2; Table 1). The higher production of E. armata in the Una beach was due to the higher curvature parameter (K) and asymptotic length (L ∞) (see Petracco et al. 2010) and, consequently, to the higher weight-specific growth rate (G).

Table 1 Somatic production (P s: g AFDW m−2 year−1), somatic turnover rate (P s/B: year−1), and life span (LS: years) of sandy beach peracarid crustaceans (Isopoda: I; Amphipoda: A) of different latitudes
Full size table

The P s/B ratios estimated for E. armata on Una beach and for E. braziliensis on Chilean (Zuñiga et al. 1985) and Uruguayan beaches (Lercari et al. 2010) are the highest values obtained for the genus Excirolana. The considerable P s/B ratios of E. armata on Una beach derived from fast growth (Petracco et al. 2010). Furthermore, the low life span of this population can explain these considerable values as P s/B ratios of peracarids were significantly correlated with life span (n = 27, r = −0.61, P < 0.001), but not with mean individual weight, seawater temperature, and latitude (Table 1). The lowest P s/B ratios of Excirolana populations (1.6–1.9 year−1) are similar to the values obtained for oniscid isopods (1.5 year−1), which showed very high life spans (4.0 years). On the other hand, the higher P s/B ratios of Excirolana (2.2–3.8 year−1) were similar to most of the estimates of amphipods (Table 1).

Life history traits of E. armata show latitudinal variation with a significant decrease in mortality from subtropical to temperate beaches (n = 11, r = −0.86, P < 0.001), while the life span and the mean individual weight follow the inverse pattern (n = 11, r = 0.89, P < 0.001 and n = 9, r = 0.97, P < 0.0001; Table 2). When four estimates of Z of E. armata (Lozoya and Defeo 2006:1.50–1.95 year−1) were used as proxies of P s/B ratios, the latter decreased significantly with the latitude and life span (n = 11, r = −0.67, P < 0.05 and r = −0.79, P < 0.01), but was not related with mean individual weight. Similarly, considering all Excirolana populations (Table 2), significant relations between latitude and the biotic variables, mortality, life span, and mean individual weight were observed. Similarly to E. armata, Excirolana P s/B was negatively correlated with life span (n = 20, r = −0.52, P < 0.05), but not related to the mean individual weight. However, Excirolana P s/B was not related to the latitude either. The relationships between P s/B ratios and life span verified for the peracarids, E. armata and Excirolana populations confirm that, for marine macroinvertebrates, life span is the variable which is most related to P s/B ratios (Cusson and Bourget 2005). As life span is strongly and negatively correlated with mortality and the latter is equal to the P s/B ratio (Allen 1971), a negative and strong relation between P s/B and life span is expected.

Table 2 Life span (LS: years), turnover rate (P s/B ratio: year−1), and instantaneous rate of mortality (Z: year−1) estimated for males (M), females (F) and populations of Excirolana armata (Souza 1998; Lozoya and Defeo 2006; Lercari et al. 2010; this study) and E. braziliensis (Zuñiga et al. 1985; Fonseca et al. 2000; Veloso et al. 2003; Defeo and Martínez 2003; Caetano et al. 2006; Lercari et al. 2010)
Full size table

Besides latitude, some factors acting on a local scale, such as inter and intraspecific interactions as well as the morphodynamic state of the beach, influence the growth and mortality of populations (Defeo and Martínez 2003; Lozoya and Defeo 2006) and, consequently, the P s/B ratios. For instance, at similar latitudes on Uruguayan beaches, contrasting P s/B estimates for E. braziliensis were obtained (Lercari et al. 2010: P s/B = 0.95 and 3.78 year−1). The low value of P s/B ratio of E. braziliensis can be attributed to the interspecific interaction with E. armata, which decreases the growth rate of both species (Defeo et al. 1997; Defeo and Martínez 2003). However, as verified for the other life history traits, P s/B ratios of Excirolana populations were expected to undergo a latitudinal variation, as verified for Donax populations (Cardoso and Veloso 2003; Herrmann et al. 2009). However, despite showing lower life span and higher mortality rate in relation to the Uruguayan Excirolana populations, unexpected lower and conservative P s/B estimates were observed for most of the Excirolana populations of Brazilian sandy beaches (P s/B ≤ 2 year−1) (Tables 1, 2). The productivity of Brazilian Excirolana populations were obtained from the weight-specific growth rate method (WSGRM), based on length-frequency distributions (LFDs), as opposed to the estimates for Uruguayan Excirolana populations, which were obtained by balanced ecosystem models (Lercari et al. 2010). Thus, it is possible to refer to the LFDs to understand these conservative P s/B ratios. The LFDs of E. armata (Fig. 2a, b, c, d) and E. braziliensis (Petracco 2000; Defeo and Martínez 2003) show low proportions of juveniles. This low proportion of juveniles and the consequent decrease in P s/B ratio can be explained by: (1) fast growth of juveniles, which makes it difficult to accurately represent the abundance of these individuals by monthly sampling, and (2) sampling selectivity against smaller organisms due partly to their possible escape through the mesh (Defeo, pers. com.).

According to Allen (1971), the instantaneous mortality rate (Z) of a population is equal to the P s/B ratio of such population if the individual growth of population is describable by the von Bertalanffy growth function. However, Z estimates calculated with the length-converted catch curve (LCCC) (Pauly 1983) are higher than the P s/B ratios obtained with the WSGRM when the population shows gear selectivity against smaller individuals (Brey, pers. com.). The high values of r 2 (0.88–0.99) of the LCCC fitted for E. armata and E. braziliensis (Defeo and Martínez 2003; Caetano et al. 2006; Petracco 2008) show that the single negative exponential model is appropriate to estimate the mortality of these species and provides reliable estimates of Z. Specifically for E. armata on Una beach, the similarity of the initial number of individuals calculated from the LCCC of males and females (330) and the number of individuals estimated from the annual number of embryos (450) reinforce this idea. The Z estimated for Excirolana populations were significantly higher than the P s/B ratios calculated with the WSGRM (paired t test: t 1,14 = 4.12, P < 0.001; Table 2). Moreover, the relationship between these two variables described by a significant linear model shows a significantly lower slope than that of Allen’s relation (Z = P s/B, slope = 1; t 1,22 = 2.66, P < 0.05; Fig. 4). The difference between the P s/B ratio and Z reinforces the hypothesis of underestimation of P s/B ratios for Brazilian Excirolana populations due to a sampling selectivity and/or an inadequate (monthly) sampling regime.

Fig. 4
figure 4

P/B ratios obtained with the weight-specific growth rate method plotted against the instantaneous mortality rate (Z) (dashed line) obtained for males (filled circles) and females (open circles) of Excirolana armata (Souza 1998; this study), and for males (filled square), females (open square) and populations (triangle) of E. braziliensis (Zuñiga et al. 1985; Fonseca et al. 2000; Veloso et al. 2003; Caetano et al. 2006); Allen’s relationship: P/B = Z (full line)

Full size image

From the negative relationship between the monthly P s/B ratios of females and ovigerous female abundance, it is possible to characterize two distinct annual periods of P s/B ratio of females. In the period with lower abundance of ovigerous females, females show lower lengths and, consequently, higher P s/B ratios than in the period characterized by higher abundance of ovigerous females.

Despite the fact that fecundity increases with the length of females (Petracco et al. 2010), higher gonad production occurred at 5.5 and 6.0 mm length classes, due to the higher abundance of ovigerous females in these classes (mean length of ovigerous females = 5.87 mm). However, the higher proportion of gonad production compared with the somatic production at larger length classes (6.0–7.5 mm) derived from low somatic production in these classes because older individuals invest more energy in reproduction than in growth.

The contribution of gonad production, compared with the total population production (somatic + gonad production) estimated for E. armata (6%), was lower than those observed for the cirolanids Cirolana imposita (15%; Shafir and Field 1980) and C. harfordi (13%; Johnson 1976a). However, when comparing the gonad production of E. armata to the total production of females calculated from the length reached by females at sexual maturity (length ≥ 4 mm), a considerable contribution of gonad production (ca 20%) was observed. According to Shafir and Field (1980), the high percentage of productive energy channeled into reproduction of C. imposita was strongly supported by the fact that ovigerous females were present in samples throughout the year. C. harfordi also showed ovigerous females throughout the year and in high percentage (Johnson 1976b). In E. armata, low abundance of ovigerous females was observed over a long period of time (December 2003 to May 2004), although ovigerous females occurred throughout the year and with a considerable mean percentage (25%). This fact contributed to the smaller relative contribution of gonad production of E. armata when compared with the other two cirolanids.

The P g/B ratios of E. armata (0.24 year−1) and C. harfordi (0.30 year−1; Johnson 1976a) were similar to each other but much lower than that estimated for C. imposita (0.86 year−1; Shafir and Field 1980). This similarity between E. armata and C. harfordi P g/B ratios is not in accordance with the positive relation between P g/B ratio and temperature (Clarke 1987; Brey 1995). C. harfordi and C imposita live at a seawater temperature (15°C), which is lower than the temperatures measured on Una beach (23°C). However, the fact that E. armata and C. harfordi live in the intertidal zone of sandy beaches, while C. imposita occurs in kelp beds in depths between 12 and 16 m, may contribute to the difference in P g/B ratios. In populations of intertidal invertebrates, energetic costs increase with aerial and wave exposure (Sebens 2002). Moreover, food availability influences positively the P g/B ratio (Clarke 1987; Brey 1995). A kelp bed is a high productive ecosystem, mainly when compared with the intertidal zone of sandy beaches. A higher P s/B ratio estimated for C. imposita (4.80 year−1), when compared with E. armata, despite the higher seawater temperature on Una beach, indicates fast growth, supported by high food availability typical of the environment of kelps.

In summary, the high production and P s/B ratios of E. armata in relation to other Excirolana populations suggest that this species is an important component of the energy flow on Una beach. However, like other Excirolana populations of lower latitudes, the productivity of E. armata on Una beach is probably underestimated due to the low percentage of juveniles in the population. Therefore, fortnightly sampling frequency and mesh with smaller aperture than 0.5 mm must be employed in studies on the production of this genus, mainly on tropical/subtropical beaches where the individuals recruit with smaller length and grow faster than their relatives on temperate beaches (Cardoso and Defeo 2003, 2004; Petracco et al. 2010). The considerable contribution of gonad production (~20%) compared with the total production of females at sexual maturity (length ≥4 mm) shows that gonad production must be included in production studies of species of the genus Excirolana.