Body Size, Growth and Life Span: Implications for the Polewards Range Shift of Octopus tetricus in South-Eastern Australia

Understanding the response of any species to climate change can be challenging. However, in short-lived species the faster turnover of generations may facilitate the examination of responses associated with longer-term environmental change. Octopus tetricus, a commercially important species, has undergone a recent polewards range shift in the coastal waters of south-eastern Australia, thought to be associated with the southerly extension of the warm East Australian Current. At the cooler temperatures of a polewards distribution limit, growth of a species could be slower, potentially leading to a bigger body size and resulting in a slower population turnover, affecting population viability at the extreme of the distribution. Growth rates, body size, and life span of O. tetricus were examined at the leading edge of a polewards range shift in Tasmanian waters (40°S and 147°E) throughout 2011. Octopus tetricus had a relatively small body size and short lifespan of approximately 11 months that, despite cooler temperatures, would allow a high rate of population turnover and may facilitate the population increase necessary for successful establishment in the new extended area of the range. Temperature, food availability and gender appear to influence growth rate. Individuals that hatched during cooler and more productive conditions, but grew during warming conditions, exhibited faster growth rates and reached smaller body sizes than individuals that hatched into warmer waters but grew during cooling conditions. This study suggests that fast growth, small body size and associated rapid population turnover may facilitate the range shift of O. tetricus into Tasmanian waters.


Introduction
The distribution and abundance of marine species depends on their functional traits and associated biotic factors, i.e. population genetic structure and gene flow [1], physiological limits [2][3][4], phenotypic plasticity [5], dispersal ability [6,7], and intra and inter-specific interactions [8,9]. These functional traits and biotic factors are in turn modulated by abiotic factors such as temperature, oxygen and pH [2,10]. Temperature is by far the easiest abiotic factor to record and therefore the most studied environmental variable. Moreover, all aspects of ectotherm behaviour and physiology are sensitive to environmental temperature [11], and species changes in distribution in response to climate change are thought to be largely driven by fluctuations in temperature [12]. However, there are substantial inter-specific differences in the magnitude of responses to such temperature variability [4], and we have little knowledge about the processes responsible for the vast variation in species responses. Some studies have suggested that in response to ocean warming, marine species with short lifespans, high genetic diversity, high dispersal capacity, e.g. with a planktonic larval stage or high migration potential, and that live near their upper thermal limit may be more able to change their distribution as they track their optimum thermal conditions [4,11,13,14].
Long-term data sets appropriate to examine the response in life history parameters of long-lived species that may be undergoing climate-driven range shifts are rarely available [15]. In contrast, ecologically and commercially important cephalopods [16] may facilitate the examination of such life history parameters as a function of their generally short lifespan [17]. Life histories of cephalopods are extremely flexible under changing environmental conditions [18], largely due to the effect of temperature on growth [19], size at maturity [20], hatchling size [21], as well as social and behavioural aspects of courtship, mating, and egg-laying [22]. The combination of temperature-driven flexibility in life-history and the short lifespan of cephalopods may be critical for their capacity to thrive under ocean warming.
The gloomy or common Sydney octopus, Octopus tetricus, is a merobenthic species with a planktonic paralarval stage of 2.260.01 SE mm at hatching size (Ramos et al. unpublished data) that is subjected to ocean currents. The duration of the paralarval stage before settlement is unknown for O. tetricus but it is assumed to be similar to that of closely related species [23], i.e. 35-60 days for O. vulgaris reared under laboratory conditions [24,25]. Octopus tetricus reaches an approximate arm-span of 2 m [26,27]. This species is commonly distributed in temperate waters of the east coast of mainland Australia, from southern Queensland to southern New South Wales as suggested by scientific surveys [27,28]. However, its distribution has extended polewards to south-eastern Australia, along the coasts of Victoria (see [29]) after 2000 approximately, and eastern Tasmania in 2006 (as reported by fisheries data [30] and supported by citizen science monitoring using scientist-verified and geo-referenced photographs [31,32]; Fig. 1). This polewards shift in distribution, like many others in the same area (e.g. see [32,33]), is thought to be related to the southern extension of the warm East Australian Current (EAC) [34,35] and is consistent with expected changes in distribution promoted by climate driven warming [12]. The EAC flows from the southern Coral Sea and reaches the south-east coast of mainland Australia [36]. Over the past 60 years the EAC has extended approximately 350 km further south, along the relatively cool east coast of Tasmania [34,35]. This extension of the EAC has resulted in the southern Tasman Sea warming at a rate of three to four times the global average, with the ocean temperatures in the region projected to increase by 3uC by 2070 [37].
Like other cephalopods that are key components of trophic webs [16], O. tetricus may play an important ecological role in transition of the energy flux from low to high trophic levels. For example, by competing with other octopus species for ecologically and commercially important prey species [38][39][40]. Furthermore, O. tetricus is an important target of the octopus fishery in its historical distribution [26] and now also in the leading edge of its range extension, contributing 14% of the octopus catch. Thus, O. tetricus provides a good opportunity to examine the influence of environmental variability on life history characteristics, in particular growth, body size and lifespan; and how such characteristics may influence its capacity to become established in the new sections of its range, i.e. if growth rate and population turnover are fast, the establishment of the population will be favoured in the range extension.
Therefore, this study aims to examine the effect of environmental variables on the life history characteristics of O. tetricus at the southern edge of the recent range shift, in particular the size structure, growth rates and lifespan, and discuss how these characteristics may potentially influence the current and future establishment of this species in Tasmanian waters.

Ethics statement
This research was conducted under the University of Tasmania Animal Ethics Committee, permit approval no. A11591. No specific collection permits were required given that Octopus tetricus is not an endangered or protected species, and specimens were provided by commercial fishers. The whole animals were frozen on board at 220uC. Specimens collected during February and May 2011 were preserved on board in 80% ethanol. Individuals were dissected in the laboratory and eviscerated total wet weight (TW) and mantle weight (MW) (g) were recorded. Mantle weight was preferred over TW because missing and incomplete arms from many individuals provided an underestimate of TW. However, TW was recorded to facilitate comparison with previous studies. Measurements and weights were recorded to the nearest 0.1 cm and 0.01 g respectively.

Collection of wild caught specimens
To correct weights of 220uC frozen and 80% ethanol preserved individuals, sections of mantle tissue of 1 cm width 62 cm length of different wild caught individuals were weighed fresh and frozen at 220uC (n = 100) or preserved in 80% ethanol (n = 86). Weight was recorded again after the same period of time that passed between collection of wild caught specimens and weighing in the lab.
Sex and maturity (immature, mature and spent) were determined based on the macroscopic characteristics of the gonads. The maturity scale was modified from previous studies [41,42].

Age estimation
Stylets, the vestigial shells of octopods, have recently been used with success to estimate age and growth rates [43][44][45][46][47][48]. Stylets were removed from the mantle of fresh, frozen or 80% ethanol preserved specimens and stored in 70% ethanol. Stylets of all specimens (n = 527) were cut, embedded, ground, and polished following [46] with slight modifications: Two to five pictures, depending on section diameter, were sequentially taken from the nucleus to the edge of the section at either 6100, 6200 or 6400 magnification (Fig. 2) using the software Leica Application Suite (LAS) v. 3.6.0 (Leica Microsystems, Switzerland) with a transmit- ted-light microscope Leica DM LB2 connected to a digital camera Leica DFC420. Pictures were sequentially stitched together and daily increments identified following [43]. Two non-consecutive increment counts were made by one reader using key counter software (KeyCounter v. 1.1.0) and a third count was carried out by a second reader. Recorded number of growth increments was considered as the mean of the three counts. Stylet sections were discarded (n = 313) if growth increments were not clear along the section, if more than 10% of the section was unable to be counted, and if the three counts differed by more than 10%. The daily periodicity of growth increments was assumed in this study as it has been validated or assumed for holobenthic [43] and merobenthic octopods [46][47][48].

Oceanographic data
Monthly average Sea Surface Temperature (SST) and Chlorophyll-a (Chl-a) concentration of the sampling area was obtained from the MODISA satellite imagery at a 4 km scale (http://oceandata.sci. gsfc.nasa.gov/MODISA/Mapped/Monthly/4 km/). Sea surface temperature was considered a valid temperature estimate of the habitats occupied by Octopus tetricus because this species was collected at shallow depths (35-46 m) where wind driven mixing is high and tidal currents are strong [49]. Chlorophyll-a concentration is an estimator of primary productivity often with strong links to the abundance of higher trophic levels [50,51].

Data analyses
Paired sample t-tests were used to assess significant differences between fresh and frozen, and fresh and 80% ethanol preserved samples. Model II linear regressions were conducted when necessary to adjust frozen and 80% ethanol weights so they were comparable to fresh weights. All data was Box-Cox transformed when necessary using the ''car'' package in R v. 3.0.1 [52,53]. Normality was determined using Shapiro-Wilk's test and homogeneity of variances evaluated by visual inspection of residual plots.
Hatch month of each individual was back-calculated, by subtracting an individuals estimated age (days) from its date of capture. Hatch months were grouped into a 'warm season' including December to May (17.660.38uC SE and 0.7060.04 mg m 23 SE, n = 12) and a 'cool season' of June to November (13.260.30uC SE and 0.9060.07 mg m 23 SE, n = 12) (following [54]). Maximum life span was considered a proxy of population turnover.
The Gompertz, exponential, power and linear growth models were generated for mantle weight of females and males pooled by season of hatching. The 3-parameter Gompertz growth model had the smallest Akaike Information Criterion (AIC) and Akaike weight (wAIC) closest to 1 [55,56] using the package ''qpcR'' in R v. 3.0.1 [53,57] and was identified as the model that best fitted the size (MW) at age data ( Table 1). The Gompertz growth model was constructed using the non-linear weighted least square method following [58]: Where m is mantle weight (g); a is age (days); m ' is the asymptote parameter in m(a) (g); c is the shape parameter in m(a); g 1 is the rate coefficient parameter in m(a) (day 21 ); m 1 and m 2 are location parameters in m(a) (g), or predicted mantle weight at minimum or maximum observed age; a 2 is the maximum observed age; m is the inflection point parameter in m(a) (g) and SE is the standard error. The three parameters to estimate are m ' , c and g 1 . The 95% confidence intervals for the coefficients of the Gompertz growth model were estimated by bootstrapping using the package ''car'' in R v. 3.0.1 [52,53]. The F-statistics was calculated through an analysis of residual sum of squares (ARSS) to compare Gompertz growth models between genders and among warm and cool hatching seasons [59]. The instantaneous relative rate of growth (G) for the Gompertz model was estimated following [58]: G~ĉ cĝ g 1 e {g 1 a A two-way ANOVA was used to compare differences in MW of mature females, and mature and spent males, as well as to compare differences in age of mature and spent females and mature and spent males among warm and cool seasons of hatching. All statistics and models were carried out using R v. 3.0.1 [53].
Growth rates, body size and life span at hatch seasons Individuals collected in 2011 hatched throughout 2010 and 2011. Greatest numbers of these animals hatched in 2010 when SST was at coolest and Chl-a concentration highest (Fig. 4). A second hatching peak was observed in January 2011 when SST was increasing and Chl-a concentration was decreasing. The estimated average instantaneous relative growth rateĜ G of all O.
tetricus in north-eastern Tasmania was 0.01460.0006 SE day 21 , n = 214 ( Table 2). The ARSS indicated that growth models   differed between hatching seasons (F (3,211) = 7.03, P,0.0001; Fig. 5). For instance, instantaneous relative growth rate of cool hatched animals was significantly faster than the instantaneous relative growth rate of warm hatched animals (

Discussion
This study demonstrates that Octopus tetricus has a fast growth rate, small body size and a short lifespan of approximately 11 months, even at the cooler leading edge of its polewards range extension. These characteristics correspond to an r-selected life history strategy, which would facilitate the apparent rapid population expansion of this species and assist the 'invasion' into new environments [60,61]. Fast growth rates and short lifespan, combined with successful reproduction, i.e. mating, high fecundity and production of viable embryos (Ramos et al. unpublished data), may underpin a capacity for O. tetricus to quickly increase the size of the emerging population in the zone of the range extension. Additionally, such a short lifespan and associated high population turnover may give O. tetricus, most likely an efficient generalist predator at the population level like most other octopus species [62,63], a competitive advantage in the short term (see [64]) over the longer-lived species already found within the new range area.
If food supply is not limited, octopuses from cooler waters are expected to grow slower during the exponential phase of growth and reach maturity at larger sizes compared to octopuses from warmer waters [65,66]. In contrast, most individuals collected in the relatively cooler Tasmanian waters during 2011 (annual average 15.362.4uC SD, n = 12 months), were quite small (, approx. 400 g TW) with the maximum of 2.3 kg TW measured, compared with larger individuals (.3 kg of TW, S. Montgomery pers comm) reported from the warmer New South Wales waters (annual average 20.362uC SD, n = 12 months). Life history characteristics, i.e. growth rates, body size and life span, may differ over the distribution of a species, and may even diverge at the extension of the species distribution [61] possibly a function of reduced genetic diversity [67], or altered as an adjustment to the new physical environment [11], or to different community interactions [68]. For example, body size may be smaller in the region of range extension even though life-history theory predicts body size should be larger in cooler waters. Alternatively, the use of shelter pots may have led to aggregation of mature females or limited the body size of O. tetricus collected in Tasmanian waters. In contrast, the use of trawl nets in New South Wales would not lead to aggregation of mature females or limit the body size of captured octopuses. However, additional evidence suggests that the age at sexual maturity (206626 days SD, n = 214; Ramos et al. unpublished data) and time for egg laying and embryo develop-  A short lifespan can facilitate rapid population turnover. Selection acts on biological traits of every generation [72]; thus, favoured genotypes are likely to be selected more often in species with shorter generation times [73,74] due to greater probability of occurrence of mutations or formation of new gene complexes [75]. In this sense, it is possible that the combined effects of small body size, short life span, and likely rapid adaptation to environmental changes and biotic pressures may allow exploitation of niches, which may facilitate the establishment of O. tetricus in the leading edge of the range shift into Tasmanian waters.
The 3-parameter Gompertz growth model was an appropriate fit for size at age of O. tetricus. This model has adequately described non-linear relationships for growth estimations for other cephalopods taxa, e.g. squids [58,76], and is simpler than other models, i.e. the 4-parameter Schnute growth model [58]. To our knowledge, this is the first study that suggests the Gompertz model as the best fit for growth of an octopod. Therefore, it is not possible to compare with growth models of other octopods. In contrast, the instantaneous growth rate estimated in this study is comparable only to those estimated during the exponential growth of octopods using the equation (G = (lnW 2 -lnW 1 )/(t 2 -t 1 )) by [19,58]. Octopus tetricus shows similar growth rates compared to wild caught octopods in their historical distribution, e.g. 0.01160.003 SE day 21 , n = 628 for O. vulgaris [69], and octopods in captivity, e.g. 0.01460.0004 SE day 21 , n = 18 for O. pallidus [77], 0.01860.002 SE day 21 for O. maya, n,40 [78], and 0.03660.005 SE day 21 , n = 84 for O. bimaculoides [65]. Similarity of growth rates suggests that the growth rate of O. tetricus in the area of the range extension may not be negatively impacted, still allowing fast growth rates and promoting a short life span and rapid population turnover. In this sense, fast growth rates may facilitate the establishment of O. tetricus at the range extension in Tasmanian waters.

Influence of environmental factors on growth rates
The estimated frequency of hatched individuals may have been masked by gaps in the collection of specimens during some months, in addition to natural processes such as predation, natural mortality, etc. that were not accounted in this study. With this in mind, our results show that O. tetricus hatched throughout the year, with greater number of hatched individuals during the cool and highly productive season. Individuals that hatched in cool and under higher Chl-a concentration experienced warming conditions later in their life. Those individuals grew faster and achieved smaller body sizes than individuals that hatched in warm conditions, under low Chl-a concentration, and grew during cooling conditions (Fig. 6). Thus, the initial greater peaks of Chl-a (and inferred greater availability of food), combined with increasing temperatures after hatching may be related to the faster growth rate of individuals hatched in cool conditions. Similarly, reproductive events of O. vulgaris seem to be synchronized with local events of high productivity [79][80][81] that eventually may benefit hatchlings with greater availability of resources [51]. Likewise, squids have experienced faster growth rates in cool waters and this has been attributed to increased productivity or availability of food (Todarodes angolensis [82], Loliolus noctiluca [83], Loligo opalescens [84]).
Size at age variability was observed as aged increased. Individual growth variability has also been noted in other studies [48,58,71,85]. Such variability in response to environmental factors, particularly to temperature, has been extensively studied in squids [19]. Yet, the relationship of growth in octopods and environmental variability is far from understood [86] and other biological factors such as gender, reproduction, genetics and physiological traits must be considered. For instance, differential growth of females and males has been observed during the slower phase of growth with the start of sexual maturity [19], which may explain the different growth rates observed between females and males that hatched in the warm season. This suggests that SST, availability of food, and probably other environmental and biotic factors influenced growth rates at different levels. Therefore, growth must be studied considering the interaction of environmental and biotic factors, in order to disentangling their individual effects.
This study has provided biological information on age and growth of O. tetricus at the leading edge of a recent and rapid range shift. The synchrony of hatching events with environmental conditions, such as availability of higher food concentration and warmer temperatures, appears to have a large influence on growth rates, body size and life span of this octopus. Instantaneous growth rates of O. tetricus were similar to those of other octopods within their historical range of distribution or reared in captivity. The estimated life span of 11 months may allow O. tetricus to cope with environmental variability and possibly facilitate exploitation of available niches. Additional studies on population linkages, reproductive biology (e.g. Ramos et al. unpublished data), trophic ecology, thermal physiology, and dispersal or migration capacity are essential to develop a more complete understanding of the capacity of a species to alter its range and comprehend the biological and ecological mechanisms that underpin that extension.