This article was downloaded by: [203.109.163.22] On: 10 March 2015, At: 18:58 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Invertebrate Reproduction & Development Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tinv20 Spatial variation in reproduction in southern populations of the bivalve ventricosa (Veneroida: ) Kendall Gadomskia & Miles Lamarea a Department of Marine Science, University of Otago, PO Box 56, Dunedin 9054, New Zealand Published online: 13 Feb 2015.

Click for updates

To cite this article: Kendall Gadomski & Miles Lamare (2015): Spatial variation in reproduction in southern populations of the New Zealand bivalve (Veneroida: Mesodesmatidae), Invertebrate Reproduction & Development, DOI: 10.1080/07924259.2015.1007216 To link to this article: http://dx.doi.org/10.1080/07924259.2015.1007216

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions Invertebrate Reproduction & Development, 2015 http://dx.doi.org/10.1080/07924259.2015.1007216

Spatial variation in reproduction in southern populations of the New Zealand bivalve Paphies ventricosa (Veneroida: Mesodesmatidae) Kendall Gadomski and Miles Lamare* Department of Marine Science, University of Otago, PO Box 56, Dunedin 9054, New Zealand (Received 25 November 2014; accepted 9 January 2015)

Paphies ventricosa is a large surf endemic to New Zealand, and whose populations have substantially declined dur- ing the past century owing to overfishing and habitat degradation. Poor recruitment is now evident, and therefore, under- standing the reproductive patterns of P. ventricosa is a key to developing and implementing conservation strategies for the species. This study examines the reproductive cycle of P. ventricosa over one year in a population at Oreti Beach, Southland, the southernmost known extent of the species. At the same beach, we quantify spatial variation in reproduc- tion among four sites using quarterly surveys. Reproductive status is quantified from body indices and histological exam- ination of gonads. Based on changes in oocyte sizes, gametogenic stages and condition index, we observed a species with a primary spawning in spring and a second spawning event in autumn, with no resting phase but minimal reproduc- tive activity over winter. Seasonal reproduction corresponded with warmer sea surface temperature and a peak in chloro- phyll-a concentrations in the region. Small-scale (<15 km) variation in the timing of spawning was also evident along Oreti Beach, and these patterns maybe an important consideration when identifying areas that may be considered for conserving source populations. Keywords: gametogenesis; surf clam; condition index; histology; sea temperature

Introduction substantial overfishing in the first half of the twentieth Paphies ventricosa (Veneroida: Mesodesmatidae) Gray century (Rapson 1954), habitat degradation (Beentjes fi 1843, known by the Māori name toheroa, is a large surf et al. 2006) and disturbance from beach traf c are con- clam (up to 150 mm shell length) endemic to New Zealand. tributors (Brunton 1978; Williams et al. 2013). fi The species has a broad latitudinal distribution ranging Recruitment has been identi ed as a key issue for the between 34.5° and 46.3°S, although only three substantial species (Beentjes et al. 2006; Morrison & Parkinson populations now exist, these being physically separated by 2008), and the production and supply of P. ventricosa up to 700 km (Beentjes et al. 2006). The species is larvae may be limited due to the small and fragmented restricted to high-energy dissipating beaches that are on nature of present-day populations. A key aspect in under- western-facing fully exposed coasts made of fine sand standing recruitment in bivalves is quantifying processes (~0.25 mm average grain size) (Rapson 1952;Cassie controlling larval supply, which includes a detailed Downloaded by [203.109.163.22] at 18:58 10 March 2015 1955). Adult size classes (>90 mm) are concentrated at understanding of the spatial and temporal patterns in mid-tide levels, sub-adults (40–90 mm) are higher on the reproductive cycles and gamete production. beach, whilst the juveniles (<40 mm) have a wider and Reproduction has been examined in a range of more even vertical distribution across the tidal range but are temperate bivalve orders such as Veneroida (Newell & most abundant higher up the beach (Beentjes et al. 2006; Bayne 1980; Harvey & Vincent 1989; Gaspar et al. Beentjes & Gilbert 2006). 1999), Mytiloida (Wilson & Hodgkin 1967), Ostreoida P. ventricosa population sizes throughout New (Ruiz et al. 1992), and Myoida (Zaidman et al. 2012). Zealand have decreased substantially over the past cen- This includes a number of bivalve species inhabiting tury. For example, Greenway (1972) estimated that the New Zealand soft sand environments, such as northern population has declined from an estimate of largillierti (Gribben et al. 2001), Panopea zelandica 10,000,000 individuals in 1964 to 1,000,000 in 1971. (Gribben et al. 2004) and Zenatia acinaces (Gribben Southern populations have also experienced similar 2005). For the New Zealand species, reproduction fi declines, with Beentjes et al. (2006) reporting population follows relatively well-de ned annual cycles with spawn- sizes reduced from 2.2 million in the mid-1970s to ing occurring within the spring and summer period. 78,000 individuals larger than 100 mm by 1990. The Of interest in this study is the family Mesodesmati- reasons for the decline are poorly understood, but dae (: Veneroida), in which the reproduction has

*Corresponding author. Email: [email protected]

© 2015 Taylor & Francis 2 K. Gadomski and M. Lamare

been described in a small number of species including threatened bivalve species such as P. ventricosa, where the Australian Donacilla cuneata (Roberts 1984), Asian knowledge of spatial and temporal patterns of reproduc- Coecella chinensis (Kim et al. 2013), the South tion can be used in conservation strategies to maximise American donacium (Peredo et al. 1987; future recruitment by identifying larval source popula- Riascos et al. 2009; Uribe et al. 2012) and Mesodesma tions (Gaines et al. 2010). mactroides (Herrmann et al. 2009), and the tropical spe- cies Atactodea striata (Baron 1992). Reproduction in New Zealand mesodematids, all in the Paphies, Materials and methods has been described for P. ventricosa (Redfearn 1974), Study site and collection P. australis (Hooker & Creese 1995), Paphies donacina Reproduction in P. ventricosa was examined at four sites (Marsden 1999) and Paphies subtriangulata (Grant & along Oreti Beach, Southland, New Zealand. Oreti Beach Creese 1995). is a south-west facing, 26 km long, dissipating surf Most of what it known about P. ventricosa reproduc- beach on the shores of Foveaux Strait (Figure 1). Each tion comes from the North Island. Hoby (1933) found end of the beach is bordered by two river/estuary sys- – μ that egg diameter increased from June (59 65 m) to tems. – μ October (75 80 m) and that sperm activity increased To examine reproduction over one year, 20 mature “ ” “ from fairly active at the end of July to extremely were collected from a single site (Site B) every ” active in October. Additionally, Hoby (1933) found that month between January 2011 and December 2011, at the beginning of October few animals had spawned, excluding April 2011 (Figure 1). The following year but by the end of the month, most animals had spawned. (2012), spatial and seasonal patterns of reproduction in fi Redfearn (1974) identi ed four stages of gametogenic P. ventricosa along Oreti Beach was examined at four fi development in P. ventricosa, classi ed as early active sample sites along the length of the beach (Site A–D) (EA), late active (LA), mature and partially spawned/ for autumn (March), winter (June), spring (September) spent. The study found two to three spawning events and summer (December) (Figure 1). For this, 20 animals occurred during the summer season, with an initial were collected per site per season. All animals collected maturing period in September followed by a second were greater than 95-cm shell length, exceeding the size maturing period in December/January when mean sea of sexual maturity (76-mm shell length) reported by surface temperature (SST) was the highest. Smith (2003) Rapson (1952). All extractions were made by hand, fi fi classi ed ve gametogenic developmental stages (EA, packed in local sand and seawater for transportation to LA, ripe, partially spawned/spent and rapid redevelop- the laboratory for dissection within 2 days. Collections ment) and also observed two ripening/spawning events, in both years were made at low tide, ±2 h. one in August/September, and a second from December to March. No published accounts exist for reproduction of P. ventricosa in southern New Zealand, although an Individual condition index (CI) unpublished account by Mandeno (1999) examined reproduction of the South Island population from a sin- For each animal, live individuals were blotted dry and gle site. Mandeno (1999) classified southern P. ventrico- the total wet mass (Mt) was determined to the nearest Downloaded by [203.109.163.22] at 18:58 10 March 2015 sa into the same four gametogenic stages as Redfearn 0.1 g, following which the animals were dissected and (1974) and identified two major ripening events, one in the combined weight of the visceral mass and foot deter- September, and a second in February. mined (Mv). Using these weights and methods described The current study aims to increase our understanding by (Herrmann et al. 2009; after de Villiers 1975)aCI of the reproduction of P. ventricosa, particularly in terms for each individual was calculated as:  of temporal and spatial patterns at Oreti Beach, South- Mv CIð%Þ¼100 land, New Zealand. Here, a detailed description of the Mt Mv reproduction of P. ventricosa is made using a condition index (CI) and detailed histological study of the gonad over a period of two years at four different sites (one site in 2011, and four sites in 2012) at Oreti Beach. Repro- ductive activity over the period was compared with Histological preparation physical environmental conditions from the study sites. The gonad tissue of P. ventricosa is diffused within the This study is the first to detail the reproductive cycle body; therefore, an intact 15-mm cross-sectional portion in southern populations of P. ventricosa and provides of foot, gonad and gut was excised adjacent to the ante- insight into the spatial variation in reproduction, both rior adductor muscle and immediately fixed in 10% tetra- −1 across latitude, but at smaller scales (i.e. <5 km). This sodium borate (30 g L ) buffered formalin in seawater. information is required for the effective management of Following fixation, a section of gonad tissue (<15 mm Invertebrate Reproduction & Development 3

Figure 1. Location of study sites at Oreti Beach, Southland for 2011 (Site B (46.4790°S, 168.2522°E)) and 2012 (Sites A (46.4943°S, 168.2604°E), B (46.4790°S, 168.2522°E), C (46.3935°S, 168.1970°E), and D (46.3701°S, 168.1673°E).

side length, surrounded by foot tissue to stabilise the Redfearn 1974; Grant & Creese 1995; Hooker & Creese gonad tissue) was removed and placed into tissue cas- 1995; Mandeno 1999). settes for tissue processing. Any portion of the crystal- line style was removed from the sections, and care was taken to avoid gut tissue in the preparations. Oocyte diameter and density Excised sections were dehydrated via an alcohol ser- For each female, the size of up to 50 oocytes sectioned Downloaded by [203.109.163.22] at 18:58 10 March 2015 ies, and re-saturated with Paraplast® (Sigma-Aldrich) through the nucleus was measured as the average of two wax. Sections were then embedded in Paraplast®, sec- perpendicular diameter measurements of the longest (d1) tioned at 7–10 μm thickness and mounted on chrome and shortest (d2) axes. From this, the oocyte size fre- alum gelatine-coated slides. Sections were first stained quency and average oocyte size calculated for each with Meyer’s haemotoxylin, counterstained with eosin female at each month. In addition, for each month, all and mounted using DPX Mountant (Sigma-Aldrich). oocyte size measurements were pooled to obtain the size-frequency distribution. All measurements were made using the software ImageJ (NIH, USA). For each section, Determination of sex and reproductive stage the average free oocyte density (per mm2 of gonad Mounted gonad sections were photographed using an tissue) was estimated for each slide. Olympus BX51 compound microscope fitted with an Olympus ColorviewIII camera, controlled by the soft- ware Analysis LS (Olympus Corporation). Animals were Environmental measurements sexed and assigned to one of four reproductive stages SST and salinity at each site was collected using a (Figure 2, Table 1), namely: EA, LA, ripe (RI) and HORIBA U-50 multi-parameter water quality checker partially spawned/spent (SP). Definitions of the stages (HORIBA, Ltd) at each sampling time in 2011 and were adapted from earlier bivalve studies (Ropes 1968; 2012. Regional SST was derived from the monthly 4 K. Gadomski and M. Lamare Downloaded by [203.109.163.22] at 18:58 10 March 2015

Figure 2. Histological sections of female and male Paphies ventricosa gonads. Females stages: EA (A); LA (B); ripe (C), and; par- tially spawned/spent (D). (Aw = Alveoli wall; Pg = protogonia; PO = pre-vitellogenic oocyte; VO = vitellogenic oocyte; Lm = lumen) Males stages: EA (E); LA (F); ripe (G), and; partially spawned/spent (H). (Aw = Alveoli wall; Sg = Spermatogonia; Sd = Spermatid; S = Spermatozoa; Ct = connective tissue; Lm = lumen). See Table 1 for detailed explanations of female and male reproductive stages. Scale bars = 100 μm. Invertebrate Reproduction & Development 5

Table 1. Description of four reproductive stages assigned to female and male Paphies ventricosa gonad sections. Abbreviations refer to Figure 2.

Stage Females Males Early active The alveoli walls (Aw) are thick and contain The alveoli walls (Aw) are thin and have a layer of numerous protogonia (Pg). Many pre-vitellogenic spermatogonia (Sg). Spermatids (Sd) fill the lumen oocytes (PO) are present, attached to the aveoli (Lm). A few spermatozoa (S) still remain in many walls by broad cytoplasmic bases. Free lumen centres vitellogenic oocytes (VO) may remain in the lumen (Lm) from the last ripe cycle

Late active The alveoli walls (Aw) are less dense. Vitellogenic The alveoli walls (Aw) are lined with oocytes are larger and quite numerous. Remaining spermatogonia (Sg). Roughly one-third of the basal attachments are much thinner in pre- lumen (Lm) is filled with spermatids (Sd). Lumen vitellogenic oocytes (PO) centres are loosely packed with radiating bands of spermatozoa (S). Little lumen is visible

Ripe The alveoli walls (Aw) are thin and may contain The lumen (Lm) is mostly filled with densely protogonia. Vitellogenic oocytes (VO) are at their packed spermatozoa (S). A thin layer of spermatids largest and fill the lumen of mostly follicles. Pre- (Sd) fill the rest of the lumen space vitellogenic oocytes (PO) are few and are broadly attached to the alveoli walls

Partially spawned/spent The alveoli walls (Aw) are thin and contracted, The alveoli walls (Aw) may be slightly contracted; with some partially ruptured. Some pre- have a few spermatogonia (Sg); and have a vitellogenic oocytes (PO) remain attached to the marbled appearance. Spermatids form a thin layer alveoli walls. Few vitellogenic oocytes (VO) may around the outer edges of the lumen (Lm). remain in the lumen (Lm) Remaining spermatozoa (S) are distributed throughout the lumen (Lm) in loosely packed groups

composite Aqua MODIS SST (11 μ daytime) 9 km data- water pixel values using the provided scale. Site-specific set (www.oceancolor.gsfc.nasa.gov). Regional averages calculated chlorophyll-a data were compared to regional were calculated from the same 6 by 6 pixel area for each satellite derived chlorophyll-a concentrations from the month, stretching from Colac Bay to Bluff, Southland and fully encompassing the study area, by determining Downloaded by [203.109.163.22] at 18:58 10 March 2015

Figure 3. Changes in male (- - - - -) and female (…..) CI, and the index for sexes pooled (•) for Paphies ventricosa from Oreti Beach, New Zealand, between January 2011 and Decem- ber 2011. Standard errors are given for the CI of pooled data, Figure 4. Average CI (±SE) for Paphies ventricosa at four but error bars for the individual sexes are omitted for clarity. sites along Oreti Beach, New Zealand, during four months in Significant differences in the pooled CI among months are indi- 2012. Significant differences in the pooled CI among sites cated by lower case lettering. within each month are indicated by lower case lettering. 6 K. Gadomski and M. Lamare

monthly composite Aqua MODIS chlorophyll-a concen- among months in 2011 were tested using a one-way tration 9 km dataset (www.oceancolor.gsfc.nasa.gov)in ANOVA, with significant pairwise differences between both 2011 and 2012. Regional averages were calculated months identified using a Tukey’s post hoc test. Due as mentioned above for SST. Day length was obtained to heterogeneity of variances, a Welch’s test was used for the 15th day of each collection month, and tidal range to assess statistical differences in the average oocyte/ was determined using Land Information New Zealand ova size among months in 2011. Differences in the CI tide prediction charts for Bluff, New Zealand. and oocyte density among sites and months in 2012 were tested using two-way ANOVA, and where a fi Statistical analyses signi cant interaction between the two factors occurred, differences among months within each site The sex ratio of P. ventricosa at the four sampling were tested using a one-way ANOVA. All transformed sites (months pooled) and for all samples pooled was data were tested for normality visually via normal 2 ∂ tested using the X -test. Statistical differences ( = 5%) quantile–quantile plots and homoscedasticity was con- in CI (arcsine square-root transformed) between sexes firmed by Levene’s test. All statistical analyses were and among months were tested using a two-way carried out using the JMP 10.0 statistical package ANOVA, whilst the average oocyte/ova size (ln (x) (SAS Institute). transformed) and abundance (ln (x) transformed) Downloaded by [203.109.163.22] at 18:58 10 March 2015

Figure 5. Monthly changes in the percentage of female (A) and male (B) Paphies ventricosa in each of four gametogenic stages from Oreti Beach, New Zealand, between January 2011 and December 2011. Invertebrate Reproduction & Development 7

Results (Supplementary Table 1). Similarly, there was no signifi- Sex ratio cant difference in the male:female ratio for the four sam- ple sites in 2012 (Supplementary Table 1), or when all The ratio of males and females sampled over 2011 at fours sites are pooled despite a bias towards males (304 Site B (103 Male: 93 Females) was not significantly Male: 264 females). different (X2 = 0.5102, p = 0.4751) from 1:1 Downloaded by [203.109.163.22] at 18:58 10 March 2015

Figure 6. The percentage of female (left column) and male (right column) Paphies ventricosa in each gametogenic stage from four sites along Oreti Beach, New Zealand, at four months in 2012. 8 K. Gadomski and M. Lamare

Condition indices among months (F(10, 194) = 9.4460, p < 0.001), with a ’ The average monthly CI during 2011 ranged from 41.6 Tukey s post hoc comparison of means indicating that to 73.7% for males and 43.9% to 72.6% for females the CI in January (42.9%) and December (43.1%) were fi (Figure 3). Two-way ANOVA indicated that there was signi cantly lower than all other months (Figure 3). The greatest CI occurred in October (73.2%), although no significant difference in the CI between sexes (F(1, the difference in the CI from May to October was 194) = 1.125, p = 0.291) across all months (Sex × generally non-significant. Month, F(10, 194) = 1.052, p = 0.402). When sexes were pooled, there was a significant difference in the CI Downloaded by [203.109.163.22] at 18:58 10 March 2015

Figure 7. Monthly oocyte size-frequency distributions for Paphies ventricosa from Oreti Beach, New Zealand, between January 2011 and December 2011. n = number of females sampled /number of oocytes measured on each sampling date. Invertebrate Reproduction & Development 9

The CI for P. ventricosa collected during 2012 70% of the sample), but were most common in September (Figure 4) varied significantly among sites (F(3, 288) = and October (64 and 70%, respectively). There was a 20.989, p < 0.001), although the effects of site depended smaller proportion of ripe females in February (40%), on the month of collection (F(9, 288) = 7.582, p < 0.001). which was followed by a decrease in the proportion ripe In this respect, there was no significant difference among to 14% by June. The proceeding early active and late sites in September (F(3, 79) = 2.527, p = 0.063) and active gametogenic stages were most abundant in the December (F(3, 78) = 1.271, p = 0.291), with the CI rang- months between the ripe maxima, making up to 86% (EA ing between 32.2 and 45.6% during these times. The CI and LA pooled) in June and 73% in December. Spawned was greater in March and June, ranging between 47.0 females were present in May (20%), but were most and 75.1%, and was significantly lower at Site A than common from September to December, making up 50% Site D in both March (F(3, 52) = 13.853, p < 0.001) and of the females in November. For males, histological June (F(3, 72) = 16.22, p < 0.001). examination of spermatogenic stages over a 1-year period (Figure 5(B)) indicated a clearer pattern in the abundance of ripe individuals, with peaks in February/March (90 Gametogenic cycle and 93%, respectively) and through October and For females, histological examination of oogenic stages November (100 and 83%, respectively). Early active and over a 1-year period in 2011 (Figure 5(A)) indicated that late active stages were predominant in the remaining ripe females occurred in all months (ranging from 9 to months, whilst spawned males were present in May and Downloaded by [203.109.163.22] at 18:58 10 March 2015

Figure 8. (A) Average oocyte/ova size (±SE) and (B) Average oocyte density (±SE) at monthly intervals for Paphies ventricosa from Oreti Beach, New Zealand, between January 2011 and December 2011. The sample size for each month is given in Figure 6. For both figures, significant differences in the oocyte measurements among months are indicated by lower case lettering. 10 K. Gadomski and M. Lamare

August, but most common in December when they made September. In the following three months, the distribu- up 44% of males. tions decreased in size and breadth, with few oocytes lar- Spatial variation in the gametogenic stages was ger than 50 μm by December. examined at four sites along Oreti Beach for four months During the 2011 period, the average oocyte size var- in 2012 (Figure 6). There was a substantial amount of ied significantly (Welch’s test, F(10, 1479.5) = 66.26, variation among sites for both sexes, particularly in p < 0.001) among months (Figure 8(A)), with two dis- March and December, the time likely to be near or fol- tinct peaks in average size, the first in January/March lowing spawning, and suggests a degree of spawning (37.8–38.8 μm) and a second in September (36.3 μm). variability along Oreti Beach. For both sexes, a greater The smallest average oocyte sizes occurred during the degree of gametogenic synchrony was evident among winter months when oocyte average size was as small as sites in June and September, although there was a ten- 29.3 μm, and in December when oocyte was at a mini- dency for individuals to be more advanced gametogeni- mum (25.8 μm). Tukey’s post hoc tests indicated signifi- cally with increasing distance north (i.e. from site A to cant differences among the months of maximum oocyte site D). This is reflected in typically a greater proportion diameter with those of minimum sizes (Figure 8(A)). of ripe and late active stages, and fewer spawned indi- viduals at site D in most months. Oocyte density Oocyte density over 2011 ranged from 13.9 oocytes Oocyte/ova size frequencies mm−2 in December to 47.7 oocytes mm−2 in July (Fig- Oocyte and ova size-frequency distributions over the ure 8(B)). Densities were significantly different among 2011 sampling period (Figure 7) were unimodal, and months (F(10, 92) = 8.605, p < 0.001), with Tukey’s post typically right-skewed. During the months of January to hoc comparisons indicating that densities were signifi- March, the distributions were relatively stable in shape cantly higher in July, August and September than in the although there was a progression of the distribution into following three months. Oocyte densities in the four sites bigger size classes. A clear change in size-frequency measured in 2012 (Figure 9), varied significantly among occurred in May with a broader distribution (≈12– sites (ln (x) transformed, F(3, 135) = 7.294, p = 0.002), 60 μm), and a reduction in the abundance of larger although the effects of site depended on the month of oocytes and ova. From June to August, there was a collection (Site × Month, F(9, 135) = 5.239, p < 0.001). progressive change from a smaller and right-skewed size Significant differences among sites were evident in all distribution to a larger and broader size range by months: March (F(3, 33) = 5.422, p < 0.001), June (F(3, 32) = 3.166, p = 0.013), September (F(3, 32) = 5.542, p = 0.004) and December (F(3, 35) = 5.895, p = 0.003). With the exception of March, bivalves at site B generally had significantly lower oocyte densities (Figure 9), whilst those at site A and C had significantly higher oocyte densities in March and December, respectively. Downloaded by [203.109.163.22] at 18:58 10 March 2015 Physical environment Average SSTs in the Oreti Beach region of the Foveaux Strait (MODIS satellite measurements, averaged over a 30 km × 30 km area) for 2011 ranged from 15.4 °C in February to a minimum winter temperature of 9.6 °C in July (Figure 10(A)). In situ measurements of temperature in the surf zone immediately next to the P. ventricosa beds were similar in the spring months, but ≈2.3 °C cooler and 2.1 °C warmer than remote measurements for the winter and spring months, respectively (Fig- ure 10(A)). Chlorophyll-a concentrations for the same region of the Foveaux Strait ranged from 1.14 mg m−3 in June to a maximum of 5.13 mg m−3 in October, repre- Figure 9. Average oocyte density (±SE) for Paphies ventrico- senting the peak spring bloom for 2011 (Figure 10(B)). sa at four sites along Oreti Beach, New Zealand, during four months in 2012. Significant differences in the oocyte density Day length ranged from 15 h 46 min in December to 8 h among sites within each month are indicated by lower case let- 36 min in June (Figure 10(C)). The decrease in CI and tering. spawning in P. ventricosa during 2011 was associated Invertebrate Reproduction & Development 11

Figure 10. SST (A), chlorophyll-a concentration (B) and day length (C) for Oreti Beach during 2011. Sea temperature and chloro- phyll-a concentrations are graphed for both satellite estimates and in situ measurements. The estimated P. ventricosa spawning periods for 2011 are indicated by horizontal bars. Downloaded by [203.109.163.22] at 18:58 10 March 2015

with periods of changing temperature, namely an ≈3°C resting phase with continuous spawning punctuated by warming from September to December, and an ≈4°C 2–3 main spawning periods occurring over the warmer cooling from March to May, whilst the main spawning summer months (December to February). In this study, event in spring commenced at the time of the September P. ventricosa at its southern location also had an annual chlorophyll-a concentration peak. reproductive cycle that lacked a resting phase and two main spawning periods, the first in spring and the second in late summer. Histological examination of gametogene- Discussion sis indicated that whilst ripe females were found Gametogenic cycle throughout the year, the proportions were greatest in Annual and spatial reproductive biology of P. ventricosa September to November (up to 70%) and February was examined over two years along Oreti Beach, a loca- (40%), with intervening months dominated by early and tion representing the southernmost known extent of the late active ovaries (73%). The male cycle was more pro- – species range. The only published account of the repro- nounced, with ripe testes prevalent (83 100%) in Octo- ductive cycle in P. ventricosa by Redfearn (1974) ber/November and February/March. Redfearn (1974) described an annual reproductive cycle in individuals described a similar gametogenic pattern in P. ventricosa from the species most northern extent, which lacked a ovaries, with two to three periods of predominantly 12 K. Gadomski and M. Lamare

mature stages over the spring to late-summer period, sep- associated with spawning events. Marsden (1999) noted arated by actively growing ovaries with no resting period that the poor correlation would result from simultaneous during the annual cycle. In contrast to the present study, changes in somatic tissue, shell and gonad tissue over a Redfearn (1974) observed spawned individuals through- reproductive cycle. out the year, whilst the Oreti Beach population lacked spawned individuals in the winter and mid-summer months. Spatial and temporal variation An extended spawning period that involves an inter- Whilst the annual reproductive cycle of P. ventricosa vening resting or active period is often observed in showed a clear annual pattern, there was a degree of spa- bivalve species (Sastry 1979), including other Mesode- tial variation among the four sites separated by 15 km. smatids such as M. mactroides (Herrmann et al. 2009). The variation was mostly owing to variation during the Within the wider Paphies genus, gametogenesis also fol- months associated with spawning, March and December, lows an annual cycle, with spawning occurring over the with individuals at the southern end of the beach (site A) summer periods for P. donacina (Marsden 1999), P. aus- more advanced gametogenically, having the higher tralis (Hooker & Creese 1995) and P. subtriangulata oocyte densities earlier, and a lower CI in March and (Grant & Creese 1995). As for P. ventricosa, the latter June following spawning. This suggests that whilst two species have no resting phase during gametogenesis, annual cycles are broadly similar among sites, there may and for P. subtriangulata spawning occurring in spring be a degree spawning asynchrony along Oreti Beach. and late-autumn with a mid-summer active period. P. do- Spatial differences in bivalve reproduction at relatively nacina may also undergo more than one gametogenic small scales can be marked (Gosling 2004). Pecten maxi- cycle annually. However in contrast to other Paphies mus populations around coastal Isle of Man, for exam- species, it has an inactive resting phase during winter. ple, showed significant variation in gonad condition, Monthly changes in the oocyte size frequency distri- fecundity and spawning onset at small horizontal scales butions and associated average oocyte size and density of < 5 km (Hold et al. 2013), whilst intertidal bivalves further support the suggestion that female P. ventricosa can vary reproductively across tidal heights within a undergo a major spawning in spring, followed by active population (Harvey & Vincent 1989; Azouzi et al. oocyte development leading to a second spawning in 2002). late-summer. The unimodal oocyte distributions in Latitudinal differences in reproduction can also be P. ventricosa are commonly observed in bivalves pronounced. Examples are found in a range of bivalve (Herrmann et al. 2009), with the broad oocyte size range species (Sastry 1979), including Mesodesmatidae, in observed throughout the year reflecting gametogenic which Uribe et al. (2012) noted significant reproductive processes where active and mature females are present asynchrony in populations of M. donacium along the throughout the year. The distinct loss of larger oocytes 4000-km long Humboldt Current upwelling ecosystem. and reduction in oocyte density in the spring and autumn P. ventricosa spans 12° of latitude (1350 km), and whilst months reflected two spawning episodes. the cycles are broadly similar, the reproductive cycles of P. ventricosa appear more pronounced in the southern population compared with those at its northern extent Downloaded by [203.109.163.22] at 18:58 10 March 2015 Condition index (Redfearn 1974). This is best demonstrated by the pres- A CI has been applied to a range of bivalve species to ence of spawned individuals all year round in the north, describe reproductive cycles, including mesodesmatids whilst spawned individuals were absent from winter and where the gonad and visceral tissue are closely associ- late-summer months in the southern P. ventricosa popu- ated (Herrmann et al. 2009). For P. ventricosa, the CI lation. Other Paphies species show a similar pattern of varied seasonally, with a ≈50% decrease in the index in spatial variation, with differences in spawning period in spring, consistent with the major spawning event seen in P. subtriangulata (Grant & Creese 1995) and P. australis the Oreti Beach population. The decrease in CI in late- (Hooker & Creese 1995) observed among locations. summer when spawning occurred was less clear and not Spatial differences in gametogenesis in bivalves are significant. Redfearn (1974) calculated the CI (deter- the result of the interaction of environmental factors, mined by volume) in P. ventricosa over one year, and which includes primarily sea temperature and food sup- correlated decreases in CI with spawning in spring and ply, as well as photoperiod and lunar cycles (Sastry late-summer. The application of the index to other me- 1979), whilst spawning episodes have been mainly asso- sodesmatids has been less informative. For example, ciated with temperature changes, salinity and light both Marsden (1999) and Herrmann et al. (2009) showed (Sastry 1979; Gosling 2004). Temperature is undoubtedly the CI cycled annually in P. donacina and M. mactro- influential in mesodesmatid species inhabiting seasonal ides, respectively, but neither study could consistently environments. Hermann et al. (2009) correlated the correlate changes in the CI with gametogenic changes reproductive cycle of M. mactroides with seasonal Invertebrate Reproduction & Development 13

changes in sea temperatures which ranged from 11 to bivalves (Sastry 1979; Gosling 2004), and specifically in 23 °C annually; Manzi et al. (1985) reported temperature mesodesmatids (Peredo et al. 1987; Hermann et al. influenced the gametogenic cycle in Mercenaria merce- 2009). Seasonal variations in chlorophyll-a concentration naria; and Peredo et al. (1987) discussed sea temperature are greater in southern New Zealand (Murphy et al. variation and annual reproduction in M. donacium. In 2001), which ranged from 0.2 to 0.6 mg chl-a m−3 in this study, P. ventricosa spawning was associated with northern offshore regions compared with 0.3 to > 1.0 mg periods of changing sea temperature, namely an ≈3°C chl-a m−3 in southern waters offshore of Oreti Beach. warming from September to December, and a 4 °C cool- We also noted a chlorophyll-a concentration spike ing from March to May. Redfearn (1974) discussed (>5 mg chl-a m−3) during October, coinciding with the reproductive events associated with temperature changes, main spawning of P. ventricosa. It is possible, therefore, noting spawning probably occurred in December/January that the more pronounced reproductive cycles in the following a period of gonad maturation over July to southern population reflect greater seasonality when a November when sea temperatures increased from 14 to range of environmental variables are considered. 20 °C, with a second spawning following the 22 °C sum- For other New Zealand Paphies species, Marsden mer temperature maxima. (1999) concluded that P. donacina reproduction was pro- Temperature variation may also play a role in latitu- moted by warmer summer temperatures, with high food dinal variation in reproduction in P. ventricosa although concentrations potentially extending gonad development. it is interesting that the annual variation in temperature Grant and Creese (1995) suggested that warmer tempera- at both locations (8 °C) is similar, despite the southern tures promoted spawning both seasonally and tidally, and population having a shorter and more defined spawning Hooker and Creese (1995) noted spawning during spring period. Latitudinal variation in the reproductive cycles of coincided with short-term temperature fluctuation. The New Zealand bivalves has been related to sea tempera- small-scale spatial variations in spawning that we ture. Gribben et al. (2004) concluding that the duration observed along Oreti Beach are unlikely due to tempera- of gametogenesis and spawning period in two popula- ture variation, but may be associated with gradients in tions of the New Zealand , P. zelandica sepa- food supply. Densities of P. ventricosa vary significantly rated by ≈5° of latitude was related to seasonal changes along Oreti Beach (Beentjes 2010) and are greatest at in sea temperature, with spring spawning occurring when the sites A and B in the present study. P. ventricosa dis- temperature reached 15 °C in both populations. For Pa- tribution is thought to be controlled by phytoplankton phies species, cold temperatures have been shown to abundances (Redfearn 1974), and the high density at delay gametogenesis in P. donacina (Marsden 1999), and sites A and B have been attributed to enhanced food Grant and Creese (1995) suggested that the minimum supply from the adjacent New River Estuary (Beentjes temperature threshold for spawning in P. subtriangulata 2010, Figure 1). It is likely that the animals at these sites was 15 °C. It is conceivable, therefore, that the colder are nutritionally advantaged and therefore complete temperatures in the southern population during June to gametogenesis and spawning earlier. This suggestion September (≈10 °C) suppress spawning in the species would be consistent with observations in bivalves that during the winter period. Such cold temperatures are not intra-specific variation in reproductive cycles can be experienced by the northern population, with tempera- related to small-scale food supply patterns (Hold et al. Downloaded by [203.109.163.22] at 18:58 10 March 2015 tures remaining above 14 °C throughout the year, during 2013). Salinity fluctuations can also induce spawning in which spawning is continuous. bivalves (Sastry 1979), and it is equally plausible indi- It is also possible that other environmental factors viduals at sites A and B experience fluctuations in salin- influence reproductive periodicity in P. ventricosa. ity associated with freshwater output from the estuary Indeed, for M. donacium, latitudinal variation in repro- that could induce spawning episodes. ductive cycles has been attributed to a range of factors in addition to temperature, including localised upwelling and changes in salinity and runoff (Uribe et al. 2012). Conclusions Few studies have considered the effects of day length on P. ventricosa has experienced dramatic declines in popu- reproductive cycles in bivalves as it is often difficult to lation sizes over the past century (Greenway 1972; separate from seasonal temperature changes (Gosling Beentjes et al. 2006), and the conservation and manage- 2004), but it is worth noting that seasonal changes in ment of the species require an understanding of the day length are more pronounced at the southern location reproductive biology of the species across a range of (from 8 h 34 min to 15 h 50 min, seasonally) compared spatial scales. This study is the first to show that whilst with the north (from 9 h 34 min to 14 h 36 min, season- reproductive cycles are broadly similar in the species ally). Furthermore, seasonal changes in chlorophyll-a across its latitudinal extent, southern populations have a concentrations, which tend to be more variable at higher more defined spawning period that may be associated latitudes, have been linked to seasonal reproduction in with greater environmental seasonality. Smaller-scale 14 K. Gadomski and M. Lamare

variation in reproduction was also evident within the Greenway JPC. 1972. Further data from population surveys on Oreti Beach population and may be an important consid- toheroa (: Eulamellibranchiata) on Northland bea- – eration in identifying areas for conserving source popula- ches, 1962 1971. New Zealand Marine Department Fisher- ies Technical Report. 85. Wellington; 36p. tions (Gaines et al. 2010). Gribben PE. 2005. Gametogenic development and spawning of the razor clam, Zenatia acinaces in northeastern New Zealand. New Zealand Journal of Marine and Freshwater Acknowledgements Research. 39:1287–1296. The authors would like to thank kaitiaki Michael Skerrett of Gribben PE, Creese RG, Hooker SH. 2001. The reproductive the Murihiku Marae for providing access to the Oreti popula- cycle of the New Zealand venus clam . fi – tion, Matthew Downes for histology advice, and Professor Journal of Shell sh Research. 20:1101 1108. Henrik Moller and Dr Michael Beentjes for assistance with the Gribben PE, Helson J, Jeffs AG. 2004. Reproductive cycle of research. Additional thanks to staff at Environment Southland the New Zealand Geoduck, Panopea zelandica, in two – for field support and the technicians at the Portobello Marine North Island populations. Veliger. 47:53 65. Laboratory (PML) who helped with sample preparation. Harvey M, Vincent B. 1989. Spatial and temporal variations of Finally, the comments of two anonymous reviewers greatly the reproductive cycle and energy allocation of the bivalve fl improved the manuscript. Macoma balthica (L.) on a tidal at. Journal of Experimen- tal Marine Biology and Ecology. 129:199–217. Herrmann M, Alfaya JEF, Lepore ML, Penchaszadeh PE, Lau- Supplemental data dien J. 2009. Reproductive cycle and gonad development of the Northern Argentinean Mesodesma mactroides (Bival- Supplemental data for this article can be accessed here. http:// via: Mesodesmatidae). Helgoland Marine Research. dx.doi.org/10.1080/07924259.2015.1007216. 63:207–218. Hoby KE. 1933. The anatomy of Amphidesma ventricosum (Gray) [MSc Thesis]. Wellington: Victoria University of References Wellington. 116p. Hold N, Murray LG, Hinz H, Neill SP, Lass S, Lo M, Kaiser Azouzi L, Bourget E, Borcard D. 2002. Spatial variation in the MJ. 2013. Environmental drivers of small scale spatial vari- intertidal bivalve Macoma balthica: biotic variables in rela- ation in the reproductive schedule of a commercially tion to density and abiotic factors. Marine Ecology Pro- important bivalve mollusc. Marine Environmental gress Series. 234:159–170. Research. 92:144–153. Baron J. 1992. Reproductive cycles of the bivalve molluscs Hooker SH, Creese RG. 1995. The reproductive biology of Atactodea striata (Gmelin), Gafrarium tumidum Röding pipi, (Gmelin, 1790) (Bivalvia: Mesodes- and Anadara scapha (L.) in New Caledonia. Australian matidae). I. Temporal patterns of the reproductive cycle. Journal of Marine and Freshwater Research. 43:393–402. Journal of Shellfish Research. 14:7–15. Beentjes MP. 2010. Toheroa survey of Oreti Beach, 2009, and Kim JH, Son MH, Kang H-W, Lee K-Y. 2013. Reproductive review of historical surveys. New Zealand cycle of surf clam, Coecella chinensis (Mesodesmatidae: Assessment Report 2010/6. Wellington; 40p. Bivalvia). Korean Journal of Malacology. 29:113–120. Beentjes MP, Carbines GD, Willsman AP. 2006. Effects of Mandeno MA. 1999. Gonadal cycle and spawning induction in beach erosion on abundance and distribution of toheroa toheroa (Paphies ventricosa) from Southern New Zealand (Paphies ventricosa) at Bluecliffs Beach, Southland, New [MSc Thesis]. Dunedin: University of Otago. 54p. Zealand. Journal of Marine and Freshwater Research. Manzi JJ, Bobo MY, Burrell VG. 1985. Gametogenesis in a 40:439–453. population of the , Mercenaria mercenaria (Lin- Beentjes MP, Gilbert DJ. 2006. Oreti Beach toheroa survey: naeus) in North Santee Bay, South Carolina. The Veliger. yield per recruit and review of historical surveys. New Zea- Downloaded by [203.109.163.22] at 18:58 10 March 2015 28:186–194. land Fisheries Assessment Report. 2006/36. Wellington; Marsden ID. 1999. Reproductive cycles of the surf beach clam 47p. Paphies donacina (Spengler, 1793) from New Zealand. Brunton PM. 1978. Toheroa predation by black-backed gulls on Journal of Shellfish Research. 18:539–546. Dargaville Beach, North Auckland, New Zealand. Notornis. Morrison M, Parkinson D. 2008. Distribution and abundance of 25:128–140. toheroa (Paphies ventricosa) on Ninety Mile Beach, 2006. Cassie RM. 1955. Population studies on the Toheroa, Amphi- New Zealand Fisheries Assessment Report. 2008/26. desma ventricosum Gray (Eulamellibranchiata). Australian Wellington; 27p. Journal of Marine and Freshwater Research. 6:348–391. Murphy RJ, Pinkerton MH, Richardson KM, Bradford-Grieve Gaines SD, White C, Carr MH, Palumbi SR. 2010. Designing JM, Boyd PW. 2001. Phytoplankton distributions around marine reserve networks for both conservation and fisheries New Zealand derived from SeaWiFS remotely-sensed management. Proceedings of the National Academy of Sci- colour data. New Zealand Journal of Marine and ence. 107:18286–18293. Freshwater Research. 35:343–362. Gaspar M, Ferriera R, Monteiro CC. 1999. Growth and repro- Newell RIE, Bayne BL. 1980. Seasonal changes in the physiol- ductive cycle of Donax trunculus L., (Mollusca: Bivalvia) ogy, reproductive condition and carbohydrate content of the off Faro, southern Portugal. Fisheries Research. 41:309– Caridum (= Cerastoderma) edule (Bivalvia: Cardii- 316. dae). Marine Biology. 56:11–19. Gosling EM. 2004. Bivalve molluscs: biology, ecology, culture. Peredo S, Parada E, Valdebenito I. 1987. Gametogenesis and Oxford: News Books, 443p. reproductive cycle of the clam Mesodesma donacium Grant CM, Creese RG. 1995. The reproductive cycle of the (Lamarck, 1818) (Bivalvia: Mesodesmatidae) at Quele -Paphies subtriangulata (Wood, 1828), in New Beach, Southern Chile. The Veliger. 31:55–68. Zealand. Journal of Shellfish Research. 14:287–292. Invertebrate Reproduction & Development 15

Rapson AM. 1952. The toheroa, Amphidesma ventricosum Gray Sastry, AN. 1979. Pelecypoda (Excluding Ostreidae). In: Giese (Eulamellibranchiata), development and growth. Australian AC, Pearse JS, editors. Reproduction of marine Journal of Marine and Freshwater Research. 3:170–198. invertebrates, Vol 5. Molluscs: Pelecypods and Lesser Clas- Rapson AM. 1954. Feeding and control of the toheroa (Amphi- ses. New York, NY: Academic Press; p. 113–292. desma ventricosum Gray) (Eulamellibranchiata) populations Smith S. 2003. The reproduction and recruitment of toheroa in New Zealand. Australian Journal of Marine and (Paphies ventricosa) [MSc Thesis]. Auckland: University Freshwater Research. 5:486–512. of Auckland. 106p. Redfearn P. 1974. Biology and distribution of the toheroa, Pa- Uribe RA, Oliva ME, Aguilar S, Yamashiro C, Riascos JM. 2012. phies (Mesodesma) ventricosa (Gray). Fisheries Research Latitudinal variation in the reproductive cycle of two bivalves Bulletin. 11:51p. with contrasting biogeographical origin along the Humboldt Riascos JM, Carstensen D, Laudien J, Arntz WE, Oliva ME, Current Upwelling Ecosystem. Scientia Marina. 76:713–720. Güntner A, Heilmaer O. 2009. Thriving and declining: de Villiers G. 1975. Reproduction of the white sand climate variability shaping life-history and population Donax serra Röding. South African Sea Fisheries Branch persistence of Mesodesma donacium in the Humboldt Investigational Report. 102. Cape Town; 33p. Upwelling System. Marine Ecology Progress Series. Williams J, Ferguson H, Tuck I. 2013. Distribution and abun- 385:151–163. dance of toheroa (Paphies ventricosa) and tuatua (P. subtri- Roberts D. 1984. A comparative study of Lasaea australis, angulata) at Ninety Mile Beach in 2010 and Dargaville Vulsella spongiarum, Pinna bicolor and Donacilla cuneata Beach in 2011. New Zealand Fisheries Assessment Report. (mollusca; bivalvia) from Princess Royal Harbour, Western 2013/39. Wellington; 56p. Australia. Journal of Molluscan Studies. 50:129–136. Wilson BR, Hodgkin EP. 1967. A comparative account of the Ropes JW. 1968. Reproductive cycle of the surf clam, Spisula reproductive cycles of five species of marine (Biv- solidissima, in offshore New Jersey. Biological Bulletin. alvia: ) in the vicinity of Fremantle, Western Aus- 135:349–365. tralia. Australian Journal of Marine and Freshwater Ruiz C, Martinez D, Mosquera G, Abad M, Sanchez J. 1992. Research. 18:175–203. Seasonal variations in condition, reproductive activity Zaidman PC, Kroeck MA, Oehrens Kissner EM, Morsan EM. and biochemical composition of the flat , Ostrea 2012. Reproductive pattern of the southern geoduck, Pano- edulis, from San Cibran (Galacia, Spain). Marine Biology. pea abbreviata, at El Sotana (San Matias Gulf, Patagonia, 112:67–74. Argentina). Marine Biology Research. 8:172–181. Downloaded by [203.109.163.22] at 18:58 10 March 2015