Agricultural intensity in ovo affects growth, metamorphic development ...

Ecotoxicology (2011) 20:901–911 DOI 10.1007/s10646-011-0658-5

Agricultural intensity in ovo affects growth, metamorphic development and sexual differentiation in the Common toad (Bufo bufo) Frances Orton • Edwin Routledge

Accepted: 17 March 2011 / Published online: 30 March 2011 Ó Springer Science+Business Media, LLC 2011

Abstract Pollution was cited by the Global Amphibian Assessment to be the second most important cause of amphibian decline worldwide, however, the effects of the agricultural environment on amphibians are not well understood. In this study, spawn from Bufo bufo was taken from four sites in England and Wales with varying intensities of arable agriculture. Spawn was either placed in tanks containing aged tap water (ex-situ, five replicates) or in cages at the native site (caged, five replicates). Hatching success, abnormal tadpoles, and forelimb emergence were recorded during the larval stage. Individuals were also sampled at five time points (TP) during development (5-, 7-, 9-, 12-, 15-weeks post-hatch) and analysed for morphological parameters. The thyroids (TP2) and the gonads (TP3,4,5) were also analysed histologically. With the exception of the thyroid histopathology, all analysed endpoints were significantly different between ex-situ individuals reared under identical conditions from the different sites. In addition, intensity of arable agriculture had a negative effect on growth and development. At one site, despite distinct rearing

Electronic supplementary material The online version of this article (doi:10.1007/s10646-011-0658-5) contains supplementary material, which is available to authorized users. F. Orton E. Routledge Institute for the Environment, Brunel University, Kingston Lane, Uxbridge UB8 3PH, UK E. Routledge e-mail: [email protected] Present address F. Orton (&) Centre for Toxicology, School of Pharmacy, University of London, 29/39 Brunswick Square, London WC1N 1AX, UK e-mail: [email protected]

conditions, a high level of intersex (up to 42%) and similar sex ratios were observed in both ex-situ and caged individuals. Taken together, these data suggest that maternal exposure and/or events in ovo had a much larger effect on growth, metamorphic development, and sexual differentiation in B. bufo than the ambient environment. This could have important implications for traditional exposure scenarios that typically begin at the larval stage. Intersex is reported for the first time in European amphibians in situ, highlighting the potential use of distinct populations of amphibians in fundamental research into the aetiology of specific developmental effects in wild amphibians. Keywords Amphibian Pollution Pesticides Herbicides Endocrine disruption Intersex Abbreviations TP Time point SVL Snout-vent length HLL Hindlimb length FLE Forelimb emergence DO Dissolved oxygen TO Testicular oocyte

Introduction Amphibian populations are in crisis globally with 32% of all known amphibian species highly threatened with extinction. The causes of amphibian decline are numerous and only partly understood, however, habitat loss and pollution are thought to be the biggest contributors (GAA 2004). Therefore, conversion of native habitat to agricultural land and consequent pollution due to agrochemical use may be instrumental in observed declines.

123

902

Anuran density and diversity have been shown to be deleteriously affected by proximity to agriculture (Bishop et al. 1999) and pesticides have been shown to alter community structure, ultimately resulting in reduced growth (Relyea and Diecks 2008) and increased risk of parasite infection (Rohr et al. 2008) in mesocosm studies. Field evidence from California, which has the highest pesticide use in the western world, has shown that (i) declining amphibian populations are associated with increased concentrations of organophosphates (Sparling et al. 2001), (ii) individuals from a declining population carried higher tissue burdens of DDE, c-chlordane and trans-nonachlor compared to those of a stable population (Fellers et al. 2004) and (iii) that spray drift of pesticides had a more significant impact on the decline of Rana muscosa than the presence of predatory fish in ponds (Davidson and Knapp 2007). However, other authors have found no such correlations between agriculture and deleterious effects (Knutson et al. 2004). The specific effects of agricultural compounds on hatching success, growth, timing of metamorphosis and consequent survival are less well studied in situ. However, it is well known that metamorphosis is a highly plastic process which can be affected by external stresses, and that both increased time in the larval stage and smaller metamorphosed individuals are less likely to survive in the wild (Rose 2005). Indeed, laboratory exposures to low levels of pesticides have shown negative effects on growth and metamorphosis (Greulich and Pflugmacher 2004; Hayes et al. 2006). In the latter study, developmental exposure of Rana pipiens to a mixture of pesticides retarded growth and development and, in some cases, abolished the normally positive correlation between time to metamorphosis and size. In contrast to these findings, no effects of pesticides on growth and metamorphosis have also been reported (Carr et al. 2003). Pesticide exposure has also been implicated in gonadal abnormalities in amphibians, both in the laboratory (Hayes et al. 2002; Orton et al. 2006) and in situ (Hayes et al. 2003; McCoy et al. 2008; McDaniel et al. 2008), although no effect results have also been reported (laboratory: Carr et al. 2003; Smith et al. 2005; in situ: Kloas et al. 2009). In R. pipiens, Hayes et al. (2003) reported an association between atrazine and feminisation, including intersex, and McDaniel et al. (2008) reported increased intersex from agricultural sites than non-agricultural sites, although no association with sex steroid levels was found. In Bufo marinus, McCoy et al. (2008) reported that gonadal morphology and secondary sexual characteristics were feminised in toads taken from agricultural sites, and that intersex individuals had lower circulating testosterone levels. Bidders’ organ, a rudimentary ovary found in Bufonidae species, was also better developed in intersex individuals than normal males, which is not surprising as it is known that there is a dynamic relationship between the two organs

123

F. Orton, E. Routledge

(Calisi 2005). Although the effects of intersex on populations are not known, detrimental effect of intersex on fecundity has recently been shown in fish (Jobling et al. 2002) and amphibians (Gyllenhammar et al. 2009), and thus are a cause for concern. In this study, we assessed the ex-situ and in situ (caged) impact of arable agricultural intensity on survivorship, growth, metamorphic development, and sex differentiation in the Common toad (Bufo bufo). This study formed part of a larger project ‘Assessing the impact of endocrine disrupters in the aquatic environment on British frogs and toads’, further details of which may be found on the DEFRA (Department for Environment, Food and Rural Affairs) website at: http://randd.defra.gov.uk/Default.aspx? Menu=Menu&Module=More&Location=None&ProjectID= 12752&FromSearch=Y&Publisher=1&SearchText=endo crine%20disruptor&SortString=ProjectCode&SortOrder= Asc&Paging=10#Description.

Methods Site selection Sites were selected using a combination of toad population data (provided by Froglife, registered charity: 1088255), DEFRA land use data (DEFRA 2005), and local site characteristics. Due to the magnitude of possible contaminants in the environment and little known effects of those contaminants, analysing information on land use is an accepted method of exposure assessment both in wildlife (e.g. amphibians: McCoy et al. 2008; McDaniel et al. 2008) and in human epidemiology (Nuckols et al. 2004). Four sites in England and Wales were selected; one was classified as ‘reference’ (Pant-y-llyn), one as a ‘low’ agriculture site (Yatton) and two as ‘high’ agriculture sites (Twenty-foot and Layes pool). Local characteristics were as follows: Pant-y-llyn (SN606166) is a turlough (has no groundwater inflow or outflow) located adjacent to woodland in a rough grazing area of Carmarthenshire; Yatton (ST41649) is a small pond situated on the edge of a nature reserve in a pastural area of North Somerset; Twenty-foot (TL322972) is located in a drainage ditch in the highly agricultural Fenland, which forms part of a connected network of canals; Layes pool (SO802658) is a large pond bordered by arable agriculture, pasture and woodland in the moderately agricultural district of Malvern hills (Fig. 1). Experimental design During March/April 2006 toadspawn was collected from 2 to 4 different strings (it is not known whether these originated from different females) from selected sites. The

Agricultural intensity in ovo affects the Common toad (Bufo bufo)

903

period (Yatton—29.03.06; Pant-y-Llyn—2.04.06; TwentyFoot—3.04.06; Layes Pool—5.04.06). Developmental rate of embryos and time to hatching also varied between sites, so sampling was also carried out on different dates from each site according to hatching date. Animals were sampled at five time points (TP) following hatching (5-, 7-, 9-, 12-, and 15-weeks post hatch), based on the date when all embryos had hatched (hatch date: Yatton—21.04.06, Pant-y-Llyn & Twenty-Foot—24.04.06; Layes Pool—25.04.06). TP1/2 occurred prior to metamorphosis, and therefore samples were of larval individuals. Mortality was measured at three times: (1) initial hatching success, (2) during larval period (up to TP2), (3) during metamorphic period (up to end of experiment, TP5). In addition, the incidence of forelimb emergence was recorded at TP2 (see Fig. 2). Ex-situ

Fig. 1 Local site characteristics and land use data for Pant-y-llyn (top), Yatton (second), Twenty-foot (third) and Layes pool (bottom). Blue shading in the centre of each map indicates the test site (water body). Pie charts show the proportion of each type of agriculture in the total farmed area for each district: red (first segment) arable crops, green (second segment) pasture, blue (third segment) fallow, violet (fourth segment) woodland, orange (fifth segment) other. Total area were as follows: Carmarthenshire (Pant-y-llyn), 239,500 ha; North Somerset (Yatton), 37,468 ha; Fenland (Twenty-foot), 54,645 ha; Malvern Hills (Layes pool), 57,707 ha. Maps Copyright Google 2009, Map DataÓ 2009

developmental stage of embryos at this time was between gastrulation and late cleavage (Rugh 1951). At 18°C, embryos of R. pipiens reach the late cleavage stage at 21 h post-fertilisation (Rugh 1951). Although no similar data are available for B. bufo, it is likely that the eggs collected had been laid in the previous few days. Spawn collection and sampling of caged individuals Time of spawn-deposition by amphibians is controlled mainly by temperature, and therefore detection/collection of spawn, and its deployment into cages occurred over an 8 day

Spawn were transported to the laboratory in a cooled ice box (5 9 100 eggs) and were kept in 20 L glass aquaria containing 12 L ‘aged’ tap water (tap water allowed to stand for [48 h), in a shaded outside environment (five replicates per site). During embryo development and prior to hatching, water quality [temperature, pH, dissolved oxygen (DO)], and ammonia (NH3?) levels were measured every 7 days. Ammonia values were converted to unionised ammonia (NH3), using temperature, and pH values. Conductivity (lS/ cm) was also measured once at the beginning of the experiment. At hatching (*3 weeks post collection), the number of hatched individuals and any abnormal larvae (edema, irregular swimming, bent spines) were counted. Tanks were emptied and cleaned, and tadpoles were culled to maintain good water quality, before returning larvae to their respective tanks (40 per tank). Following hatching, water change (75%) was conducted every 48 h, and water quality was measured prior to each change. After each water change, larvae were fed boiled organic spinach. To facilitate emergence of metamorphs water was drained, tanks were tilted and sphagnum moss was added when tail regression began to be observed (*8 weeks post hatch). Caged Spawn was placed in replicate cages at each site (4 9 100 eggs). Cage design was similar to that used by Cooke (1977). Square plastic plant pots (Wyevale Garden centre, 35 9 35 9 25 cm) were leached in tap water for 48 h. The middle panels on each side of the pot were removed, the pots were then lined with fine plastic mesh (*1 mm diameter) to allow free flow of water and nutrients into the cage. Upon deployment of the cages, *2 m of nylon wire was tied round each bottom corner of the cage, and the other end of this wire was tied around a breezeblock. To

123

904


Embryos

Lab -3 weeks

2 (+ 7 weeks)

Sampling Points

3 (+ 9 weeks) 4 (+ 12 weeks)

5 (+ 15 weeks)

Field

Moved to Vivariums

End

Observations

4 Cages (x 100 eggs)

Mortality

Tadpoles Culled (40/tank)

Metamorphs

1 (+ 5 weeks)

Larvae

0 Hatching

5 Tanks (x 100 eggs)

Individuals brought to Lab & moved to Vivariums

End

FLE Mortality

Mortality

Fig. 2 Experimental design of laboratory (Lab) and field site (Cage) observations and sampling points during embryonic, larval, and postmetamorphic development of Bufo bufo. See text for details. FLE forelimb emergence

avoid loss by drowning of caged individuals, they were brought to the laboratory when forelimb emergence occurred at each site, and were kept under the same regime as ex-situ individuals, except native site water was used instead of aged tap water. Native water was collected in large plastic containers and kept for a maximum of 1 week in an outside environment. In the field environment, water quality parameters (temperature, pH, DO, and conductivity) were measured (YSI 556 handheld multiparameter system, YSI Hydrodata UK) at the time of egg retrieval, hatching, and TP1, 2, and 3 (five times). Between TP2/3, metamorphosis occurred in most of the individuals, and therefore, tanks were drained and water quality measurements were not taken. Ex-situ and caged metamorphs On completion of metamorphosis, metamorphs were transferred to vivariums, which were made from plastic rat boxes (21 9 33 9 19 cm) and covered with muslin cloth (held in place with elastic). The bottom of the vivarium was covered with damp sphagnum moss and a shallow ceramic dish filled with water was placed in the corner. Animals were initially fed wingless Drosophila (Blades Biological, UK), and when they were larger, micro-crickets (Blades Biological, UK), every 48 h. The vivariums were also checked for moisture at feeding, and aged tap water was added to sphagnum moss and ceramic dishes when necessary. Morphological and histological analysis At each sampling TP, ex-situ and caged individuals were anaesthetised by immersion in MS-222 (1 g/L), and sacrificed by pithing (brain death by insertion of a needle through the skull, all animal procedures were approved by Brunel University and the Home Office). Ex-situ individuals were

123

weighed and snout-vent length (SVL) and hindlimb length (HLL) were measured with digital callipers. Samples were then fixed in neutral buffered formalin (NBF: Sigma, U.K.). Due to practicalities of taking measurements in the field environment, caged specimens were fixed in NBF prior to morphological measurements and correction factors for morphological measurements were calculated using measurements from before and after NBF treatment of ex-situ specimens. Due to size differences between ex-situ and caged individuals, HLL was expressed as a proportion of SVL (HLL/SVL) and weight as body-mass index (BMI, weight/SVL), to facilitate comparison between these groups. Specimens were dissected by opening the body cavity and the gonad–kidney complex was removed. The head was removed for thyroid analysis. Tissue was placed in Bouins’ solution (20 h), followed by 70% industrial methylated spirits (24 h), placed in a tissue processor (Shandon Citadel 2000, Thermo Fisher Scientific, UK) and embedded with paraffin wax. Thyroids from TP2 individuals were sectioned at 5 lM and two section were mounted on a glass slide every 30 lM. Gonads from TP3,4,5 individuals were sectioned longitudinally at 7 lM, and each section was mounted on a glass slide. All slides were randomised and coded (any identifying labels on slides were hidden), and analysis of all histological sections was undertaken blind. Analysis of histological sections Analysis of thyroid histopathology was undertaken using OECD guidelines developed for Xenopus laevis (OECD 2007). Briefly, sections were assessed for epithelial cell hypertrophy/hyperplasia and colloid depletion using a scoring system ranging from 1 (normal) to 5 (severe) according to the level of disruption. Bidder’s organ (Figure S1) and gonads (Figures S2–S4) were analysed. The sex of each individual was characterised as female, male, intersex, or undifferentiated. Ovaries were characterised by reduction of the medulla, leading to formation on an ovarian cavity, and the presence of oogonial cell nests or oocytes. Testes were characterised by a well developed medulla, and organisation of spermatogonia into early seminiferous tubules (Falconi et al. 2004). Undifferentiated individuals could not be characterised as female or male due to absence of features described above. Every testicular section was checked for the presence of testicular oocytes (TO) and intersex individuals were characterised as containing a minimum of one TO (Figure S4). Where found, the number of TO’s per testis were counted to assess severity of intersex. To quantify organ size, the number of sections containing the right Bidders’ organ or gonad was counted, the section containing the largest area of tissue was photographed (by observing the middle numerical section, and the preceding/proceeding


905

Table 1 Mortality at distinct timepoints at each site, ex-situ and caged Hatch failurea N (rep.)

Mean % (SE)

Larval mortalityb

Metamorph mortalityc

Totald

N (rep.)

N (rep.)

Mean % (SE)

Mean % (SE)

Mean % (SE)

Ex-situ Pant-y-llyn

500 (5)

6.2 (1.5)

200 (5)

9.5 (3.6)

131 (5)

51.0 (3.4)

60.5 (4.1)

Yatton

500 (5)

5.6 (2.3)

200 (5)

8.0 (3.5)

134 (5)

55.5 (5.7)

63.5 (6.1)

Twenty-foot

500 (5)

7.2 (1.2)

200 (5)

12.0 (2.7)

126 (5)

39.0 (5.5)

51 (5.5)

Layes Pool

500 (5)

16.2 (10.2)

200 (5)

38.0 (13.2)

78 (5)

16.5 (8.9)

54.5 (6.4)

Pant-y-llyn Yatton

400 (4) 400 (4)

15 (2.9) 69.5 (11.0)

122 (4) 340 (4)

85.5 (6.6) 61.2 (7.5)

32 (2) 18 (4)

5.1 (2.9) 5.9 (2.1)

90.6 (3.7) 67.1 (9.0)

Twenty-foot

400 (4)

76.7 (2.4)

93 (4)

36.4 (2.4)

35 (4)

3.1 (2.1)

39.5 (3.7)

Layes Pool

400 (4)

81.5 (4.8)

74 (4)

100 (0)

Cage

–

–

100

Rep. replicate a

Percentage of eggs that failed to hatch

b

In ex-situ tanks animals were culled and value is a percentage of culled total, in cages n values represent the number of successfully hatched tadpoles c

In ex-situ and cages, n values represent the number of individuals surviving post-metamorphosis (minus the number of sampled larvae = 50 in all cases, except Layes pool = 46) and the mean % is a measure of mortality from the n of larvae d

Addition of larval mortality and metamorph mortality results in the total percentage mortality post-hatch

three sections) and the number of tissue containing sections was then multiplied by the area at the largest point (Bidders’ organ/testes are ovoid and ovaries lobular, and thus the size of the organs were a relative measure for use within this study, not absolute values). To assess ovarian and Bidderian development, the size of first growth phase diplotenic oocytes, which provide a measure of the maturity of the organ (Falconi et al. 2007), were measured. To ensure standard oocyte measurements between photographs, only oocytes with a visible nucleus were measured (the nucleus was generally found in only 1 section). All slides were analysed microscopically (Olympus BXSI microscope) and photographs taken/analysed (Micropublisher 5.0 RTV camera, QCapturePro 5.1 software). Statistical analysis Sex ratio data were analysed using chi-square test. Linear regression analysis for growth data was performed using Graphpad Prism (version 4). All other data were analysed for normality (Bartlett’s test), and when normally distributed, were assessed for differences between groups using ANOVA. When not normally distributed, pairs were analysed with Wilcoxon’s test and groups with Kruskal– Wallis’s test. If significant differences were observed between groups, Dunnett’s or Tukey’s tests were used for post-hoc analysis. No differences in replicate tanks (ex-situ) or cages were observed and therefore replicate data within treatments were pooled. Student’s t-test (unpaired)

was used for comparison of the number of TO’s and percentage of intersex per tank in ex-situ or caged individuals. Data was analysed using JMPin software (version 7).

Results Water quality There were no differences in aged tap water quality between tanks containing ex-situ individuals (ANOVA p [ 0.05). Values were within the range required for normal tadpole development [mean ± SE, temperature (°C): 15.6 ± 0.02, pH: 8.03 ± 0.01, DO (% saturation): 75 ± 0.6, conductivity (lS/cm): 817.4 ± 4.2, NH3 (mg/L): 0.03 ± 0.007]. There was no difference in temperature, and pH, between laboratory conditions and field sites (ANOVA p [ 0.05). However, DO was significantly higher at Yatton and Twenty-Foot (110 ± 13 and 109 ± 6 Dunnett’s p = 0.013 and 0.008) and conductivity was significantly higher at Twenty-Foot (2485.2 ± 228 lS/cm, Dunnett’s p \ 0.0001), and lower at Yatton and Pant-y-Llyn (500.5 ± 16.3 and 212.3 ± 12.5, Dunnett’s p = 0.018 and\0.0001), compared to laboratory conditions. Mortality In ex-situ individuals, hatching failure and larval mortality were low (*6 and 10%, Table 1), with the exception of

123

906

F. Orton, E. Routledge 1000

Snout-vent length (mm)

22

900

Weight (mg)

800 700 600 500 400 300 200 100

20 18 16 14 12 10

1

2 3

4

5

1

Timepoint

2 3

4

5

Timepoint

Fig. 3 Linear regression analysis revealed significant differences in weight and SVL between sites in ex-situ individuals between time points 1–2 and 3–5 (see text for details). Black squares/black line Pant-y-Llyn, grey triangles/grey line Yatton, cross/grey dash line Twenty-foot, open circle/black dash line Layes Pool. Data points represent mean values for each tank (n = 5 individuals per tank for time points 1 and 2 from all sites, except two tanks at Layes Pool at timepoint 2, n = 4 and 2). Due to high mortality post-time point 2,

total n values and the number of remaining replicate tanks from other time points varied [Pant-y-Llyn: TP3, n = 14 (five tanks); TP4, n = 10 (four tanks); TP5, n = 5 (three tanks); Yatton: TP3, n = 13 (five tanks); TP4, n = 5 (three tanks); TP5, n = 5 (four tanks); Twenty-Foot: TP3, n = 21 (five tanks); TP4, n = 15 (five tanks); TP5, n = 10 (five tanks); Layes Pool: TP3, n = 27 (five tanks); TP4, n = 15 (five tanks); TP5, n = 5 (four tanks)]

Layes pool (16 and 38%, Table 1), which was significantly higher in the larval stage compared to Yatton (Tukey’s p = 0.03). In addition, more abnormal larvae were observed in ex-situ individuals from Layes pool (‘high’ agriculture site) compared to other sites (mean ± SE: Panty-llyn = 1.4 ± 0.7; Yatton = 1 ± 0.5; Twenty-foot = 2 ± 0.7; Layes pool = 5.2 ± 1.2, Dunnett’s p = 0.02). Metamorph mortality was high in all groups (16–55.5%), but was significantly lower at Layes pool compared to Pant-y-llyn and Yatton (Tukey’s p \ 0.01), but not Twenty-foot. In caged individuals, hatching mortality was higher at all sites compared to Pant-y-llyn (Table 1, Tukey’s p \ 0.0001) and larval mortality was higher at Layes pool compared to Yatton and Twenty-foot (Table 1, p \ 0.001), higher at Pant-y-llyn compared to Yatton and Twenty-foot (p \ 0.05) and higher at Yatton compared to Twenty-foot (p = 0.04). Metamorph mortality did not differ between groups. Due to high mortality and consequent low n values in caged individuals, only those from Twenty-foot were analysed histologically (see below).

Growth

Development Metamorphic development differed between ex-situ individuals. Increased percentage of forelimb emergence, and larger HLL was observed in individuals from the agricultural sites compared to Pant-y-llyn, which was significant at Layes pool (Tukey’s p \ 0.01). In caged individuals, no difference in forelimb emergence was observed (Tukey’s p \ 0.01), however, HLL/SVL was larger at Yatton and Twenty-foot compared to Pant-y-llyn (Tukey’s p \ 0.01). Due to 100% mortality prior to TP1 at Layes pool developmental parameters from caged individuals could not be assessed.

123

No correlation between weight or SVL and mortality per tank/cage was observed at TP 1–2 (p [ 0.6) or TP 3–5 (p [ 0.5). Morphometric comparisons between ex-situ individuals showed significantly slower growth at the ‘high’ agricultural sites, Twenty-foot and Layes pool, compared to the ‘low’ agricultural site (Yatton) and the ‘reference’ site (Pant-y-llyn). A significantly positive correlation between TP1–2 or TP3–5 and weight was observed at Pant-y-llyn [r2(slope): TP1–2 = 0.6(82.6); TP3–5 = 0.7 (228.6)], Yatton [r2(slope): TP1–2 = 0.4(58.2); TP3–5 = 0.8 (175.0)] and Twenty-foot [r2(slope): TP1–2 = 0.7(53.0); TP3–5 = 0.8(91.1)], but not Layes pool [r2(slope) TP1–2 = 0.2(-32.4); TP3–5 = 0.004(6.2)]. Comparison of slopes between Pant-y-llyn (‘reference’ site) and agricultural sites revealed a significant difference at TP3–5 at Twenty-foot (p = 0.006) and a significant difference at TP1–2 and TP3–5 at Layes pool (p \ 0.001). A significantly positive correlation between TP and SVL was observed at Pant-y-llyn [r2(slope): TP1–2 = 0.5(142.3); TP3–5 = 0.7(284.0)], Yatton [r2(slope): TP1–2 = 0.6(125.8); TP3–5 = 0.7(197.9)], Twenty-foot [r2(slope): TP1–2 = 0.7(78.8); TP3–5 = 0.6 (128.5)] and Layes pool at TP1–2 [r2(slope) = 0.6(110.7)] but not at TP3–5 [r2(slope) = 0.01(16.8)]. Comparison of slopes between Pant-y-llyn (‘reference’ site) and agricultural sites revealed significant difference at TP3–5 at Twenty-foot (p = 0.01) and Layes pool (p \ 0.001). All caged individuals were significantly smaller than their ex-situ counterparts at TP2 (weight/SVL: Pant-y-llyn: 3.5-fold/1.7-fold; Yatton: 2.3-fold/1.3-fold; Twenty-foot: 1.1-fold/1.1-fold). These difference persisted over the experimental period until TP5 at Pant-y-llyn and Yatton and TP4 at Twenty-foot (data not shown) (Fig. 3).

907 100

75

75

Sex Ratio (%)

100

50

25

0 YT

50

25

Ex-situ

TF

A Ex-situ

Cage

B TF 100

75

75

Sex Ratio (%)

100

50

25

0

50

25

0 4

3

C Ex-situ

A

0 PYL

Sex Ratio (%)

Fig. 4 Pooled sex ratio’s over all TP’s of: ex-situ individuals from each site (a), ex-situ and caged individuals from Twenty-Foot (b). Sex ratio by TP in caged individuals from Twenty-Foot (c), a significant difference between ex-situ and caged individuals over time was observed. Change in severity of intersex over time in ex-situ and caged individuals from Twenty-foot (d). Black F, dark grey I, light grey M, white U. N values: a PYL = 27, YT = 23, TF = 46; b ex-situ = 46, cage = 32; c TP3, n = 25; TP4 n = 15, TP5 n = 10; d TP3, n = 13; TP4, n = 10; TP5, n = 9

Sex Ratio (%)


5

Histopathology Using OECD guidelines for analysis of thyroid histopathology, no differences were observed between any groups (data not shown). Bidderian and gonadal development were similar to that previously reported for B. bufo (Falconi et al. 2004), except the previously reported mesogonad in males was rarely observed in this study (Figure S1–S4). Sex ratios were significantly different in ex-situ individuals (v2 p = 0.01), more undifferentiated individuals were observed at Yatton, and more intersex at Twenty-foot (Fig. 4a). There was no difference in overall sex ratio between ex-situ and caged individuals from Twenty-foot (Fig. 4b). In caged (but not ex-situ) individuals the proportion of male/ intersex increased over time and the proportion of females decreased (Fig. 4c/d, v2 p = 0.005). Over all TP’s more TO were observed per testis in ex-situ than the caged individuals (mean ± SE: ex-situ = 9 ± 2.9; caged = 3.2 ± 0.7; t-test, p = 0.06) but the percentage of intersex per tank was lower (mean percentage ± SE: ex-situ = 13.2 ± 6.2; caged = 30.2 ±5.5; t-test, p = 0.08). There were no sex-specific differences in Bidders’ organ size, or size of bidderian first growth phase oocytes within individuals sampled at each site ex-situ, except that Bidders’ organ size in undifferentiated individuals was larger than those in intersex individuals at Twenty-foot (t-test p = 0.037). Due to the minimal effect of sex on Bidders’ organ, data was pooled between sexes for further analysis

3

D Cage

Time Point

4

5

Time Point

(Table S1). No differences in Bidders’ organ could be observed between ex-situ individuals except the size of bidderian first growth phase oocytes was larger at Yatton than Pant-y-llyn or Twenty-foot (ANOVA p = 0.01). There were no differences in ovary or male/intersex testis size between ex-situ individuals. Ovary size (1.6-fold, ANOVA p = 0.05), size of ovarian first growth phase oocytes (1.6-fold, Wilcoxon p = 0.016) and testis size (1.4-fold, ANOVA p = 0.001) were larger in ex-situ individuals from Twenty-foot compared to their caged counterparts (Table S1).

Discussion Data reported here suggest that events in early embryogenesis, such as genotype, maternal transfer of pollutants and/or early exposure, had a much larger effect on growth, development and sexual differentiation in B. bufo than their subsequent ambient environment. Surprisingly, almost all the endpoints measured were different in ex-situ individuals from different sites, which may have important consequences in relation to the analysis of amphibian exposure data. The negative effects on growth and development were related to the intensity of arable agriculture, with higher mortality and slower growth from the highly agricultural. Intersex is also reported for the first time in a European amphibian species in situ.

123

908

Mortality Mortality over the experimental period was high in ex-situ and caged individuals from all sites, although this is not unusual for toad species. During the larval stage, mortality was less than 20% in ex-situ individuals, which was lower than previous reports for B. bufo (60%: Zaccanti et al. 1969; 37%: Petrini and Zaccanti 1998). Comparison of post-larval mortality was not possible as no previous data were available. In ex-situ individuals, the very short window (hours) of conversion from gill respiration to exclusive lung respiration and inter-individual variability in rate of metamorphosis, may have contributed to post-larval mortality. It is likely that not all tanks were drained prior to conversion and mortality of some individuals at this TP was an artefact of drowning. Moreover, metamorphs did not always climb onto the sphagnum moss after tank draining, which may have been due to lack of adequate terrestrial environment in the tank. In cages, larval mortality (typically around 80%) may have been partly due to predation, as invertebrates were sometimes observed in cages at Yatton and Pant-y-llyn. Cooke (1977) reported that invertebrates can quickly decimate a population of tadpoles in cages. Previous studies have reported loss of tadpole populations from field deployment of cages. In one case tadpole samples could not be retrieved from one-third of the cages deployed (Karasov et al. 2005), and in another mortality was generally low (exposure period over larval stage only), but all tadpoles disappeared from some cages (Cooke 1977). Growth and development It is well known that hatching and embryogenesis are extremely sensitive stages of development in vertebrates. In caged individuals, hatching success was higher at Panty-llyn (84%), compared to Yatton, Twenty-foot, and Layes pool (*20%), suggesting that early life stage toxicity was lower at the reference site compared to the agricultural sites. A negative correlation between hatching success and increasing industrial contaminant exposure in caged R. pipiens and Rana clamitans-melanota has previously been reported (Karasov et al. 2005). Total posthatch mortality in ex-situ individuals from Layes pool was comparable to other sites, however, it had the lowest hatching rate and higher larval abnormalities than other sites. In addition, no larvae were observed in cages at this site after hatching or outside the cages in the ambient water, one explanation could be lethal toxic effects of agrochemicals as it is known that low concentrations of pesticide mixtures cause higher mortality (35%) compared to each pesticide alone (0–7.8%) in amphibian larvae (Hayes et al. 2006).

123


Surprisingly, despite having been reared in identical conditions, there were significant differences in size of ex-situ individuals from different sites. Those from Pant-yLlyn (‘reference’) and Yatton (‘low’ agriculture) were largest, and from Twenty-Foot and Layes Pool (‘high’ agriculture) smallest, which followed the same trend as intensity of arable agriculture at field sites. These differences became more pronounced over the experimental period, suggesting that early life stage influences, possibly including agricultural contaminants, may have a significant and lasting effect on metamorphs. Possible explanations include: non-anthropogenic differences in water quality, maternal transfer of pollutants, distinct genetic groups and epigenetic effects caused by early exposure to pollutants from the ambient aquatic environment. Although conductivity and DO values differed between sites, it has been reported that values of 42–1,790 lS/cm have no effect in FETAX (Fort and Paul 2002), and it seems unlikely that the differences in DO would affect these endpoints. It has also been reported that different races of Rana temporaria display different age and size at metamorphosis in the laboratory, depending on latitude of the parent frogspawn (Laugen et al. 2003), although the range of sampling sites in the latter study was 14°N (*1,500 km) compared to 1°N latitude (*160 km) here. Maternal and genetic input have also been shown to affect age and size at metamorphosis. Plasticity in size at metamorphosis was high, and was affected by female origin, male origin, and food abundance (Laugen et al. 2002). If the same is true in B. bufo, this may have contributed to differences observed in ex-situ individuals. Hatching success and developmental abnormalities have been correlated with levels of trace metals (associated with coal combustion waste) in narrow-mouth toad (Gastrophyrne carolinensis) in adults (Hopkins et al. 2006). Although similar studies using pesticides have not been reported in amphibians, in ovo exposure to the pesticides atrazine (0.2 lg/L) or endosulfan (2 or 20 lg/L) reduced hatchling size of Caiman latirostris (Beldomenico et al. 2007). In addition, Karasov et al. (2005) reported that tadpole growth was negatively correlated with contaminant loading of sediment by industrial pollutants. A smaller size at metamorphosis may have population level consequences, as it is believed to result in a higher risk of predation, and reduced fecundity due to a smaller size at first reproduction (Harris 1999; Hayes et al. 2006). The rate of metamorphosis also differed in ex-situ individuals from the various sites, with faster development at all the exposed sites compared to Pant-y-llyn (‘reference’). It is well known that metamorphosis can be affected by temperature, pond drying, predation, and nutrition (Rose 2005), however, these factors were equivalent in all ex-situ individuals. It is possible that faster metamorphosis was a consequence of ‘stress’, a generic term for any factor that


increases corticosterone levels and thus increases metamorphic rate (Hayes 1995). Indeed, increased metamorphic rate in ‘stressed’, but smaller size at metamorphosis has been shown in the laboratory in response to a mixture of pesticides (Hayes et al. 2006). Bidderian and gonadal differentiation Previously reported sex specific differences in size of Bidders’ organ and size of first growth phase oocytes in the Bidders’ organ were not observed (Petrini and Zaccanti 1998). Ex-situ individuals from the sites differed in size, as did ex-situ and caged individuals from Twentyfoot. The relationship between gonadal development and somatic growth in amphibians is reported to be weak (Gruca and Michalowski 1961; Ogielska and Kotusz 2004) and therefore histological differences observed in gonadal development are likely to be largely independent of growth. Due to high mortality, it cannot be discounted that skewed sex ratios in ex-situ and caged individuals were a function of sex-specific mortality, however, to the authors’ knowledge this has not been previously investigated. Similarly to effects observed in morphological endpoints, there were differences in sex ratio of ex-situ individuals between sites. In addition, sex ratios of ex-situ and caged individuals from Twenty-foot were very similar, suggesting very little influence of the ambient environment. Contrary to expectation, the intersex observed at Twentyfoot did not seem to be a result of feminisation, as successively fewer females were observed over time, while the proportion of intersex and males increased. Moreover, if intersex was a stage of normal gonadal development in this race of B. bufo, as has been reported in Ranids (Witschi 1929), the percentage of females could be expected to stay constant, while the percentage of male and intersex would change. In support of these data ovary size and size of first growth phase oocytes were smaller in caged than ex-situ individuals and less TO were observed in caged than ex-situ individuals. To the authors’ knowledge there are no reports of masculinisation in response to pollutants in amphibians, however, in fish it has been linked with compounds that inhibit aromatase activity such as fadrozole (Zerulla et al. 2002), TBT (McAllister and Kime 2004), and with paper and pulp mill effluent in situ (Larsson et al. 2000). Similarly, the pharmaceutical anti-androgen cyproterone acetate caused masculinisation, perhaps due to its inhibitory effect on 3b-HSD activity in developing amphibians (Hsu et al. 1979). Although little is known about the effects of current use pesticides on steroidogenesis in amphibians in vivo and resultant effects on sex differentiation, it is known that environmentally relevant pesticides disrupt steroidogenesis in amphibian oocytes in vitro (Orton et al. 2009).

909

Conclusions Differences in growth, development and sexual differentiation in ex-situ individuals (reared under identical conditions) from the sites were unexpected, and suggest that traditional exposure scenarios starting at the larval stage are in danger of missing crucial developmental windows. The variability in the morphology and development of amphibians from the different sites ex-situ (which may be associated with differences in early environmental exposure to pollutants and genetic/epigenetic components) may partially explain some of the inconsistencies in the literature pertaining to the response (or lack of response) of amphibian species to some chemicals. Furthermore, the replication of field effects in the laboratory using this method could be a useful biomonitoring technique if repeatable. The arable agricultural intensity support the negative trends observed on growth and development, which may have implications for metamorph survival to adulthood. Finally, this is the first time intersex has been reported from a European amphibian species in situ; a discovery which may facilitate understanding on the causes (environmental or other) of intersex in native amphibians. Acknowledgments Funding from DEFRA gratefully acknowledged. Also, the generous sharing of amphibian databases from Froglife (registered charity: 1093372).

References Beldomenico PM, Rey F, Prado WS, Villarreal JC, Munoz-de-Toro M, Luque EH (2007) In ovum exposure to pesticides increases the egg weight loss and decreases hatchlings weight of Caiman latirostris (Crocodylia: Alligatoridae). Ecotoxicol Environ Saf 68(2):246–251 Bishop CA, Mahony NA, Struger J, Ng P, Pettit KE (1999) Anuran development, density and diversity in relation to agricultural activity in the Holland River watershed, Ontario, Canada (1990–1992). Environ Monit Assess 57(1):21–43 Carr JA, Gentles A, Smith EE, Goleman WL, Urquidi LJ, Thuett K, Kendall RJ, Giesy JP, Gross TS, Solomon KR, Van der Kraak G (2003) Response of larval Xenopus laevis to atrazine: assessment of growth, metamorphosis, and gonadal and laryngeal morphology. Env Tox Chem 22(2):396–405 Calisi RM (2005) Variation in Bidder’s organ volume is attributable to reproductive status in Bufo woodhousii. J Herpetol 39(4):656–659 Cooke AS (1977) Effects of field applications of herbicides diquat and dichlobenil on amphibians. Environ Pollut 12(1):43–50 Davidson C, Knapp RA (2007) Multiple stressors and amphibian declines: dual impacts of pesticides and fish on yellow-legged frogs. Ecol Appl 17(2):587–597 DEFRA (2005) e-Digest statistics about: Inland water quality and use. http://www.defra.gov.uk/environment/statistics/inlwater/iwground water.htm. Retrieved January 2007 Falconi R, Dalpiaz D, Zaccanti F (2004) Ultrastructural aspects of gonadal morphogenesis in Bufo bufo (Amphibia anura) 1. Sex differentiation. J Exp Zool A Comp Exp Biol 301A(5):378–388

123

910 Falconi R, Dalpiaz D, Zaccanti F (2007) Morphological aspects of gonadal morphogenesis in Bufo bufo (Amphibia anura): Bidder’s organ differentiation. Anat Rec Adv Integr Anat Evol Biol 290(7):801–813 Fellers GM, McConnell LL, Pratt D, Datta S (2004) Pesticides in mountain yellow-legged frogs (Rana muscosa) from the Sierra Nevada Mountains of California, USA. Environ Toxicol Chem 23(9):2170–2177 Fort DJ, Paul RR (2002) Enhancing the predictive validity of frog embryo teratogenesis assay—Xenopus (FETAX). J Appl Toxicol 22(3):185–191 GAA (2004) http://www.iucnredlist.org/initiatives/amphibians/ analysis/major-threats. Retrieved Nov 2009 Greulich K, Pflugmacher S (2004) Uptake and effects on detoxification enzymes of cypermethrin and tadpoles of embryos. Arch Env Cont Tox 47(4):489–495 Gruca S, Michalowski J (1961) The rate of development of gonads against the backdrop of the development of the whole organism in the tadpole Xenopus laevis Daud. Acta Biol Cracoviensia Ser Zool 4:191–197 Gyllenhammar I, Holm L, Eklund R, Berg C (2009) Reproductive toxicity in Xenopus tropicalis after developmental exposure to environmental concentrations of ethynylestradiol. Aquat Toxicol 91(2):171–178 Harris RN (1999) The anuran tadpole: evolution and maintenance. In: McDiarmid RW, Altig R (eds) Tadpoles: the biology of anuran larvae. University of Chicago Press, Chicago, pp 279–294 Hayes TB (1995) Interdependance of corticosterone-hormones and thyroid-hormones in larval toads (Bufo boreas). 1. Thyroid hormone-dependent and hormone-independent effects of corticosterone on growth and development. J Exp Zool 271(2): 95–102 Hayes TB, Collins A, Lee M, Mendoza M, Noriega N, Stuart AA, Vonk A (2002) Hermaphroditic, demasculinized frogs after exposure to the herbicide atrazine at low ecologically relevant doses. Proc Natl Acad Sci USA 99(8):5476–5480 Hayes T, Haston K, Tsui M, Hoang A, Haeffele C, Vonk A (2003) Atrazine-induced hermaphroditism at 0.1 ppb in American leopard frogs (Rana pipiens): laboratory and field evidence. Environ Health Perspect 111(4):568–575 Hayes TB, Case P, Chui S, Chung D, Haeffele C, Haston K, Lee M, Mai VP, Marjuoa Y, Parker J, Tsui M (2006) Pesticide mixtures, endocrine disruption, and amphibian declines: are we underestimating the impact? Environ Health Perspect 114:40–50 Hopkins WA, DuRant SE, Staub BP, Rowe CL, Jackson BP (2006) Reproduction, embryonic development, and maternal transfer of contaminants in the amphibian Gastrophryne carolinensis. Environ Health Perspect 114(5):661–666 Hsu CY, Hsu LH, Liang HM (1979) The effect of cyproterone acetate on the activity of delta5-3beta-hydroxysteroid dehydrogenase in tadpole sex transformation. Gen Comp Endocrinol 39:404–410 Jobling S, Coey S, Whitmore JG, Kime DE, Van Look KJW, McAllister BG, Beresford N, Henshaw AC, Brighty G, Tyler CR, Sumpter JP (2002) Wild intersex roach (Rutilus rutilus) have reduced fertility. Biol Reprod 67(2):515–524 Karasov WH, Jung RE, Vanden Langenberg S, Bergeson TLE (2005) Field exposure of frog embryos and tadpoles along a pollution gradient in the Fox River and Green Bay ecosystem in Wisconsin, USA. Environ Toxicol Chem 24(4):942–953 Kloas W, Lutz I, Urbatzka R, Springer T, Krueger H, Wolf J, Holden L, Hosmer A (2009) Does atrazine affect larval development and sexual differentiation of South African clawed frogs? Ann NY Acad Sci 1163:437–440

123

F. Orton, E. Routledge Knutson MG, Richardson WB, Reineke DM, Gray BR, Parmelee JR,Weick SE (2004) Agricultural ponds support amphibian populations. Ecol Appl 14(3):669–684 Larsson DGJ, Hallman H, Forlin L (2000) More male fish embryos near a pulp mill. Environ Toxicol Chem 19(12):2911–2917 Laugen AT, Laurila A, Merila J (2002) Maternal and genetic contributions to geographical variation in Rana temporaria larval life-history traits. Biol J Linn Soc 76(1):61–70 Laugen AT, Laurila A, Rasanen K, Merila J (2003) Latitudinal countergradient variation in the common frog (Rana temporaria) development rates—evidence for local adaptation. J Evol Biol 16(5):996–1005 McAllister BG, Kime DE (2004) Early life exposure to environmental levels of the aromatase inhibitor tributyltin causes masculinisation and irreversible sperm damage in zebrafish (Danio rerio) (vol 65, pg 309, 2003). Aquat Toxicol 67(3):301–302 McCoy KA, Bortnick LJ, Campbell CM, Hamlin HJ, Guillette LJ, St. Mary CM (2008) Agriculture alters gonadal form and function in the toad Bufo marinus. Environ Health Perspect 116(11): 1526–1532 McDaniel TV, Martin PA, Struger J, Sherry J, Marvin CH, McMaster ME, Clarence S, Tetreault G (2008) Potential endocrine disruption of sexual development in free ranging male northern leopard frogs (Rana pipiens) and green frogs (Rana clamitans) from areas of intensive row crop agriculture. Aquat Toxicol 88(4):230–242 Nuckols JR, Ward MH, Jarup L (2004) Using geographic information systems for exposure assessment in environmental epidemiology studies. Environ Health Perspect 112(9):1007–1015 OECD (2007) Series on testing and assessment. Guidance document on amphibian thyroid histology. Environmental Health and Safety Publications, Paris, France Ogielska M, Kotusz A (2004) Pattern and rate of ovary differentiation with reference to somatic development in anuran amphibians. J Morphol 259(1):41–54 Ohtani H, Miura I, Ichikawa Y (2000) Effects of dibutyl phthalate as an environmental endocrine disruptor on gonadal sex differentiation of genetic males of the frog Rana rugosa. Environ Health Perspect 108(12):1189–1193 Orton F, Handy RD, Carr JA (2006) Effects of nitrate and atrazine on larval development and sexual differentiation in the northern leopard frog, Rana pipiens. Environ Toxicol Chem 25(1):65–71 Orton F, Lutz I, Kloas W, Routledge EJ (2009) Endocrine disrupting effects of herbicides and pentachlorophenol: in vitro and in vivo evidence. Environ Sci Technol 43(6):2144–2150 Petrini S, Zaccanti F (1998) The effects of aromatase and 5 alphareductase inhibitors, antiandrogen, and sex steroids on Bidder’s organs development and gonadal differentiation in Bufo bufo tadpoles. J Exp Zool 280(3):245–259 Relyea RA, Diecks N (2008) An unforeseen chain of events: lethal effects of pesticides on frogs at sublethal concentrations. Ecol Appl 18(7):1728–1742 Rohr JR, Raffel TR, Sessions SK, Hudson PJ (2008) Understanding the net effects of pesticides on amphibian trematode infections. Ecol Appl 18(7):1743–1753 Rose CS (2005) Integrating ecology and developmental biology to explain the timing of frog metamorphosis. Trends Ecol Evol 20(3):129–135 Rugh R (1951) The frog: its reproduction and development. McGrawHill Book Company, Inc, New York Smith EE, Du Preez LH, Gentles A, Solomon KR, Tandler B, Carr JA, Van der Kraak GL, Kendall RJ, Giesy JP, Gross TS (2005) Assessment of laryngeal muslce and testicular cell types in Xenopus laevis (Anura Pipidae) inhabiting maize and

Agricultural intensity in ovo affects the Common toad (Bufo bufo) non-maize growing areas of South Africa. Afr J Herpet 54(1):69–76 Sparling DW, Fellers GM, McConnell LL (2001) Pesticides and amphibian population declines in California, USA. Environ Toxicol Chem 20(7):1591–1595 Witschi E (1929) Studies on sex differentiation and sex determination in amphibians. I. Development and sexual differentiation of the gonads of Rana sylvatica. J Exp Zool 52(2):235–265

911 Zaccanti F, di Grande F, Pasquini CP (1969) Effeti del testosterone sulle gonadi e sugli organi di Bidder in esemplari postmetamorfici di Bufo bufo. Academic Nazionale del Lincei Series 8 46:106–110 Zerulla M, Lange R, Steger-Hartmann T, Panter G, Hutchinson T, Dietrich DR (2002) Morphological sex reversal upon short-term exposure to endocrine modulators in juvenile fathead minnow (Pimephales promelas). Toxicol Lett 131(1–2):51–63

123