environment hormones in hula valley

This was a chapter for a book on the Hula that was never published.

Source and transport of natural and synthetic steroid hormones in Hula Valley and their  effects on fish in aquiculture.

L. Shore*, Ph.D, Kimron Veterinary Institute, Bet Dagan, POB 12, Israel

 Key words:  testosterone, ethinylestradiol, estrogen, Cyprinus carpio 

Abstract

Natural and synthetic steroid hormones are constantly being excreted into the environment.  In the Hula Valley, five steroidal compounds were identified in the surface waters – testosterone, estrogen (estradiol-17b and estrone), androstenedione, ethinyl estradiol and estriol.  The sources of the hormones were shown to be cattle pasture and treated sewage effluent and the sources could be identified based on hormone profile.  However, of the five steroids tested, only testosterone could permeate the soil to reach the groundwater.  Fish with male gonadal atrophy but normal female gonads were found in a reservoir used for aquaculture. The gonadal atrophy was associated with the presence of hormones, specifically the synthetic hormone, ethinyl estradiol, which is used in contraceptive pills.    When in the subsequent year, the same fish were grown in the absence of the hormone, the males were found to have normal gonads.  

Introduction

In the past decade, many concerns have been expressed on compounds that mimic hormones and can disrupt reproduction in animals including humans (see Lintelmann et al., 2003 for review).  Of particular concern are synthetic hormones that are specifically design to act on biological systems.  All of these compounds act against a background of naturally produced hormones such as testosterone and estradiol that are constantly release into the environment (see Shore and Shemesh, 2003 for review).  Since our initial observation of the presence of significant amounts of testosterone and estrogen in Lake Kinneret in 1993 (Shore et al. 1993), we have surveyed the Jordan Valley including the Upper Jordan River Catchment and the Southern Jordan River for the presence of steroidal hormones (Bar-El Cohen et al., 2005; Shore et al.. 2004).   The present paper will described the results of these surveys in the Hula Valley as to the determination of sources, the pathways of transport and the effects on fish. 

The primary steroid hormones excreted into the environment are estrone, estradiol-17β, testosterone and androstenedione (Fig. 1). The natural steroids of major concern are estrone and estradiol since they exert their physiological effects at lower concentrations than other steroids and can be found in the environment in concentrations above their Lowest Observed Effect Level (LOEL) for fish (increased vitellogenin) and plants (increased growth) (10 ng/l) (Christiansen, 2002; Shore et al., 1992,1995b). In rivers and soil, estradiol is converted abiotically to estrone, so for environmental studies estradiol and estrone can be considered together as “estrogens” (Colucci et al., 2001; Jürgens et al., 2002). Effluent from human sources can also contain estriol, a weak estrogen excreted in the urine of pregnant women, and synthetic estrogens such as mestranol and ethinylestradiol (Fig. 1) (Wenzel et al., 1998). These synthetic compounds are of particular concern as they have LOEL’s in the order of 1 ng/l (Christiansen et al., 2003).

To study the occurrence, origin and effects of the natural and synthetic steroids, a general survey of the Hula Valley catchment area was performed as well as more in depth studies, at two sites to (1) determine nature of export of hormones from cattle pasture and (2) determine the effect of the hormones on fish raised in aquaculture.

Materials and methods:

Extractions and assays

Manures: Extraction and determination of testosterone, and estrogen in cattle manure were done by established procedures (Shore and Shemesh, 1993; Shore et al., 1993). Androstenedione was determined using a commercial ELISA (DRG GmbH, Marburg, Germany). 

Water: All water samples were collected into acid-washed polyethylene 1 L bottles, and immediately put in a cooler before transporting to the laboratory for further processing.  Extraction and determination of testosterone, estrogen, estriol and ethinylestradiol were done as previously described (Shore et al, 2004).

Soil:  Five gram samples were extracted twice with 15 ml of ethyl acetate, the supernatants combined, evaporated and redisolved in 0.5 ml of methanol.   100 µl aliquots were evaporated to dryness and redisolved in the testing media for analysis.

Statistics: All data are in means ± SD. Student t test used for comparing two groups.  

Figure 1.  Structures of natural and synthetic environmental steroids

Site description

The Hula Valley catchment area is divided by the Jordan River into a Western and Eastern sub-catchment areas.  The Eastern portion (Golan Heights, eastern part of the Hula valley) is characterized by the confluence of the Dan, Hermon and Sneir tributaries at the Joseph Bridge to form the Jordan River, which receives the runoff from the Golan Heights until it leaves the Hula Valley at a second confluence, the Huri Bridge.   At the Joseph Bridge, about 10% of the Jordan River is diverted to the Western Canal which receives runoff from the western or Naphtali Heights.  The Western Canal rejoins the Jordan shortly before its exit from the Hula Valley. Precipitation in the Hula Valley catchment area occurs only in the rainy season from Oct. to May. The area consists of small farms, cattle pasture, fishponds with some urban development.  The water table is shallow being less than 2 meters from the surface. The fifteen sites sampled after each rain event (11 to 12 times a hydrological year) on the eastern side of the catchment are shown in Figure 2.  The sites samples on the western side were in two separate subcatchments (points P, KD on  Fig. 3)  which were sampled on two occasions (Jan. 24, Feb 2, 2005) following rain events. For the aquiculture observations, a reservoir containing 400 tons of three species of fish (grey mullet, carp and tilapia) was sampled to determine the effect on aquaculture of exposure to sewage effluent (site 404/2, fig. 3).  The density of the fish was approximately 1 fish/l.  The reservoir held 5.6 mcu³ over area of 100 acres and a depth of 15 m. The reservoir received water from three sources: sewage effluent from an urban area of about 8000, excess water flow from Lake Hula  and Jordan River water from the Western canal. The limonological properties of the reservoir have been described in detail (Gafny et al., 1999). Samples were taken on 16.02.05 before fish were introduced and 04.05.05, 04.08.05 and 19.11.05 representing 1, 5 and 8 months after the introduction. Samples were taken at depths of 1, 3 and 6 m except on the final sampling before harvesting when the depth was only 0.8 m. Samples were taken from a control reservoir without fish on the same dates.  In Nov. 2005, the remaining males were transferred to another pond containing the same water without any sewage contaminant. Gonadal somatic index (GSI) for the carp was evaluated according to Degani et al. (1996, 1998).

For the cattle pasture study, a fenced field in which cattle were not present most of the year with a pool receiving runoff from the field was studied after the introduction of the cattle for three months. The grazing intensity was 0.63 cows/ha. For edge of field drainage, one liter samples in polyethylene bottles were obtained from the receiving pool after each of three rain events. Samples were acidified and transported to the laboratory.  Single samples of soil were taken after 1 mo., 2 mo and 3 mo. after introduction of cattle at a depth of 10, 20 and 30 cm.

Figure 2.  Sites in the Hula Valley draining the Eastern side of the catchment sampled in the hydrological years 2001/2 and 2002/3. The sites are listed in Table 1 found below.

Figure 3. Two subcatchments (Di, Kd) draining cattle pasture in the western portion of Hula Valley catchment studied in hydrological year 2003/2004.  The defined cattle pasture studied for soil and manure samples was located at N2 (Site #1). The reservoir used for aquaculture was at site 404/2.

Results

Eastern part of the Hula Valley

The transport of testosterone, estrogen, ethinylestradiol and estriol was measured at fifteen sites in the eastern side of Hula Valley catchment area after major rain events (10 -12 per hydrological year).  It was found in the rain season of 2001/2002, that following a rain sequence of 131 mm/wk, there was an initial large increase in the concentration of testosterone (maximum 6 ng/l) accompanied by high estrogen (maximum 6 ng/l), which then gradually declined to non-detectable levels (<0.3 ng/l) over a period of three months. These peaks originated from runoff from cattle pasture (Fig. 5). Later peaks consisted only of testosterone that was moderately associated with sulfate (r²=0.53, P<0.05) and somewhat associated with total phosphorus (r²=0.49, P<0.1) indicating that the origin was leaching from the sulfate rich peat soil typical of the Hula Valley (Fig. 6). Testosterone in this catchment area was therefore due initially to surface runoff from cattle pasture and then as discharge from the soil.  The testosterone pulses after each rain event were more pronounced at the Joseph Bridge than at the Huri Bridge (Fig. 7).  Ethinylestradiol was consistently above 1 ng/l at three sites, two of which were known to receive sewage effluent and one had a high ammonia (>1.2 mg/l) and bacteria count (>50,000 fecal coli/100 ml), probably originating from a leaky sewage holding pond. There was no correlation (r<0.2) with the amount of testosterone, ethinylestradiol, estrogen or estradiol present in the samples with each other. However in the few sites (5/28) in which estriol was above 1 ng/l, ethinylestradiol was also above 1 ng/l (15/28).  A more detailed report on the results for 2001/2002 and 2002/2003 hydrological years can be found in Shore et al., 2004.
                       
                               

                     Ng steroid/l        

Figure 4.  Hormonal profile of runoff from cattle pasture.  

Figure 5.  Testosterone and sulfate concentrations in the runoff in the Hula Valley after rain events in the 2001/2002 hydrological year.

Table 1. Representative testosterone and ethinylestradiol concentrations at 14 sites on the Eastern side of the Hula Valley.

Site no.	Site name	Testosterone ng/l	Ethinylestradiol ng/l		Comments
1	Dan Tributary	1.1	1.3
2	Hermon Tributary	0.9	1.1	Some raw sewage
3	Sneir tributary	1.4	0.7
4	Hyun Creek	1.1	0.9
5	Joseph bridge	2.0	0.5	Confluence of sites 1,2,3,4
6	Kalil Canal	1.3	0.8
7	Yardonin Creek	2.1	1.2
8	Gonen Canal	2.2	1.6	High ammonia and fecal coli
9	Bitachon Canal	1.4	1.3
10	Lehavot Bridge	1.2	0.7	River below sites 6,7
11	Green Bridge	1.2	0.5	River below site 8
12	Hardale Bridge	1.3	0.7	River below site 8
13	Western Canal	2.1	1.2	Some treated urban effluent
14	Pkak Bridge	1.5	0.8	River below Confluence of sites 12,13 and 9.
Mean±SD (no. of samplings)		1.5±0.4 (125)	0.9±0.3 (112)

                   ng/l testosterone

Figure 6. Testosterone concentrations at the Joseph and Huri Bridges over the Jordan River. Samples were taken after each rain event of great than 30 mm/72 hr and during the dry season during which there is no precipitation (Days 225-284). Numbers on the x axis for individual rain events are the days numbered from 26.10.2001 which marked the initial rainfall.

Western part of the Hula Valley

Catchment  Streams:

Two subcatchments (Fig. 3) draining primarily cattle pastures were sampled at 15 sites on two dates following rain events representing 2 subcatchments draining primary cattle pasture. The amount of rain on the two dates sample was 21 and 36 mm/72 h respectively. On the date of the first sampling there were major spills of secondary effluent from sewage ponds.  Streams were considered contaminated with spillage effluent as the ethinylestradiol was detectable (>0.3 ng/l). In the contaminated streams, androstenedione and estrogen were significantly elevated (P<0.05) above the same streams without contamination while testosterone concentrations were not (Fig. 7). 

 Fig. 7.   Testosterone, androstenedione, estrogen, ethinylestradiol concentrations in ng/L determine in samples of water from 15 sites sampled after two rain events in two subcatchments draining primarily cattle pasture.  On one of the two dates, there were major spills of sewage water into the catchments (spillage).  Values are means ±SD.

 Studies of cattle pasture

First we measured the manure content for testosterone and androstenedione by collecting samples at 10 sites, isolated them from the cattle, and sampled the manure piles at one month and three months later. It was apparent that both compounds were excreted in about equal amounts and that there was little change over the three month period (Table 2).

Table 2.  Androstenedione (A) and testosterone (T) in cattle manure, 1 mo. and 3 mo. after being on pasture.  Ten sites were sampled on each occasion. Values are means ± SD of 10 determinations.

	mg/kg (after 1 mo.)	mg/kg (after 3 mo.)	1 mo. vs. 3 mo.
T	15.9±2.4	25.2±6.9	P<0.01
A	13.9±2.1	17.8±2.6	P<0.01
T vs. A	P<0.03	P<0.01

We then measured the amount of hormones in the soil after three months of grazing (Table 3) as well as the concentrations of testosterone and androstenedione in the runoff at the edge of the field (Table 4). It was apparent that androstenedione did not penetrate the soil as did testosterone.

 

Table 3.  Samples from a pool draining the same field as in Table 2.  Samples were taken after three rain events of at least 70 mm in the proceeding 72 h and tested for testosterone and androstenedione. Each point is a single observation done in duplicate.

Rain event	Testosterone (ng/L)	Androstenedione (ng/L)
08.01.04   6:40 10:40 11:50	2.0 2.1 1.7	2.1 1.8 1.6
24.01.04	0.7	1.8
27.01.04	0.7	2.2

Table 4. Testosterone and androstenedione extracted from soil samples at 5, 15, and 30 cm in a field use for cattle pasture, 4, 5 and 6 mo. after introduction of the cattle. Each value represents a single determination done in duplicate.

Soil depth	4 mo.	5 mo.	6 mo.
Testosterone  (ng/Kg)
5 cm	280	250	228
15 cm	130	86	56
30 cm	124	182	40
Androstenedione (ng/Kg)
5 cm	26	50	28
15 cm	14	40	8
30 cm	10	18	12

Studies of the effects on fish

Catchment streams: A survey of the 14 species of fish present in the Dan and Sneir rivers and four other sites in the Hula Valley catchment did not show any evidence of skewed sex ratio (Kroton, 2004}.  However, in a nearby stream (Nachal Kibutzim) which had a rich variety of species (13 species), most species were identified as having a skewed sex ratio. Of particular concern was Acanthobrama lissneri whose habitat is unique to the area.  In this stream river, measurable concentrations of ethinylestradiol (0.6-0.8 ng/l) were found.   The source of the ethinylestradiol appeared to be from the large number of swimmers in the small stream rather than contamination with sewage effluent.

Aquaculture reservoir

Hormone profiles: In the aquaculture pond, the concentration of steroidal hormones in the pond water increased with maturation (Table 5).  In the first months the amount of testosterone was comparable to that of androstenedione, but at the time the spawning (8 months), androstenedione was three fold higher than testosterone.

Prior to introduction of the fish, the reservoir was found to contain appreciable amounts of ethinylestradiol, medroxyprogesterone (both components of contraceptive pills); and benzodiazepines but was negative for barbiturates (Table 6.). After three months in the presence of the fish the levels of these compounds were essential non-detectable (<0.5 ng/l).  In contrast, in a control reservoir without fish, the level of medroxyprogesterone and ethinylestradiol remained between 1.3 and 1.7 ng/l during the same time period. 

Fish gonads: There were three species of fish in the reservoir, Oreochromis aureus (Tilapia), Cyprinus carpio (carp), Mugil cephalous (mullet).  The Tilapia had been treated with methyltestosterone so no ovaries were expected. However the fish should have had masculine gonads, which were absent.  The mullet usually take two years or three years to mature so no gonads were expected in the mullet. However the carp should have had gonads weighing about 800 grams at eight months and the ratio of the GSI should have been in the order of  4 %.  We found that the GSI in the females was between 0.4 to 3% and in the males, the testes were rudimentary.  Histological examination of the ovaries indicated the ovaries were fully matured and there was no evidence of intersex.  Some of the male carp were transferred in Nov. 2006 to a pond.  In the second year, the same males were placed in same water without exposure sewage effluent.  All of the nine male fish examined developed normal sized gonads after 4 months of growth.   

Table 5.  Steroid hormones in an aquaculture pond containing about 400,000 fish.  The pond was sampled 1, 3 and 8 months after introduction of the hatchlings.  Values represent the mean of three samplings at 1, 3 and 6 meters.

Months	Testosterone ng/l	Estrogen  ng/l	Androstenedione ng/l
1	1.62	0.97	2.37
3	3.75	2.20	4.67
8	5.73	7.07	18.23

Table 6:  Pharmaceutical concentrations in ng/l in reservoir water.

	Ethinyl-estradiol	Medroxy-progesterone	Benzo-diazepenes
Before introduction  of fish	1.43	0.5	Weak
One month after	1.20	0.5	Negative
Three months after	<0.5	<0.5	Negative
Eight months after	0.6	<0.5	Negative

Discussion

The major observations on steroidal hormones in the Hula Valley are summarized in Figure 8.

Figure 8.  Summary of observations on source, transport and effects

 of natural and synthetic hormones in the Hula Valley

In the Jordan River and its tributaries, hormone pulses were seen after rain events, particularly in the early part of a hydrological season.  In the dry season, hormone levels were essentially undetectable. At the start of the hydrological season, testosterone, androstenedione and estrogen pulses were observed due to runoff from cattle grazed fields and effluent from fishponds. These initial testosterone and androstenedione pulses dissipate over a three and five month period respectively. The absence of an estrogen pulse is due to estrogen binds tightly to the soil. The longer androstenedione pulse is due to testosterone penetrating the soil while the androstenedione remains on the surface but unlike estrogen is not tightly bound to the soil. The rest of the season is characterize by testosterone pulses which presumably come from soil washout as the pulses correlate with sulfate and phosphorus which are released from peat soils during the same rain events.  

These data suggest the hypothesis that there are four patterns of testosterone transport in the environment: (1) Testosterone associated with estrone, ethinylestradiol and estriol which are characteristic of sewage effluent; (2) Testosterone associated with estrone and estradiol which is characteristic of runoff from cattle pasture and manure fertilized fields; and (3) Testosterone alone, characteristic of leaching from soil and baseflow and (4) androstenedione in a much higher concentration in relation to testosterone which is characteristic of fishpond effluent (Kolodziej et al., 2004).

 Aquaculture

Testosterone, estrogen and androstenedione began rising the fourth month of development.  Androstenedione reached a maximum at the time of spawning.   The high androstenedione may be related to the large increase in androstenedione characteristic of spawning carp species (Sorensen et al., 2005). 

The male carp in the first year exposed to the ethinylestradiol in the range of 1.2-1.4 ng/l and did not develop gonads.  On the other hand, the female carp developed normal ovaries. This indicated that the effect was from an endocrine disruptor like ethinylestradiol which is well documented to affect male gonads in such low concentrations (Christiansen et al., 2002). However this would be the first report in non-laboratory setting that complete failure to develop testes was observed, apparently with just three months exposure to the compound.  Interestingly, in the second year, when the same fish were grown without exposure to ethinylestradiol, the gonads developed normally.  The effect therefore was not permanent. 

References

Barel-Cohen, K., Shore, L.S., Shemesh, M., Wenzel, A., Mueller, J. & Kronfeld-Schor, N. 2005. Monitoring of natural and synthetic hormones in a polluted river.  Journal of Environmental Management 78:16-23.

Casey, F.X.M., Hakk, H. & Simunek, J. & Larsen, G.L.  2004. Fate and transport of testosterone in agricultural soils. Environmental Science and Technology 38:790-798.

Colucci, M.S., Bork H. & Topp, E., 2001. Persistence of estrogenic hormones in agricultural soils: I. 17β-estradiol and estrone. Journal of Environmental Quality 30:2070-2076.

Christiansen, L.B., Winther-Nielsen, M. & Helwig C. 2002. Feminisation of fish. The effect of estrogenic compounds and their fate in sewage treatment plants and nature, Environmental Project No. 729. Danish Environmental Protection Agency. Available at: http://www.mst.dk/udgiv/publications/2002/87-7972-305-5/html/default_eng.htm

Degani, G., Boker, R., Jackson K., 1996. Growth hormone, gonadal development and steroid levels in female carp (Cyprinus carpio). Comp. Biochem. Physiol. C: Pharmacol Toxicol Endocrinol 115:133-140

Degani, G., Boker, R., Jackson K., 1998. Growth hormone, sexual maturity and steroids in male carp (Cyprinus carpio). Comp. Biochem. Physiol. C: Pharmacol Toxicol Endocrinol 120:433-430.
Gafny, S., Gasith, A. & Wise, G.S., 1999 - The Enan reservoir. In: Reservoirs for wastewater storage and reuse. Ecology, performance and engineering design. (eds M. Juanico and I. Dor). Springer, Environmental Science Series, Berlin, pp. 369-387.

Jürgens, M.D., Holthaus, K.I.E., Johnson, A.C., Smith, J.J.L., Hetheridge, M. & Williams, R.J., 2002. The potential for estradiol and ethinylestradiol degradation in English rivers. Environmental Toxicology & Chemistry 21:480-488.

Kolodziej, E.P., Harter, T. & Sedlak, D.L. 2004. Dairy wastewater, aquaculture, and spawning fish as sources of steroid hormones in the aquatic environment. Environmental Science & Technology 38:6377-6384.

Kroton, Y.  2004. Biodiversity and community structure of Jordan system fishes and their relationship to the habitat structure and anthropogenic impact. Thesis M.Sc, Dept. of Zoology, Tel Aviv University.

Lintelmann, L., Katayama, A., Kurihara, N., Shore, L. & Wenzel, A., 2003. Endocrine disruptors in the environment (IUPAC Technical Report). Pure and Applied Chemistry 75:631-681.

Segner, H., Caroll, K., Fenske, M., Janssen, C.R., Maack, G., Pascoe, D., Shהfers, C., Vandenbergh, G.F., Watts M. & Wenzel A., 2003. Identification of endocrine-disrupting effects in aquatic vertebrates and invertebrates: report from the European IDEA project. Ecotoxicological and Environmental Safety 54:302-314.

Shore, L.S., Correll D. & Chakroborty P.K., 1995a. Fertilization of fields with chicken manure is a source of estrogens in small streams, in Steele K., ed., Animal Waste and the Land-Water Interface, Lewis Publishers, Boca Raton, Florida, pp. 49-56.

Shore, L.S., Gurevich M. & Shemesh M., 1993. Estrogen as an environmental pollutant. Bull. Environmental Contamination and Toxicology 51: 361-366.

Shore, L.S., Hall D.W. & Shemesh, M., 1997. Estrogen and testosterone in ground water in the Chesapeake Bay Watershed. Dahlia Greidinger Inter. Symp. on Fertilization and the Environment, pp. 250-255, Technion, Haifa, Israel.

Shore, L.S., Harel‑Markowitz E, Gurevich M, Shemesh M.  Factors affecting the concentration of testosterone in poultry litter.  Environ Sci Health Part A.: 1993; A28:1737-1749.

Shore, L.S, Kapulnik, Y., Ben-Dov, B., Fridman, Y., Wininger, S. & Shemesh, M. 1992. Effects of estrone and 17-estradiol on vegetative growth of Medicago sativa. Physiologia Plantarium 84: 217-222.

Shore, L.S., Kapulnik, Y., Gurevich, M., Wininger, S., Badamy, H. & Shemesh, M., 1995b. Induction of phytoestrogens production in Medicago sativa leaves by irrigation with sewage water. Environmental & Experimental Botany 35:363-369.
Shore,L.S., Reichman, O., Shemesh, M., Wenzel, A.&  Litaor, M.I., 2004. Washout of accumulated testosterone in a watershed.  Science of the Total Environment 332:193-202

Shore, L.S. & Shemesh, M., 2003. Naturally Produced Steroid Hormones and their Release into the Environment. Pure and Applied Chemistry 75:1859–1871.

Sorensen P.W., Pinillos, M. & Scott, A.P.  2005. Sexually mature male goldfish release large quantities of androstenedione into the water where it functions as a pheromone. General & Comparative Endocrinology 140:164-75.

Wenzel, A., Kuechler, Th. & Mueller, J., 1998,  Konzentrationen oestrogen wirksamer Substanzen in Umweltmedien. Report. Project sponsored by the German Environmental Protection Agency; Project No 216 02 011/11 (In German).

.

Ph.D. with P.H.D. (Piled High and Deep)