- Email: [email protected]
Boron in siliceous materials as a paleosalinity indicator∗
Geochimica er Cosmochimico Acre Vol. 45. pp. 1 lo 13 0 Pcrgamon Press Ltd 1981. Printed in Great Brttrin
Boron in siliceous materials as a paleosalinity indicator” MARIAN
J.
FURST?
Division of Geological and Planetary Sciences, California Institute of Technology, Pasadena, CA 91125. U.S.A. (Received 1 June 1979; accepted in revisedform 20 August 1980)
Abstract-The “B(n, a)‘Li nuclear reaction has been used with alpha-sensitive plastic track detectors to determine boron concentrations in siliceous live-collected and fossil sponge spicules. This radiographic technique allows B determinations with 56% uncertainties on objects 2&25pm in diameter and for concentrations as low as 0.5 ppm. Boron concentrations in spicules from different specimens from the same location agreed to within 10% when the spicules were not: (1) smaller than 20 pm in diameter, (2) from dictyonine skeletons, (3) the extremely large root-like spicules found in some soft substrate hexactinellids, or (4) microscleres. These criteria also applied to spicules found in sediment samples. Spicules from live-collected sponges exhibited a taxonomy-independent correlation of B concentrations with water salinity for samples from regions of low water temperature and high productivity. Measured concentrations ranged from nearly 0 ppm B (freshwater sponges) to 500-700 ppm (marine sponges), with intermediate values for brackish-water specimens. However, spicules from tropical, low-productivity marine locations contained markedly less boron than spicules from temperate, high-productivity regions. Thus, water temperature and/or food supply also seem to influence B concentrations. Pleistocene spicules from deep-sea cores contained less boron than was expected in comparison with live-collected spicules based on present water temperatures and nutrient supplies, but B concentrations did not vary with depth in the cores. Infrared spectroscopy, electron microprobe analysis and visual inspection revealed no evidence for chemical or mineralogic alteration. It is not clear whether the lower B concentrations of the Pleistocene samples are the result of diagenetic processes or the as yet undefined effects of differences in food supply and/or environmental conditions.
sponges
INTRODUCFION IN 1932 GOLDSCHMIDT and F’ETERS published emission spectroscopic analyses of biogenic carbonates and opals showing that radiolarian oozes and marine
sponges tend to have significantly more boron than marine carbonate shells. They also reported low boron concentrations in freshwater diatomites (GOLDSCHMIDT, 1954). More recent work has confirmed the high boron contents of marine biogenic opal (GROSS, 1967). The relationship between boron concentrations in opal skeletal material and water salinity was investigated with the goal of finding a suitable type of organism for use in paleosalinity determinations. The greater difference between concentrations in marine and freshwater opal may allow better discrimination of salinities than is possible using carbonate shells (Fuasr et al., 1976). Skeletal remains of several extant groups of opalprecipitating organisms are abundant in the fossil record, including silicoflagellates, diatoms, radiolarians and sponges. The skeletons of the first three taxa tend to be very delicate, porous, and small, presenting difficulties for analysis, particularly the analysis of individual
specimens.
Both
diatoms
*Caltech contribution No. 3221. tPresent address: Schlumberger-Doll Box 307, Ridgefield, CT 06877, U.S.A. ci.c.*.45’ I --A
and
in freshwater
and
marine
environ-
EXPERIMENTAL The radiographic technique employed in this study is described in FURST et al. (1976). Boron determinations were made by placing a plastic detector which is sensitive to alpha particles over each sample. Neutron irradiation results in the “B(n, a)‘Li reaction. Alpha particles which are produced in the surface layers of the sample enter the plastic detector and cause radiation damage. After irradiation, the detector is chemically etched. The radiationdamaged areas etch faster than the bulk plastic, resulting in a series of conical holes or ‘tracks’ which can be counted with a scanning electron or optical microscope. The plastic employed was Kodak-Pathi cellulose nitrate CA%)-15. which is commercially produced specifically for alpha radiography. The background track density was equivalent to about 0.3 ppm B and provides a limit to the sensitivity of the technique. The sources of these tracks are the “G(n, a) reaction, fast neutron interactions, and possibly a small contribution from the 14N(n, n)W reaction. The only significant interference m B beierminations with this track technique is from the ‘Li(n, a)3H reaction. Spicules from several sponges were analyzed for lithium by atomic absorption spectroscopy and found to contain less than 10 ppm Li, too low a concentration for a significant contri-
siliceous
Research,
occur
ments. Because some representatives of each of these two groups can tolerate salinities other than normal marine, and because they occur abundantly in the fossil record, these two groups were chosen for investigation.
P.O. 1
2
MARIAN J. Funsr
bution to the observed spicule track densities; no correction was made for lithium in spicules. Silica glass and pIastic blanks were included in each irradiation. The B Standards used in this work were epoxy-mounted polished sections of National Bureau of Standards SRM 610 and 612. Due to uncertainties in B and Li contents of the two glasses, the accuracy of B determinations is also limited to about 67: (Fuasr et al., 1976). Two types of samples were analyzed (1) Fresh sponges were preserved in 70-7574 ethanol or dried. Small pieces were cut off with a clean scalpel and soaked in 6% NaClO solution (commercial Chlorox) in plastic beakers until the soft parts were completely oxidized and bubbling had ceased (2-4 days). The spicules were collected in a tine sieve (20 or 37pm mesh) and washed repeatedly with distilled water and then finally with spectroscopic quality methanol. They were then potted in E-7 epoxy and polished with aluminium oxide powder. Several specimens had only very small spicules and individual spicules could not be analyzed. Suspensions of these spicules were collected on filter paper and analyzed as bulk samples. Epoxy-mounted and filter-paper-collected spicules from the same sponge had identical B contents for two sponges with large enough spicules to allow the comparison. (2) Core samples were suspended in distilled water and seived. Large spicules were hand-picked from the > 70 pm portion and potted in epoxy. One core, V-19-29, from 3” 35’ S., 83” 56’ W., and 3157 m depth, was thoroughly cemented with carbonate and required dissolution in dilute HCI before sieving In order to verify that the Chlorox treatment did not change the boron content of the spicules due to contamination or leaching in the basic solution, two experiments were conducted. First, a sponge was divided into two parts and one part was treated with Chlorox. Spicules from the treated and the untreated parts were mounted and analyzed separately, yielding B contents which agreed within counting statsitics. Second, several samples were treated with hydrogen peroxide, which is an acidic solution. The boron concentrations also agreed well with those found in Chlorox-treated sponges of the same species from the same location. Subsequent analyses were carried out on Chloroxtreated spicules because of the shorter time required to oxidize the soft parts of the sponges (peroxide digestion required one week or more).
areas counted, they agree to well within counting statistics. 2. Two mounts of sponge spicules were reanalyzed; the second analysis of both samples agreed with the 6rst in both cases to well within the 6% uncertainty of the determinations. The reproducibility of thd optically counted track data was established by repolising and re-irradiating a number of samples on different occasions. Results showed consistency comparable to the SEM data. generally agreeing to better than the 6% standard deviation of each analysis. Duplicate analyses of the same detector also agreed well in most cases. Even when they did not agree to within counting statistics, all analyses of a given sample were averaged. Such discrepancies may have been due to sample inhomogeneities, but could also have represented track counting errors. B. Uniformity of boron in spicules A typical SEM traverse across a single large spicule is shown in Fig. 2. The spicule edge is very clearly
400
3/14/77
390
5000 x
380 370 E 500-g
210-T zoo-
"
RESULTS
Samples and standards usually had B concentrations within a factor of 2-3 of each other and should have similar reproducibilities. The reproducibility of track analyses, determined from SEM photographs at nominal magnifications of 5000 and 7000, was established in two ways. 1. Several SRM 610 (351 ppm) standards were included in each irradiation. Typical results for the standards from two irradiations are shown in Fig. 1. All of the standards from the 3/77 irradiation have track densities within one standard deviation (based on counting statistics) of the mean of all standards counted at the same magnification, and all of the standards from the 4/77 irradiation have track densities within two standard deviations from the mean for that irradiation. Further, when the average track densities of the 3/77 standards measured at different magnifications are normalized for the different
I 500-l
I
I
500-5
300-a
,
1
I
410
-
3/14/77
1
:
7000x
I
_
T
no-' 180 -
A. Reproducibility
m-7
I
500-10 500-12 1
1
420390-v
I
4125177
3801 ,,
7000 x 370 360:[,,,: 350. 500-9 500-10 5onoo-12 Fig 1. Reproducibility of SEM track data. Track densities are plotted for three groups of standards from two irradiations. Error bars represent one sigma standard deviations based on counting statistics. The average track density for each irradiation group is indicated by a horizontal line. When corrected for the different areas of the fields of view, the average track densities for the 3/14/77 data at 5000x and at 7000x agree to within I track per field of view. The neutron fluences are not controlled. so that track densities for different irradiations cannot be compared.
Boron in siliceous materials as a paleosalinity indicator
1
I I 1 lionverse Profile
I I Longhlinal
I
3
I 1 Proftie
3 410.0, ; 390i
-u 5 370x 2350c p 330-
W
_
__-
--_-
--
1
I
310- ?‘7’ 0
25
t 50
1 75
,
1:
I,,,tip
lo0
Distance from edge microns adjacent fields of view
middle
tip
Interval 75~
Fig. 2. A typical traverse across a single large spicule. The transverse profile begins near the boundary of the spicule and goes more than half-way across the spicule. The error bars represent one standard deviation based on counting statistics. In the longitudinal profile, the dashed line indicates the average of the four points plotted, and the four high-density points from the transverse profile have been averaged
and plotted as the third point on the longitudinal profile.
defined in the transverse profile and the boron concentration is uniform within counting statistics across the diameter of the spicule. The longitudinal profile is representative and is not the most uniform case observed. The slight decrease in boron concentration near one end of the spicule may be due to inadequate opal thickness at the spicule tip. The boron concentration is quite uniform (certainly better than 10%) within individual spicules on a scale of 5-10pm. Because the spicules are composed of many layers of opal and organic material, it is likely that the boron is inhomogeneously distributed in these layers on a finer scale than can be resolved using tracks. The agreement between B concentrations in untreated and Chloroxed spicules supports the assumption that the B is located in the inorganic layers. As shown in Fig. 3, the boron contents of different spicules from the same sponge are uniform within counting statistics in most live-collected specimens. However, several sponges were found which had different boron concentrations in spicules originating in different parts of the sponge or in spicules of different morphologies. Spicules from the finger-like projections which supported a specimen of Polymastia robusta (Bowerbank) (YH 10) above a soft substrate had 15% more boron than spicules from the main mass of the sponge. Two specimens of Geodia from Barbados (BB3 and BB4) each had concentrations which differed by a factor of three in different types of spicules. In both specimens, the spicules which had lower B contents were spherical or ellipsoidal in shape and had distinctive surface patterns (sterrasters); these spicules also were the only microscleres encountered which were large enough to analyze. It may be that there is a fundamental difference between boron concentrations in megascleres and microcleres. Because of their distinctive morphologies, most microscleres
are easily recognized and can be excluded from analyses of individual spicules. Subject to the criteria discussed in Section C, it can be concluded that there is a uniform boron concentration in spicules from a single sponge, and that analysis of only one or a few
25765-4s .z _ s r 21)0g e Q
ZZO-
Fig. 3. Plot of track densities for several spicules from each of two different sponges. Each point represents one spicule, and error bars represent standard deviations based on counting statistics. The differences in track densities for the two sponges are due to different irradiation conditions and different unit areas; PV-2-l was counted optically and 25765-4a was counted using the SEM. Average track densities for each group of spicules are indicated with dashed lines.
4
MARIAN J.
spicules will result in a representative boron concentration. C. Selection criteria for boron data from sponge spicules
The following criteria for spicule selection were necessary to obtain reproducible boron concentrations which agreed with other spicules from the same sponge and/or other sponges from the same location. 1. Spicules had to have diameters greater than 20pm. Narrower spicules gave low track densities, presumably because the portions remaining after polishing were thinner than the range of the alpha particles (6-7 pm). 2. Spicules from a dictyonine skeleton were excluded. 3. Root-like spicules, which anchor and support some soft-substrate sponges, were also excluded; early analyses of several of these spicules yielded low B concentrations relative to spicules from the main portions of sponges. However, the low B contents may actually have been related to other factors [see discussion of criterion (5) below] and it is possible that this criterion is not really needed. Also, these very large (around 1 mm dia) spicules did not polish well and the analyses could have been excluded on that basis. 4. Sterrasters and other morphologically distinctive microscleres were excluded. 5. The spicule surface must be well-polished and free of pits and cracks because extraneous material collects in holes and extensive pitting is evidence for dissolution of the spicule. The area analyzed must be from the interior of the spicule; material adsorbed or cemented on the exteriors of spicules is thereby excluded. 6. The central canal of the spicule must be present but not significantly enlarged. If the canal is greatly enlarged due to solution of opal after the sponge died, the walls of the spicule may be too thin and the same problem with low B analyses occurs as in criterion (1). All of these criteria can readily be applied to fossil spicules. It should be emphasized that the first four criteria deal with readily observable morphological features of spicules and present no problem for selection of spicules from sediment samples. Indeed, objects smaller than 20pm in diameter are difficult to handle individually. The spicules selected for boron analyses from core samples were subjected to more stringent selection criteria than .the live-collected specimens. The track densities found in a number of spicules from a single core and irradiated at the same time are plotted against spicule diameter in Fig. 4. For spicules wider than 25 pm, the B concentration is independent of spicule diameter, but several spicules in the 20-25 pm range gave low B results (less than 1600 tracks/field of view). It was concluded that slightly wider spicules
FURST
were required for reliable B determinations. The greater spicule diameter required for fossil samples is probably due to the partial dissolution of the spicules along the central canals. After polishing which removes material from the exterior of the spicule, the remaining wall may be thin relative to the alpha range. In order to identify spicules which had undergone chemical alteration, electron microprobe analyses were obtained for all fossil spicules whose B analyses are included in this work and for spicules of several live-collected specimens also. Spicules from livecollected sponges are known to contain 5-20% water and are composed of otherwise very pure SiO*; no other cations or anions were detected in significant quantities (>O.l%) by the microprobe. Spicules with non-hydrous portions greater than 80% of the total oxide sum, with SiOZ comprising 99.6 + 0.4% of the non-hydrous fraction, and which met the optical criteria discussed above were selected for B analysis, No correlation was found between boron concentration and oxide sum. Only about 5% of the fossil spicules analyzed had to be rejected because of low SiOZ content. The oxides present in addition to Si02 in measurable quantities were A120s, MgO, TiOz, Fe0 and BaO, never in excess of lo/;, of the total. High A1203 concentrations may be due to the presence of grains of A1203 on the surface from polishing. When the first probe analysis of a spicule had an oxide total of less than 80514,the analysis was repeated on a different spot. Usually the second analysis had a total
‘5 .p EOOO-
z
X .!i
2
1600-
“0 c ,.
1200-
.Z s
0" _=
800-
e k 400 t
P
0 70 cm 0 90cm 0 470cm 1
Fig. 4. Plot of track density vs spicule diameter for spicules from the 76, 90-, and 470-cm. depths of core V23-42. All data were obtained from samples irradiated together and track densities are directly comparable to each other. Rep resentative one sigma error bars are based on counting statistics. A track density of 2OOO/field of view is equivalent to about 450 ppm B.
Boron in siliceous materials as a paleosalinity indicator Table 1. Comparison of boron concentrations in ditlerent sponges from the same marine locations Genus and species Haliclona Polymastia
oculata rohusra
XrstosporIgia Terhya crqra
muta
Geodia sp. Geodia sp.
Syringella sp.
Myxilla sp. lophon pattersoni Syringella sp. Thalysias juniperina Iotrochata hirorulata Age/as clathrodes Neotihularia nolitungere Slaurocalyptus Sp. (Hexactinellida) Lissodendoryx sp.
(Hexactinellida) (Hexactinellida) (Demospongia)
At least 4 genera. not identified
WI
Sample YH 7 YH 10
66& 32 604 2 30 > 236 +_ 13 402 i 22
YH 1 YH 3
308 + 30 505 f 19 141 f 1 18 k I1 103 f 25
BB 1 BB 2 BB 3 BB 4 BB 8 FP I FP 2 PV 2 PV 3 ws 1 MV 1
506 f 30 521 k+ 31 430 26
lOOk8 1 353 f 19 125 +9 282 k 17
25765- 1 25765-3
596 599 514 569
25165-4 25766-2 25767-2 26601
Sponges of different species from the same marine location tend to have the same boron concentrations, but there are some notable exceptions. Brackish-water sponges and sponges from tropical locations have variable boron contents. Marine sponges from higher latitudes usually have boron concentrations ranging about ten percent from the mean for each location (6% standard deviations for each determination). When the data for a sample did not agree with other samples from the same location, the sample was rejected only if reexamination demonstrated that there was good &use. Several samples were re-irradiated and gave results in agreement with the values obtained the first time. Table 1 compares boron concentrations from sponges from each of several locations ‘from which two or more species met the selection criteria. Representatives from temperate
+ k f+
34 14 25 30
693 k 53 523 f 19 481 + 31 476 _+ 30 435 + 28 468 + 30 538 + 36 590 + 36
AA 1 AA 2 AA 3 AA 4 AA 5 AA6
D. Intersponge variation in boron concentration
Massachusetts Jamaica
Barbados
Washington
499 * 30 514 +* 35 34 1 581
Cl c2 c3 c4
greater than 800/, and it was assumed that the first analysis included some epoxy due to the difficulty of locating the beam on the narrow spicules. About 2% of the spicules had low second analyses and were rejected. These criteria were applied to all samples discussed below.
Location
Honduras
California
Antarctica
locations show good agreement whereas those from tropical sites do not. Thus, it can be concluded that the observed boron concentrations are independent of species, but appear to be a function of some environmental factor. The groups of sponge samples from Barbados and Southern California included both Demospongia and Hexactinellida, so it can be concluded that the boron concentration is independent of genetic controls at the class level. *Table 2 lists boron concentrations in spicules from representatives of the same genera from different locations. When locations which are both tropical or both temperature are compared, the concentrations are in good agreement; poor agreement is found between tropical and temperate specimens, as in the case of Haliclona. The mediterranean location is slightly hypersaline and intermediate between ‘temperate and ‘tropical’ in climate. Correlarion of the boron content of he-collected spicules with saliniry and orher environmental factors
E.
The results of all boron analyses of siliceous materials are listed in Table 3. Samples of the same species which were collected from one location (Askii Island, Sweden; Woods Hole, Massachusetts; and
MARIANJ. FUR~T Table 2. Comparison of B concentrations
in sponges from the same genera but different locations
PI Location
Sample
Genus and species
ppm
FP 1 ws 1
San Juan Islands, Washington* San Juan Islands, Washington*
587 + 35 506 * 30 >
Syringelia
c2 YH 9
Gulf of Mexico-Honduras Bahamas
353 + 19 345 + 20 >
Iotrochata birotulata
Massachusetts Bay Puerto Scotia Rico Nova
660 f 32 334 + _+33 191 670
Haliclona
Spain-Mediterranean (hypersaline) California (small spicules)
484 k 25 485 * 55 >
Axinella
YH4 YH 7 YH 11 25766-l YH 5
* The two sites in the San Juan Islands are separated by about 10 km.
Table 3. B concentrations
Salinity x
Sample and location LIVE-COLLECTED
in sponge spicules CBI’ ppm
Classification
SPONGES 28-32
San Juan Islands, Washington FPl FP2 PV2 PV3 PV4 WSI MVI Egg1 San Nicholas Island, California
0
Catalina Island, California
596 690 699 599 574 143 293 540 429 569 693 572 177
Staurocalyptus sp. Aphrocallistes vastus
Iophon pattersoni Lyssodendoryx firma
Syringella sp. Spongilla
Fam. Craniellidae Lissodendoryx sp. Axinella sp. Axinella sp. AxineUa sp.
33.5 267 523
26601 26601 -36
Magdalena Bay, Mexico
399 367 621 648 398
4545-l 4545-2 4081-i 4081-2 4544-la 4544-2
Massachusetts Bay, Massachusetts
Syringella sp. Myxilla sp.
33.5 25765-l 25765-2 25765-2 25765-3 25765-4 25765-5 25766-l 25766-l 25766-l 25766-2 25767-2 25767-3 25767-3
Cedros Island, Mexico Northumberland Str., Nova Scotia Eastport, Maine
587 506 521 430 334 499 574 c 1.5
332 -35
YH4 YHl2 YH7 YHlO YHlOa
406
Single Single Single Single Single Single Single
spicules spicules spicules spicules spicules spicules spicules
32 k 1
670
Haliclona oculata
32 k 1 31.6 k 0.2
505
Isodictya palmata
660 604
Haliclona oculata Polymastia robusta Polymastia robusta
698
Boron in siliceous materials as a paleosalinity indicator Table 3 (cont.) Salinity 0, ‘00
Sample and location Woods Hole, Massachusettst WHl (Hz(&)
PI* ppm
Classification
550 573 550 236 213 253 217
Microciona sp. Microciona sp. Microciona sp. Halichondria sp. Halichondria sp. Halichondria sp. Hafichondria sp.
20-30
~712 WzW WH3 (H,O,) WHl3 (HsO,) WH14 (H202) WH15 (HsOz) W H 13~ (Chlorox) Block Island, Rhode Island New Haven. Connecticut
YHS
841
Suberities @us
YH2 YH6
15-32
402 258
Cliona celata Microciona prolifera
Gloucester Point, Virginia
GPl GP2
lC-25
287 428
Halichondria bowerbanki Lissodendoryx carolinensis
North Carolina
CBS5
0
Nassau, Bahamas
YH9
31 + 1
Cl c2 c3 c4
36 + 1
YHl YH3
Honduras
Jamaica Puerto Rico
YHll
Barbados
BB-l-1 BB-1-2 BB-l-4 BB-2-2 BB2 BB-3-4 BB-3-4 BB3 BB4 BB8
Bermuda
3.8 Spongilla sp, 345
lotrochata birotulata
:: 125 282
Thalysias juniperina lotrochata birotulata Age/as clarhrodes Neotibularia nolitangere
36 f 1
236 402
Xestospongia muta Tethya crypta
34+1 36.s36.75
334
Haliclona rubens
round
315 227 331 288 505 288 96 147 78 103
Geodia sp.
Geodio sp. Geodia sp. Geodia sp. Calliospongia uqinalis
Ber-1
37 f 1
61
Ask6 Island. Sweden
AI1 AI2 AI3
5.67.1
29.1 32.3 30.0
Ephydatia Juoiatilis
Cataluna, Spain
YHS
?( > 35)
484
Axinella pollapoides
Arthur Harbor. Antarctica
AA1 AA2 AA3 AA4 AAS AA6
32.5 + 0.5
487 476 435 468 538 590
450 cm
marine
361
CORE SAMPLES RCl3-263 South Atlantic V19-20 E. Equatorial Pacific
V23-42 North Atlantic
0 cm-3 -4 120 cm-3 -4 850 cm-3 -4 OCm 10 20 30 30 40 50 50 60 10
117 241 172 252 217 243 marine
340 344 273 262 546 434 608 524 470 559
8
MARIANJ. FURST Table 3 (conr.)
CBI’
Salinity 0
Sample and location 90 430 450 470 160cm
ppm
Classification
marine
437 477 408 425 649
SOCAL (Gulf of California)
marine
342
MSN Antarctic
marine
384
RC12-249 North Central Atlantic
* Standard deviations of boron concentrations are usually 6% or less. t The Woods Hole, Massachusetts samples were treated with hydrogen peroxide or Chlorox as indicated. All other sponges were treated with Chlorox.
Islands, Washington) have boron concentrations which agree within counting statistics. The average boron concentrations for each location are plotted in Fig. 5. Fresh-water spicules contain 5 ppm B or less, and brackish-water spicules have boron contents between the fresh water and the group of temperate marine values. The obvious environmental difference between tropical and temperate locations is surface water temperature. Boron concentrations have been plotted relative to published surface water temperatures for the marine specimens in Fig. 6. The grouping of the warm- and cold-water points is striking. (Most of the sponges were collected in water San Juan
I
d
,m 5
,
L
10
15
shallower than 100 m). There does not appear to be a simple correlation between temperature and boron concentration; for temperatures less than 15°C the boron concentrations are uniformly high, and for temperatures greater than 20” the boron concentrations are lower and more variable. The high concentration of 84Oppm in the Block Island, Rhode Island sample may reflect polluted water. Water samples collected near population centers may have markedly increased boron concentrations due to the presence of industrial wastes and household detergents (AHL and JONSSON,1972; MATTHEW$ 1974). Long Island Sound is surrounded by a region of high
I
I
20 Salinity,
25
I
30
8
35
I 40
?L
Fig. 5. Plot of B concentration versus salinity for sponge sample locations. Each point represents the average of the boron concentrations of all samples analyzed from each location. The error bars represent the range in boron concentrations for all samples analyzed and the ranges in water salinity. When error bars for salinities are absent, the variations are small (k I”&,or less). The closed circles represent areas of low water temperature (less than 15’C) and high nutrient supply; open circles represent tropical regions (greater than 15°C) with low nutrient supply. The dashed lines indicate a range for the trend of boron concentrations with salinity for cold water sponges.
Boron in siliceous materials as a paleosalinity indicator
It'"".
0'
I
I
I
,
I
I
1
I
-5
0
5
10
15
20
25
30
35
40
Temperature(“C) Fig. 6. Plot of boron concentrations versus surface water temperature for live-collected sponge spicules from normal marine localities. Each point represents analyses of spicules from a single sponge. Representative error bars indicate standard deviations in B conozntrations based on counting statistics and the ranges in surface water temperatures, except the San Nicholas Island specimens. for which the actual temperature at the collecting site was measured. The temperature data for all locations except San Nicholas Island are from U.S. Department of Commerce (1970, 1973). The symbols represent the following locations: O--Arthur Harbor, Antarctica: O-Block Island, Rhode Island; ,&--Massachusetts Bay, Massachussetts; *San Nicholas Island, California: l -Eastport, Maine; m---San Juan Islands Washington; 0-Jamaica; V-British Honduras: M-Barbados; A-Bermuda ; O-Bahamas; O-Puerto Rico.
9
come directly from seawater or indirectly through particulate food. The possibility that the boron concentration in the spicules is a reflection of the boron/silicon ratio in seawater where the sponge grew, was also considered. The sponges from the areas with the largest silicon concentrations (Antarctica, Southern California and Washington) have high boron concentrations in their spicules. The boron concentration of the open sea varies by only lo’%, independent of ocean, latitude, or water depth (MOBERGand HARDING, 1933; RAKESTRAW and MANHCKE,1935; IGELSRUD,et al., 1938; GASSAWAY,1967; RYABININ,1972; UPPSTROM,1974). If the sponges were unable to discriminate between the two elements, one would expect that specimens from high-Si water would have lower spicule B contents than those from low-Si water, contrary to the data presented here. Moreover, since the silicon concentration in the water column increases with depth for the first several hundred meters (CALVERT,1974), one would expect that sponges from deeper water (California) might have a lower B/Si ratio in their spicules, also contrary to observation. It can be concluded that boron and silicon can be discriminated by sponges. F. Boron analyses offossil spicules
The primary purpose in studying the core samples, which are all from normal marine environments, was to test for diagenetic effects on spicules in sediment samples. Ages were determined on the basis of uniform sedimentation rates inferred from radiometric ages of several tug layers in nearby cores (V23-42) (RUDDIMAN,1978), or by oxygen isotopic composition in relationship to microfauna (COOKE, 1978). Visual examination with a microscope revealed a trend toward increased solution pitting on exterior surfaces and enlargement of the central canal with depth in cores (age). Infrared spectra were obtained for spicules from core V23-42 positions corresponding to ages of approximately 0, 10,000 and about 100,000 yr. There was no evidence for any alteration of opal to cristobapopulation density. The wide variation in B concentrations found in the Barbados sponges may be due to lite or quartz. Similarly, electron microprobe data did localized mixing of seawater and freshwater which not indicate any significant alteration or dehydration travelled downward from the surface of the island except in a few spicules; there were no core positions through the porous reef carbonate (SENN, 1940; with abundant large spicules (greater than 25 m dia) from which a significant proportion of the spicules FISCHER,private communication, 1979). However, water temperature is not the only en- had to be excluded on the basis of chemical altervironmental factor which differs for the temperate and ation. Boron contents of individual fossil spicules are tropical locations from which these specimens were plotted against age in Fig. 7. For two cores, V23-42 obtained. All of the tropical specimens are from water and V19-29, data were obtained from several posof low productivity and all of the temperate speci- itions in each core. V23-42 samples were selected to mens are from areas of high productivity (LISITZIN, cover a time interval extending backward through the 1977). By productivity is meant the number or mass last major glaciation, investigating the changes in of planktonic organisms found per unit volume of boron concentrations in spicules with changes in clinear-surface water; the number of organisms is mate (Fig. 7b). There is significantly more boron in dependent on the available supply of nutrients, par- the spicules from the glacial interval (9,000-l 3.000 yr) ticularly nitrogen, phosphorus, and silicon. Further, it and in the older spicules than in the more recently is not known how the sponges obtain the boron deposited material. The peak of the Wisconsin glacial which is incorporated into their spicules. It could period was about 11,500 yr ago. For the first 7500 yr
MARIAN J. Furts’r
10
700-
: 600-
E 2
c s 5 "
.
500-
i
z
g
.
.
I 400-I
al
.
.
I I 300-,
. .
.
I8 '. ZOO-1
I
9
.
. . RCQ-249 4 AC13-263 . v19-29
.
I9 loo-;,
I
I 100,000
1
olooo
w@J
0, l,oOo.ooo
Agelyeors)
700-
600-
I
0
1
O1000
10,coo
100.000
Age(years)
Fig. 7. (a) Plot of boron concentration versus age for spicules from cores V19-29, RC12-249, and RCl3-263. Each point represents one spicule, selected according to the criteria described in the text. The heavy line along the left edge of the figure indicates the range of boron concentrations observed in live-collected marine spicules. Temperate, high-productivity values are shown with a solid line; tropical, low-productivity values are shown with a dashed line. The core locations were V19-29-west coast of Ecuador; RC12-249-north central Atlantic; RC13-263-south Atlantic. (b) Plot of boron concentration versus age for spiculea from core V23-42, from the Atlantic Ocean between Iceland and Greenland. Each point represents a single spicule; all spicules were selected according to the criteria discussed in the text.
there is no significant change in boron contents of spicules, and there is no significant change in spicules from 10,000 to 100,000 yr old. V19-29 spicules exhibit no significant differences in boron concentration over the last 120,OOOyr,reflecting the equatorial location of this core. The fossil spicules tend to have less boron than
live-collected spicules with the exception of RC12-249; the fossil data are compared with livecollected data in Fig. 8. Since the salinity changes were probably negligible at the core locations during the last lo6 yr, the fossil spicules may have undergone rapid diagenetic loss of some of their B in a manner which has not produced other changes detectable by
11
Boron in siliceous materials as a paleosalinity indicator
900
t 800
_
i
RCl2-249
V23-42 * RC13-263 n
E =
0 SOCAL A v19-29 . MSN
700-
% .= 600‘E $ 6 ” m
soo400300zoo0 100 k
I 0
0
'
/
0
0
/
0
0
0
0 /-
0 L/
I
I
5
10
15
I
I
25
I
?BO
1
35
J
40
SoliZty, %O Fig 8. Plot of boron concentrations versus salinity for fossil spicules. The areas within the dashed lines represent the regions in which the live-collected data fall (see Figs 5 and 6). Each point indicates the average of all spicules analyzed for a single core position. Representative error bars show one sigma standard deviations of the values for individual spicules about the mean. Solid symbols represent cold, high-productivity surface water and open symbols represent warm, low-productivity surface water for the core locations at present. Data are shown from two locations which were not included in Fig. 8 because the sample ages (Holocene or Pleistocene) are not known precisely.
In three cases, BB3, BB4, and YHlO (Table 3), analysis of two types of spieules from distinctly separate parts of the sponges yielded different B concentrations (see discussion in Results Section B). In BB3 part of the boron is retained. It is also possible that the different boron concentrations do represent real and BB4 the spicules were differentiated into megaand microseleres with higher B concentrations in the environmental differences between sponges now and megascleres, while in YHlO the spicules from the armsponges in the past. like supports had greater boron concentrations than those from the main part of the sponge. Despite the relative independence of function in sponge cells, it DISCUSSION appears that boron metabolism in sponges can depend on the location within the sponge (e.g. ease of Relationships between environment and food supply for access to food and nutrients). Further, the boron consponges tent of the spicules correlates better with the environIf sponges obtain boron from their food, a more ment in which the food originates than with the endetailed consideration of their food supply is in order. vironment of the sponge itself, suggesting that Sponges are all filter-feeders, and all digestion is intrasponges obtain B from their food. cellular. They are indiscriminate, taking in any parIt would be of great interest to determine which ticles which are small enough. Thus their diet is part of the food supply provides the most boron to believed to consist of bits cf organic detritus, bacteria the sponges. It is known that diatoms and other algae and other cells, and dissolved organic compounds (DELAUBENFELS, 1955; REISWIG, 1971; BERGQUIST, require boron and cannot reproduce in its absence, 1978). The bulk of this food originates in the surface and some types of algae concentrate B in the organic parts of their cells (MCILRATH and SKOK, 1958; layers of the water, where the primary producers LEWRJ, 1965, 1966% b; YAMAMOTD,ef al., 1.971; reside, and rains down to the bottom, possibly being cycled through one or more additional organisms (e.g. LEWIN and CHEN, 1976). Diatoms also concentrate B zooplankton and bacteria). It is believed that the only in their frustrules. Pleistocene core samples (> 50% diatoms) from the Gulf of California, the Antarctic differences in food consumption between different Ocean, and the Bering Sea have 95-110ppm B and sponges are due to possible differences in maximum particle size tolerated (taxonomy-related) and differ- marine diatomites from Southern California contain 90-180 ppm B. ences in the nature of the food available (dependent Most other types of plankton for which analyses mainly on surface conditions. Thus, the food supply is have been found appear to be low in B content a link between the surface water and the sponge.
infrared spectroscopy or microprobe analysis. HOWever, the lack of systematic changes in B content in the deeper core samples (Fig. 7b) suggests that at least
12
MARIANJ. FURS
(NICHOLLS et al., 1959; BOWEN
and GALJCH, 1966; 1973). Boron analyses were conducted on several dried mixed zooplankton samples. Two samples which had been collected near the California coast between Los Angeles and San Diego (high productivity) contained 55 f 6 ppm B and two samples from the central Pacific, about 1600 km north of Hawaii (low productivity) contained 23 f 3 ppm. These boron concentrations are low relative to those found in sponge spicules or diatom frustrules, but the trend of increased boron with increased productivity is in agreement with the observations made with sponges. However, the high-productivity samples were collected near the shore and may contain more B-rich clay than the low-productivity samples. Nothing is known about the boron content of the very small particles which the sponges ingest-the bacteria and detritus. The differences between the food supplies for high- and low-temperature sponges could be due to (1) different kinds of planktonic populations which include different abundances of boronconcentrating organisms; or (2) organisms of the same type which accumulate boron differently in response to water temperature, nutrient supply, or some other variable. If primary producers gather B for sponges, they may exhibit a boron-salinity correlation with fewer complications than have been encountered with sponge spicules. YAMAMOTO et al.,
SUMMARY AND CONCLUSIONS A correlation between water salinity and boron concentration has been demonstrated for livecollected siliceous sponge spicules from temperature regions of high productivity. However, boron concentrations also reflect either water temperature or nutrient supply available to the sponge or to organisms at earlier stages in the food chain. At present there are insufficient data available to determine the detailed nature of the relationship between the boron content of spicules and environmental conditions other than salinity. Since B is a function of several variables, paleosalinity determination with sponge spicules is ditBeult. Fossil spicules tend to contain less boron than is found in live-collected material but no clear trend with age was found. More study is required to evaluate the possibility of subtle shortterm diagenetic effects. It is also possible that the lower boron concentrations mental conditions.
reflect different
environ-
WILLARDD. HARTMAN,Yale University; WILLIAMF. RUDDIMAN,PETERTHOMPSON, DAVID W. COOKEand FLOYD MCCOY,Lamont-Doherty Geological Observatory; GARY ZUMWALT,University of California, Davis; HEINZ LOWENSTAM.California Institute of Technology. JOHN PLATENAK assisted with the laboratory work. Neutron irradiations were conducted at the UCLA Nuclear Energy Laboratory with the assistance of TONY ZANE.This research was supported by NSF contract number EAR7684422AOl and represents part of the Ph.D. work of the author. The samples from Lamont-Doherty Geological Observatory were supplied under contracts-from the Office of Naval Research ONR-NOOO14-15-C-02 and NSF-0CE7618.
REFERENCES Am ‘III. and JONS~CJNE. (1972) Boron in Swedish and Norwegian fresh waters. Am& 1, 66-70. BERGQUIST P. R. (1978) Sponges, pp. 27-35, Univ. of California Press. BOWEX,J. E. and GAUCH H. B. (1966) Nonessentiality of boron in fungi and the nature of its toxicity. Planr. Phy siol. 41, 319-324. CALVERTS. E. (1974) Deposition and diagenesis of silica in marine sediments. In Pelagic Sediments: On Land and Under the Sea (eds K. J. Hsu and H. C. Jenkyns), Vol. 1. pp. 273-299. Spec. Publ. Int. Assoc. Sedimentol. DELAUBENFELS M. W. (1955) Part E: Porifera, in Treatise on Invertebrate Paleontology (ed. R. C. Moore). Geol. Sot. Am. and Univ. of Kansas Press. FURSTM., LOWEN~TAM H. and BURNETTD. (1976) Radiographic study of the distribution of boron in recent mollust shells. Geochim. Cosmochim. Acta 40, 1381-1386. GA~~AWAYJ. D. (1967) New method for boron determination in sea water and some preliminary results. Int. J. Oceonol. LimnoL 1, 85-90. GOLD~CHM~DT V. M. (1954) Geochemistry, 730 pp. Clarendon Press. GOLDSCHM~DT V. M. and PETERSC. (1932) Zur Geochemie des Bors. II. Nachr. Ges. Wiss. Goettinaen. Muth. Phvs. Kl. 402,528-545.
G~0s.s M. G. (1967) Concentrations of minor elements in diatomaceous sediments of a stagnant fjord. Am. Assoc. Adtt. Sci. Spec. Publ. 83, 273-282.
INGELSRUDI., THOMPSONT. G. and ZWICKER B. M. G. (1938) The boron content of sea water and of Marine organisms. Am. J. Sci. 5th Ser. XXXV, 47-63. LEWW J. (1965) The boron requirement of a marine diatom. Naturwissenschaften St, 70. LEW~NJ. (1%6a) Physiological studies of the boron requirement of the diatom, Cylindrotheca fusiformis Reimann and Lewin. J. Exp. Rot. 17,473-479. LEWINJ. (1966b) Boron as a growth requirement for diatoms. J. Phycol. 2, 160-163. LEWIN J. and CHEN. C.-H. (1976) Effects of boron deficiency on the chemical composition of a marine diatom. J. Exp. Rot. 27, 916-921.
LISITZINA. P. (1977) Biogenic sedimentation in the oceans and zonation. Lithol. Mineral. Rcs. 12, 1-17. MA~-~HEWS P. J. (1974) A survey of the boron content of certain waters of the greater London area using a novel analytical method. Water Res. 8, 1021-1028. MCILRATHW. J. and SKOKJ. (1958) Boron requirement of
Acknowledgements-The author wishes to acknowledge the assistance of DONALDS. BURNETTand HEINZ A. LOWENChlorella vulgaris. Bat. Car. 119, 231-233. MOBERGE. G. and HARDINGM. W. (1933) Boron content STAM. Samples were provided by the following people: KEN WEBB,Virginia Institute of Marine Science; Carolina Bioin sea water. Science 77, 510. NICHOLLSG. P., CURL H. JR and BOWENV. T. (1959) Speclogical Supply, Burlington, NC; GEORGECLEVELAND, Calitrographic analyses of marine plankton. Limnol. fornia Division of Mines and Geology; Caltech Mineral Oc&&gr. 4,472278. Collection; TOM WALSHand MIKE M:LLIN, Scripps InstiRAKE~TRAW N. W. and MAHNCKEH. E. (1935) Boron contution of Oceanography; SVEN ANKAR, University of tent of sea water of.the north Atlantic coast. Ind. Eng. Stockholm: JOHN VALOIS,Woods Hole Oceanographic Institute; GERALDBAKU~,University of Southern California; Chem. Anal. Ed. 7, 425.
Boron in siliceous materials as a paleosalinity indicator REISWIGH. M. (1971) Particle feeding in natural populations of three marine Demosponges. Biol. Bull. 141, 568-591. RUDDIMANW. F. (1978) Private communication. RYABININA. I. (1972) Boron in the tropical zone of the Atlantic Ocean. Geochem. lnt. 9, 597-603. SENN A. (1940) Paleogene of Barbados. Am. Assoc. Pet. Geol. Bull. 24, 1548-1610. United States Department of Commerce, National Ocean Survey (1970) Surface water temperature and density: Pacific coast of North and South America and Pacific Ocean islands. N.O.S. Publ., 31-3, 3rd edn.
13
United States Department of Commerce, National Ocean Survey (1973) Surface water temperature and density: Atlantic coast and South America. N.O.S. Publ. 31-3, 4th edn. UPPSTROML. F. (1974) The boron/chlorinity ratio of deepsea water from the Pacific Ocean. Deep-Sea Res. 21, 161-162. YAMAMOTO T.. YAMAOKA T.. FUJ~TAT. and IS~DA C. (1971) Chemical studies on the seaweeds (26): Boron content in seaweeds. Rec. Oceanogr. Works Jpn 11, 7-13. YAMAMOTO T., YAMAOKA T., FUJITAT. and ISODA C. (1973) Boron content in marine plankton. Rec. Oceanogr. Works Jpn 12, 13-21.
Boron in siliceous materials as a paleosalinity indicator” MARIAN
J.
FURST?
Division of Geological and Planetary Sciences, California Institute of Technology, Pasadena, CA 91125. U.S.A. (Received 1 June 1979; accepted in revisedform 20 August 1980)
Abstract-The “B(n, a)‘Li nuclear reaction has been used with alpha-sensitive plastic track detectors to determine boron concentrations in siliceous live-collected and fossil sponge spicules. This radiographic technique allows B determinations with 56% uncertainties on objects 2&25pm in diameter and for concentrations as low as 0.5 ppm. Boron concentrations in spicules from different specimens from the same location agreed to within 10% when the spicules were not: (1) smaller than 20 pm in diameter, (2) from dictyonine skeletons, (3) the extremely large root-like spicules found in some soft substrate hexactinellids, or (4) microscleres. These criteria also applied to spicules found in sediment samples. Spicules from live-collected sponges exhibited a taxonomy-independent correlation of B concentrations with water salinity for samples from regions of low water temperature and high productivity. Measured concentrations ranged from nearly 0 ppm B (freshwater sponges) to 500-700 ppm (marine sponges), with intermediate values for brackish-water specimens. However, spicules from tropical, low-productivity marine locations contained markedly less boron than spicules from temperate, high-productivity regions. Thus, water temperature and/or food supply also seem to influence B concentrations. Pleistocene spicules from deep-sea cores contained less boron than was expected in comparison with live-collected spicules based on present water temperatures and nutrient supplies, but B concentrations did not vary with depth in the cores. Infrared spectroscopy, electron microprobe analysis and visual inspection revealed no evidence for chemical or mineralogic alteration. It is not clear whether the lower B concentrations of the Pleistocene samples are the result of diagenetic processes or the as yet undefined effects of differences in food supply and/or environmental conditions.
sponges
INTRODUCFION IN 1932 GOLDSCHMIDT and F’ETERS published emission spectroscopic analyses of biogenic carbonates and opals showing that radiolarian oozes and marine
sponges tend to have significantly more boron than marine carbonate shells. They also reported low boron concentrations in freshwater diatomites (GOLDSCHMIDT, 1954). More recent work has confirmed the high boron contents of marine biogenic opal (GROSS, 1967). The relationship between boron concentrations in opal skeletal material and water salinity was investigated with the goal of finding a suitable type of organism for use in paleosalinity determinations. The greater difference between concentrations in marine and freshwater opal may allow better discrimination of salinities than is possible using carbonate shells (Fuasr et al., 1976). Skeletal remains of several extant groups of opalprecipitating organisms are abundant in the fossil record, including silicoflagellates, diatoms, radiolarians and sponges. The skeletons of the first three taxa tend to be very delicate, porous, and small, presenting difficulties for analysis, particularly the analysis of individual
specimens.
Both
diatoms
*Caltech contribution No. 3221. tPresent address: Schlumberger-Doll Box 307, Ridgefield, CT 06877, U.S.A. ci.c.*.45’ I --A
and
in freshwater
and
marine
environ-
EXPERIMENTAL The radiographic technique employed in this study is described in FURST et al. (1976). Boron determinations were made by placing a plastic detector which is sensitive to alpha particles over each sample. Neutron irradiation results in the “B(n, a)‘Li reaction. Alpha particles which are produced in the surface layers of the sample enter the plastic detector and cause radiation damage. After irradiation, the detector is chemically etched. The radiationdamaged areas etch faster than the bulk plastic, resulting in a series of conical holes or ‘tracks’ which can be counted with a scanning electron or optical microscope. The plastic employed was Kodak-Pathi cellulose nitrate CA%)-15. which is commercially produced specifically for alpha radiography. The background track density was equivalent to about 0.3 ppm B and provides a limit to the sensitivity of the technique. The sources of these tracks are the “G(n, a) reaction, fast neutron interactions, and possibly a small contribution from the 14N(n, n)W reaction. The only significant interference m B beierminations with this track technique is from the ‘Li(n, a)3H reaction. Spicules from several sponges were analyzed for lithium by atomic absorption spectroscopy and found to contain less than 10 ppm Li, too low a concentration for a significant contri-
siliceous
Research,
occur
ments. Because some representatives of each of these two groups can tolerate salinities other than normal marine, and because they occur abundantly in the fossil record, these two groups were chosen for investigation.
P.O. 1
2
MARIAN J. Funsr
bution to the observed spicule track densities; no correction was made for lithium in spicules. Silica glass and pIastic blanks were included in each irradiation. The B Standards used in this work were epoxy-mounted polished sections of National Bureau of Standards SRM 610 and 612. Due to uncertainties in B and Li contents of the two glasses, the accuracy of B determinations is also limited to about 67: (Fuasr et al., 1976). Two types of samples were analyzed (1) Fresh sponges were preserved in 70-7574 ethanol or dried. Small pieces were cut off with a clean scalpel and soaked in 6% NaClO solution (commercial Chlorox) in plastic beakers until the soft parts were completely oxidized and bubbling had ceased (2-4 days). The spicules were collected in a tine sieve (20 or 37pm mesh) and washed repeatedly with distilled water and then finally with spectroscopic quality methanol. They were then potted in E-7 epoxy and polished with aluminium oxide powder. Several specimens had only very small spicules and individual spicules could not be analyzed. Suspensions of these spicules were collected on filter paper and analyzed as bulk samples. Epoxy-mounted and filter-paper-collected spicules from the same sponge had identical B contents for two sponges with large enough spicules to allow the comparison. (2) Core samples were suspended in distilled water and seived. Large spicules were hand-picked from the > 70 pm portion and potted in epoxy. One core, V-19-29, from 3” 35’ S., 83” 56’ W., and 3157 m depth, was thoroughly cemented with carbonate and required dissolution in dilute HCI before sieving In order to verify that the Chlorox treatment did not change the boron content of the spicules due to contamination or leaching in the basic solution, two experiments were conducted. First, a sponge was divided into two parts and one part was treated with Chlorox. Spicules from the treated and the untreated parts were mounted and analyzed separately, yielding B contents which agreed within counting statsitics. Second, several samples were treated with hydrogen peroxide, which is an acidic solution. The boron concentrations also agreed well with those found in Chlorox-treated sponges of the same species from the same location. Subsequent analyses were carried out on Chloroxtreated spicules because of the shorter time required to oxidize the soft parts of the sponges (peroxide digestion required one week or more).
areas counted, they agree to well within counting statistics. 2. Two mounts of sponge spicules were reanalyzed; the second analysis of both samples agreed with the 6rst in both cases to well within the 6% uncertainty of the determinations. The reproducibility of thd optically counted track data was established by repolising and re-irradiating a number of samples on different occasions. Results showed consistency comparable to the SEM data. generally agreeing to better than the 6% standard deviation of each analysis. Duplicate analyses of the same detector also agreed well in most cases. Even when they did not agree to within counting statistics, all analyses of a given sample were averaged. Such discrepancies may have been due to sample inhomogeneities, but could also have represented track counting errors. B. Uniformity of boron in spicules A typical SEM traverse across a single large spicule is shown in Fig. 2. The spicule edge is very clearly
400
3/14/77
390
5000 x
380 370 E 500-g
210-T zoo-
"
RESULTS
Samples and standards usually had B concentrations within a factor of 2-3 of each other and should have similar reproducibilities. The reproducibility of track analyses, determined from SEM photographs at nominal magnifications of 5000 and 7000, was established in two ways. 1. Several SRM 610 (351 ppm) standards were included in each irradiation. Typical results for the standards from two irradiations are shown in Fig. 1. All of the standards from the 3/77 irradiation have track densities within one standard deviation (based on counting statistics) of the mean of all standards counted at the same magnification, and all of the standards from the 4/77 irradiation have track densities within two standard deviations from the mean for that irradiation. Further, when the average track densities of the 3/77 standards measured at different magnifications are normalized for the different
I 500-l
I
I
500-5
300-a
,
1
I
410
-
3/14/77
1
:
7000x
I
_
T
no-' 180 -
A. Reproducibility
m-7
I
500-10 500-12 1
1
420390-v
I
4125177
3801 ,,
7000 x 370 360:[,,,: 350. 500-9 500-10 5onoo-12 Fig 1. Reproducibility of SEM track data. Track densities are plotted for three groups of standards from two irradiations. Error bars represent one sigma standard deviations based on counting statistics. The average track density for each irradiation group is indicated by a horizontal line. When corrected for the different areas of the fields of view, the average track densities for the 3/14/77 data at 5000x and at 7000x agree to within I track per field of view. The neutron fluences are not controlled. so that track densities for different irradiations cannot be compared.
Boron in siliceous materials as a paleosalinity indicator
1
I I 1 lionverse Profile
I I Longhlinal
I
3
I 1 Proftie
3 410.0, ; 390i
-u 5 370x 2350c p 330-
W
_
__-
--_-
--
1
I
310- ?‘7’ 0
25
t 50
1 75
,
1:
I,,,tip
lo0
Distance from edge microns adjacent fields of view
middle
tip
Interval 75~
Fig. 2. A typical traverse across a single large spicule. The transverse profile begins near the boundary of the spicule and goes more than half-way across the spicule. The error bars represent one standard deviation based on counting statistics. In the longitudinal profile, the dashed line indicates the average of the four points plotted, and the four high-density points from the transverse profile have been averaged
and plotted as the third point on the longitudinal profile.
defined in the transverse profile and the boron concentration is uniform within counting statistics across the diameter of the spicule. The longitudinal profile is representative and is not the most uniform case observed. The slight decrease in boron concentration near one end of the spicule may be due to inadequate opal thickness at the spicule tip. The boron concentration is quite uniform (certainly better than 10%) within individual spicules on a scale of 5-10pm. Because the spicules are composed of many layers of opal and organic material, it is likely that the boron is inhomogeneously distributed in these layers on a finer scale than can be resolved using tracks. The agreement between B concentrations in untreated and Chloroxed spicules supports the assumption that the B is located in the inorganic layers. As shown in Fig. 3, the boron contents of different spicules from the same sponge are uniform within counting statistics in most live-collected specimens. However, several sponges were found which had different boron concentrations in spicules originating in different parts of the sponge or in spicules of different morphologies. Spicules from the finger-like projections which supported a specimen of Polymastia robusta (Bowerbank) (YH 10) above a soft substrate had 15% more boron than spicules from the main mass of the sponge. Two specimens of Geodia from Barbados (BB3 and BB4) each had concentrations which differed by a factor of three in different types of spicules. In both specimens, the spicules which had lower B contents were spherical or ellipsoidal in shape and had distinctive surface patterns (sterrasters); these spicules also were the only microscleres encountered which were large enough to analyze. It may be that there is a fundamental difference between boron concentrations in megascleres and microcleres. Because of their distinctive morphologies, most microscleres
are easily recognized and can be excluded from analyses of individual spicules. Subject to the criteria discussed in Section C, it can be concluded that there is a uniform boron concentration in spicules from a single sponge, and that analysis of only one or a few
25765-4s .z _ s r 21)0g e Q
ZZO-
Fig. 3. Plot of track densities for several spicules from each of two different sponges. Each point represents one spicule, and error bars represent standard deviations based on counting statistics. The differences in track densities for the two sponges are due to different irradiation conditions and different unit areas; PV-2-l was counted optically and 25765-4a was counted using the SEM. Average track densities for each group of spicules are indicated with dashed lines.
4
MARIAN J.
spicules will result in a representative boron concentration. C. Selection criteria for boron data from sponge spicules
The following criteria for spicule selection were necessary to obtain reproducible boron concentrations which agreed with other spicules from the same sponge and/or other sponges from the same location. 1. Spicules had to have diameters greater than 20pm. Narrower spicules gave low track densities, presumably because the portions remaining after polishing were thinner than the range of the alpha particles (6-7 pm). 2. Spicules from a dictyonine skeleton were excluded. 3. Root-like spicules, which anchor and support some soft-substrate sponges, were also excluded; early analyses of several of these spicules yielded low B concentrations relative to spicules from the main portions of sponges. However, the low B contents may actually have been related to other factors [see discussion of criterion (5) below] and it is possible that this criterion is not really needed. Also, these very large (around 1 mm dia) spicules did not polish well and the analyses could have been excluded on that basis. 4. Sterrasters and other morphologically distinctive microscleres were excluded. 5. The spicule surface must be well-polished and free of pits and cracks because extraneous material collects in holes and extensive pitting is evidence for dissolution of the spicule. The area analyzed must be from the interior of the spicule; material adsorbed or cemented on the exteriors of spicules is thereby excluded. 6. The central canal of the spicule must be present but not significantly enlarged. If the canal is greatly enlarged due to solution of opal after the sponge died, the walls of the spicule may be too thin and the same problem with low B analyses occurs as in criterion (1). All of these criteria can readily be applied to fossil spicules. It should be emphasized that the first four criteria deal with readily observable morphological features of spicules and present no problem for selection of spicules from sediment samples. Indeed, objects smaller than 20pm in diameter are difficult to handle individually. The spicules selected for boron analyses from core samples were subjected to more stringent selection criteria than .the live-collected specimens. The track densities found in a number of spicules from a single core and irradiated at the same time are plotted against spicule diameter in Fig. 4. For spicules wider than 25 pm, the B concentration is independent of spicule diameter, but several spicules in the 20-25 pm range gave low B results (less than 1600 tracks/field of view). It was concluded that slightly wider spicules
FURST
were required for reliable B determinations. The greater spicule diameter required for fossil samples is probably due to the partial dissolution of the spicules along the central canals. After polishing which removes material from the exterior of the spicule, the remaining wall may be thin relative to the alpha range. In order to identify spicules which had undergone chemical alteration, electron microprobe analyses were obtained for all fossil spicules whose B analyses are included in this work and for spicules of several live-collected specimens also. Spicules from livecollected sponges are known to contain 5-20% water and are composed of otherwise very pure SiO*; no other cations or anions were detected in significant quantities (>O.l%) by the microprobe. Spicules with non-hydrous portions greater than 80% of the total oxide sum, with SiOZ comprising 99.6 + 0.4% of the non-hydrous fraction, and which met the optical criteria discussed above were selected for B analysis, No correlation was found between boron concentration and oxide sum. Only about 5% of the fossil spicules analyzed had to be rejected because of low SiOZ content. The oxides present in addition to Si02 in measurable quantities were A120s, MgO, TiOz, Fe0 and BaO, never in excess of lo/;, of the total. High A1203 concentrations may be due to the presence of grains of A1203 on the surface from polishing. When the first probe analysis of a spicule had an oxide total of less than 80514,the analysis was repeated on a different spot. Usually the second analysis had a total
‘5 .p EOOO-
z
X .!i
2
1600-
“0 c ,.
1200-
.Z s
0" _=
800-
e k 400 t
P
0 70 cm 0 90cm 0 470cm 1
Fig. 4. Plot of track density vs spicule diameter for spicules from the 76, 90-, and 470-cm. depths of core V23-42. All data were obtained from samples irradiated together and track densities are directly comparable to each other. Rep resentative one sigma error bars are based on counting statistics. A track density of 2OOO/field of view is equivalent to about 450 ppm B.
Boron in siliceous materials as a paleosalinity indicator Table 1. Comparison of boron concentrations in ditlerent sponges from the same marine locations Genus and species Haliclona Polymastia
oculata rohusra
XrstosporIgia Terhya crqra
muta
Geodia sp. Geodia sp.
Syringella sp.
Myxilla sp. lophon pattersoni Syringella sp. Thalysias juniperina Iotrochata hirorulata Age/as clathrodes Neotihularia nolitungere Slaurocalyptus Sp. (Hexactinellida) Lissodendoryx sp.
(Hexactinellida) (Hexactinellida) (Demospongia)
At least 4 genera. not identified
WI
Sample YH 7 YH 10
66& 32 604 2 30 > 236 +_ 13 402 i 22
YH 1 YH 3
308 + 30 505 f 19 141 f 1 18 k I1 103 f 25
BB 1 BB 2 BB 3 BB 4 BB 8 FP I FP 2 PV 2 PV 3 ws 1 MV 1
506 f 30 521 k+ 31 430 26
lOOk8 1 353 f 19 125 +9 282 k 17
25765- 1 25765-3
596 599 514 569
25165-4 25766-2 25767-2 26601
Sponges of different species from the same marine location tend to have the same boron concentrations, but there are some notable exceptions. Brackish-water sponges and sponges from tropical locations have variable boron contents. Marine sponges from higher latitudes usually have boron concentrations ranging about ten percent from the mean for each location (6% standard deviations for each determination). When the data for a sample did not agree with other samples from the same location, the sample was rejected only if reexamination demonstrated that there was good &use. Several samples were re-irradiated and gave results in agreement with the values obtained the first time. Table 1 compares boron concentrations from sponges from each of several locations ‘from which two or more species met the selection criteria. Representatives from temperate
+ k f+
34 14 25 30
693 k 53 523 f 19 481 + 31 476 _+ 30 435 + 28 468 + 30 538 + 36 590 + 36
AA 1 AA 2 AA 3 AA 4 AA 5 AA6
D. Intersponge variation in boron concentration
Massachusetts Jamaica
Barbados
Washington
499 * 30 514 +* 35 34 1 581
Cl c2 c3 c4
greater than 800/, and it was assumed that the first analysis included some epoxy due to the difficulty of locating the beam on the narrow spicules. About 2% of the spicules had low second analyses and were rejected. These criteria were applied to all samples discussed below.
Location
Honduras
California
Antarctica
locations show good agreement whereas those from tropical sites do not. Thus, it can be concluded that the observed boron concentrations are independent of species, but appear to be a function of some environmental factor. The groups of sponge samples from Barbados and Southern California included both Demospongia and Hexactinellida, so it can be concluded that the boron concentration is independent of genetic controls at the class level. *Table 2 lists boron concentrations in spicules from representatives of the same genera from different locations. When locations which are both tropical or both temperature are compared, the concentrations are in good agreement; poor agreement is found between tropical and temperate specimens, as in the case of Haliclona. The mediterranean location is slightly hypersaline and intermediate between ‘temperate and ‘tropical’ in climate. Correlarion of the boron content of he-collected spicules with saliniry and orher environmental factors
E.
The results of all boron analyses of siliceous materials are listed in Table 3. Samples of the same species which were collected from one location (Askii Island, Sweden; Woods Hole, Massachusetts; and
MARIANJ. FUR~T Table 2. Comparison of B concentrations
in sponges from the same genera but different locations
PI Location
Sample
Genus and species
ppm
FP 1 ws 1
San Juan Islands, Washington* San Juan Islands, Washington*
587 + 35 506 * 30 >
Syringelia
c2 YH 9
Gulf of Mexico-Honduras Bahamas
353 + 19 345 + 20 >
Iotrochata birotulata
Massachusetts Bay Puerto Scotia Rico Nova
660 f 32 334 + _+33 191 670
Haliclona
Spain-Mediterranean (hypersaline) California (small spicules)
484 k 25 485 * 55 >
Axinella
YH4 YH 7 YH 11 25766-l YH 5
* The two sites in the San Juan Islands are separated by about 10 km.
Table 3. B concentrations
Salinity x
Sample and location LIVE-COLLECTED
in sponge spicules CBI’ ppm
Classification
SPONGES 28-32
San Juan Islands, Washington FPl FP2 PV2 PV3 PV4 WSI MVI Egg1 San Nicholas Island, California
0
Catalina Island, California
596 690 699 599 574 143 293 540 429 569 693 572 177
Staurocalyptus sp. Aphrocallistes vastus
Iophon pattersoni Lyssodendoryx firma
Syringella sp. Spongilla
Fam. Craniellidae Lissodendoryx sp. Axinella sp. Axinella sp. AxineUa sp.
33.5 267 523
26601 26601 -36
Magdalena Bay, Mexico
399 367 621 648 398
4545-l 4545-2 4081-i 4081-2 4544-la 4544-2
Massachusetts Bay, Massachusetts
Syringella sp. Myxilla sp.
33.5 25765-l 25765-2 25765-2 25765-3 25765-4 25765-5 25766-l 25766-l 25766-l 25766-2 25767-2 25767-3 25767-3
Cedros Island, Mexico Northumberland Str., Nova Scotia Eastport, Maine
587 506 521 430 334 499 574 c 1.5
332 -35
YH4 YHl2 YH7 YHlO YHlOa
406
Single Single Single Single Single Single Single
spicules spicules spicules spicules spicules spicules spicules
32 k 1
670
Haliclona oculata
32 k 1 31.6 k 0.2
505
Isodictya palmata
660 604
Haliclona oculata Polymastia robusta Polymastia robusta
698
Boron in siliceous materials as a paleosalinity indicator Table 3 (cont.) Salinity 0, ‘00
Sample and location Woods Hole, Massachusettst WHl (Hz(&)
PI* ppm
Classification
550 573 550 236 213 253 217
Microciona sp. Microciona sp. Microciona sp. Halichondria sp. Halichondria sp. Halichondria sp. Hafichondria sp.
20-30
~712 WzW WH3 (H,O,) WHl3 (HsO,) WH14 (H202) WH15 (HsOz) W H 13~ (Chlorox) Block Island, Rhode Island New Haven. Connecticut
YHS
841
Suberities @us
YH2 YH6
15-32
402 258
Cliona celata Microciona prolifera
Gloucester Point, Virginia
GPl GP2
lC-25
287 428
Halichondria bowerbanki Lissodendoryx carolinensis
North Carolina
CBS5
0
Nassau, Bahamas
YH9
31 + 1
Cl c2 c3 c4
36 + 1
YHl YH3
Honduras
Jamaica Puerto Rico
YHll
Barbados
BB-l-1 BB-1-2 BB-l-4 BB-2-2 BB2 BB-3-4 BB-3-4 BB3 BB4 BB8
Bermuda
3.8 Spongilla sp, 345
lotrochata birotulata
:: 125 282
Thalysias juniperina lotrochata birotulata Age/as clarhrodes Neotibularia nolitangere
36 f 1
236 402
Xestospongia muta Tethya crypta
34+1 36.s36.75
334
Haliclona rubens
round
315 227 331 288 505 288 96 147 78 103
Geodia sp.
Geodio sp. Geodia sp. Geodia sp. Calliospongia uqinalis
Ber-1
37 f 1
61
Ask6 Island. Sweden
AI1 AI2 AI3
5.67.1
29.1 32.3 30.0
Ephydatia Juoiatilis
Cataluna, Spain
YHS
?( > 35)
484
Axinella pollapoides
Arthur Harbor. Antarctica
AA1 AA2 AA3 AA4 AAS AA6
32.5 + 0.5
487 476 435 468 538 590
450 cm
marine
361
CORE SAMPLES RCl3-263 South Atlantic V19-20 E. Equatorial Pacific
V23-42 North Atlantic
0 cm-3 -4 120 cm-3 -4 850 cm-3 -4 OCm 10 20 30 30 40 50 50 60 10
117 241 172 252 217 243 marine
340 344 273 262 546 434 608 524 470 559
8
MARIANJ. FURST Table 3 (conr.)
CBI’
Salinity 0
Sample and location 90 430 450 470 160cm
ppm
Classification
marine
437 477 408 425 649
SOCAL (Gulf of California)
marine
342
MSN Antarctic
marine
384
RC12-249 North Central Atlantic
* Standard deviations of boron concentrations are usually 6% or less. t The Woods Hole, Massachusetts samples were treated with hydrogen peroxide or Chlorox as indicated. All other sponges were treated with Chlorox.
Islands, Washington) have boron concentrations which agree within counting statistics. The average boron concentrations for each location are plotted in Fig. 5. Fresh-water spicules contain 5 ppm B or less, and brackish-water spicules have boron contents between the fresh water and the group of temperate marine values. The obvious environmental difference between tropical and temperate locations is surface water temperature. Boron concentrations have been plotted relative to published surface water temperatures for the marine specimens in Fig. 6. The grouping of the warm- and cold-water points is striking. (Most of the sponges were collected in water San Juan
I
d
,m 5
,
L
10
15
shallower than 100 m). There does not appear to be a simple correlation between temperature and boron concentration; for temperatures less than 15°C the boron concentrations are uniformly high, and for temperatures greater than 20” the boron concentrations are lower and more variable. The high concentration of 84Oppm in the Block Island, Rhode Island sample may reflect polluted water. Water samples collected near population centers may have markedly increased boron concentrations due to the presence of industrial wastes and household detergents (AHL and JONSSON,1972; MATTHEW$ 1974). Long Island Sound is surrounded by a region of high
I
I
20 Salinity,
25
I
30
8
35
I 40
?L
Fig. 5. Plot of B concentration versus salinity for sponge sample locations. Each point represents the average of the boron concentrations of all samples analyzed from each location. The error bars represent the range in boron concentrations for all samples analyzed and the ranges in water salinity. When error bars for salinities are absent, the variations are small (k I”&,or less). The closed circles represent areas of low water temperature (less than 15’C) and high nutrient supply; open circles represent tropical regions (greater than 15°C) with low nutrient supply. The dashed lines indicate a range for the trend of boron concentrations with salinity for cold water sponges.
Boron in siliceous materials as a paleosalinity indicator
It'"".
0'
I
I
I
,
I
I
1
I
-5
0
5
10
15
20
25
30
35
40
Temperature(“C) Fig. 6. Plot of boron concentrations versus surface water temperature for live-collected sponge spicules from normal marine localities. Each point represents analyses of spicules from a single sponge. Representative error bars indicate standard deviations in B conozntrations based on counting statistics and the ranges in surface water temperatures, except the San Nicholas Island specimens. for which the actual temperature at the collecting site was measured. The temperature data for all locations except San Nicholas Island are from U.S. Department of Commerce (1970, 1973). The symbols represent the following locations: O--Arthur Harbor, Antarctica: O-Block Island, Rhode Island; ,&--Massachusetts Bay, Massachussetts; *San Nicholas Island, California: l -Eastport, Maine; m---San Juan Islands Washington; 0-Jamaica; V-British Honduras: M-Barbados; A-Bermuda ; O-Bahamas; O-Puerto Rico.
9
come directly from seawater or indirectly through particulate food. The possibility that the boron concentration in the spicules is a reflection of the boron/silicon ratio in seawater where the sponge grew, was also considered. The sponges from the areas with the largest silicon concentrations (Antarctica, Southern California and Washington) have high boron concentrations in their spicules. The boron concentration of the open sea varies by only lo’%, independent of ocean, latitude, or water depth (MOBERGand HARDING, 1933; RAKESTRAW and MANHCKE,1935; IGELSRUD,et al., 1938; GASSAWAY,1967; RYABININ,1972; UPPSTROM,1974). If the sponges were unable to discriminate between the two elements, one would expect that specimens from high-Si water would have lower spicule B contents than those from low-Si water, contrary to the data presented here. Moreover, since the silicon concentration in the water column increases with depth for the first several hundred meters (CALVERT,1974), one would expect that sponges from deeper water (California) might have a lower B/Si ratio in their spicules, also contrary to observation. It can be concluded that boron and silicon can be discriminated by sponges. F. Boron analyses offossil spicules
The primary purpose in studying the core samples, which are all from normal marine environments, was to test for diagenetic effects on spicules in sediment samples. Ages were determined on the basis of uniform sedimentation rates inferred from radiometric ages of several tug layers in nearby cores (V23-42) (RUDDIMAN,1978), or by oxygen isotopic composition in relationship to microfauna (COOKE, 1978). Visual examination with a microscope revealed a trend toward increased solution pitting on exterior surfaces and enlargement of the central canal with depth in cores (age). Infrared spectra were obtained for spicules from core V23-42 positions corresponding to ages of approximately 0, 10,000 and about 100,000 yr. There was no evidence for any alteration of opal to cristobapopulation density. The wide variation in B concentrations found in the Barbados sponges may be due to lite or quartz. Similarly, electron microprobe data did localized mixing of seawater and freshwater which not indicate any significant alteration or dehydration travelled downward from the surface of the island except in a few spicules; there were no core positions through the porous reef carbonate (SENN, 1940; with abundant large spicules (greater than 25 m dia) from which a significant proportion of the spicules FISCHER,private communication, 1979). However, water temperature is not the only en- had to be excluded on the basis of chemical altervironmental factor which differs for the temperate and ation. Boron contents of individual fossil spicules are tropical locations from which these specimens were plotted against age in Fig. 7. For two cores, V23-42 obtained. All of the tropical specimens are from water and V19-29, data were obtained from several posof low productivity and all of the temperate speci- itions in each core. V23-42 samples were selected to mens are from areas of high productivity (LISITZIN, cover a time interval extending backward through the 1977). By productivity is meant the number or mass last major glaciation, investigating the changes in of planktonic organisms found per unit volume of boron concentrations in spicules with changes in clinear-surface water; the number of organisms is mate (Fig. 7b). There is significantly more boron in dependent on the available supply of nutrients, par- the spicules from the glacial interval (9,000-l 3.000 yr) ticularly nitrogen, phosphorus, and silicon. Further, it and in the older spicules than in the more recently is not known how the sponges obtain the boron deposited material. The peak of the Wisconsin glacial which is incorporated into their spicules. It could period was about 11,500 yr ago. For the first 7500 yr
MARIAN J. Furts’r
10
700-
: 600-
E 2
c s 5 "
.
500-
i
z
g
.
.
I 400-I
al
.
.
I I 300-,
. .
.
I8 '. ZOO-1
I
9
.
. . RCQ-249 4 AC13-263 . v19-29
.
I9 loo-;,
I
I 100,000
1
olooo
w@J
0, l,oOo.ooo
Agelyeors)
700-
600-
I
0
1
O1000
10,coo
100.000
Age(years)
Fig. 7. (a) Plot of boron concentration versus age for spicules from cores V19-29, RC12-249, and RCl3-263. Each point represents one spicule, selected according to the criteria described in the text. The heavy line along the left edge of the figure indicates the range of boron concentrations observed in live-collected marine spicules. Temperate, high-productivity values are shown with a solid line; tropical, low-productivity values are shown with a dashed line. The core locations were V19-29-west coast of Ecuador; RC12-249-north central Atlantic; RC13-263-south Atlantic. (b) Plot of boron concentration versus age for spiculea from core V23-42, from the Atlantic Ocean between Iceland and Greenland. Each point represents a single spicule; all spicules were selected according to the criteria discussed in the text.
there is no significant change in boron contents of spicules, and there is no significant change in spicules from 10,000 to 100,000 yr old. V19-29 spicules exhibit no significant differences in boron concentration over the last 120,OOOyr,reflecting the equatorial location of this core. The fossil spicules tend to have less boron than
live-collected spicules with the exception of RC12-249; the fossil data are compared with livecollected data in Fig. 8. Since the salinity changes were probably negligible at the core locations during the last lo6 yr, the fossil spicules may have undergone rapid diagenetic loss of some of their B in a manner which has not produced other changes detectable by
11
Boron in siliceous materials as a paleosalinity indicator
900
t 800
_
i
RCl2-249
V23-42 * RC13-263 n
E =
0 SOCAL A v19-29 . MSN
700-
% .= 600‘E $ 6 ” m
soo400300zoo0 100 k
I 0
0
'
/
0
0
/
0
0
0
0 /-
0 L/
I
I
5
10
15
I
I
25
I
?BO
1
35
J
40
SoliZty, %O Fig 8. Plot of boron concentrations versus salinity for fossil spicules. The areas within the dashed lines represent the regions in which the live-collected data fall (see Figs 5 and 6). Each point indicates the average of all spicules analyzed for a single core position. Representative error bars show one sigma standard deviations of the values for individual spicules about the mean. Solid symbols represent cold, high-productivity surface water and open symbols represent warm, low-productivity surface water for the core locations at present. Data are shown from two locations which were not included in Fig. 8 because the sample ages (Holocene or Pleistocene) are not known precisely.
In three cases, BB3, BB4, and YHlO (Table 3), analysis of two types of spieules from distinctly separate parts of the sponges yielded different B concentrations (see discussion in Results Section B). In BB3 part of the boron is retained. It is also possible that the different boron concentrations do represent real and BB4 the spicules were differentiated into megaand microseleres with higher B concentrations in the environmental differences between sponges now and megascleres, while in YHlO the spicules from the armsponges in the past. like supports had greater boron concentrations than those from the main part of the sponge. Despite the relative independence of function in sponge cells, it DISCUSSION appears that boron metabolism in sponges can depend on the location within the sponge (e.g. ease of Relationships between environment and food supply for access to food and nutrients). Further, the boron consponges tent of the spicules correlates better with the environIf sponges obtain boron from their food, a more ment in which the food originates than with the endetailed consideration of their food supply is in order. vironment of the sponge itself, suggesting that Sponges are all filter-feeders, and all digestion is intrasponges obtain B from their food. cellular. They are indiscriminate, taking in any parIt would be of great interest to determine which ticles which are small enough. Thus their diet is part of the food supply provides the most boron to believed to consist of bits cf organic detritus, bacteria the sponges. It is known that diatoms and other algae and other cells, and dissolved organic compounds (DELAUBENFELS, 1955; REISWIG, 1971; BERGQUIST, require boron and cannot reproduce in its absence, 1978). The bulk of this food originates in the surface and some types of algae concentrate B in the organic parts of their cells (MCILRATH and SKOK, 1958; layers of the water, where the primary producers LEWRJ, 1965, 1966% b; YAMAMOTD,ef al., 1.971; reside, and rains down to the bottom, possibly being cycled through one or more additional organisms (e.g. LEWIN and CHEN, 1976). Diatoms also concentrate B zooplankton and bacteria). It is believed that the only in their frustrules. Pleistocene core samples (> 50% diatoms) from the Gulf of California, the Antarctic differences in food consumption between different Ocean, and the Bering Sea have 95-110ppm B and sponges are due to possible differences in maximum particle size tolerated (taxonomy-related) and differ- marine diatomites from Southern California contain 90-180 ppm B. ences in the nature of the food available (dependent Most other types of plankton for which analyses mainly on surface conditions. Thus, the food supply is have been found appear to be low in B content a link between the surface water and the sponge.
infrared spectroscopy or microprobe analysis. HOWever, the lack of systematic changes in B content in the deeper core samples (Fig. 7b) suggests that at least
12
MARIANJ. FURS
(NICHOLLS et al., 1959; BOWEN
and GALJCH, 1966; 1973). Boron analyses were conducted on several dried mixed zooplankton samples. Two samples which had been collected near the California coast between Los Angeles and San Diego (high productivity) contained 55 f 6 ppm B and two samples from the central Pacific, about 1600 km north of Hawaii (low productivity) contained 23 f 3 ppm. These boron concentrations are low relative to those found in sponge spicules or diatom frustrules, but the trend of increased boron with increased productivity is in agreement with the observations made with sponges. However, the high-productivity samples were collected near the shore and may contain more B-rich clay than the low-productivity samples. Nothing is known about the boron content of the very small particles which the sponges ingest-the bacteria and detritus. The differences between the food supplies for high- and low-temperature sponges could be due to (1) different kinds of planktonic populations which include different abundances of boronconcentrating organisms; or (2) organisms of the same type which accumulate boron differently in response to water temperature, nutrient supply, or some other variable. If primary producers gather B for sponges, they may exhibit a boron-salinity correlation with fewer complications than have been encountered with sponge spicules. YAMAMOTO et al.,
SUMMARY AND CONCLUSIONS A correlation between water salinity and boron concentration has been demonstrated for livecollected siliceous sponge spicules from temperature regions of high productivity. However, boron concentrations also reflect either water temperature or nutrient supply available to the sponge or to organisms at earlier stages in the food chain. At present there are insufficient data available to determine the detailed nature of the relationship between the boron content of spicules and environmental conditions other than salinity. Since B is a function of several variables, paleosalinity determination with sponge spicules is ditBeult. Fossil spicules tend to contain less boron than is found in live-collected material but no clear trend with age was found. More study is required to evaluate the possibility of subtle shortterm diagenetic effects. It is also possible that the lower boron concentrations mental conditions.
reflect different
environ-
WILLARDD. HARTMAN,Yale University; WILLIAMF. RUDDIMAN,PETERTHOMPSON, DAVID W. COOKEand FLOYD MCCOY,Lamont-Doherty Geological Observatory; GARY ZUMWALT,University of California, Davis; HEINZ LOWENSTAM.California Institute of Technology. JOHN PLATENAK assisted with the laboratory work. Neutron irradiations were conducted at the UCLA Nuclear Energy Laboratory with the assistance of TONY ZANE.This research was supported by NSF contract number EAR7684422AOl and represents part of the Ph.D. work of the author. The samples from Lamont-Doherty Geological Observatory were supplied under contracts-from the Office of Naval Research ONR-NOOO14-15-C-02 and NSF-0CE7618.
REFERENCES Am ‘III. and JONS~CJNE. (1972) Boron in Swedish and Norwegian fresh waters. Am& 1, 66-70. BERGQUIST P. R. (1978) Sponges, pp. 27-35, Univ. of California Press. BOWEX,J. E. and GAUCH H. B. (1966) Nonessentiality of boron in fungi and the nature of its toxicity. Planr. Phy siol. 41, 319-324. CALVERTS. E. (1974) Deposition and diagenesis of silica in marine sediments. In Pelagic Sediments: On Land and Under the Sea (eds K. J. Hsu and H. C. Jenkyns), Vol. 1. pp. 273-299. Spec. Publ. Int. Assoc. Sedimentol. DELAUBENFELS M. W. (1955) Part E: Porifera, in Treatise on Invertebrate Paleontology (ed. R. C. Moore). Geol. Sot. Am. and Univ. of Kansas Press. FURSTM., LOWEN~TAM H. and BURNETTD. (1976) Radiographic study of the distribution of boron in recent mollust shells. Geochim. Cosmochim. Acta 40, 1381-1386. GA~~AWAYJ. D. (1967) New method for boron determination in sea water and some preliminary results. Int. J. Oceonol. LimnoL 1, 85-90. GOLD~CHM~DT V. M. (1954) Geochemistry, 730 pp. Clarendon Press. GOLDSCHM~DT V. M. and PETERSC. (1932) Zur Geochemie des Bors. II. Nachr. Ges. Wiss. Goettinaen. Muth. Phvs. Kl. 402,528-545.
G~0s.s M. G. (1967) Concentrations of minor elements in diatomaceous sediments of a stagnant fjord. Am. Assoc. Adtt. Sci. Spec. Publ. 83, 273-282.
INGELSRUDI., THOMPSONT. G. and ZWICKER B. M. G. (1938) The boron content of sea water and of Marine organisms. Am. J. Sci. 5th Ser. XXXV, 47-63. LEWW J. (1965) The boron requirement of a marine diatom. Naturwissenschaften St, 70. LEW~NJ. (1%6a) Physiological studies of the boron requirement of the diatom, Cylindrotheca fusiformis Reimann and Lewin. J. Exp. Rot. 17,473-479. LEWINJ. (1966b) Boron as a growth requirement for diatoms. J. Phycol. 2, 160-163. LEWIN J. and CHEN. C.-H. (1976) Effects of boron deficiency on the chemical composition of a marine diatom. J. Exp. Rot. 27, 916-921.
LISITZINA. P. (1977) Biogenic sedimentation in the oceans and zonation. Lithol. Mineral. Rcs. 12, 1-17. MA~-~HEWS P. J. (1974) A survey of the boron content of certain waters of the greater London area using a novel analytical method. Water Res. 8, 1021-1028. MCILRATHW. J. and SKOKJ. (1958) Boron requirement of
Acknowledgements-The author wishes to acknowledge the assistance of DONALDS. BURNETTand HEINZ A. LOWENChlorella vulgaris. Bat. Car. 119, 231-233. MOBERGE. G. and HARDINGM. W. (1933) Boron content STAM. Samples were provided by the following people: KEN WEBB,Virginia Institute of Marine Science; Carolina Bioin sea water. Science 77, 510. NICHOLLSG. P., CURL H. JR and BOWENV. T. (1959) Speclogical Supply, Burlington, NC; GEORGECLEVELAND, Calitrographic analyses of marine plankton. Limnol. fornia Division of Mines and Geology; Caltech Mineral Oc&&gr. 4,472278. Collection; TOM WALSHand MIKE M:LLIN, Scripps InstiRAKE~TRAW N. W. and MAHNCKEH. E. (1935) Boron contution of Oceanography; SVEN ANKAR, University of tent of sea water of.the north Atlantic coast. Ind. Eng. Stockholm: JOHN VALOIS,Woods Hole Oceanographic Institute; GERALDBAKU~,University of Southern California; Chem. Anal. Ed. 7, 425.
Boron in siliceous materials as a paleosalinity indicator REISWIGH. M. (1971) Particle feeding in natural populations of three marine Demosponges. Biol. Bull. 141, 568-591. RUDDIMANW. F. (1978) Private communication. RYABININA. I. (1972) Boron in the tropical zone of the Atlantic Ocean. Geochem. lnt. 9, 597-603. SENN A. (1940) Paleogene of Barbados. Am. Assoc. Pet. Geol. Bull. 24, 1548-1610. United States Department of Commerce, National Ocean Survey (1970) Surface water temperature and density: Pacific coast of North and South America and Pacific Ocean islands. N.O.S. Publ., 31-3, 3rd edn.
13
United States Department of Commerce, National Ocean Survey (1973) Surface water temperature and density: Atlantic coast and South America. N.O.S. Publ. 31-3, 4th edn. UPPSTROML. F. (1974) The boron/chlorinity ratio of deepsea water from the Pacific Ocean. Deep-Sea Res. 21, 161-162. YAMAMOTO T.. YAMAOKA T.. FUJ~TAT. and IS~DA C. (1971) Chemical studies on the seaweeds (26): Boron content in seaweeds. Rec. Oceanogr. Works Jpn 11, 7-13. YAMAMOTO T., YAMAOKA T., FUJITAT. and ISODA C. (1973) Boron content in marine plankton. Rec. Oceanogr. Works Jpn 12, 13-21.