Chlorine Compounds and Stratospheric Ozone

Table 1. Summary of noradrenaline concentration and T-maze performance data; S.E.M., standard error of mean. Group Control One-stage Two-stage One-stage total lesion T wo-stage lesion Partial Total N Noradrenaline, cortex and hippocampus (ng/g) (mean± S.E.M.) with food reinforcement to habituate them to the general experimental situ- ation. Tht: animals were then given 12 trials eacu day in a T-maze in which entrance mto an arm with an olfactory stimulus ( 1 percent amyl acetate solu- tion) led to food reward ( 4). The criterion was 22 correct responses in 24 trials. Animals were again given free access to food upon reaching cri- terion, and behavioral testing ended with measurement of activity by means of an Animex meter (5) over a 90- minute period. Throughout the study each lesion group was tested concur- rently with a nonlesion control group (3). The animals were killed by spinal stem were isolated for subsequent paraf- concussion; the cerebellum and brain- fin embedding and staining with thio- nin, and the cortex and hippocampus were removed for noradrenaline assay The histological and chemical anal- yses (Table 1) demonstrated extensive destruction of the locus coeruleus in most animals. In the groups classified as having total lesions, only two ani- mals had any identifiable locus coerul- eus cells remaining. The reduction in cortical noradrenaline ranged from 67 to 90 percent in these animals. T he one-stage bilateral lesions markedly debilitated the animals; ataxia, tempo- rary aphagia, and urogenital disorders were observed (7) . Postoperative re- covery in the animals with bilateral two-stage lesions was both more rapid and more complete. During the habituation trials, all ani- mals increased their running speeds so as to complete the 120-cm runway within 5 seconds. The running speeds for the animals with lesions were con- sistently lower than those for the con- trol animals; however, we attribute this to motor variables. Motor difficulties observed (7) in some animals in all the lesion groups may have restricted ·t heir maximum running speeds. Activity, as measured during the 90-minute test period, was positively correlated with T-maze, days to criterion Median Range noradrenaline content (group with single-stage lesions and corresponding controls, r = .82; group with two-stage lesions and controls, r = .67) . T he motor disturbances and decreased ac- tivity in the animals with lesions may account for their slower speeds in the runway. Running speed in the T-maze was not a limiting variable because the criterion was based on entrance into the goal arm rather than speed of traversal. There was no initial preference for the olfactory cue. The animals with bi- lateral single-stage lesions did not differ from their controls with respect to days to criterion (Table 1) [Mann-Whitney: U = 11, P = .92 (8)]. There were no significant differences for this measure for the group with total two-stage lesions, the group with partial two- stage lesions, and their controls [Krus- kal-Wallis: H e= 2.13, P > .1 (8)]. These results fail to demonstrate a learning deficit in a T -maze discrimina- tion despite a substantial decrease in cortical noradrenaline caused by bi- lateral locus coeruleus lesions. If the nucleus locus coeruleus is the substrate for a reinforcement system operative in appetitive learning, de- struction of the nucleus should lead to deficits in any learning task with food reward. Since animals with complete bilateral lesions of the locus coeruleus can learn a T-maze discrimination, we conclude that the structure is not an essential component of a common rein- forcement system. D AVID G. AMARAL Department of Psychology and Center for Brain R esearch, University of Rochester, Rochester, New York 14627 JoHN A. Foss Department of Psychology, University of Rochester References and Notes I. T. J. Crow, Na111re (Lond.) 219, 736 (1968); Psycho/. Med. 3, 66 (1973). 2. G .. M. Anlezark, T . J. Crow, A. P. Greenway, Sc1ence 181, 682 (1973). 3. Coordinates were P 1.8, L 1.4, and V 6. 7; the interparietal bone was made horizontal, and (0, 0, 0) was the intersection of the midline with a line extrapolated through the lateral lambdoid suture at the surfa:e or the skull. Lesions were made by passing I ma of current for 10, 15, or 20 seconds between the tip of the electrode and a rectal cathode. In the two- st age group, lesions were made I week apart. Controls for the one-stage group were anesthe- tized and an incision was made over the skull; U1ose for the two-stage group had no surgical treatment. 4. The apparatus is described by C. J. Long and J. T. Tapp [J. Comp. Physio/. Psycho/. 12, 435 (1970)]. The position of the olfactory cue var- ied according to a schedule of sequences [B. J. Fellows, Psycho/. Bull. 61, 81 (1967)]. S. J. Maj, B. Durek, W. Palider, J. Phann. PharmacI none of the group with two-stage le- sions died in this period. Those surviving ani- mals in the one-stage group continued to lose weight for a median of S days after surgery (for the two-stage group: 4.S days after the first lesion and 3 days after the second). An initial nystagmus and continued excessive tear- ing were also observed. Five animals were ataxic at the start of training (three in the two- stage group). Two surviving animals broke or loosened o ne of their upper incisors (four in the two-stage group). Urogenital disturbances, including bleeding from the penis, incontinence, ' and infection, were observed in 8 of the initial 14 animals in the one-stage group (none in the two-stage group). 8. S. Siegel, Nonparametric Srarisr/cs (McGraw· Hill, New York, 1956), pp. 116-127, 184-193. 9. We thank C. Kellogg, J. S. S: hwartzbaum, and G. J. T homas for their comments. Supported by NSF grant GB 38384 and PHS grant NS 19 July 1974 Chlorine Compounds and Stratospheric Ozone T he potentially serious impact of man- niade chlorofluoromethanes (CF.,Cl 11 ) on stratospheric 0 3 is now documented in the scientific literature (1-3). In our recent report (2) we showed that the potential size of this atmospheric per- turbation is large, so large that chemical control of the stratosphere will eventu- ally pass to the chlorine oxides (ClO.,) that arise from CF <&Cl 11 usage. Ac- cording to our calculations, the CLO., sink for stratospheric O:i can be ex- pected to dominate natural sinks for 0 3 by 1985 or 1990. This time dependence is striking. Time scales of decades arise from our current knowledge of atmo- spheric mixing rates, and from the present belief (I-3) that the only sig- nificant mechanism by which nature can break apart CF 2 Cl 2 and CFCl 3 is by stratospheric photodissociation brought about by ultraviolet radiation. One purpose of this technical com- ment is to faci litate the readers' efforts SCIENCE, VOL. 188

with food reinforcement to habituate them to the general experimental situation.Tht: animals were then given 12 trials eacu day in a T-maze in which entrance mto an arm with an olfactory stimulus ( 1 percent amyl acetate solution) led to food reward ( 4).The criterion was 22 correct responses in 24 trials.Animals were again given free access to food upon reaching criterion, and behavioral testing ended with measurement of activity by means of an Animex meter (5) over a 90minute period.Throughout the study each lesion group was tested concurrently with a nonlesion control group (3).The animals were killed by spinal stem were isolated for subsequent parafconcussion; the cerebellum and brainfin embedding and staining with thionin, and the cortex and hippocampus were removed for noradrenaline assay (6).noradrenaline content (group with single-stage lesions and corresponding controls, r = .82;group with two-stage lesions and controls, r = .67) .T he motor disturbances and decreased activity in the animals with lesions may account for their slower speeds in the runway.
Running speed in the T-maze was not a limiting variable because the criterion was based on entrance into the goal arm rather than speed of traversal.There was no initial preference for the olfactory cue.The animals with bilateral single-stage lesions did not differ from their controls with respect to days to criterion (Table 1) [Mann-Whitney: U = 11, P = .92(8)].There were no significant differences for this measure for the group with total two-stage lesions, the group with partial twostage lesions, and their controls [Kruskal-Wallis: H e= 2.13, P > .1 (8)].
These results fail to demonstrate a learning deficit in a T -maze discrimination despite a substantial decrease in cortical noradrenaline caused by bilateral locus coeruleus lesions.
If the nucleus locus coeruleus is the substrate for a reinforcement system operative in appetitive learning, destruction of the nucleus should lead to deficits in any learning task with food reward.Since animals with complete bilateral lesions of the locus coeruleus can learn a T-maze discrimination, we conclude that the structure is not an essential component of a common reinforcement system.  1) demonstrated extensive destruction of the locus coeruleus in most animals.In the groups classified as having total lesions, only two animals had any identifiable locus coeruleus cells remaining.The reduction in cortical noradrenaline ranged from 67 to 90 percent in these animals.T he one-stage bilateral lesions markedly debilitated the animals; ataxia, temporary aphagia, and urogenital disorders were observed (7) .Postoperative recovery in the animals with bilateral two-stage lesions was both more rapid and more complete.

Chlorine Compounds and Stratospheric Ozone
During the habituation trials, all animals increased their running speeds so as to complete the 120-cm runway within 5 seconds.The running speeds for the animals with lesions were consistently lower than those for the control animals; however, we attribute this to motor variables.Motor difficulties observed (7) in some animals in all the lesion groups may have restricted • their maximum running speeds.Activity, as measured during the 90-minute test period, was positively correlated with 378 T he potentially serious impact of manniade chlorofluoromethanes (CF.,Cl 11 ) on stratospheric 0 3 is now documented in the scientific literature (1-3).In our recent report (2) we showed that the potential size of this atmospheric perturbation is large, so large that chemical control of the stratosphere will eventually pass to the chlorine oxides (ClO.,) that arise from CF <&Cl 11 usage.According to our calculations, the CLO., sink for stratospheric O:i can be ex-pected to dominate natural sinks for 0 3 by 1985 or 1990.This time dependence is striking.Time scales of decades arise from our current knowledge of atmospheric mixing rates, and from the present belief (I-3) that the only significant mechanism by which nature can break apart CF 2 Cl 2 and CFCl 3 is by stratospheric photodissociation brought about by ultraviolet radiation.
One purpose of this technical comment is to faci litate the readers' efforts SCIENCE, VOL.188 to verify our calculations (2).Several statistics and definitions require clarification, and one minor error should be noted by those who desire to check our work exactly.The CF.,Cl 11 mixing ratio shown for the year 2005 in figure 2 of (2) does not result from a steady model 2 emission rate of 2 x 107 cm-2 sec -1 at ground as the text implies.In fact, the text should read "5 X 107" on page 1166, column I , line 46.Also we erroneously used a doubling time of 2.64 years in our computer computations, not 3.5 years as we stated.Consequently, we projected the total CF .,Cl11 flux at ground level to be 5.2 X 10 7 cm -~ sec-1 in late 1975 instead of the proper value, 4.2 x 10 7 cm-2 sec-1.
We have repeated our calculations for model 3 with the proper 3.5-year doubling time, a 5.05-year e-folding time.Although model 3 is more strongly affected than models I and 2, we find with these corrections only small changes in the model 3 results: the column Oa destruction rate peaks around 1990, as before, at a value of about 1.55 X 1031 molecule sec-1 integrated over the globe.In our report (2) we showed a peak value (figure 1) of about 1.81 x 103 1 molecule sec-1 .
It should be noted that our presumed present stratospheric CIX (total concentration of gaseous chlorine compounds including Cl, ClO, and HCl) background accounts for more than hall of the model 3 0 3 destruction rate and that stratospheric CIX concentrations are not well known ( 4).On the other hand, we must emphasize that uncertainties in the present ClX background do not affect the conclusions of our report (2).rt is extremely important to learn present ClX values.Otherwise, future CIX increases will be impossible to document, whether due to chlorine release from CF.,.Clu, CIX injection by volcanoes, or HCI from the proposed Space Shuttle system (5).

APRIL 1975
Also our reference 11 in (2) presents 1973 CF 2 Cl 2 and CFC1 8 production statistics as current.These q.uantities, 4.5 X 10s kg and 2. 7 x l os kg, respectively, are DuPont's estimates of the 1973 annual global production.Presumably, the I 974 figures are larger.
The three paragraphs immediately above should assist others in repeating our calculations.Because of the potential seriousness of this problem, such checks are desirable.Similar reexamination of Crutzen's recent assessment of the expected 0 3 depletion (3) are also to be encouraged.The sensitivity of model calculations to parameters such as the characterization of atmospheric transport through eddy diffusion coefficients, chemical reaction rates, and assumed background concentrations of certain species (for example, OH) must be ascertained.Researchers must reexamine stratospheric photochemical schemes for completeness: H ave important gas-phase or heterogeneous reactions been overlooked?The search for alternative natural sinks (6) for CP.,Ciy should continue as we indicated (2) in reference 10.Researchers should also investigate the atmospheric fate of the lluorine atoms released from CF"'C1 1 r No direct effect on stratospheric 0 3 is expected (J-3), but reactions involving fluorine atoms may disrupt natural cycles of hydrogen-containing species and they may present possibilities for monitoring the CF .,Cl11 effect.
Perhaps more important, atmospheric measurements of background gaseous concentrations are needed for OH, atomic oxygen, the nitrogen oxides, and the chlorine oxides, and, obviously, continued monitoring of 0 3 is necessary.Short of observing 0 : 1 decreases directly, the most telling measurement would be• of CIO, its present abundance and its rate of increase.To assess the impact of HCl from the Space Shuttle, stratospheric OH data would also be useful.

Table 1 .
Summary of noradrenaline concentration and T-maze performance data; S.E.M., standard error of mean.
Stratospheric profiles of CF 2 Cl 2 , CFCl 3 , and CCl 4 are also needed immediately.As all these studies are formulated and completed, the appropriate industrial and governmental bodies may decide intelligently on questions of the regulation and control of CF .,Cl11 production.Each Space Shuttle launch will deposit 5 X 10 4 to 105 kg of HCJ in the stratosphere ["Environmental Statement for the Space Shuttle Program" (NASA, Washington, D.C., July 1972)].At the projected launch rate of 50 shuttles per year, we calculate that the globally averaged HCI input rate is at least 100 times smaller than the corresp<>nding chlorine injection rate due to continued CF, CI. usage at the 1973 levels.6. P. S. Liss and P .G. Slater [Na111re (Lo11d.)247, 181 (1974)] have established that the Rux of CFCl 3 from the atmosphere to the ocean is about 2 percent of the industrial production rate.7. We thank Dr. S. C. Liu for his extremely careful reading of our report (2), and for several helpful discussions on how to assist other readers.We acknowledge similar assistance from Prof. D. D. Davis.This research was supported by NASA grant NGR 23-005-616 and by the Climatic Jmpact Assessment Program, Department of Transportation, through NSF grant GA-43326.