1 1 DIVISION OF ADMINISTRATIVE HEARING DEPARTMENT OF ADMINISTRATION, STATE OF FLORIDA 2 Case Nos. 92-3038, 92-3039, 923040 3 SUGAR CANE GROWERS COOPERATIVE OF ) 4 FLORIDA, a Florida agricultural ) cooperative marketing association; ROTH ) 5 FARMS, INC.; and WEDGEWORTH FARMS, INC.,) ) 6 and ) ) 7 FLORIDA SUGAR CANE LEAGUE, INC.; UNITED ) STATES SUGAR CORPORATION; and NEW HOPE ) 8 SOUTH, INC., ) ) 9 and ) ) 10 FLORIDA FRUIT AND VEGETABLE ASSOCIATION,) LEWIS POPE FARMS, W.E. SCHLECHTER & ) 11 SONS, INC., and HUNDLEY FARMS, INC., ) ) 12 Petitioners, ) ) 13 vs. ) ) 14 SOUTH FLORIDA WATER MANAGEMENT DISTRICT ) an Agency of the State of Florida, ) 15 ) Respondent, ) 16 ) and ) 17 ) THE UNITED STATES OF AMERICA, ) 18 MICCOSUKEE TRIBE OF INDIANS, the ) FLORIDA DEPARTMENT OF ENVIRONMENTAL ) 19 REGULATION, the FLORIDA WILDLIFE ) FEDERATION, et al. ) 20 ) Respondent-Intervenors. ) 21 ) ________________________________________/ 22 DEPOSITION 23 OF 24 JAMES T. McCLAVE 25 Taken December 10, 1992 Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 2 1 STATE OF FLORIDA ) 2 COUNTY OF ALACHUA ) 3 Deposition of JAMES T, McCLAVE, taken at the 4 instance of respondent herein, pursuant to Notice of 5 Taking Deposition, at 305 Southeast Second Avenue, 6 Gainesville, Florida, on Thursday, December 10, 1992, at 7 9:00 o'clock a.m., before Harrisse S. Coffee, C.S.R., 8 C.P., Registered Professional Reporter and a Notary Public 9 in and for the State of Florida at Large. 10 - - - 11 APPEARANCES: 12 GARY PERKO, Esquire, of the law firm of Hopping, Boyd, Green & Sams, 123 South Calhoun Street, Post Office 13 Box 6526, Tallahassee, Florida 32314, (904) 222-7500, appearing for Petitioners Sugar Cane Growers Cooperative 14 of Florida, Roth Farms, Inc., and Wedgeworth Farms, Inc. 15 ROBERT H. BLANK, Esquire, of the law firm of Peeples, Earl & Blank, One Biscayne Tower, Suite 3536, Two 16 South Biscayne Boulevard, Miami, Florida 33131, (305) 358-3000, appearing for Petitioners Florida Sugar Cane 17 League, Incl, United States Sugar Corporation, and New Hope South, Inc. 18 GEOFFREY GARVER, Esquire, U.S. Department of 19 Justice, Environment & Natural Resources Division, General Litigation Section, P.O. Box 663, Washington, D.C. 20 20044-0663, (202) 272-4692, appearing for Respondent-Intervenor United States of America. 21 ALSO PRESENT: 22 DOUGLAS S. ROBSON, Ph.D., Statistical Consultant. 23 - - - 24 25 Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 3 1 I N D E X 2 WITNESS: PAGE: 3 JAMES T. McCLAVE 4 Direct Examination by Mr. Garver 5 5 - - - 6 EXHIBITS 7 DEPOSITION EXHIBIT FOR IDENTIFICATION 8 No. 1 - Proposal for water analyses 43 9 No. 2 - Affidavit 79 10 No. 3 - Document notes 111 11 No. 4 - Notes on a review 112 12 No. 5 - Critique 119 13 No. 6 - Appendix E 126 14 No. 7 - SAS printout 149 15 No. 8 - Correlation coefficients 150 16 No. 9 - Step-wise regression analysis 159 17 No. 10 - Cluster analysis 160 18 No. 11 - Cluster analysis 162 19 No. 12 - Regression model 164 20 No. 13 - Second order model 172 21 No. 14 - Station specific model 181 22 No. 15 - Database 184 23 No. 16 - Regression analysis model 186 24 No. 17 - Model of four quadrants 189 25 No. 18 - Observations 191 Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 4 1 EXHIBITS (CONTINUED) 2 DEPOSITION EXHIBIT FOR IDENTIFICATION 3 No. 19 - Models 194 4 No. 20 - Regression models 197 5 No. 21 - Analyses of Maffei data 199 6 No. 22 - Results of Info Tech modeling 201 7 No. 23 - Correlation analyses 204 8 No. 24 - Info Tech graphics 206 9 No. 25 - Plot 209 10 11 Certificate of Reporter 218 12 - - - 13 14 15 16 17 18 19 20 21 22 23 24 25 Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 5 1 JAMES T. McCLAVE, 2 having been produced and first duly sworn, testified as 3 follows: 4 DIRECT EXAMINATION 5 BY MR. GARVER: 6 Q Please state your name, and spell your last 7 name, please. 8 A My name is James T, McClave, M-c-C-l-a-v-e. 9 Q This is Sugar Cane Gowers Cooperative of 10 Florida, et al. versus South Florida Water Management 11 District, et al., Cases No. 92-3038, 92-3039, 92-3040. 12 And I am Geoffrey Garver. I'm an attorney with 13 the United States Department of Justice. I'm representing 14 the United States in this proceeding. 15 Dr. McClave, you've been listed by the Sugar 16 Cane Cooperative as an expert witness who will testify on 17 the interim and long term phosphorus concentration limits 18 for the Loxahatchee National Wildlife Refuge and 19 Everglades National Park. Is that consistent with your 20 understanding of your expected testimony in this matter? 21 A It is. 22 Q Are there any other areas you anticipate giving 23 expert testimony on in this proceeding? 24 A The only other area that's most recently been 25 brought to my attention, and it may or may not be Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 6 1 encompassed in what you've said, but just to be sure that 2 I've mentioned it, is the issue of settling velocities. 3 And, in particular, I think there's been some statistical 4 analyses done by Dr. Walker in a recent report that I've 5 just received and haven't had a chance to review at this 6 point, but I've been told I may be asked to review the 7 statistical aspects of that report. 8 Q To this point have you done any work with 9 respect to settling velocities? 10 A I have not. 11 Q Have you ever had your deposition taken before, 12 Dr. McClave? 13 A Yes. 14 Q Then you understand that I'll be asking you a 15 series of questions, and I am entitled to your complete 16 and honest answers to those questions. Unless your 17 attorney tells you otherwise, you must answer the 18 questions that I ask you. 19 A I understand. 20 Q If you don't understand the question, which is 21 not at all improbable, given the nature of this 22 deposition, please let me know and I'll attempt to 23 rephrase it. 24 I'm certainly not an expert in this area, and 25 sometimes my phraseology may be a little awkward. So if Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 7 1 that happens, please let me know and I'll do my best to 2 get back on track. 3 A All right. 4 Q Dr. McClave, have you done anything to prepare 5 for this deposition? 6 A Yes. 7 Q What have you done? 8 A I have reviewed my files and met with 9 Mr. Perko yesterday in preparation. 10 Q Okay. Have you read the notice of deposition 11 for this deposition? 12 A I did at some point, yes; probably when I first 13 received it. 14 Q Okay. Are you familiar with the list of 15 documents that was attached to the back of that deposition 16 notice, asking you to produce certain documents? 17 A Yes. 18 Q And have you produced all the documents that you 19 have in your possession that are responsive to the list 20 attached to the deposition notice? 21 A I intended to. I think I have, yes. 22 MR. PERKO: For the record, I'd note that we 23 have withheld I think it's six documents on the basis 24 of privilege that's indicated in the letter that I 25 sent to you last week. Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 8 1 MR. GARVER: Okay. Mr. Perko, I have received 2 that list, yes. 3 BY MR. GARVER: 4 Q Can you just describe to me, in general, what 5 documents you have produced in connection with the 6 following categories. 7 The first category was all documents that 8 describe, explain or relate to statistical analysis of 9 SWIM plan interim and the long term phosphorus 10 concentration limits or levels for the Loxahatchee 11 National Wildlife Refuge and/or Everglades National Park. 12 A May I look at what you're reading from? That 13 would help me. That was a long sentence. 14 Q (Document handed to witness.) 15 A (Witness perusing document.) 16 I would say that the documents that respond to 17 number one are a number of printouts, a large number of 18 what I'll call statistical computer printouts, some of 19 which are just data, many of which are statistical 20 analyses that were performed on these data. 21 I produced, I believe, documents that the 22 District -- contained the District's statistical analyses. 23 And I guess a broad reading of this would -- it doesn't 24 say just my analyses, so I guess that would respond to 25 that as well. Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 9 1 And I believe that I produced at least one 2 affidavit that I had written and several by others, 3 including Dr. Walker, and I believe Dr. Pollman, that were 4 in my possession that related to statistical analyses, at 5 least in part. 6 Q Are there any other documents that you can 7 recall at this time that you produced, responsive to 8 category number one? 9 A Other than to say that, at least in my mind, 10 most of what I produced would fall under category one and 11 to the extent I haven't described them all, there may be 12 others. But I -- but just so you understand, the way I 13 view it, most of what I produced had to do with 14 statistical analyses in this area. 15 Q Okay. With respect to the second category, 16 which includes all documents regarding Lake Okeechobee, or 17 Everglades water quality and/or water quantity, including 18 but not limited to those issues raised in the Sugar Cane 19 Growers Cooperative, et al. petition for formal 20 administrative proceedings pursuant to Section 120.57(1) 21 Florida Statutes? 22 A I would say that -- I'm not sure what that 23 means. I'll respond as best I can, that the data -- much 24 of the data involves water quality, so the raw data that I 25 produced I think might be responsive to that. There may Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 10 1 be some redundancy between one and two in that regard. 2 And there were certain reports that I produced 3 that I had looked at or relied upon that had to do with 4 studies of what I remember of water quality studies in the 5 Everglades or Lake Okeechobee. 6 Q Are these reports that you're referring to 7 reports including such things as District technical 8 publications -- 9 A That's -- 10 Q -- and other publicly available literature? 11 A Yes. That's what I had in mind, yes. Certain 12 University of Florida, I think, publications and District 13 publications. 14 Q All right. With respect to the third category, 15 which includes any and all documents utilized or relied 16 upon in preparing, formulating, developing, authoring, 17 co-authoring, reviewing or organizing anticipated expert 18 testimony in this action, what documents have you produced 19 in that category? 20 A Well, again I would put the reports, including 21 certainly Dr. Walker's reports, in that category, as well 22 as the reports we just talked about under number two. 23 Certainly District documents that had to do with the SWIM 24 plan would, I guess, fall under that -- under number three 25 as well. Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 11 1 That's what comes to mind. 2 Q Have you yourself authored or co-authored any 3 reports or other documents in preparing for your 4 anticipated testimony? 5 A I think to this point the only thing that I've 6 authored, the only thing that comes to my mind right now, 7 is an affidavit that was submitted sometime back having to 8 do with Walker's analysis of the ENP. 9 Q When you say ENP you mean Everglades National 10 Park? 11 A That's right. 12 Q And with respect to the fourth category, all 13 technical publications, technical memoranda, or other 14 publications you've authored or co-authored, or the 15 preparation of which you participated in, relating to the 16 Everglades protection area, the Everglades agricultural 17 area, or any other subject matter in this action, or 18 documents that you produced in connection with that 19 category? 20 A Again, there may be others, but the one that I'm 21 remembering right now is the affidavit we just talked 22 about. I don't think there are any others. 23 Q Okay. Dr. McClave, what is your occupation? 24 A I'm a statistician. 25 Q And how long have you been a statistician? Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 12 1 A Well, I began the study of statistics in 1966. 2 I received a Ph.D. in that area in 1971. 3 Q And how are you employed? 4 A I am currently president of Info Tech, 5 Incorporated. 6 Q I believe you just stated you obtained a Ph.D. 7 in 1971. Is that correct? 8 A Yes. 9 Q Do you have any other professional degrees? 10 A I have a Bachelor of Science in physics as 11 well. 12 Q Where did you get your Bachelor of Science in 13 physics? 14 A At Bucknell University. 15 Q Did you study statistics while earning your 16 Bachelor's? 17 A Yes, I did take a number of courses on the 18 undergraduate level in statistics; yes. 19 Q Do you recall how many courses you took on 20 statistics in your undergraduate education? 21 A Roughly half a dozen. 22 Q Do you recall the names of any of those courses? 23 A Too long ago. They would generally have been in 24 the area of introductory statistics. 25 I took just about every statistics course that Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 13 1 Bucknell University had to offer so it would have ranged 2 from introductory statistics up through regression 3 analysis and analysis of variants and the like. 4 Q Did you take enough statistics to have what 5 would be called a minor in statistics while you were at 6 Bucknell? 7 A I think that I had enough to qualify for a minor 8 in mathematics. We did not have a degree in statistics. 9 Bucknell did not offer a degree in statistics. I believe 10 that I ended up with a minor in math, primarily as a 11 result of all the math and statistics that I had taken. 12 Q What course work did you take in statistics 13 while earning your Ph.D? 14 A Gosh. I would need a transcript to tell you 15 that. Basically every course that the University of 16 Florida offered in statistics. I can start and try to 17 list as many as I can, but it would have been Theory of 18 Statistics, at least four semesters; four semesters of 19 what we called Applied Statistics, Design of Experiments, 20 Analysis of Variants, Regression Analysis and the like, 21 Time Series Analysis, Nonbarometric Statistics, Sequential 22 Analysis. 23 I'm sure there are many more that I'm 24 forgetting, but basically the catalogue for the University 25 of Florida would tell you, because it's just about every Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 14 1 course that they offered I'm sure that I took during those 2 five years. 3 Q So you got your Ph.D. at the University of 4 Florida; is that correct? 5 A I did, yes. 6 Q Did you have any teaching or research assistant 7 positions while getting your Ph.D? 8 A I did, yes. 9 Q Can you describe those positions for me, please? 10 A I participated both in teaching and consulting 11 as a graduate student. I had an assistantship the entire 12 time, as I recall, and by the time I got my Ph.D. I was 13 teaching at least half time. I was teaching large section 14 statistics courses by 1970 and '71. 15 So most of it was teaching, as I recall, but in 16 some semesters I would also do some consulting around 17 campus in various areas in the agricultural school and 18 medical school, for example. 19 Q Do you recall what courses you were teaching 20 specifically? 21 A Primarily it would have been -- we have several 22 introductory undergraduate statistics courses, and what I 23 recall is teaching those, those courses. I may have 24 occasionally in the summer, for example, taught a little 25 more advanced level, but primarily while I was earning my Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 15 1 Ph.D. I was teaching introductory statistics. 2 Q Were you a research assistant at any time while 3 getting your Ph.D? 4 A I don't know what the University of Florida 5 called it. It may have been called a research 6 assistantship, but I just remember it as being an 7 assistantship and we did both -- our duties included 8 whatever we were told to do, but primarily it was teaching 9 and consulting. 10 Q Did you conduct any research while you were 11 getting your Ph.D? 12 A Well, other than my Ph.D. research, you mean? 13 Outside of the research I was doing for my Ph.D., or 14 including that? 15 Q Why don't we start with outside of your Ph.D. 16 A Other than research that would have been 17 involved -- in other words, when a statistician consults 18 with another researcher, there's a joint effort involved 19 there in research. So I would have done some research 20 with others, usually faculty members who needed 21 statistical help. 22 I did not conduct any primarily statistical 23 research, other than what I was doing for my Ph.D., during 24 that period. 25 Q What research did you conduct towards getting Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 16 1 your Ph.D? 2 A The area in which I got my Ph.D. was known as 3 time series analysis, and so I conducted several years of 4 research in the general area of time series analysis and, 5 in particular, in time series modeling. 6 Q Can you explain to me generally what time series 7 analysis is? 8 A Time series analysis is the area of statistics 9 that explicitly looks at data which are collected in some 10 sequence over time, and tries to take into account the 11 fact that -- the time factor in the data. Whereas many 12 data analyses would not be considering time as a factor, 13 time series analysis explicitly considers time as a 14 factor. 15 Q And can you describe to me, just in general at 16 this point, what would be involved in time series 17 modeling? 18 A In general - and this is very general, but to 19 just give you the flavor, most of the time when data are 20 modeled, we assume that the data are independent of one 21 another and don't have any kind of correlation with one 22 another. In a -- you know, as far as assumptions are 23 concerned. 24 In time series analysis, it's acknowledged that 25 there's probably going to what be what we call serial Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 17 1 dependence. If you're talking about a business cycle, for 2 example, you're either in an up cycle or a down cycle, and 3 observations in the up cycle are likely to be related, and 4 so are those in the down cycle. 5 So in a very general way I would say time series 6 analysis and time series models consider explicitly the 7 serial dependence of data, whereas most statistical 8 analyses or conventional statistical analyses do not. 9 Q And I believe you said that the main difference 10 would be that in those other conventional analyses, 11 different factors or variables would be assumed to be 12 independent. Is that the main difference? 13 A There's usually an underlying assumption of 14 independence of the underlying random structure of the 15 data in a conventional analysis, and that assumption is 16 what goes away usually, is not made in a time series 17 analysis. 18 Again, that's a pretty gross oversimplification, 19 because time series, as you might imagine, involves -- has 20 involved many person years of research over the last -- 21 it's really only about twenty-five or thirty years old as 22 a serious science, but -- there's much more to it than 23 that, but that does, I hope, give you some idea of the 24 difference between conventional analyses and time series 25 analyses. Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 18 1 Q Did you take any -- do any course work in 2 stochastic processes? 3 A Yes. 4 Q What kind of course work did you take in that? 5 A Well, stochastic processes, the concept of a 6 process is usually related to time. It doesn't have to 7 be. And so stochastic processes is the theoretical 8 underpinnings, in a sense, at least as I look at it, to 9 time series. 10 And so at Florida we had to take or we did take 11 - we had to take, I think, two years - at least one year 12 that I remember for sure of theoretical stochastic 13 processes before getting into the time series analysis and 14 time series research. 15 Q Okay. 16 A And, in fact, now that you mention it, I also 17 taught -- there was a course that we taught over TV to 18 engineers down at Cape Canaveral - at that time it was 19 Cape Canaveral - on stochastic processes, and I taught 20 that course as I neared the completion of my Ph.D., so 21 that was one graduate level course. They were earning 22 graduate level credits, and I taught the stochastic 23 processes course to them. 24 Q Was your assistantship tied to a particular 25 professor at the University of Florida? Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 19 1 A The assistantship? 2 Q Yes. 3 A No, not that I remember. It would have been 4 various professors over time. 5 Q Did you have a particular major professor that 6 you worked with during -- while getting you Ph.D? 7 A Yes, I did. 8 Q Who was that? 9 A Dr. Kamal, K-a-m-a-l, Chanda, C-h-a-n-d-a. 10 Q What was involved in -- you wrote a dissertation 11 in connection with your getting your Ph.D.; correct? 12 A I did. 13 Q What was involved in putting together your 14 dissertation? What did you do? 15 A It involved at least a year of background 16 research, in addition to course work. We've already 17 talked about course work. It involved about a year's, as 18 I recall, independent research with Dr. Chanda, looking 19 for a topic, becoming familiar with the general area that 20 I wanted to do my research in, which was time series 21 modeling, and then about a year of proving theorems and 22 getting results an writing it up. 23 So this was about two solid years, as I recall, 24 of research and development of the thesis. 25 Q What specific type of research was involved in Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 20 1 honing in on your topic? 2 A We were working in the area of time series 3 modeling. More specifically we were doing logistic 4 process modeling, which is a specific kind of time series 5 model. So much of it, as I recall, involved reading every 6 piece of research that had been done in that area, and 7 finding out whether the problem that we thought hadn't 8 been solved indeed had been solved, or what progress had 9 been made on the problem, and then taking what had been 10 done forward another step. 11 Q Okay. And what was the specific area of time 12 series analysis or logistics? 13 A There -- I'm sorry. 14 Q What was -- 15 A Logistic processes? 16 Q Logistic processes that you honed in on to do 17 research and write your dissertation on? 18 A It had to do with developing better estimates of 19 the parameters or the aspects of the model that were used 20 for forecasting in the future. 21 In other words, most time series models have at 22 least two objectives. One is to describe the process that 23 underlies what's called the past, the historic past. And 24 then two is to, once you believe you've made that 25 description, to use that to forecast the future of the Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 21 1 process. 2 And again, a little bit of an over 3 simplification, but most of my research as I recall at 4 this point, twenty years later, involved the -- developing 5 better estimates of the parameters of the model, and thus 6 better forecasts of the future. 7 Q When you say "parameters of the model," what 8 parameters are you talking about? 9 A A parameter, to be -- to take a simple example, 10 if you had a straight line, a straight line has two 11 parameters: the intercept an the slope. Those are what we 12 mean by parameters. 13 Now, in a logistic time series model, there may 14 be three or four or five different parameters of the 15 model. Parameters are constants which, if you were 16 graphing the process, would put the model in two or three 17 or four dimensions, if you could draw it in two or three 18 or four dimensions. 19 So every -- every statistical model has some 20 number of parameters associated with it, and one of the 21 objectives of modeling is to develop the best estimates of 22 those parameters that you can. 23 Q And when you're -- when you're attempting to 24 estimate those parameters and -- no, let me back up. 25 We'll get to that later, maybe. Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 22 1 When you say "best estimates" of parameters, 2 best in what sense do you mean? 3 A Best in a statistical sense. And typically what 4 that mens is an estimate that's not biased; it's not 5 tending to either miss the parameter on the high or low 6 side on a consistent basis; and one that has minimum 7 variance. That is an estimate that varies around the true 8 parameter as little as possible, falls as close to the 9 true parameter -- In other words, we think of a model 10 out there that's got some true value of the parameter that 11 we don't know, and we're trying to use data to estimate 12 that. 13 So we're trying to fall as close to it as we 14 can, to the real parameter, and to not miss it 15 consistently on one side or the other. That's a 16 typically, again, a little bit of over simplification, but 17 typically what we mean by best in statistics. That's what 18 we're trying to do. 19 Q When you were doing your Ph.D. research, what 20 specifically were you doing to try and come up with the 21 best estimates of parameters, the best way of estimating 22 parameters? 23 A We were considering -- all estimates are derived 24 from data. There are formulas that involve data that give 25 an estimate - that result in an estimate of a parameter. Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 23 1 And we were working on formulas that were different than 2 formulas that had been heretofore utilized for certain 3 parameters, trying to improve the estimates in the senses 4 that I just gave you, making them less biased and making 5 them smaller variants. 6 And, in particular, in the area of logistic 7 process modeling, there had not been -- the parameter 8 estimates that were being utilized at that point in time 9 were relatively simple, and we were using somewhat more 10 complex estimates. 11 Q And were you -- you were successful? Were you 12 successful in developing methods for getting better 13 estimates of the parameters? 14 A Yes. My recollection is we advanced -- the 15 whole idea, I think, of - at least now looking back and 16 having directed Ph.D. dissertations myself, is to make 17 some advance forward, and I think we did that. The 18 results ended up being published and there's been a lot 19 done since then to still further improve. 20 But, yes, at that point in time I think we were 21 successful. 22 Q Okay. And in what way did your research and 23 your dissertation, your Ph.D. work, make improvements over 24 what had been done before? 25 A At that point in time, this is late 1960s and Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 24 1 early '70s, there still was a lot of time series 2 estimation that was done - what I call by hand. That is, 3 it was not taking full advantage of the computer power 4 that was just then becoming available. 5 And I think the main thing that we did was make 6 it evident that these complicated processes could be 7 computerized, and we could take full advantage of complex 8 formulae to estimate parameters; that we did not need to 9 rely on simple scratch pad kinds of calculations. So as I 10 look back, one of the main things we did was, in that 11 particular area of time series, bring it into the computer 12 age. And since then almost all the analyses have been 13 computer driven. 14 Q Okay. You mentioned, I believe, that your 15 thesis was published. Is that correct? 16 A Yes. I published several papers within the next 17 three or four years after getting my Ph.D. that were 18 derived from the work in my dissertation. 19 Q Was your dissertation itself published? 20 A You mean in toto? 21 Q Yes. 22 A No, other than just in the usual way. I mean 23 it's published for the purpose of the University of 24 Florida Libraries, but not in any journal or anything, 25 no. Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 25 1 Q You mentioned three or four publications that -- 2 of your work. Is that correct? 3 A Right. In other words, I kept working in that 4 area after I got my degree. And I would say there are 5 several publications, probably three or four that, if I 6 looked at my vitae, I would say that in some way was 7 derived from my dissertation. 8 Q Did one of those publications include a 9 publication Biometrika, entitled "On the bias of 10 autoregressive approximations to moving averages"? 11 A Yes. I would classify it in that category. 12 Q Do you recall any of the other publications 13 where your dissertation work was published? 14 A Not without looking at my vitae, I don't, no. 15 It's been long enough ago that I've forgotten the -- I 16 think there was one in the Journal of the American 17 Statistical Association, but other than those two, I don't 18 recall. 19 Q Is that an article entitled "Choosing the order 20 of an autoregressive process: the max chi-square 21 Technique"? 22 A That's right. 23 Q What did you do after -- well, when did you get 24 your Ph.D? 25 A 1971. Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 26 1 Q Okay. What did you do after obtaining your 2 Ph.D? 3 A I accepted a postdoctoral position at the State 4 University of New York at Buffalo. 5 Q And how long were you there? 6 A I was there for one academic year. 7 Q And what years -- what year was that? 8 A That was 1971-'72, as I recall. It might have 9 been '72-'73 but I think it was '71-'72. 10 Q What did you do during that postdoctoral 11 period? 12 A There was an individual there by the name of 13 Emanuel Parzen, P-a-r-z-e-n, who was and is one of the 14 foremost researchers in time series analysis. And I 15 wanted to continue my study of time series under someone 16 of his stature, and had the opportunity to do that and did 17 that for a year. 18 Q What kind of work did you do with Dr. Parzen? 19 A I continued my time series research that I had 20 started with my dissertation, began to expand it into the 21 areas that led to the publications that we just talked 22 about, plus some others, as I recall. I attended his 23 advanced time series class for that year, and just had the 24 opportunity to talk and do research with him for a year. 25 Q Did you do any teaching during that year? Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 27 1 A I did, yes. 2 Q What did you do following your year at Buffalo? 3 A I returned, as a faculty member, to the 4 University of Florida. 5 Q And for how long were you a professor at the 6 University of Florida? 7 A Assuming I've got my years right, and I think I 8 do, I came in 1972 and taught basically full time, with 9 some leaves here and there and sabbaticals here and there 10 until 1989, so I guess eighteen years; seventeen, eighteen 11 years. 12 Q Your vitae states that you are now an adjunct 13 professor at the University of Florida. Is that correct? 14 A That's correct, yes. 15 Q What is the nature of your position as an 16 adjunct professor? 17 A I was with the college of business the last 18 seven or eight years of my tenure at the University of 19 Florida. And when I decided to leave, I was asked to 20 remain as an adjunct with a center known as the Public 21 Policy Research Center at the University of Florida. 22 And I still go back and lecture occasionally to 23 them. And it's just an informal appointment with the 24 University of Florida through that center. 25 Q All right. Do you ever teach full courses any Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 28 1 more? 2 A I have -- the intent, when we talked about the 3 adjunct, was to do that on occasion, but it's only been 4 two years now, and that hasn't happened yet. It's just 5 been occasional teaching assignments. 6 Q Okay. While you were teaching either at the 7 SUNY Buffalo or at the University of Florida, did you 8 direct any graduate research? 9 A Yes. 10 Q And how many Ph.D. candidates' work did you 11 direct or -- 12 A Actual major professor direction, only a couple; 13 probably two. As far as being on committees, Ph.D. 14 committees and having some directional responsibility, 15 probably twenty. 16 Q Do you recall the subject matter of the Ph.D. 17 work in which you were the major professor? 18 A It was time series analysis in both cases. 19 Q Was there a specific aspect of time series 20 analysis involved in those Ph.D. programs? 21 A Yes. It was considered continuing in the same 22 line as my research had been for the last - at that point, 23 ten years of modeling, parameter estimation, and the like. 24 (Off-record discussion.) 25 (Brief recess.) Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 29 1 BY MR. GARVER: 2 Q Well, now that we've taken a very short break, 3 Dr. McClave, I'll just ask you to let me know any time you 4 feel you need a break. 5 A Thank you. I will. 6 Q I'll be glad to accommodate that. 7 A Okay. 8 Q I believe you were going over the areas that -- 9 the specific areas or subareas of time series analysis 10 involved in the Ph.D. work that you directed as a major 11 professor. Is that correct? 12 A Right. 13 Q Sorry for having you do this, but could you just 14 go over all that? 15 A Yes. It was an extension of research that I had 16 been doing for ten years which, again, in general, 17 involved time series modeling, parameter estimation, use 18 of computers in developing better estimates for time 19 series parameters and that would have been the same area 20 that my students worked in. 21 Q Did those students, those two or so students 22 whose Ph.D. work you were directing as a major professor, 23 did they obtain their Ph.Ds? 24 A Yes. 25 Q I believe you mentioned earlier that you're the Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 30 1 president of Info Tech, Incorporated. Is that correct? 2 A Yes. 3 Q What is Info Tech? 4 A Info Tech is a statistical and econometric 5 consulting firm, as well as a software development firm. 6 Q And when was Info Tech founded? 7 A In 1977, I believe, we became a corporation. 8 Q And were you the founder of Info Tech? 9 A Yes. 10 Q Did you start Info Tech on your own? 11 A I believe, in the very beginning, there were two 12 other professors involved, and we formed a partnership. 13 That was probably about 1975. 14 By the time that it became a corporation, I was 15 full owner, sole owner. 16 Q Okay. Who were the other professors that you 17 were in partnership with? 18 A When it was a partnership, there was Dr. Richard 19 Scheaffer, that's S-c-h-e-a-f-f-e-r, and Dr. Ray Littell, 20 L-i-t-t-e-l-l. 21 Q And when you started Info Tech, were they 22 involved at all in starting that business up? 23 A By the time it became a corporation? 24 Q Yes. 25 A No. We had done several jobs together as Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 31 1 partners, and they both decided to focus on other things. 2 Q Did you have any other employees at Info 3 Tech when you began? 4 A I may have had a -- well, I'm sure my wife would 5 be upset if I didn't say that she was an employee, because 6 she was a big factor in helping me get started. But it 7 wasn't very long before we hired a receptionist as well, 8 secretary as well. 9 But I think when I first started -- in '77 when 10 it became a corporation, it was probably just my wife and 11 I. 12 Q How many people are employed by Info Tech today? 13 A Approximately forty. 14 Q What is the organizational structure of Info 15 Tech now? 16 A I'm sorry, I'm not sure I understand. 17 Q You're the president; is that correct? 18 A Yes. 19 Q Are there any vice-presidents? 20 A Oh, yes. I have a vice-president in charge of 21 the software development division. His name is Dr. Tom 22 Rothrock, R-o-t-h-r-o-c-k. He's the only other management 23 level, or executive management level, individual. 24 We have some team leaders and the like. We're 25 organized roughly into teams, and each team has a team Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 32 1 leader. 2 Q How many different teams are there? 3 A Five or six, I believe. 4 Q Do the teams have different subject areas that 5 they focus on? 6 A Yes, they do. 7 Q What are those different areas? 8 A We have one that we call a consulting team that 9 does the statistical and econometric consulting. We have 10 a team involved with -- most of the rest of the teams are 11 involved with software development, and we have what we 12 call our development team. We have a team of customer 13 support people. We have an accounting team. We have 14 decision support system team, called the DSS team. We 15 have a PC, a personal computer operations and development 16 team. 17 I believe that's it. 18 Q How many of your forty or so employees are 19 professional level employees? 20 A Probably about seventy-five percent; twenty-five 21 or thirty I'd say, of them. 22 Q How many of those employees are Ph.D. level 23 employees? 24 A At least three that I can think of. I think 25 it's three. There may be several others. I'm not sure. Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 33 1 Q What type of software does Info Tech develop? 2 A Primarily what we have developed is a software 3 to perform statistical and econometric analyses in the 4 detection of white collar crime and, more specifically, 5 antitrust violations. And most of the software we've 6 developed has been installed in state and federal agencies 7 who are purchasing through the sealed bid process. And so 8 our software assists in analyzing those bids for potential 9 violations. 10 Q Your vitae lists a number of consulting projects 11 you've done, and one of those projects is some work you 12 did for the Florida Electric Power Consulting Group, 13 involving statistical analysis of water quality data to 14 evaluate compliance with Department of Environmental 15 Regulations Secondary Standards. Is that correct? 16 A Yes. 17 Q Can you describe to me that project you did for 18 Florida Electric Power Consulting Group? 19 A I don't remember many details, other than the 20 fact that I think it had to do with the measurement of 21 water quality at wells on the boundaries of the plants, 22 and the standards that the water in those wells had to 23 meet and the statistical analysis of those data. Quite 24 frankly, I don't remember what parameters we were looking 25 at even, at this point in time. That's about all I Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 34 1 remember about it. 2 Q Were you involved personally in that project? 3 A Yes, um-hmm. 4 Q Do you recall what types of statistical analysis 5 you did in connection with those data? 6 A I'm sorry, I don't. I'd have to review our 7 files on it. I just don't remember. 8 Q Okay. And do you remember when that project was 9 conducted? 10 A The vitae should have a year or set of years 11 there. I don't remember. 12 Q It says 1986-1987. Does that -- 13 A That sounds right. 14 Q Do you recall a project you did for 15 Environmental Science and Engineering and the 16 Environmental Protection Agency, Pesticides BAT Phase II 17 Project, involving the analysis of discharge of pesticide 18 plants to determine best available technology for 19 controlling pollutants? 20 A In general, yes. I remember doing that kind of 21 work. 22 Q Can you describe that project for me? 23 A I did a number -- there are a number of entries 24 on there that have to do with effluent guidelines, 25 guidelines for various industries, including pesticide, Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 35 1 timber, and other industries, most of which came -- I 2 served or Info Tech served as a subcontractor to ESE, who 3 was contracted by EPA. 4 And what I remember doing for those was there 5 were a lot of data involved and I did the statistical 6 analysis and statistical support for those guideline 7 development projects. 8 Q Do you recall what type of statistical analysis 9 was involved generally in those effluent guideline type 10 projects? 11 A As I recall, it ranged from the very simple 12 descriptive analyses calculation, meaning the standard 13 deviations, to some pretty sophisticated regression 14 modeling, time -- it did involve some time series 15 modeling, as I recall. 16 So there was a very wide range, as I recall, of 17 statistical sophistication used in those analyses. 18 Q Do you recall specifically any of the projects 19 in which you were using regression modeling to develop 20 guidelines? 21 A I would answer you that I think I ended up using 22 regression in all of them, but that may be wrong. It 23 seems like most of them ended up involving some regression 24 analysis, regression modeling. And several of them, 25 probably not all of them, involved time series modeling. Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 36 1 I don't remember the specifics of which. 2 Q Do you recall -- why in some cases did you do 3 time series analysis, and at other times not do it? 4 A It gets back to the serial correlation issue 5 that we discussed earlier. There were some cases where 6 the data were collected at regular intervals over time, 7 and it was obvious, on analyzing the data, there were time 8 series effects, serial correlation effects, that needed to 9 be taken into account. And my recollection is that, when 10 that was the case, we used time series modeling. 11 Q Do you recall any times when you had time series 12 data and did not do a time series analysis? 13 A No, I don't. 14 Q Do you recall specifically which projects you 15 used time series analysis on? 16 A No. I, again, would have to look back at the 17 files of those. Most of those have been ten or so years 18 ago, I think, if you look at the dates, and I just don't 19 recall. 20 Q Have you provided consulting services in 21 connection with litigation over water quality in the 22 Everglades? 23 A Not that I recall, no, not prior to this issue 24 here. I don't think so. 25 Q But you have been approached to provide Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 37 1 consulting services in this litigation, dealing with water 2 quality of the Everglades; is that correct? 3 A Yes. 4 Q When were you first approached to provide 5 consulting services in connection with litigation over the 6 water quality in the Everglades? 7 A I don't recall the exact date. My best estimate 8 would be sometime in -- I don't think it was 1990. I 9 think it was sometime in 1991, early 1991. But again, I'd 10 have to look at our contracts, and so forth, to know for 11 sure. 12 Q Do you remember who first approached you? 13 A I believe I was first approached by Dr. Curt 14 Pollman, P-o-l-l-m-a-n, with KBN, who was already working 15 on the project, and my recollection is he called me about 16 doing some statistical work on the project. 17 Q Do you recall what you discussed with him at 18 that time? 19 A I think that the issue at that time was the BMP 20 rule, and the statistical analyses that were being -- had 21 been performed or were being proposed by the District. I 22 don't remember the name of the document, but I do remember 23 it had to do with BMP rule, and I was asked to review 24 those analyses. 25 I think that was the first thing I ever saw in Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 38 1 this litigation. 2 Q Do you recall when you first saw this report or 3 approach with regard to the BMP rule? 4 A It would have been whenever that first contact 5 was, and again I'd have to look in my file. I don't 6 remember. It seems sitting here that it's been more than 7 a year ago, and I think it has, but I'm not positive about 8 that. 9 Q Did you actually do a review of some aspect of 10 the District's proposed BMP rule? 11 A I did, yes. 12 Q Did you provide any documentation of the work 13 you did with respect to the BMP rule in connection with 14 this deposition? 15 A I don't believe so, and the reason I didn't is I 16 viewed that as basically a separate project from the one 17 that we're working on now; perhaps mistakenly, but that's 18 the way we've been viewing it at Info Tech, anyway. 19 Q What type of analysis did you do in reviewing 20 the District's proposed BMP rule? 21 A The district had performed an analysis, as I 22 recall, regression analysis, on the data that were being 23 utilized to develop the BMP rule, and I analyzed those 24 same data to determine whether or not there were better 25 ways or different ways of modeling those same data, Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 39 1 whether or not -- I was asked to determine whether or not 2 I agreed with the statistical approach that was being 3 taken. 4 I eventually considered some other data besides 5 those that were being used by the District. In 6 particular, I remember rainfall records going back much 7 further than had been utilized by the District. 8 So it was just generally statistical analysis of 9 the data involved in the BMP rule. 10 Q Do you recall what types of data were involved 11 in the BMP rule? 12 A Well, there certainly was phosphorus data. I 13 remember that. Rainfall data. And as I recall, the basic 14 concept was to relate phosphorus concentrations to 15 rainfall. That was the basic premise behind the rule. 16 Q What types of phosphorus data were involved? 17 A I'm sorry, I don't understand. 18 Q Where did the phosphorus data come from and what 19 locations were they taken from, and that sort of thing? 20 A I'm sorry, I don't recall any longer where the 21 -- I'd have to look back at the documents. I don't know. 22 MR. GARVER: Mr. Perko, just for the record, I 23 believe documentation regarding that work by 24 Dr. McClave would fall under the notice of 25 deposition. Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 40 1 MR. PERKO: We can provide you that 2 information. 3 It does not relate to his anticipated testimony 4 in this case, however. 5 MR. GARVER: That's not the portion of -- 6 MR. PERKO: I understand. 7 MR. GARVER: Okay. 8 Dr. McClave, I may have further questions 9 on that, but it would probably save time if I had the 10 documentation to look at and review before I do that, 11 so we'll move on. 12 BY MR. GARVER: 13 Q Aside from the BMP rule that you did some 14 statistical analysis with respect to, were there any other 15 consulting services you were asked to provide in 16 connection with litigation over water quality in the 17 Everglades? 18 A Other than what we've talked about in terms of 19 my anticipated testimony here? 20 Q No, no. Other than the BMP rule, and I'm 21 including any other work you've done including work 22 involving your anticipated testimony. 23 A I'm sorry. I still don't understand your 24 question. 25 Q What I'm trying to do is get a chronology of Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 41 1 what you -- the work you performed, the consulting 2 services you provide in connection with Everglades water 3 quality issues. 4 A Um-hmm. 5 Q You've mentioned, now, some statistical analysis 6 you did with respect to the BMP rule and I'm trying to 7 move ahead now in time to other work you performed. 8 A Moving ahead, then, would be to this matter, in 9 the way I'm thinking about it. In other words, the only 10 other thing I've done other than BMP rule work involves 11 the matter that we're here about today. 12 Q Okay. And when you say this matter, the matter 13 we're here about today, what are you talking about? 14 A I'm talking about, from my point of view, 15 anyway, the statistical work that involves data in the 16 Loxahatchee and data in the ENP, Everglades National Park. 17 Q Okay. When did you first begin working at data 18 from Loxahatchee and Everglades National Park? 19 A It was not necessarily after the BMP rule 20 analysis, but it was after I had already begun the BMP 21 rule analysis. Again, the way I look at it, this sort of 22 separate and distinct matter came up, the Loxahatchee and 23 ENP analyses. 24 So it would have been some months, several 25 months, after my first engagement. Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 42 1 Q And who approached you with respect to doing an 2 analysis of the Loxahatchee an Everglades National Park 3 data? 4 A It would have been some combination of Curt 5 Pollman and the attorneys for the co-op. 6 Q What, specifically, were you asked to do in 7 terms of analyzing data from Loxahatchee Refuge and 8 Everglades National Park? 9 A The first thing I was asked to do was to review 10 any analyses that had already been performed by the 11 District or consultants who were working for the District 12 or for the federal government in this area. 13 Q And when did you begin conducting that 14 evaluation and analysis? 15 A Again, I'm sorry I don't have these dates in 16 mind, but it was several months after -- if I'm right that 17 it was early 1991, it would have been several months 18 later. 19 Q Did you prepare a proposal to do the work 20 involving statistical analysis of park and refuge data? 21 A Yes. 22 Q Do you recall specifically which statistical 23 analysis or other analysis that was performed by the 24 District or other consultants that you were asked to 25 review? Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 43 1 A Yes. In the ENP area analyses, it was primarily 2 Dr. Walker's time series analysis of ENP data that I was 3 asked to review and then make a proposal for further 4 analyses. 5 In the Loxahatchee, I don't know who performed 6 the analysis, but I think it was primarily District 7 documents that contained the analysis that I was asked to 8 review. That included, I think, a report dated August 9 1991. I guess it would have to have been after that that 10 I reviewed it, and appendix E to the SWIM plan. 11 MR. GARVER: Ms. Court Reporter, can you please 12 mark this as McClave No. 1. 13 (The document last above referred to was marked 14 for identification as McClave Deposition Exhibit No. 1.) 15 BY MR. GARVER: 16 Q Dr. McClave, I'm handing you what's been marked 17 as Exhibit McClave No. 1. 18 Do you recognize this document? 19 A Yes. 20 Q What is it? 21 A This is a proposal for analyses to be conducted, 22 by KBN and Info Tech, on the Everglades National Park 23 water quality data that were analyzed by Dr. Walker. 24 Q And this document is dated August 6th, 1991; is 25 that correct? Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 44 1 A Yes. 2 Q Were you involved in the preparation of this 3 proposal? 4 A Yes. 5 Q Was anybody else involved in the preparation of 6 the proposal? 7 A Yes. 8 Q Who else? 9 A Dr. Pollman, and possibly others at KBN, but I 10 only know of Dr. Pollman. And at Info Tech, an assistant 11 consultant that works with me, by the name of Cindy 12 Hewitt, H-e-w-i-t-t. 13 Q What portions of this proposal did you prepare, 14 Dr. McClave? 15 A I prepared the tasks -- In general, I prepared 16 the tasks that involved the statistics and database 17 instructions. So task number one, database construction. 18 And task number two, descriptive analyses, I prepared 19 that. Task three, time series analyses, I prepared that. 20 So I would have been involved in -- Most of this 21 is my work. 22 Q How about on the second page, the section 23 entitled General Comments And Cost Estimates? Was that 24 something you also prepared? 25 A Yes. I either wrote it or reviewed it. Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 45 1 Q Prior to preparing this proposal, had you -- had 2 you done any work, commenced work on evaluating any of the 3 documents you mentioned earlier, being Dr. Walker's 4 analysis of park data, or District documents involving 5 analysis of refuge or park data? 6 A I specifically remember having reviewed Walker's 7 -- Dr. Walker's analysis prior to preparing this 8 proposal. I don't know whether - I don't think this 9 proposal involves any District analyses, so I don't 10 remember whether my review of those documents was before 11 or after this. 12 Q Okay. How much of the work that's described in 13 the first two pages of Exhibit No. 1 has actually been 14 performed at this time? 15 A I would say, as far as this proposal is 16 concerned, none of it has been performed at this time. 17 I have prepared an affidavit, but not as a 18 result of analysis with regard to this proposal. So I'm 19 referring to 4B where it says "Provide affidavits." I 20 mean that -- what I meant by that was after doing this 21 analysis, and I have not done this analysis yet. 22 Q When you say "this analysis," what analysis are 23 you talking about? 24 A The analyses that are described in tasks two and 25 three. Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 46 1 Q Why has that analysis not been performed? 2 A The primary reason is that the data involved in 3 these analyses, insofar as we know, has not been provided 4 in an analyzable form. 5 We have received data in various forms, hard 6 copy and many, many PC diskettes that includes data that 7 might be involved in Dr. Walker's analyses. But it's been 8 our belief there must be a database that he utilized to do 9 his analysis and, at least that I know of, we have not 10 been - or Info Tech has not been provided that database. 11 It was our view we would spend an awful lot of 12 time and probably wasted effort reconstructing a database 13 that may or may not end up being the same one he used. So 14 we are still awaiting the provision of those data. 15 Q When you say the data -- that data has not been 16 provided in analyzable form, why is the data that has been 17 provided not analyzable? 18 A What I mean by that is, for example, we use 19 something called SAS, the Statistical Analysis System, and 20 SAS has certain data -- certain formats for its databases 21 that are readable by SAS. 22 Every statistical package has a similar 23 characteristic that the data reside in a particular format 24 that's utilized by that package. That, as opposed to just 25 raw data, data that -- records might include, for example, Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 47 1 all sorts of water quality parameters in addition to, say, 2 phosphorus, if that's the one we want to look at. So data 3 in sort of a much -- I was going to use the word "random" 4 and I don't mean random, but a much more loose form, free 5 form format that may or may not be complete, as well. 6 And the intent of this, as the proposal 7 indicates, is to analyze exactly the same data that were 8 analyzed by Walker. And so, again, it's been my feeling 9 that, since the database has to exist in a statistically 10 analyzable format, that there's going to be a -- if we 11 can't get it in that format, there's going to be an awful 12 lot of time an effort spent, and not necessarily 13 successfully spent, in trying to reconstruct -- trying to 14 put the data into that format, trying to pick the data out 15 of the free form format, if indeed we have it in that 16 format, and put it into a statistically analyzable form. 17 There's a step there. 18 Q Is the specific problem that you're talking 19 about that you don't have Dr. Walker's format for the data 20 and compilation of data, so that you can't do exactly -- 21 you can't work with exactly the same thing that he worked 22 with? Is that the basic problem? 23 A That's exactly the basic problem, yes. In other 24 words, I am -- it's my understanding his deposition has 25 not been taken yet, or at least I haven't read it, if it Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 48 1 has. And it's my supposition at this point that, if 2 asked, he would say: Yes, I used so-and-so statistical 3 package, and the data exists in the following format for 4 that package. 5 I don't know what the answers to those questions 6 are, and they're very important answers to me as a 7 statistical analyst. 8 Q On page two of Exhibit No. 1, the first sentence 9 says: 10 "In general the statistical analyses 11 performed by Walker appear to be thoughtful and 12 careful." 13 Is that a sentence that you wrote? 14 A Yes. 15 Q What about Walker's analysis leads you to the 16 conclusion that it was thoughtful and careful? 17 A At that point, this is reminding me that I must 18 have read some of the other analyses at Loxahatchee or 19 some of the District analyses. 20 It's my view, overall, in this whole process of 21 having reviewed five or ten statistical analyses that have 22 been done at one time or another in the Everglades, that 23 many of them use insufficient data, poor data or appear to 24 be done quickly and not in a very sophisticated manner. 25 It was my view that the Walker analysis stood Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 49 1 apart, in the sense that it seemed to have -- and I don't 2 have the data yet, but I know that it utilized a lot of 3 data over a relatively long time period, relative to the 4 other analyses, that is; not necessarily in an absolute 5 sense, but a relatively long time period. It utilized or 6 attempted to utilize -- haven't been able to check it -- 7 some modeling tools that were more than just simple 8 regression analyses, which were done in many of the other 9 analyses. 10 So I was contrasting it here with some of the 11 other analyses that I've seen done by the District, other 12 BMP rule analysis, for example, which I felt to be rather 13 simplistic. 14 This one did not seem to me to be simplistic. 15 It seemed to me that a lot of time had been taken and he 16 attempted to do a careful analysis. 17 Q Other than the BMP rule analysis, do you recall 18 which other analysis you were comparing Walker's analysis 19 to? 20 A Well, at least now, and I think when I wrote 21 this, the Loxahatchee analysis would fall into the area of 22 analyses that I would call simplistic. The analyses that 23 went into the SWIM plan, Appendix E of the SWIM plan. 24 Q The next paragraph states: 25 "Careful" -- excuse me, "A careful time Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 50 1 series analysis of the database constructed by 2 Walker will be difficult and time consuming. 3 However, until such analyses are performed, the 4 effects of the problems identified with the 5 Walker analysis remains speculative." 6 Did you write that section, as well? 7 A Yes. 8 Q Now, you -- I believe you testified a moment ago 9 that the time series analysis has not yet been performed. 10 Is that correct? 11 A I haven't performed a time series analysis. I 12 don't know if somebody else has, but I haven't. 13 Q Okay. Is it safe to conclude, then, or fair to 14 conclude that at this time the effects of problems 15 identified with the Walker analysis remain speculative? 16 A Yes, certainly in a sense -- what I was -- the 17 particular meaning I had with the word "speculative" here 18 is until you do the analysis, you don't know what the 19 results are going to be. Although you might have a hunch 20 or even a strong opinion, you don't know until you do 21 them. And so, in that sense, they're speculative. 22 Q Do you intend -- I'm sorry if I've already 23 stated this, but do you intend to cover -- to conduct a 24 time series analysis of Dr. Walker's data, once you obtain 25 it? Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 51 1 A Yes. 2 Q And do you intend to conduct that analysis in 3 accordance with the proposal in Exhibit 1? 4 A Yes. I can't think of anything that has changed 5 my proposal or would have changed it since August of '91, 6 so yes. 7 THE WITNESS: Short break? 8 MR. GARVER: Yes. 9 (Brief recess.) 10 BY MR. GARVER: 11 Q All right. Dr. McClave, we were just talking 12 about the proposal that you did in August of 1991. 13 I want to go back to -- to get back on track 14 with the chronology, you stated that you had -- I believe 15 you stated that the first thing you did with respect to 16 statistical analysis of park and refuge data was to review 17 work done by others. 18 Have you completed that task? 19 A The review of work by others task? Is that what 20 you're asking? 21 Q Yes. 22 A No. I think I mentioned before there was a 23 report, a set of reports, but one I remember in particular 24 by Dr. Walker, on the settling velocity issue, that I've 25 just -- that has just been made available to me. And I Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 52 1 haven't reviewed -- I've been told that I will probably be 2 asked to review it, but I haven't done it yet. 3 Q Okay. Speaking specifically now about 4 Dr. Walker's report on the water quality trends in the 5 park and the District's analysis and District documents of 6 data in Loxahatchee and in the park, including Appendix E, 7 have you concluded your review and evaluation of those 8 documents? 9 A I certainly have, as far as the Loxahatchee 10 documents are concerned. 11 I would say that part of this proposal is 12 certainly going to require continued review and renewed 13 review, if you will, statistical review of Dr. 14 Walker's work. So I would say that's not complete. 15 But the work in the Loxahatchee, as far as I 16 know, is at this point. 17 Q You have conducted some kind of an evaluation of 18 Dr. Walker's report on water quality trends; is that 19 correct? 20 A Oh, yes. I did a review and evaluation of that, 21 in order to prepare this proposal, and in order to prepare 22 an affidavit subsequent to this proposal. All I'm saying 23 to you is, once I get into the data analysis, I'm sure I'm 24 going to have to look a lot more deeply into that report. 25 Q Okay. The review and evaluation that you Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 53 1 already have conducted of Dr. Walker's report, during what 2 time period did you conduct that review and evaluation? 3 A I'm going now by the date of this proposal, 4 which is August '91, so I would assume it would have been 5 done in the summer of '91. 6 Q And was it concluded upon the completion of your 7 affidavit? Is that fair to say? 8 A That's fair to say, yes. 9 MR. PERKO: Just for the record, I want to make 10 clear that the review, the initial review -- you're 11 speaking of the initial review of the Walker trends 12 analysis, correct? With the understanding that 13 additional analysis may be performed in accordance 14 with this proposal? 15 THE WITNESS: Yes. I thought that was what the 16 question was about, the initial review. 17 MR. GARVER: That was my -- right. 18 BY MR. GARVER: 19 Q Of the review and evaluation you've conducted of 20 Dr. Walker's report, and I'm now talking about the period 21 leading up to your affidavit, what did you do in order to 22 evaluate his report? 23 A I read it several times; thought about it; 24 probably consulted some time series -- or not probably, I 25 did consult some time series textbooks and research Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 54 1 articles, probably. If I didn't consult them I at least 2 tried to recall what they were about. 3 Those are probably the main tasks that were 4 involved in that review. 5 Q Do you recall what textbooks you consulted? 6 A I always consult one by Fraser, F-r-a-s-e-r. I 7 think the name is something more exciting than Time Series 8 Analysis, but I don't recall the exact title. That's one 9 I've used and taught from as a reference from time to 10 time. 11 I may have looked at the work by Box and 12 Jenkins. Again I don't remember what the title is, 13 something to do with time series analysis. 14 Those two I recall specifically. There may have 15 been others. 16 Q Do you recall any particular portion of the 17 Fraser text that you mentioned that you were consulting, 18 in reviewing Walker's report? 19 A I don't. It would have had to do with time 20 series modeling, but I don't remember what chapters or 21 sections that I looked at. 22 Q Did anyone assist you in conducting your 23 evaluation of Walker's report leading up to your 24 affidavit? 25 A Yes. Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 55 1 Q Who was that? 2 A Ms. Hewitt, again, would have been assisting me 3 in that review. 4 Q Okay. What is Ms. Hewitt's position? 5 A She is a senior consultant with Info Tech. 6 Q Are you familiar about her educational 7 background? 8 A Yes. She has a Master's in statistics. 9 Q And do you recall where she got her Master's 10 from? 11 A She got her Master's degree from the University 12 of Florida. 13 Q And do you recall when she got her Master's? 14 A No, I don't know. 15 Q How long has she been working at Info Tech? 16 A Approximately ten years; probably a little less 17 than ten years, but almost ten years. 18 Q Was there anyone else at Info Tech, at the 19 professional level, who assisted in the evaluation of Dr. 20 Walker's report during that time period in 1991? 21 A Not that I recall. 22 Q Did you prepare any other documentation, other 23 than your affidavit, in connection with your review and 24 evaluation of Dr. Walker's report in the summer and fall 25 of 1991? Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 56 1 A I probably prepared -- I don't remember if I did 2 it in written form, or just on my computer, but I usually 3 keep notes as I'm reading through a report of things to 4 do, or thoughts as they occur to me. So I probably had 5 some notes prior to actually writing the affidavit. 6 Q Okay. Prior -- did you produce those notes in 7 connection with this deposition? 8 A I don't know. If they were in written form, I 9 did. If they were on the computer, they may not have even 10 been kept. In other words, often what I'll do is, because 11 of limitation of disk space, if I produce a final 12 document, I'll eliminate the notes that led up to that 13 final document. And I don't remember. 14 Q When you were producing documents in connection 15 with this deposition, did you check your computer files -- 16 A (Nods head.) 17 Q -- for those sorts of documents, and produce 18 them if they existed? 19 A Yes. 20 Q If I recall correctly, the other documents you 21 said you reviewed in connection with Everglades water 22 quality issues were District documents relating to 23 analysis of refuge and park data, including Appendix E of 24 the SWIM plan; is that correct? 25 A Yes. Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 57 1 Q Do you recall specifically which documents, 2 other than Appendix E, you reviewed? 3 A The one I remember that was very similar to 4 Appendix E, and even parts of it I think were just copied, 5 I think was the August 19, 1991 report, that had the 6 Loxahatchee analyses as well as several others contained 7 in it. 8 Q Was this August 19, 1991 report, did it appear 9 to be similar to a draft of Appendix E? 10 A I guess I would call it more statistical support 11 or data analysis in support of Appendix E. As I recall, 12 Appendix E had more nonstatistical element to it, in 13 addition to the statistics, whereas, at least as I recall 14 the August 19th report, it was primarily data and 15 statistical analysis. 16 Q And that August 19, 1991 report came from the 17 South Florida Water Management District; is that right? 18 A I think so, yes. 19 Q Were there any other documents you recall in 20 connection with your review of park and refuge data? 21 A Those are the ones I remember reviewing. There 22 have been other documents, but those are the ones I 23 remember carefully reviewing. 24 Q Okay. Do you recall what time period -- what 25 the time period was during which you conducted the review Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 58 1 of Appendix E in that August 1991 report? 2 A I believe that those reviews began at sometime 3 after I wrote the affidavit on Dr. Walker's ENP analysis, 4 and continued -- continued in some manner up to the 5 present, or up to the time when I turned documents over; 6 would have been completed only in the last month or two. 7 Q But that review is -- and evaluation is complete 8 at the present time; is that correct? 9 A As far as I'm concerned, it is. Obviously there 10 may be -- as a result of this depositions, may be more 11 data supplied in the Loxahatchee. 12 One of the concerns I have that I'm sure you're 13 going to ask about at some point is that the data are of 14 poor quality and not very complete and so on. If 15 something is done about that, then I may do more work. 16 But as far as the data as they are now, I don't anticipate 17 doing any more work on them. 18 Q Your affidavit discusses some aspect of the 19 derivation of phosphorus limits and levels in the park and 20 refuge, doesn't it? If you recall? 21 A I'm sorry, I don't understand your question. 22 Q Your affidavit includes an evaluation and review 23 of not only Dr. Walker's report but also other reports, 24 doesn't it, if you recall? 25 A I don't recall. I remember it being primarily Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 59 1 about Dr. Walker's. I may, by then, have read -- at least 2 read the other analyses and made some comment on them, now 3 that you mention it, but what I recall primarily in that 4 affidavit is Dr. Walker's -- a review of Dr. Walker's 5 analyses. 6 Let me put it this way. If it does mention the 7 other analyses, again, it was at a preliminary stage, 8 before I had actually done any analyses of my own. 9 Q Okay. What did you do in order to review and 10 evaluate Appendix E in the August 1991 report? 11 A Began, as before, by reading them, trying to 12 understand what was done, and then there were not very 13 many data, and the data were presumably listed in Appendix 14 E, supposedly listed in Appendix E, that were used in the 15 development of the limits, and so we input those into the 16 computer and did a large number of analyses on those data, 17 both to try to reproduce the analysis of the District and 18 to do some of our own. 19 Q Do you recall, in general, without going into 20 detail, what types of analysis you conducted on your own, 21 other than trying to reproduce the results in Appendix E? 22 A In general, to determine whether using a 23 modeling approach that Info Tech might use, or might 24 adopt, a modeling approach of my own, how that might 25 differ from the District's; also to determine whether the Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 60 1 phosphorus inputs from stations S5 and S6 were 2 statistically related to phosphorus levels in the 3 Loxahatchee was another analysis that I undertook that at 4 least in Appendix E and the analyses that I had read had 5 not been undertaken by the District. 6 Q With respect to developing your own modeling 7 approach to the data in Appendix E, what specific analyses 8 or process did you apply and conduct in that task? 9 A Appendix E attempts to develop what I call a 10 time and stage model, a model relating phosphorus 11 concentrations to time and to stage levels in the 12 Loxahatchee. 13 I began with that as a constraint, that I would 14 relate and try to develop a model that involved time and 15 stage, but I had certain reservations about the approach 16 taken by the District, and -- I guess I will describe it 17 as somewhat simplistic in my view, and I wanted to see 18 whether a more sophisticated or complete or statistical 19 modeling approach would lead to similar results or not. 20 Q And what specifically did you do to try and find 21 out if there was a more sophisticated technique available? 22 A Fed a large number of models, including not just 23 straight line or linear models but also second order or 24 curvilinear models into those data; explored the 25 relationship between time and stage. The District had Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 61 1 viewed the time and stage effects as separate. I looked 2 at what we called interaction between time and stage as 3 well. 4 The District had taken a what I consider to be a 5 very subjective approach to the removal of outliers or 6 anomalous observations. I looked at a statistical 7 approach to determining outliers. The District had 8 removed two complete dates, sampling dates, due to I 9 believe what they described as stages being too low. I 10 did not do that, I included those data. Similarly, they 11 removed, I think, two stations, stations one and two, as I 12 recall, for similar reasons, I guess, because they had a 13 number of outliers -- or, no, I'm sorry, because they had 14 a lot of missing values, a lot of dates on which the 15 values were missing. And again I wanted to look at the 16 entire data set, and did so. 17 Q And what did you do, in order -- in attempting 18 to reproduce the analysis that was reflected in Appendix 19 E? 20 A I simply put the data in as -- and tried to do 21 exactly the analysis that's described in Appendix E, and 22 then - they reported the results, and compared my results 23 to the reported results. 24 Q Were you successful in reproducing the results 25 in Appendix E? Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 62 1 A Not exactly. Very close, but it's my feeling 2 that it may be one or two data points that perhaps were 3 misprinted in there, or there's some kind of rounding that 4 done in one case. We did no rounding, just did it all in 5 the computer. I don't know whether the District did some 6 rounding at some point, but -- Very close but not exact, 7 no. 8 Q Okay. In doing your own separate analyses, 9 different than what was done in Appendix E, did you use 10 any data other than the data that were listed in Appendix 11 E? 12 A Yes. I mentioned S5 and S6 concentrations, for 13 example. In those analyses, we also used some rainfall 14 data at S5, used some flow data at S5. 15 As far as inside the Loxahatchee, I think the 16 answer to your question is no. But in the S5 and S6 area, 17 we did use data that at least I didn't see in Appendix E. 18 It may be there, but we didn't get it from Appendix E. 19 Q Okay. Where did you obtain the additional data 20 from? 21 A The District. 22 Q And all of the data you got from the District; 23 is that correct? 24 A I beg you pardon. I think there also were some 25 water quality analyses. I misspoke a minute ago. There Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 63 1 were some other, additional. 2 In a report by Maffei, M-a-f-f-e-i, I think, 3 there were some additional water quality data for those 4 same stations. When I say "additional," they were 5 presumably the same samples, the same dates and same 6 samples and same stations, but instead of just reporting 7 phosphorus there were calcium and chloride and sodium, 8 cation and anion, et cetera, a number of water quality 9 parameters. That data came from Maffei. We did this 10 analysis of those water quality data as well. 11 So there were other data inside the Loxahatchee 12 that we analyzed. 13 Q So when you say the same stations, you're 14 talking about the internal stations in Loxahatchee; is 15 that correct? 16 A The sixteen internal stations. I think they're 17 numbered CA1 through CA16. That's what I'm talking about 18 when I say the stations. 19 Q Did anyone assist you in conducting your 20 analysis of Appendix E? 21 A Yes. 22 Q Who? 23 A Ms. Hewitt. 24 Q Did anybody else at a professional level? 25 A She may have utilized programmers whom she Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 64 1 supervises. I don't recall whether she did or not. I'm 2 not sure I know whether she did or not. I think she did 3 most of it herself. 4 Q Did you do any of this work in conjunction with 5 anybody from KBN? 6 A Yes. All of this work was done in conjunction 7 with Dr. Curt Pollman at KBN. When I say "all of this 8 work," I mean he was aware at each stage of what we were 9 doing and, if we had questions that involved environmental 10 science, he was the individual that we relied upon. 11 Q Do you recall any questions regarding 12 environmental science that you turned over to Dr. Pollman 13 about -- during the course of conducting this evaluation? 14 A Sure. One of the things that I can remember as 15 a specific example is when we got the Maffei data, the 16 stations did not, in all cases, line up with the stations 17 as reported in Appendix E. 18 In other words, we could find all the phosphorus 19 levels or values, but what might be reported as station 5 20 value on a particular date, might have been under station 21 6 in a phase report. So we would have consulted with Dr. 22 Pollman to try to see if he could help figure out why the 23 two data sets didn't seem to agree. That would be one 24 example. 25 He certainly supplied any of the physical or Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 65 1 chemical science formulas for calculations in calculating 2 things like conductivity, or cation, anion calculations, 3 and so on. So anytime there was any kind of physical or 4 chemical formula to be utilized, he was the one that 5 supplied it to us. So those would be examples. 6 Q Were you successful with respect to the problem 7 in shifting columns? Were you successful in resolving any 8 of those discrepancies? 9 A Well, no, not really. I mean we made the 10 assumption that Appendix E was correct for the time being, 11 and that if we had to shift columns in the Maffei 12 document, we did so. 13 So to this day we -- it's one of the concerns I 14 have about the quality of the data that are utilized in 15 Appendix E. It did not seem to be in agreement with the 16 only other document that I know of that represents those 17 same data in them. 18 Q Did your analysis or evaluation of Appendix E 19 include analysis of chemical parameters other than 20 phosphorus? 21 A Yes. 22 Q What was your objective in looking at parameters 23 other than phosphorus? 24 A I had made it clear to Dr. Pollman and the 25 others involved in this case, the attorneys and so on, Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 66 1 that I had grave reservations about the quality of the 2 data in Appendix E. And Dr. Pollman's -- I think it was 3 Dr. Pollman's idea was one of the ways we could check 4 quality was to look at the other parameters and see if 5 certain physical and chemical relationships appeared to be 6 satisfied. 7 That's what led to our trying to get those data 8 and subsequently getting the Maffei data, and then doing a 9 number of analyses, under Dr. Pollman's direction, to 10 determine whether these various relationships -- I 11 remember graphs of anions versus cations, and sodium 12 versus chloride and a number of others, to try to test the 13 quality, the internal quality of these data. 14 Q Do you recall any of the conclusions from those 15 quality tests on the data, using other chemical 16 parameters? 17 A Yes. 18 Q What were those conclusions? 19 A That the data are extremely poor, that these 20 relationships that are supposed to -- as I understand it, 21 are supposed to hold up if the data are to be relied upon 22 are not holding up; that there are some very serious 23 problems with the data. 24 Q Do you recall specifically what those problems 25 are? Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 67 1 A Again, specifics, I do remember the sodium and I 2 think it's sodium and chloride relationship not looking -- 3 not falling within the bounds that Dr. Pollman, at least, 4 thought they ought to. I think an anion/cation -- again, 5 graph that didn't fall within the bounds. And there were 6 several others, but I don't recall. 7 I'm not familiar with that science and did not 8 get into it other than to note that there was -- that he 9 was telling me that the relationships ought to ought to be 10 within fairly tight bounds, and they were not. They 11 looked very random, so from a purely statistical point of 12 view, looking at the data, I understood what he meant when 13 he said the quality were poor. 14 Q How specifically did you coordinate your 15 analysis involving other chemical parameters with 16 Dr. Pollman? 17 A How -- I'm sorry, how specifically, is that what 18 you asked me? 19 Q Specifically, how did you coordinate your 20 analysis? 21 A That was strictly under the direction of 22 Dr. Pollman. We had the database an he would say to us: 23 please calculate the following parameters and then plot Y 24 versus X, whatever Y an X would be, or print out a listing 25 of the following parameters for me over time so that I can Coffee & Vincent Gainesville, Florida 32601 (904) 373-4374 68 1 look at it. Most of them, as I recall, were plots, 2 however, of the data. And so he directed all of that. 3 Q So -- 4 A That involved no statistics, per se. It was 5 just doing some computations and plots. 6 Q Am I correct in assuming, then, that the type of 7 work you did in conjunction with Dr. Pollman, using other 8 chemical parameters was, in a statistical sense, fairly 9 simplistic? 10 A In a statistical sense it was, yes. That's 11 right. 12 Q Just for purposes of review, at the present time 13 your ongoing -- what are the ongoing projects you have 14 relating to water quality in the Everglades. 15 I believe you mentioned the settling velocities 16 work, which hasn't really begun. Is that correct? 17 A That's co