concurring in part and dissenting in part.
¶65 I concur with the Court’s statement of the standard of review and with its resolution of Issues One, Three and Four. I dissent from the Court’s decision with respect to Issue Two because, in my view, both the Water Court and now this Court have misapplied the standards of review applicable to the Water Master’s findings.
¶66 The historic capacity of the Thomas Ditch flume was an important issue in this case because, as noted by the Court, “the capacity of the diversion” measures an appropriator’s water right when his needs exceed that capacity. Opinion, ¶ 55; see also McDonald v. State, 220 Mont. 519, 530, 722 P.2d 598, 605 (1986). The Water Court found error on the basis of its conclusion that the Master “incorporated flume dimensions in his calculations [of flume capacity] from a later version of the flume, not the dimensions of the original flume” as recorded in the 1912 Bestor report.
¶67 The Water Court agreed with the Master, and this Court has affirmed, that the historic documents prepared by Pondera were appropriately admitted despite the objection that they were prepared in anticipation of litigation of the water rights on Dupuyer Creek. Opinion, ¶¶ 35-36. As the Court observes (Opinion, ¶ 33), the Master recognized the self-serving nature of the documents and their “tendency to minimize competing appropriations.” Because of the documents’ origin, the Master determined that their self-serving characteristics were a matter of “weight or credibility.” The Water Court did not disturb this determination and this Court has affirmed the Water Court’s ruling.
¶68 The Master’s findings reflect that he took Pondera’s measurements into consideration but did not accept the varying measurements at face value in assessing the flume’s capacity. His decision finds substantial support in the evidence. Experts on both sides acknowledged that it is impossible to be certain what the size of the flume was during the period in question, given that the only available information was very old and open to multiple interpretations. Although the Pondera evidence showed slight increases in the size of the flume between 1912 and 1922, the Master *347found that these discrepancies “may reflect some increase in the size of the flume or a range of error in the size estimates.” Noting that “1912 is simply the date of the first document referencing the flume,” the Master found no reason to give the 1912 Bestor report more weight than other measurements taken between 1912 and 1931. The Master found that, based on the evidence presented regarding measurements of the flume, a size of twenty-four inches wide by eleven inches deep was reasonable. This number was within the range of measurements found in the Pondera reports taken between 1912 and 1931.
¶69 The Water Court does not have unfettered discretion in reviewing a master’s findings. The reviewing court must regard the master’s findings as “presumptively correct.” Charles Alan Wright & Arthur R. Miller, Federal Practice and Procedure, vol. 9C, § 2585 (3d ed., West 2008). Differences of opinion or interpretation regarding evidentiary issues cannot constitute clear error. We have observed that, “[w]hen a reader reasonably can deduce two or more inferences from the facts, the reviewing court lacks power to substitute its deductions for those of the finder of fact.” Weinheimer Ranch, ¶ 19. The clear error standard “is a deferential one,” requiring a trial court’s findings to be affirmed when its account of the evidence is “plausible in light of the record viewed in its entirety....” Amadeo v. Zant, 486 U.S. 214, 223, 108 S. Ct. 1771, 1777 (1988). Stated more colorfully, a clearly erroneous finding of fact must strike the reviewing court as wrong “with the force of a five-week-old, unrefrigerated dead fish.” Parts & Elec. Motors, Inc. v. Sterling Elec., Inc., 866 F.2d 228, 233 (7th Cir. 1988); U.S. v. Bussell, 504 F.3d 956, 962 (9th Cir. 2007) (where, after quoting this “often repeated” comparison, the Ninth Circuit held, “No such stench lingers here.”).
¶70 The Water Court’s decision that the 1912 Bestor measurements controlled over all others shown in the Pondera data substitutes its view of the evidence for the Master’s determination that the variation in measurements did not necessarily reflect expanding flume capacity. “Although conflicts may exist in the evidence presented, it is the duty of the trial judge to resolve such conflicts.” Desaye, 250 Mont. at 324, 820 P.2d at 1287. Because the Master was acting as the trial court, the Water Court was not allowed to substitute its own view regarding the value of disputed evidence.
¶71 Importantly, both the Water Court and this Court agree with the Master’s decision that only one flume size should be used to quantify the claimants’ rights for the period of time extending from the flume’s first construction to its substantial reconstruction in approximately *3481931. It is incongruous for the Court to then limit the claimants’ water rights to the 1912 measurement on the rationale that the evidence demonstrated that the flume grew in size from 1912 to 1931. Opinion, ¶ 40. If, as a matter of fact, the flume’s capacity steadily was increasing, then there is no basis for refusing to grant the claimants a corresponding increase in their water rights for each documented measurement.
¶72 Instead of looking at all of the evidence available during the 1912 to 1931 period, the Water Court limited its approach to one date: 1912. Although the Water Court accepted the Master’s use of just one flume size for the entire time period, it determined inconsistently that the flume had been expanded during that time period and then awarded the smallest amount of water that could be supported by the 1912-1931 Pondera documents. The Water Court did not acknowledge the Master’s reasoning regarding the relative weight of the evidence and the unreliable nature of the various measurements. This is not a proper application of the clearly erroneous standard of review. Substantial evidence supports the Master’s finding that the 1912 Bestor measurements were not necessarily more accurate than the other measurements reflected in the historic data if only one flume size was to be used to quantify water rights for the entire period.
¶73 The Water Court’s approach to the Master’s determinations regarding the 1931 flume is flawed for similar reasons. The court’s decision to use the inlet-controlled method of calculation relied exclusively on the absence of determinative slope data in 1931, disregarding the fact that every expert witness and most of Pondera’s reports applied Manning’s equation.
¶74 To address the variability of the data, the claimants’ expert, Brian Casne, was asked by counsel to calculate the capacity of the flume using several different values for slope and area. He calculated the capacity using both the inlet- and outlet-controlled methods, although he used inlet when he thought the flume was only 120 feet long and switched to using Manning’s equation exclusively after encountering records indicating that the flume was much longer — one record measured the flume at 400 feet in 1912 and another at 337 feet in 1921.
¶75 At one point, the Master asked Casne why the length “is such a factor here.” Casne then explained how the length impacts the hydraulics of a flume and therefore significantly affects which formula gives an accurate result. He explained that he used a computer to test which method is appropriate based on the data. He noted that the *349flume must be very long if a slope from the record of0.01007 was used (a slope factor acknowledged by all parties to be mistaken), and this prompted his use of other slope data from the record, including 0.00157 and the 0.003 slope used by the Master.
¶76 Even Pondera’s expert, Anderson, admitted that “the Manning’s result and the inlet controls are about the same for a fairly flat surface. And this is a fairly flat slope analysis.” In his September 13, 2010 report to Pondera, Anderson provided his estimate for the flume’s capacity using only Manning’s equation without any mention of the inlet-controlled method. During the hearing before the Water Master, Anderson was asked whether he believed “that using the Manning’s equation to calculate as you did for the ditch flume capacity, that that was a valid exercise towards giving us at least some type of estimate for what a flume could carry?” He replied, “Yeah, I mean realize you have to exercise some professional judgment in the matter and if you’re in the more flat slope scenario here, yeah, I don’t get too worried about looking at a Manning’s versus an inlet control. If it’s on a steeper slope, [then] I have some — I have a different opinion for you.” Anderson acknowledged that, in forming his opinion, he relied on Manning’s equation.
¶77 The Water Court ignored evidence in the record that the flume was of sufficient length to apply Manning’s equation. In considering this question, the court was obligated to give appropriate deference to the Master’s decision based on evidence of the flume’s length and on the fact that “both [experts] stated calculating flume capacity for an outlet controlled flume was acceptable.” “[W]hen a finding is attacked as being unsupported, the power of the appellate court begins and ends with a determination as to whether, considering the whole record, there is substantial evidence which supports the conclusion reached by the trier of fact ”Fegles Const. Co. v. McLaughlin Const. Co., 205 F.2d 637, 639 (9th Cir. 1953) (cited approvingly in Weinheimer Ranch, ¶ 19). This Court’s focus on the slope variable (Opinion, ¶ 41) misses the larger point: evidence in the record supported the finding that Manning’s equation was the correct test to apply based on the length of the flume. Given the testimony of the experts, simply because the historical data left gaps in the available information is not a basis for finding clear error in the Master’s resolution of the conflicts. See e.g. Koeppen v. Bolich, 2003 MT 313, ¶¶ 49-53, 318 Mont. 240, 79 P.3d 1100 (in the face of conflicting testimony and historical evidence, “it is within the province of the trier of fact to weigh the evidence and determine the credibility of witnesses; we will not substitute our *350judgment for that of the trier of fact on such matters.”) (citation omitted). The Master’s calculation was “plausible in light of the record viewed in its entirety....” Amadeo, 486 U.S. at 223, 108 S. Ct. at 1777. The standard of review requires that we respect the Master’s judgment.
¶78 In summary, the most that may be said is that the evidence supported different conclusions. That is not enough to demonstrate clear error. DeSaye, 250 Mont. at 324, 820 P.2d at 1287. The Master’s findings are based on substantial evidence, as the record contains “relevant evidence which a reasonable mind might accept as adequate to support a conclusion.” Montanans v. State, 2006 MT 277, ¶ 79, 334 Mont. 237, 146 P.3d 759 (quotation omitted). The Master did not misapprehend the effect of that evidence and the record does not support the Water Court’s conclusion that a mistake was committed. I would reinstate the Master’s findings with respect to the historic flume capacities and remand for recalculation of the parties’ water rights.