Simmons v. BOARD OF COM'RS, BOSSIER LEVEE DISTRICT , 624 So. 2d 935 ( 1993 )


Menu:
  • 624 So.2d 935 (1993)

    Kenneth Don SIMMONS, et al.
    v.
    BOARD OF COMMISSIONERS OF the BOSSIER LEVEE DISTRICT, et al.
    Victor E. GROSJEAN, et al.
    v.
    BOARD OF COMMISSIONERS OF the BOSSIER LEVEE DISTRICT, et al.

    Nos. 25,093-CA, 25,094-CA.

    Court of Appeal of Louisiana, Second Circuit.

    September 22, 1993.

    *938 F.Q. Hood, Bossier City, for Bossier Levee Dist.

    William J. Doran, Jr., Norman L. Sisson & Sharon F. Lyles, Baton Rouge, for State of Louisiana, D.O.T.D.

    Sam N. Gregorio, Shreveport, James T. Adams, Bossier City, for Simmons, et al.

    Booth, Lockard, Politz, LeSage & D'Anna, Shreveport by Nyle A. Politz, for Grosjean, et al.

    Before MARVIN, LINDSAY and VICTORY, JJ.

    MARVIN, Chief Judge.

    Defendants, the State of Louisiana, DOTD, and the Board of Commissioners of the Bossier Levee District (Board) appeal a judgment totaling about $1.75 million that was awarded to the 33 plaintiff owners of residential lots in Bossier Parish, for property loss and damage caused by the defendants' dredging of drainage canals along the rear of plaintiffs' lots that began in the fall of 1985 and continued until early 1986.

    We amend and affirm.

    Within a day or so after the dredging occurred behind each respective lot, the canal banks, consisting of wooded land, developed crevices and dropped off in large sections toward the water in the canal. The banks continued to crack and drop off intermittently, albeit less dramatically thereafter, and had not completely stabilized at the time of trial in May 1992, more than six years after the dredging ceased.

    Plaintiffs dismissed other named defendants, including the dredging contractor and the individual Board members, shortly before trial. The State stipulated liability under various provisions of the state constitution, including Art. 1, § 4 (compensation for taking or damaging private property for a public purpose), and under CC Art. 667 (strict liability for damage to neighboring property). The Board stipulated only to Art. 667 strict liability, reserving its right to have the court determine whether it was liable on constitutional grounds.

    The trial court found that the land and trees actually lost when the canal banks failed were "taken" in the constitutional sense, and that the taking caused and will continue to cause severance damages to the remaining property. The court awarded compensation for the value of the land and trees lost, mental anguish damages, and severance damages for the present and expected future damage to the remaining property.

    The court awarded attorney fees under LRS 13:5111 in the amount of one-third of the awards for land and tree loss, exclusive *939 of mental anguish and severance damages. The court's awards, including attorney fees, range from about $37,000-$214,000 per lot, averaging about $103,000 per lot.

    The court assessed legal interest from date of judicial demand and expert witness fees for plaintiffs' experts in amounts less than the experts' total charges.

    Defendants appealed, seeking a reduction in the awards for actual losses, severance damages and expert witness fees, and a reversal of the awards for mental anguish and attorney fees.

    Plaintiffs answered the appeal, claiming that the attorney fee awards should be increased to one-third of their entire recovery, including severance damages, and that legal interest should run from the date of the taking. Plaintiffs also seek an increase in the expert witness fees and ask that legal interest be awarded on all costs, including expert witness and attorney fees, from the date of the judgment fixing the fees. In their appellate brief or in oral argument, plaintiffs have expressly abandoned their claims for an increase in the damage awards.

    FACTS

    Plaintiffs own 19 of 24 contiguous lots in the Hickory Ridge Estates and Carriage Oaks East subdivisions. The Grosjean and Doucet plaintiffs each own two adjoining lots, using one for their homes and the other for recreational structures. For simplicity, we regard plaintiffs as owning 17 lots, as did the litigants and the expert witnesses in the trial court.

    Plaintiffs' lots front on Hickory Ridge Drive, at the eastern and more rural end of Bossier City. The lots are bounded in the rear by Flat River and Benoit Bayou, separate bodies of water which effectively form a single stream behind plaintiffs' property, notwithstanding that Flat River flows in a southerly direction and Benoit Bayou (sometimes called Macks Bayou) flows northerly. The streams converge about 700 feet from the southernmost line of plaintiffs' property, with Flat River continuing to run to the east. The Coy Road Bridge crosses Flat River at the northernmost end of plaintiffs' property, while the Hickory Ridge Bridge crosses Benoit Bayou to the south. This sketch, not to scale, depicts the general area:

    *940

    *941 The lots are large, originally measuring about 100 × 300 feet in Carriage Oaks East and about 240 × 420 feet in Hickory Ridge Estates. The rear boundary of each lot is the center line of the stream. The subdivision plats show a 100-foot Board servitude across the rear of the lots.

    Before the dredging occurred, the lots sloped gently and gradually toward the water. The rear of each lot was covered with trees that blocked plaintiffs' view of the opposite bank, which was undeveloped. Plaintiffs testified that they bought the property because of the privacy and natural beauty it offered, allowing them to enjoy the tranquility of country living while still being near city conveniences.

    DREDGING OPERATIONS

    The dredging behind plaintiffs' property was done by the Board, under the State's supervision, as part of a project to improve drainage in other sections of Bossier Parish. The bank slopes were relatively flat before the dredging began, averaging about 5 feet horizontally to one foot vertically, designated as a 5:1 slope ratio. The project plans called for a much steeper slope of 2:1 after excavation. The State did not take soil samples to determine the characteristics of the soil before the work began on Flat River, downstream and to the east of plaintiffs' property.

    In early October 1985, as the dredging moved upstream on Flat River toward plaintiffs' property, the contractor, Guinn Brothers, Inc., began having difficulty achieving the channel depth called for by the project plans. Guinn's principal, George Guinn, Jr., testified that the soil became sandy or "soupy" about four or five feet before the bottom grade called for by the plans was reached, and that continued digging caused some relatively minor bank failures without increasing the channel depth. Guinn immediately reported the soil problem to the State's engineers, both orally and in writing, and sought permission to deviate from the project plans in the sandy areas to prevent additional bank failures. The State instructed Guinn follow the plans.

    The more severe bank failures began on or about October 25, 1985, the day after the excavation reached the rear of plaintiff Parker's property, at the confluence of Flat River and Benoit Bayou. The canal banks were intact when the dredging stopped around 11:00 p.m. on October 24. By 7:00 the next morning, the ground near the stream had cracked and dropped off or slid about 8-12 feet toward the stream, leaving many crevices across the rear of the lot, the largest of which measured about three feet in width and about six feet in depth. Guinn immediately notified the State's engineers, one of whom visited the site and commented that the damage "was still within the Board's [100-foot] easement." The State's supervising engineer again told Guinn to "keep digging" according to the project plans, saying the State would "determine procedures to repair the slide area" later.

    The dredging, and the immediate bank failures, continued as the work progressed upstream on Flat River and then moved to Benoit Bayou. The Coy Road Bridge, which crosses Flat River at the northernmost end of plaintiffs' property, was heavily damaged by the dredging and had to be closed for repairs. The State thereafter instructed Guinn not to dredge in the immediate area of the Hickory Ridge Bridge across Benoit Bayou at the opposite or southernmost end of plaintiffs' property.

    On January 18, 1986, shortly before the dredging reached the area of the Coy Road Bridge on Flat River, the State instructed Guinn to return to several locations downstream, behind plaintiffs' property, to do additional digging, apparently in an attempt to reach the depth called for in the project plans. The additional digging exacerbated the bank failures in those areas and did not achieve the desired depth. When Guinn moved upstream of and beyond plaintiffs' property on Flat River on or about January 23, 1986, the State, for the first time, authorized a change in the project plans and allowed Guinn to raise the channel depth by four feet and to increase the channel width from 40 to 60 feet.

    In early February 1986, Guinn returned to the confluence of Flat River and Benoit Bayou and began dredging upstream toward the *942 Hickory Ridge Bridge on Benoit Bayou, behind some of the plaintiffs' lots. In this area, the State authorized Guinn to raise the channel depth by one foot over the project plans. Guinn testified that this design change "helped, but it didn't solve the problem."

    Throughout the time that the dredging and bank failures occurred behind their homes, the plaintiffs complained to various workers at the site and asked them to stop the work that was obviously damaging plaintiffs' property. The workers told them that were under contract with the State and had been instructed to continue dredging, even after the damage was reported to and seen by State and Board personnel. Several plaintiffs also registered their complaints at Levee Board meetings.

    In February 1986, after being informed that the bank drop-offs were posing hazards, particularly to young children, the Board agreed to erect a "caution" fence along the rear of plaintiffs' property. The bank failures ultimately progressed beyond the original location of the fence on many of the lots, requiring that the fence be moved one or more times.

    INVESTIGATION OF BANK FAILURES

    In June 1986, the Soil Conservation Service of the U.S. Department of Agriculture (SCS) investigated the bank failures and issued a report finding that the dredging probably caused the failures because the type of soil in the area, called overconsolidated or Roebuck or "buckshot" clay, is known to crack and slide when stabilizing pressure is removed from the soil mass, such as when the toe of the bank is cut during excavation, particularly when the soil is exposed to water. As the soil closest to the stream slides away, this removes the stabilizing pressure from the soil behind it, which then also cracks and slides. The successive failures continue until the banks return to their natural, pre-construction slope.

    The SCS report notes that the prediction of slope stability in this type of soil "is one of the more inexact aspects of soil mechanics." The report contains pre-construction cross sections of the banks in certain areas of plaintiffs' property, showing that the natural slope in those areas averaged about five feet horizontally to one foot vertically (5:1), much flatter than the 2:1 slope called for in the project plans.

    The SCS concluded that the project should not have been designed and undertaken without soil testing, and should not have been allowed to continue once the bank failures began. The report states that the failures "endanger improvements in a subdivision," and recommends that the cost of several repair alternatives be explored "so that plans may proceed for remedial treatment."

    In September 1986, at the State's direction, Guinn attempted to repair the failed banks along two of the plaintiffs' lots by resloping and reseeding the banks. Within a week, cracks appeared and the repaired areas ultimately dropped off or slid into the stream. No further repairs were attempted.

    In March 1987, the Board hired a private engineering firm, Demopulos & Ferguson (D & F), to investigate the bank failures and offer solutions. D & F issued a report in July 1987, reaching the same conclusions about the cause of the failures as did SCS. D & F also agreed that the failures would continue until the banks returned to their natural slope, and that the successive failures "will threaten existing homes and other developments and further reduce the channel's hydraulic capacity."

    D & F detailed several repair alternatives, noting that the most economical approach would be to flatten the slopes on the opposite, undeveloped bank to a ratio of 5.5:1 to stabilize them against slope failures and to use some of the soil removed in that operation, with lime added to stabilize it, to rebuild the banks on the developed side of the stream. This approach was intended to shift the channel away from the developed areas. D & F recommended that the Board begin repairs, using any one of the described repair alternatives, "as soon as practicable, ... to avoid further property damage."

    No repairs were made, according to this record.

    *943 PLAINTIFFS' TESTIMONY

    Plaintiffs testified that the banks had continued to crack and slough or drop off, although not as severely as they did initially, as recently as two weeks before trial began in May 1992. Plaintiffs described a period of a year or so starting around 1990 when the failures appeared to have stopped, but said the failures resumed when heavy rains fell in the spring of 1992. Plaintiffs introduced photographs showing the "caved-in" areas of the banks at and since the time of their initial failures, and the gradual movement of the failed areas inland over the years. The sections nearest the water now have steep drop-offs of about 8-10 feet, rather than the gentle, gradual slopes they had before.

    Before 1985, the rear of plaintiffs' lots was densely covered with mature trees that blocked the opposite bank of the stream from view. Plaintiffs could not see or hear traffic on Shed Road, a major thoroughfare running beyond and roughly parallel to the stream. The trees enhanced the natural beauty, privacy and tranquility of the lots, each of which had a large back yard. Plaintiffs mowed their yards to the tree line, and used and maintained their property to the high bank of the stream. Each testified that the presence of the trees on the lots was the primary reason for buying the property.

    The bank failures caused some of the trees on plaintiffs' lots to fall and die. Other trees did not fall immediately but began leaning as the soil supporting their root systems weakened. These trees eventually died. Many of the trees remaining at the time of trial were leaning and were expected to die. Plaintiffs now can see the opposite bank of the stream and can see and hear traffic on Shed Road. Their photographs show the contrast in the density of the trees on their property before and after the dredging operations. Each plaintiff estimated the number of trees lost, which ranged from 25-150 per lot.

    Because of the sharp drop-offs and large cracks in the canal banks, plaintiffs are no longer able to maintain the rear portions of their lots that have become overgrown with high grass around dead trees and limbs. In plaintiffs' words, the area is now "a barren eyesore" and looks "devastated," "like hell," "like a miniature Grand Canyon."

    Each plaintiff described the structural damage that has developed in his or her home after the dredging, such as cracks in windows, walls, fireplaces, and patios, and the damage to other structures such as fences, swimming pools, oxidation ponds and septic systems. Plaintiffs are aware of their legal obligation to disclose property defects to any potential buyer of their property, and fear that their disclosures, as well as the current condition of the property, will prevent them from selling it.

    Plaintiffs said that they were extremely upset, sad and angry over the damage to their property, which caused them to "lose their reason for buying it." They felt particularly angry and frustrated because defendants, ignoring their repeated and numerous complaints, continued the work, oblivious to the obvious damage it was causing, and took virtually no remedial action thereafter.

    EXPERT ENGINEERING TESTIMONY

    The trial judge viewed the property when the trial began and heard expert testimony from two civil engineers, John Cagle for plaintiffs and Mark Morvant for defendants. The experts, who reviewed but were not involved in the prior studies of the bank failures by SCS and D & F, agreed that the banks will continue to seek their natural slope but disagreed about when the banks will stabilize and whether plaintiffs' homes will or may be damaged in the future.

    Cagle is a structural engineer whose work with structures regularly involves the use of soil analysis data furnished by soil testing laboratories. Morvant is employed by the State, DOTD as a geotechnical design engineer; he works exclusively with soil mechanics. Both witnesses were accepted as experts in the field of civil engineering, which encompasses both structural and geotechnical engineering.

    Morvant, age 34 at trial, has worked for DOTD since his graduation from college in 1981 but was not involved in designing the drainage project. He viewed the bank failures in 1987, when the State took soil borings *944 to investigate the failures, but was not asked to study or give an opinion about the failures until two weeks before trial. He then returned to view the failed banks but was denied permission to enter plaintiffs' property.

    Cagle, age 70 at trial, has worked as an engineer since his graduation from college in 1949. He has held his present position with a private engineering firm for 23 years. He visited and inspected each of plaintiffs' lots.

    For each lot, Cagle calculated and plotted on a diagram the natural slope of the banks, the present high bank at the time of trial, and the projected future location of the high bank that Cagle predicted using a mathematical formula based on Cagle's estimate of the soil's unconfined compressive strength and its angle of internal friction. Cagle also plotted the location of the banks using the 5.5:1 slope recommended by D & F to stabilize the opposite bank.

    On portions of about half of the lots, and in both contiguous and noncontiguous places, Cagle's present high bank line was beyond, or farther from the stream than, any of the other projected lines, indicating that the banks in those areas had not yet stabilized at any of the various locations predicted as their maximum failure line. According to Cagle, this illustrated the truth of the statement in the SCS report that predicting slope stability of this type of soil "is one of the more inexact aspects of soil mechanics."

    For the remaining portions of those lots and for the other lots, the maximum failure line was often, but not always, the natural slope line as calculated by Cagle from the pre-construction bank cross-sections in the SCS report and from other pre-construction cross-sections that Cagle obtained from the State. Cagle used the flow line of the stream as his reference point for calculating and plotting the natural slope line, and stood by his calculations even after Morvant testified that the proper reference point was the toe of the slope.

    When Cagle visited the lots shortly before trial, he saw that cracks or fissures had formed beyond the present high bank line on each lot. Some cracks were within a foot of the present bank, while others were as much as ten feet beyond it. Lay and expert testimony was in agreement that the areas where cracks develop will eventually slide or drop off.

    Cagle also obtained soil borings from plaintiffs' property. He testified that they showed the soil at the lower levels of the bank to be very weak, meaning that it could be easily moved under high water conditions and this movement would trigger additional failures at higher levels of the bank.

    Cagle opined that the banks would continue to fail beyond all present and projected lines for each lot on his diagram, regardless of which line was farthest from the stream on the diagram. He based his opinion on these facts: the present high bank line had already exceeded the predicted maximum failure lines on some of the lots; there were cracks beyond the present high bank lines on all lots, indicating future failures would occur in those areas; and the soil borings indicated that the lower levels of the banks could be easily disturbed, causing additional failures.

    Defendants' expert, Morvant, also calculated and plotted the present high bank line and the natural slope of the banks. His calculations of the present high bank differed only slightly from Cagle's. Morvant's natural slope line also differed from Cagle's, perhaps because Morvant used the toe of the slope rather than the stream flow line as his reference point. In some areas Morvant's natural slope line was closer to the stream than Cagle's, while in other areas it was farther from the stream than Cagle's.

    Morvant also plotted the 5.5:1 slope line from the D & F report, which he considered to be the maximum failure line of the banks. Morvant disagreed with the mathematical formula that Cagle used to calculate his projected failure line based on the soil's unconfined compressive strength and its angle of internal friction. According to Morvant, this formula "would not be utilized by a soils engineer ... today ... [because] the slide is just too complex."

    On Morvant's diagram, the present high bank line did not exceed or go beyond the natural slope line or the projected maximum *945 failure line. Morvant opined that any future failures would not exceed the 5.5:1 line, and that they would not pose any danger to the plaintiffs' homes because the backs of the homes were generally about 80-90 feet from the 5.5:1 line, with the closest home being 70 feet from that line and the farthest being 140 feet.

    Morvant examined cross-sections of the bank slopes that were taken just before trial and compared them to cross-sections taken in 1987. He found very little change in the location of the bottom of the slope, from which he concluded that the lower portions of the banks had attained their natural slope and had stabilized. Morvant agreed that failures may continue at the upper levels of the banks until they attain their natural slope, but did not believe that future soil movement, whether from the top or the bottom of the banks, would damage plaintiffs' homes because "all the movement is in the slide itself.... It's just the slide on the slope that's moving[,] not the entire land mass between the house and the slide. That's not moving."

    Morvant conceded that water conditions in the streams could cause some erosion in lower areas of the banks but opined such erosion would cause only "a slight meandering" of the lower banks, "not enough to cause a slide."

    Both Cagle and Morvant calculated the amount of land lost from the rear of plaintiffs' lots. Cagle calculated a "minimum" and a "maximum" loss for each lot. For the "minimum land loss," Cagle measured from the rear property line, or the center of the stream, to the present high bank line. These losses, measured in square feet for each lot, range from about 6,000-35,000 sq. ft. per lot, and total about 300,000 sq. ft. or about seven acres. For the "maximum land loss," Cagle measured from the rear property line to the farthest line from the stream on his diagram, even though he expected the bank failures to continue beyond that line. These losses range from 7,000-60,000 sq. ft. per lot, and total about 378,000 sq. ft. or about 8.75 acres.

    Morvant calculated the land loss as the area between the pre-construction high bank line and his projected maximum failure line. These losses range from about 3,300-13,000 sq. ft. per lot, and total 133,800 sq. ft., or about three acres. Morvant did not measure from the rear property line because that line had not changed, and the land between the pre-1985 high bank and the rear property line in the stream, although owned by plaintiffs, was not usable before the bank failures.

    Cagle inspected each of the plaintiffs' homes and other structures (garages, fences, swimming pools, etc.) for structural damage. In some of the homes, he noted cracks in walls and fireplaces, separation of molding from the walls and ceilings, difficulty in closing inside doors, and cracks in or fogging of windows. Cagle explained that the fogging of the double-pane windows was caused by a break in the thermal seal, which allowed moisture to enter the space between the two panes.

    Outside of some of the homes, Cagle found cracks in and soil subsiding from patios, garages and driveways; a fence pulling away from a house, preventing the gate from closing properly; a swimming pool with cracks in and tiles missing from the apron surrounding it; and an oxidation pond with its back wall caved in.

    For each structural defect that he noted inside and outside the homes, Cagle expressed his opinion whether the defect was caused by the bank failures. He attributed most, but not all, of the defects, to the bank failures, finding that the rest were more likely due to normal settling and shifting that occurs in all homes.

    Cagle opined that the bank failures had already affected most of the homes and would continue to affect them as long as the banks keep moving. He opined that the soil movement will have some effect on every home. For each structure, including homes, swimming pools, sheds, etc., Cagle assessed its probability of future uncontrollable structural damage as either low, moderate or high, based on its proximity to the present high bank line. The "high probability" structures were generally within 50 feet of Cagle's high bank line.

    Cagle opined that the three homes in the "high probability" category (Carroll, Yoblonsky *946 and Harris) would sustain foundation damage that could not be prevented by repair efforts (soil stabilization and addition of "footings" or underpinnings to the foundation).

    For the remaining homes with low or moderate probability of structural damage, Cagle opined that their foundations could be repaired at a cost of $8,000-12,000 each.

    Assuming that the Board servitude over plaintiffs' property exists for 100 feet from the present high bank of the streams, and is not limited to its pre-1985 location, Cagle plotted the location of the servitude from his present high bank line and determined that the servitude line will approach or enter some of plaintiffs' homes. This "adverse servitude" was considered by plaintiffs' appraiser and by the trial court as an element of severance damages.

    QUANTUM TESTIMONY

    Plaintiffs' appraiser, Robert Russell, used Cagle's calculations and assessments to quantify plaintiffs' losses, considering both the actual losses of land, trees, fences, etc., and the expected "severance damage" to the remaining property, including the adverse effect on its marketability arising from the bank failures. Two realtors also testified about the marketability of the property.

    Defendants' appraiser, Mark Montgomery, valued only the actual losses for each plaintiff (land and trees lost, damage to fences, septic lines, etc.), and the future losses that Montgomery deemed "possible," but not "probable," for some plaintiffs. Montgomery's "possible losses" include foundation repair for eight of the 17 lots and repair of fences, swimming pool areas, and barns that are not presently damaged.

    The trial court accepted Russell's opinion about the extent and measure of damages in most respects.

    Russell used the market data approach to calculate the fair market value of plaintiffs' property in 1985, when the dredging began. These values range from $92,000-$274,000 per lot and average about $135,000.

    For plaintiffs with a high probability of future structural damage according to Cagle, Russell assessed their losses as the 1985 market value, plus demolition and moving costs. The trial court used a different method of evaluating these losses, which will be discussed below.

    For plaintiffs whose homes had a low or moderate probability of future structural damage, Russell assessed their losses as including these elements:

    —Value of land lost, using Cagle's square footage calculation of "maximum land loss," at a price of $.45/sq. ft. for the larger lots and $.20/sq. ft. for the smaller lots. Russell opined that the lots were actually worth $.50 and $.22/sq. ft. respectively, but discounted the full value by ten percent because plaintiffs did not have use of the portion of their land between the former high bank and the rear property line in the stream even before the bank failures.

    —Cost to replace some but not all lost trees (20 trees on larger lots, 10 on smaller lots, @ $500). Because so many trees were lost, Russell did not consider it economically feasible for plaintiffs to replace all of them.

    —Value of lost aesthetics, or the effect of the land and tree loss on the value of the remaining property. Russell estimated this loss as 5 percent of the 1985 market value.

    —Value of items actually damaged or with a high probability of future damage, per Cagle (fences, swimming pools, etc.).

    —Value of items with a moderate or low probability of future damage, which Russell valued at 4 percent and 2 percent of the 1985 market value, respectively.

    —Cost to stabilize soil and repair slab, fixed at $5.00/sq. ft. based on Cagle's estimate of $8,000-12,000 per house, which Russell deemed reasonable. Most of thee houses contain about 2,000 sq. ft.

    —"Foundation stigma," which Russell valued at 10 percent of 1985 market value, for the adverse effect on marketability due to the fact that a particular house has had a foundation problem, even if the foundation has been repaired. Russell opined that a potential buyer who sees identical houses, one of which has had a foundation problem and the other of which has not, will buy the *947 one without the problem "a significant percentage of the time."

    Russell valued this loss at 10 percent of market value based on his statistical analysis of the price per square foot in the comparable sales he analyzed to determine the 1985 value of plaintiffs' property. The average price was $54/sq. ft. and the standard deviation of that sample was $4.09, meaning that about 70 percent of the homes will probably be valued at $54 plus or minus $4.09/sq. ft., or from about $50-58/sq. ft. Fifteen percent will probably be valued at more than $58/sq. ft. and another 15 percent at less than $50/sq. ft. The standard deviation figure of $4.09 is about 7.6 percent of the average value of the sample ($54). Russell considered the foundation stigma to be "significant" and explained:

    ... when you get to where you have a stigma or an adjustment you need to make... that you feel is significant, you've got to go beyond that range of 7.6%. I subjectively picked 10% because it is a statistically significant different number ... when you deviate by 10%. And ... for the major subjective adjustments I had to make, I felt like the adjustment needed to be more than 7.6% and 10%, of course, is a round number and that's subjectively how it was arrived at.

    Russell testified that it would have been reasonable for him to have picked any number from 8-15 percent.

    —"Project stigma" was another element of loss itemized by Russell, for the adverse effect on marketability due to the awareness of potential buyers that a specific neighborhood has had foundation and structural problems due to shifting soils. Russell testified that because of this litigation and the press coverage of it, "there's a lot of information floating around about the possibility of things that [may] happen ... [with] any house that backs [up] to the Flat River [or Benoit Bayou]..."

    Russell valued the project stigma at 10 percent of 1985 market value, using the same analysis described above, and considered it a separate element of loss from the foundation stigma.

    —"Adverse servitude," for those lots that would be adversely affected by the proximity of the Board servitude to the homes, per Cagle's diagram and assumption that the servitude runs from the present high bank. Russell valued this element of loss at 5 percent of 1985 market value if the servitude line enters the home and at 2 percent of market value if it approaches but does not enter the home.

    —"Non-compensated amount at risk," for the likelihood that a potential buyer will offer less for property with known defects because of the risk that future damage will be even greater than expected. In other words, property that Cagle estimates as having a low or moderate probability of future structural damage may actually have more damage than Cagle predicted. Russell opined that a buyer would want some type of discount for this potential risk, in addition to a price reduction for the actual or predicted risks.

    For each lot in the low or moderate probability categories, Russell calculated the amount of loss he would assign if the property were in the high probability category (i.e., 1985 market value plus demolition and moving costs). He then subtracted his estimate of the losses as they are predicted to occur under the "low" or "moderate probability" analysis from his "high probability" loss figure. He applied a "present worth factor" of 6 percent to this figure for five years, times 50 percent for "low probability" property, or times 75 percent for "moderate probability" property, estimating this as the amount of discount a potential buyer would expect for the non-compensated amount at risk.

    The trial court found this element of damage to be "overly subjective" and awarded only one-fourth of the amounts calculated by Russell.

    Russell's assessments of "just compensation" for the low and moderate probability properties, considering all the elements of loss he itemized, range from about $76,000-238,000 per lot and average about $108,000, or about 80 percent of the 1985 market value.

    Defendants' appraiser, Montgomery, like defendants' expert engineer, was retained *948 only a few weeks before trial and did not personally inspect plaintiffs' property. From the depositions of Cagle and Russell, Montgomery assessed plaintiffs' actual losses as "land lost," "trees lost," and "amenities damaged."

    With respect to "land lost," Montgomery opined that plaintiffs had only lost the "utility" of their land, and not the land itself, since they still owned to the center of the stream. He valued the land at $.45/sq. ft. for the larger lots and $.20/sq. ft. for the smaller lots (compared to Russell's values of $.50 and $.22/sq. ft.), and calculated the value of the lost utility of the land by applying 100 percent of these values to Cagle's "minimum loss" square footage figures (from center of stream to present high bank) rather than to Morvant's land loss calculations (from pre-construction high bank to projected maximum failure line), which apparently were not available to Montgomery when he prepared his report. In the report, Montgomery notes that his land loss values "would be subject to change" if the court ruled that the land loss calculations should be based on the pre-construction high bank rather than the center of the stream. Montgomery did not discount the land values by ten percent, as did Russell, because he opined that plaintiffs had lost 100 percent of the utility of the land.

    Montgomery calculated the tree loss at 10, 20 or 40 trees per lot, @ $100, and did not attempt to place a value on the number of trees actually lost, agreeing with Russell that it would not be economically feasible to replace all of the lost trees.

    For "amenities damaged," Montgomery assigned values to fences, oxidation ponds, septic lines, etc. that were actually damaged.

    Montgomery then assessed "possible losses," consisting of slab repairs and damages to fences, pools, etc. Because the expert engineers differed about the likelihood that the bank failures would cause future structural damage, Montgomery opined that these losses were only "possible," but not "probable," as Russell and Cagle had characterized them.

    Montgomery allocated slab repair costs to eight of the 17 lots, at a price of $1.75-2.50/sq. ft., apparently based on his assessment of the structural damages on these lots that were noted by Cagle, such as soil subsiding from patios and driveways. He considered this a possible loss if it was determined to be causally related to the bank failures. Montgomery was not aware of any existing foundation cracks.

    In Montgomery's opinion, and based on his experience in selling lots with repaired slabs, the fact of repair does not deter potential buyers because some consider the repaired slab to be better than the original slab. Montgomery also questioned Russell's method of valuing the "foundation stigma" at 10 percent of market value based on the standard deviation of unit prices in his comparable sales. According to Montgomery, the proper method would be to compare "before and after" sales of comparable property with foundation problems. Montgomery admitted it would be difficult to locate such comparables.

    Montgomery also disputed Russell's opinion of "project stigma," for reasons discussed below.

    Montgomery's losses, both actual and possible, total about $3,800-26,000 per lot.

    Plaintiffs called two realtors, Ron Hutchinson and Stephanie Morris, who testified as to the detailed disclosures their firms require from sellers about the condition of the property, including matters such as soil sliding and settling, cracks in walls, and foundation cracks and repair efforts.

    Based on their inspection of some of plaintiffs' homes, and the condition of the rear portions of the lots, these witnesses opined that none of plaintiffs' properties are marketable, notwithstanding that some engineers may believe the banks have stabilized, because some structural problems already exist and prospective buyers will be concerned about the possibility of future bank failures. They agreed that a "neighborhood stigma" already exists, even though some homes may have only relatively minor problems, because potential buyers will fear that those homes may develop the more severe problems that have appeared in other homes in the area.

    Defendants' appraiser, Montgomery, did not agree with plaintiffs' experts that plaintiffs' *949 property is unmarketable, and that a "neighborhood" or a "project" stigma exists because of the bank failures. Montgomery explained that there were six sales of property contiguous to plaintiffs' (in the same subdivisions and backing up to the same streams) from 1986-1991, including sales to two plaintiffs in 1986, as well as two current listings of contiguous property for sale. The parties later stipulated that the owner of one of the listed properties had recently taken it off the market.

    TRIAL COURT AWARDS

    The trial court found that the rear portion of plaintiffs' property had been "taken" in the constitutional sense, noting that the court in State through DOTD v. Chambers Inv. Co., 595 So.2d 598, 602 (La.1992) defined a "taking" as "any substantial interference with the free use and enjoyment of property." The court accepted Cagle's opinion about the likelihood of future damage to the remaining property over Morvant's, describing Cagle's testimony as "believable, accurate and corroborated by other evidence." The court accepted Russell's valuation of plaintiffs' actual losses and severance damages in most respects. Each plaintiff was awarded mental anguish damages ranging from $1,500-5,000 per person, on findings that the bank failures constituted a "continuing nuisance."

    The court awarded attorney fees under LRS 13:5111, which allows recovery of reasonable attorney fees "for the taking of property by [the state or other governmental entity], other than through an expropriation proceeding." The attorney fee awards are in the amount of one-third of the awards for land and tree loss, excluding the awards for mental anguish and severance damages.

    The court effectively grouped the plaintiffs into these categories, which we number:

    (1) Those who owned the property when the banks failed in 1985 and at trial in 1992, and whose property has a low or moderate probability of structural problems. (Most plaintiffs are in this category.)

    (2) Those who owned the property in 1985 and 1992, and whose property has a high probability of structural problems (Carroll and Yoblonsky).

    (3) Those who owned the property in 1985 and sold it before trial (Harris and Vandling).

    (4) Those who acquired the property after 1985 (Lawrence and Romero).

    (5) Those who acquired and built on the property after 1985 (Cox).

    Category 1 Plaintiffs

    For these plaintiffs, the trial court awarded the value of the property actually lost or "taken" (land and trees, numbering from 25-150 per lot), attorney fees of one-third of the awards for land and tree loss, and mental anguish damages ranging from $2,000-5,000 per person.

    The land loss awards are based on Cagle's "maximum loss" estimate (from center line of stream to farthest failure line on Cagle's diagram), at Russell's valuation of $.45/sq. ft. for the larger lots and $.20/sq. ft. for the smaller lots.

    The tree loss awards are based on the actual number of trees lost, which the court valued at $300 per tree for the first 20 trees, $100 each for trees numbering 21-100, and $50 per tree for trees in excess of 100.

    These "actual loss" awards, including mental anguish damages and attorney fees, range from about $21,000-42,000 per lot and average $31,000.

    The court also awarded these plaintiffs "severance damages," ranging from about $57,000-180,000 per lot, averaging about $88,000 and consisting of:

    —Value of land lost, as calculated by Cagle and Russell (awarded again, in the same amount that was awarded for actual loss).

    —Cost to replace some but not all lost trees (10 trees on smaller lots, 20 on larger lots, @ $500, per Russell).

    —Other elements of severance damages itemized by Russell, at the values he assigned, including value of items actually damaged, value of items with moderate or low probability of future damage, value of lost aesthetics, cost to stabilize soil and repair slab, foundation stigma, project stigma, and adverse servitude. For Russell's "non-compensated *950 amount at risk," the court awarded only one-fourth of Russell's loss estimates.

    The total awards to the Category 1 plaintiffs range from about $78,000-214,000 per lot, and average about $118,000 per lot.

    Category 2 Plaintiffs

    The Carroll and Yoblonsky plaintiffs owned their lots in 1985 and at trial, but their property has a high probability of future problems, according to Cagle. Russell estimated their severance damages at an amount greater than the 1985 market value of the property (market value of about $140,000 plus demolition and moving costs). The trial court awarded them 70 percent of the market value, plus the value of land and trees actually lost, attorney fees of one-third of the awards for land and tree loss, and mental anguish damages of $5,000 each.

    These awards total $120,660 (Carroll) and $149,189 (Yoblonsky).

    Category 3 Plaintiffs

    The Harris and Vandling plaintiffs owned their property in 1985 but sold it before trial, in 1988 and 1989, respectively. The 1985 values were $110,000 and $123,000, respectively. The Harris property sold for $96,000 and the Vandling property for $92,629. The court awarded these plaintiffs their land and tree losses, one-third of those losses as attorney fees, mental anguish damages of $2,000 each, and the difference between the 1985 value and the sale price of their property ($14,000 for Harris and $30,371 for Vandling). The judgment expressly recognizes their right to seek indemnification from defendants if they are sued in redhibition by the buyers.

    These awards total $36,767 (Harris) and $58,404 (Vandling).

    Category 4 Plaintiffs

    The Lawrences and the Romeros bought their property in 1986, after the initial bank failures. The court found that 75 percent of the land loss from their property occurred before they acquired it, and 25 percent occurred thereafter. The court's award for trees actually lost was based on the number of trees that plaintiffs testified they had lost since they acquired the property.

    The court reduced Russell's estimate of their land loss by 75 percent, and calculated the rest of their actual losses and their severance damages using the same analysis that was used for the Category 1 plaintiffs. Each was awarded $2,000 for mental anguish.

    These awards total $67,302 (Lawrence) and $64,151 (Romero).

    Category 5 Plaintiffs

    Mr. and Mrs. Cox bought a vacant lot next to Mrs. Cox's parents, the Grosjeans, in 1986 and built a house on it in 1991, several years after the Grosjeans and other plaintiffs had filed suit alleging the damage arising out of the 1985 dredging operations. The trial court calculated the Coxes' actual losses and severance damages using the same analysis as for the Category 1 plaintiffs, and awarded the Coxes $1,500 each for mental anguish, but reduced their total recovery by 50 percent, finding them "partially at fault" for building on the property after they filed suit.

    The Coxes were awarded about $67,000.

    The court assessed legal interest on all awards from date of judicial demand (February 1987 for the Simmons plaintiffs and July 1987 for the Grosjean plaintiffs). The court awarded expert witness fees of $25,000 for Cagle and $12,000 for Russell, or about two-thirds of the charges that plaintiffs incurred for their services ($40,456 for Cagle and $18,612 for Russell).

    All parties filed motions for new trial limited to reargument on certain issues, including many of the issues raised on appeal. Except for minor editing and correcting several calculation errors, the court denied the motions.

    ARGUMENTS ON APPEAL

    Defendants argue that their actions, while admittedly the cause of the bank failures, did not constitute a "taking" because plaintiffs were not divested of title to their property and because the rear of the property was subject to a 100-foot legal servitude under LRS 38:113. Resolution of these issues affects other arguments of the parties as to damages, legal interest and attorney fees.

    Defendants concede that plaintiffs are entitled to recover for the value of the land, trees *951 and appurtenances that were lost or damaged before trial, and for the resulting loss of aesthetics. They dispute the other awards, particularly those based on the premise that the bank failures will continue in the future and will cause structural damage to plaintiffs' homes and a substantial decline in their market value.

    Defendants argue that attorney fees should not have been awarded because no taking occurred, and that mental anguish damages should not have been awarded because defendants did not commit a trespass. Defendants contend the expert witness fee awards for Cagle and Russell are excessive.

    By answer to the appeal, plaintiffs seek to have attorney fees assessed on the severance damage awards as well as the actual losses, and on the legal interest that runs on the awards, as well as their principal amounts. Plaintiffs contend legal interest should run from the date of the taking rather than from date of judicial demand.

    Plaintiffs ask that the expert witness fees for Cagle and Russell be increased to the amounts actually charged and incurred for their services, and seek recovery of about $15,000 in other expenses, mostly for paralegal services and evidence reproduction costs.

    Citing Cajun Elec. v. Owens-Corning Fiberglass, 616 So.2d 645 (La.1993), plaintiffs argue that legal interest should be assessed on all costs, including expert witness and attorney fees, from the date of the judgment fixing the fees.

    WAS THERE A TAKING?

    Defendants argue that the dredging activities and bank failures do not constitute a "taking" of plaintiffs' property under LSA-Const. Art. 1, Sec. 4, which provides that property shall not be "taken or damaged" by the state or its political subdivisions except for public purposes and with just compensation paid to the owner. Defendants note that plaintiffs have not been divested of title to their property and still own to the center of the stream, in accordance with the property descriptions in their deeds. Defendants claim there was "at most ... damage and deprivation of the character of use of certain parts of the property." Because the State stipulated to liability under Art. 1, Sec. 4, we address this issue only as it relates to the Board.

    The trial court correctly defined a "taking" as "any substantial interference with the free use and enjoyment of property," citing State through DOTD v. Chambers Inv. Co., supra. Our emphasis. While the divesting of title is, of course, a substantial interference with property rights, and may occur when the state or political subdivision files a formal expropriation proceeding, a taking may occur under circumstances where title has not passed and formal expropriation proceedings have not been brought. The pivotal issue is whether the governmental interference with the free use and enjoyment of private property is substantial. See and compare Dept. of Transp. & Dev. v. Traina, 537 So.2d 792 (La.App. 2d Cir.1989), writ denied; Dickie's Sportsman's Centers v. D. of Transp., 477 So.2d 744 (La.App. 1st Cir. 1985), writ denied; and Daspit v. State, Department of Highways, 325 So.2d 368 (La. App. 3d Cir.1975).

    The trial court's finding that the interference here was substantial and constituted a taking is unquestionably supported by the record.

    LEVEE BOARD SERVITUDE

    The Board claims it owes no constitutional compensation whether for taking or damaging plaintiffs' property because it had a legal servitude of 100 feet across the rear of the property under LRS 38:113 in 1985 when the dredging was done and the initial bank failures occurred within the bounds of the servitude. Plaintiffs argued in the trial court that defendant did not prove the existence of the § 113 servitude in 1985, and argued alternatively that if the servitude existed at the time of trial, it ran from the present high bank line, in accordance with § 113 as amended in 1991, rather than from the center of the stream or the former high bank line. Plaintiffs asserted that the servitude line, measured from Cagle's high bank line, entered or approached some of plaintiffs' homes, as depicted on Cagle's diagram.

    *952 Defendants introduced in evidence an extract of a Board resolution providing that the § 113 servitude "will remain in the same location as it existed on August 15, 1977" with respect to plaintiffs' lots. Plaintiffs argued that the Board's purported waiver of the servitude beyond its pre-1985 location, which was made on the eve of trial, was invalid.

    The trial court made no express findings as to the existence of the legal servitude and the validity of the waiver, but implicitly found that the servitude existed, and that the waiver was invalid, by awarding "adverse servitude" damages for the proximity of the servitude line to plaintiffs' homes at the time of trial, as diagrammed by Cagle.

    We conclude that the trial court's finding that the § 113 servitude existed, in or since 1985, is clearly wrong on this record. We do not address the validity of the purported waiver of the § 113 servitude beyond its pre-1985 location.

    Before its amendment in 1991, § 113 provided:

    The various levee and drainage districts shall have control over all public drainage channels within the limits of their districts and for a space of one hundred feet on each side of the channel, selected by the district and recommended and approved by the Department of Public Works, whether the drainage channels have been improved by the levee or drainage district, or have been adopted without improvement as necessary parts of or extensions to improved drainage channels, and may adopt rules and regulations for preserving the efficiency of the drainage channels.

    The emphasized portion of the statue was amended in 1991 to provide that the servitude runs "on both sides of the banks of such channels," whereas before it was "on each side of the channel."

    It has been held that the statute is not self-operative, and that the servitude will be recognized only when the levee or drainage district proves three things: that the drainage channel was previously improved by the district or adopted without prior improvement as a necessary part of or extension to improved drainage channels; that the drainage channel is a public channel; and that the channel was selected by the district and recommended and approved by the Office of Public Works. Whipp v. Bayou Plaquemine Brule Drainage Bd., 476 So.2d 1042 (La.App. 3d Cir.1985); Ortego v. First American Title Ins. Co., 569 So.2d 101 (La.App. 4th Cir. 1990).

    Plaintiffs contend, and defendants do not seriously dispute, that the Board did not prove these matters in this record. The trial court's implicit finding that the § 113 servitude existed, both in 1985 and, in a different location, at the time of trial, is clearly wrong. We shall amend the judgment to delete the "adverse servitude" damages that were awarded to some of the plaintiffs.

    In any event, and even if we were to find that the § 113 servitude existed in 1985, when the dredging and the initial bank failures occurred within the bounds of the servitude, we note that the purpose of the § 113 servitude is to allow the local governing body to prevent channel obstructions and to perform ordinary maintenance necessary for the efficient operation of the drainage system. Terrebonne Parish Police Jury v. Matherne, 405 So.2d 314 (La.1981), U.S. cert. denied. Neither the existence of the servitude, nor the fact that the levee district's activities have occurred entirely within the bounds of the servitude, will defeat a claim for constitutional compensation when those activities exceed ordinary maintenance and constitute a taking or damaging of private property, as they did here. Matherne, supra; Scott v. Red River-Bayou Pierre Levee & D. Dist., 7 So.2d 429 (La.App. 2d Cir.1942).

    Notwithstanding the deficiencies regarding the § 113 servitude, the plaintiffs were on notice of the 100-foot servitude over the rear of plaintiffs' lots since the two subdivisions were platted and developed in the 1970's. The "servitude," however, is limited to the location shown on the subdivision plats and does not extend 100 feet from the present high bank. As a practical matter, only the exact location of the drainage easement is disputed.

    *953 Plaintiffs introduced the subdivision plats for Hickory Ridge Estates and Carriage Oaks East, which plats are recorded in the public records and which show a 100-foot servitude across the rear of each lot. On the Carriage Oaks East plat, the servitude is labeled "100' Levee Board Easement" and appears to run from the high bank of the stream as it existed in 1977. Other servitudes, labeled "Utility Easements," are also shown. The property owner dedicated "to the public use in perpetuity the streets and the easements for utilities and for drainage shown on this Subdivision plat." The dedication describes in some detail what actions may be taken in exercising the "Easements for Utilities," but does not do so for the "Levee Board Easement."

    The Hickory Ridge Estates plat shows a servitude of 100 feet across the rear of each lot, and other 10- and 20-foot servitudes, none of which are expressly labeled as to their purpose. The 100-foot servitude appears to run from the center line of the "bayou" (Flat River) as it existed in 1972. The owner dedicated for public use "the Easements for Utilities and Drainage" shown on the plat.

    The lots in both subdivisions were sold with reference to the recorded subdivision plats, on which the 100-foot servitudes for drainage were shown and dedicated. Plaintiffs thus acquired the property subject to the servitudes shown on the plats. See and compare Tate v. South Cent. Bell Tel. Co., 386 So.2d 139 (La.App. 3d Cir.1980) and Triangle Development, Inc. v. Burns, 469 So.2d 29 (La.App. 1st Cir.1985).

    If the title to a predial servitude is silent as to the extent and manner of its use, the intention of the parties is to be determined in the light of the servitude's purpose. CC Art. 749. Doubt as to the extent or manner of exercising the servitude is to be resolved in favor of the servient estate. CC Art. 730.

    The titles to these drainage servitudes are silent as to the extent and manner of the servitude's use. There is no evidence in this record suggesting that the servitudes were intended for anything other than drainage canals and ordinary maintenance thereof. The dredging activities here clearly exceeded ordinary maintenance, rendering the Board liable for taking or damaging the property without compensation. See and compare Bentley v. Industrial Fire Protection Co., 338 So.2d 1177 (La.App. 2d Cir.1976) and Wright v. Department of Highways, 342 So.2d 230 (La.App. 1st Cir.1976), writ denied.

    EXTENT AND AMOUNT OF COMPENSATION

    Defendants do not dispute that the measure of plaintiffs' recovery under the constitution is "the full extent of [their] loss." Art. 1, Sec. 4. Defendants argue, however, that plaintiffs' losses extend only to the actual loss of land and trees, damage to appurtenances and loss of aesthetics which occurred before trial. Defendants contend the trial court's awards for future losses, such as foundation repair, foundation stigma, etc., are based solely on speculation because no major structural problems have occurred in the homes in the six years before trial and because the defense expert, Morvant, a civil engineer with experience in soil mechanics, opined that the banks have stabilized to the point where any future soil movement will not damage the homes, the closest of which is 70 feet from Morvant's predicted maximum failure line of the banks.

    When only part of a tract of land is taken, and the owner proves to a reasonable certainty that the remaining property is damaged by the taking, the owner may recover severance damages. State, through Dept. of Transp. v. Townsend, 473 So.2d 99 (La.App. 3d Cir.1985), writ denied.

    The owner must prove severance damages by a preponderance of the evidence and may do so by informed and reasoned expert testimony, corroborated by the facts in the record, especially when that testimony is accepted by the trier of fact. State, Dept. of Hwys. v. Denham Springs Dev. Co., Inc., 307 So.2d 304 (La.1975); Louisiana Power & Light Co. v. Mobley, 482 So.2d 706 (La.App. 1st Cir.1985), writ denied. The trial court's factual findings as to severance damages, and its weighing of expert testimony as the basis for that finding, will not be disturbed on *954 appeal unless shown to be clearly wrong. Townsend, supra.

    The trial court personally viewed plaintiffs' property and heard conflicting expert opinions about the expected point of bank stability and the likelihood of future structural damage to the homes, which we have summarized above. The court accepted the opinion of plaintiffs' expert, Cagle, over that of Morvant, on findings that Cagle's testimony was "believable, accurate and corroborated by other evidence." Neither defendants' argument nor the record affords a basis for us to say that the trial court was clearly wrong in this respect.

    The court also heard conflicting expert testimony about the effect of the bank failures on the marketability of the remaining property, even if the failures do not progress any farther toward the homes, including testimony about the likelihood that prospective purchasers will decline to buy the property, or will offer substantially less for it, because of their fear of future structural damage to the homes. This fear, which the trial judge obviously found to be legitimate and reasonable after having viewed the property, may be considered in the severance damage analysis. See and compare Mobley, supra; Texas Pipe Line Co. v. National Gasoline Co., 203 La. 787, 14 So.2d 636 (1943); and Mathis v. City of DeRidder, 599 So.2d 378, 388 (La.App. 3d Cir.1992).

    Defendants argue that no awards should have been made for "adverse servitude," based on the proximity of the Board servitude under LRS 38:113 to plaintiffs' homes when the servitude is measured from the present high bank, because the Board has waived the servitude beyond its location in 1977. Having found that the § 113 servitude was not proved, either in or after 1985, and again without addressing the validity of the waiver, we agree that these awards must be reversed. See discussion of Levee Board Servitude, supra.

    We also reverse, as duplicative, the trial court's awards for the value of the land lost and for the cost to replace some, but not all, of the lost trees as an element of severance damages. In this respect we note that the court awarded the same amount for lost land, and a greater amount for the actual number of trees lost, in its awards for the value of the property that was taken, and additionally awarded plaintiffs five percent of the market value of their property as severance damage to the remaining property for its loss of aesthetics resulting from the loss of land and trees.

    Plaintiffs are clearly owed the value of the land and trees actually lost or taken, and may recover for the effect of those losses on the value of the remaining property. City of Lafayette v. Dore, 460 So.2d 755 (La.App. 3d Cir.1984). Plaintiffs may not, however, recover the value of the land and trees twice, in addition to the award for lost aesthetics. We shall amend the judgment to delete the duplication.

    Defendants claim no awards should have been made for mental anguish because defendants stayed within the Board's 100-foot servitude shown on the subdivision plats and did not trespass on plaintiffs' property during the dredging operations. Plaintiffs argue that the element of trespass is present in "many cases which award mental anguish [damages]." If recovery is allowed for mental anguish, defendants argue that the trial court's awards of $1,500-5,000 per person are excessive. Defendants admit that there was some damage to the property but claim the plaintiffs greatly exaggerated the extent of their mental anguish.

    Recovery of damages for mental anguish arising from property damage is not limited to cases involving a trespass. Such damages may be awarded to persons whose property was damaged by intentional or illegal acts, or by acts giving rise to strict or absolute liability, or by acts amounting to a continuing nuisance, or where the property owner was present when or shortly after the damage was negligently inflicted and suffered a psychic trauma similar to a physical injury as a direct result of the incident. Farr v. Johnson, 308 So.2d 884 (La.App. 2d Cir.1975), writ not considered.

    The trial court found that defendants' acts amounted to a continuing nuisance because the initial bank failures severely *955 damaged plaintiffs' residential property, and they have since "had the unpleasant task of constantly monitoring the rear portion of their lots" for cracks or fissures, which indicate the imminent loss of more land and trees. Although not mentioned by the trial court, defendants' strict liability under CC Art. 667, established by stipulation, affords an additional basis for awarding mental anguish damages. See Barr v. Smith, 598 So.2d 438 (La.App. 2d Cir.1992), writ denied; and J.B. LaHaye Farms v. La. Dept. of Highways, 377 So.2d 1286 (La.App. 3d Cir.1979), writ denied.

    The record supports recovery of mental anguish damages on either or both bases. Plaintiffs adequately proved that they have experienced significant mental distress, far exceeding the usual worry over the consequences of property damage, for up to six years. Compare Elston v. Valley Electric Membership Corp., 381 So.2d 554 (La.App. 2d Cir.1980), and J.B. LaHaye Farms, cited supra. The trial court's awards vary in amount, logically reflecting that the mental anguish of the individual plaintiffs differed in degree and duration. We do not find the awards to be excessive.

    With respect to the remaining awards, we find no error or abuse of the trial court's discretion in determining the extent and amount of recovery that will compensate plaintiffs to the full extent of their loss. See and compare Townsend and Mobley, both cited supra.

    In the light of the stipulated liability under CC Art. 667, we find no clear error in the trial court's allocating 50 percent fault to reduce the loss and damage awards to the Cox plaintiffs. Defendants ask that we increase this allocation.

    Allocation of fault, a factual finding, is subject to the clearly wrong standard of review. Towns v. Georgia Cas. & Sur. Co., 459 So.2d 124 (La.App. 2d Cir.1984). The Coxes purchased their lot in 1986 after some dredging damage occurred, to be next door to Mrs. Cox's parents. They filed suit for the damage and thereafter in 1991 built their home on the lot. Whether the award is based on either or both constitutional composition and CC Art. 667 damages, the measure of recovery by the Cox plaintiffs is something less than the awards to plaintiffs who owned their lots before the dredging began in 1985. On this record we cannot say this reduction in damages by the trial court was clearly wrong.

    The Coxes initially asked that we reduce the allocation to something less than 50 percent but agreed at oral argument that the trial court's allocation was a factual finding which was not clearly wrong.

    ATTORNEY FEE ISSUES

    LRS 13:5111 provides in part:
    A court of Louisiana rendering a judgment for the plaintiff, in a proceeding brought against the state of Louisiana, a parish, municipality or other political subdivision or an agency of any of them, for compensation for the taking of property by the defendant, other than through an expropriation proceeding, shall determine and award to the plaintiff, as a part of the costs of court, such sum as will, in the opinion of the court, compensate for reasonable attorney fees actually incurred because of such proceeding....

    Plaintiffs have contracted to pay attorney fees of one-third of their total recovery. The trial court awarded them attorney fees in the amount of one-third of the amounts awarded for the value of the land and trees that were actually lost or taken, and declined to assess attorney fees on the severance damage awards.

    Because we have found a "taking" occurred, we find no merit in defendants' contention that attorney fees should not have been awarded.

    Plaintiffs contend the trial court erred as a matter of law in assessing attorney fees only on the awards for the property that was taken, noting that the trial court's attorney fee awards constitute only five percent of plaintiffs' total recovery. Plaintiffs argue that attorney fees should be assessed on the principal amount of all awards, and on the amount of legal interest that accrues on those awards. While conceding that a court is not bound to award attorney fees in the *956 percentage stipulated in an attorney-client fee contract, plaintiffs maintain that an award of one-third of their total recovery is reasonable on this record.

    Attorney Fees on Severance Damages?

    LRS 13:5111 allows recovery of attorney fees in an action "for compensation for the taking of property ... other than through an expropriation proceeding ..." Our emphasis. Compensation for the taking of property includes both the value of the property actually taken and the severance damages that result to the remaining property. Gray v. State, Department of Highways, 250 La. 1045, 202 So.2d 24 (1967); State, through Dept. of Transp. v. Townsend, supra.

    The trial court correctly found that the taking of a portion of plaintiffs' property caused severance damages to the remaining property. Having made these findings, which we affirm, and having awarded compensation for both the "land and trees taken" and the "severance damages" caused by the taking, the court did not have discretion under § 5111 to limit the attorney fee awards to the value of the property that was taken. The court should have assessed attorney fees on the total of the awards for the land and trees taken and the severance damages. See, e.g., Pillow v. Board of Com'rs, 425 So.2d 1267 (La.App. 2d Cir.1982), writ recalled; Mathis v. City of DeRidder, cited supra; Guion v. State Dept. of Transp., 391 So.2d 1367 (La.App. 4th Cir.1980), writ denied; and Daspit v. State, Department of Highways, cited supra.

    The trial court's narrow construction of the statute was based in part on its concern that if the severance damages to the property not taken were included as a basis for attorney fees, "every type of property damage resulting from the state's activity would amount to a ``taking' and therefore render them liable for attorneys fees." We cannot agree. If the State's activities do not constitute a substantial interference with the free use and enjoyment of property, then no taking has occurred and attorney fees are not available under § 5111. See State through DOTD v. Chambers Inv. Co., cited supra. If, however, a trial court finds that a taking has occurred and has caused severance damages, as the court did here, attorney fees are owed on the severance damages as part of the constitutional compensation for the taking.

    Some cases have denied recovery for nonpecuniary losses such as mental anguish and inconvenience as part of the constitutional compensation owed for the taking of property, even under the 1974 constitution. Compare Key v. Louisiana, Dept. of Highways, 357 So.2d 1230 (La.App. 2d Cir.1978), writ not considered, which we decided under the 1921 constitution, with the third circuit's rulings under the 1974 constitution in State, Dept. of Highways v. Johnson, 369 So.2d 191 (La.App. 3d Cir.1979) and Mathis v. City of DeRidder, cited supra. We need not and do not consider or attempt to reconcile those decisions, however, because the mental anguish damages awarded here are recoverable under defendants' stipulation of CC Art. 667 strict liability.

    Plaintiffs have cited no cases where mental anguish damages were awarded solely under a constitutional theory of recovery. While we amend to award attorney fees on the severance damage, we shall exclude attorney fees on the award of mental anguish damages.

    Attorney Fees on Principal and Interest?

    Plaintiffs contend attorney fees should be assessed on both the principal amounts of the compensation awards and on the legal interest that accrues on those awards. As authority, they cite Pillow and Guion, both cited supra; Dept. of Transp. & Dev. v. Williamson, 557 So.2d 731 (La.App. 2d Cir.1990), writ denied; and State, DOTD v. Dietrich, 598 So.2d 649 (La.App. 3d Cir. 1992).

    Williamson and Dietrich were "quick-taking" actions in which attorney fees were recoverable under LRS 48:453 "if the amount of compensation deposited in the registry of the court is less than the amount of compensation awarded in the judgment." The inclusion of both principal and interest as the basis for the attorney fee awards in Williamson and Dietrich was authorized by § 455, which, before its amendment in 1988, expressly provided that legal interest on the *957 excess award was "part of the just compensation" for the taking. There is no statutory language analogous to § 455 that applies to "inverse condemnation" cases such as this one, where a taking occurs without formal expropriation proceedings.

    Both Pillow and Guion were inverse condemnation cases. In Guion, the fourth circuit upheld the trial court's award of attorney fees on principal and interest, finding "that a fee of one-third of the amount of principal and interest is justified." 391 So.2d at 1371. No authority was cited for including interest in the attorney fee calculation.

    The trial court in Pillow assessed "attorneys' fees of 1/3 of the principal and interest under the judgment plus 12% interest on the attorneys' fees until paid." 425 So.2d at 1269. We reversed the assessment of interest on the attorney fee award after concluding that "the one-third fee on the principal awarded is reasonable, was actually incurred, and should be maintained." p. 1283; emphasis added. In the decree, we ordered the defendant to pay the principal amount of $332,914.70, with legal interest thereon until paid, "together with thirty-three and one-third (33-1/3%) of the principal or $110,971.57 as attorneys' fees ..." p. 1284; emphasis added. The issue of plaintiffs' entitlement to attorney fees on interest as well as principal was not raised or discussed in the opinion.

    In Naquin v. Dept. of Transp. and Dev., 604 So.2d 62, 69 (La.App. 1st Cir.1992), writ denied, the court also limited attorney fees to one-third of the principal awards for inverse condemnation, after the trial court had assessed the fees on principal and interest, without discussing the issue.

    Mathis v. DeRidder, supra, was also an inverse condemnation action. The third circuit affirmed the trial court's conclusion "that a fee of twenty-five percent of the actual compensation (but not including interest) awarded plaintiffs is reasonable in this case and will be allowed." 599 So.2d at 392; emphasis added. Again, the issue whether attorney fees should be assessed on principal and interest was not otherwise discussed.

    Notwithstanding that the issue of entitlement to attorney fees on interest as well as principal has arisen only obliquely in inverse condemnation cases, we find little support in the case law, and none in the statute, for assessing attorney fees on both principal and interest. We shall limit the basis for attorney fees here to the principal amounts awarded as compensation for the taking and severance damages.

    Amount of Reasonable Attorney Fees

    The trial court included the attorney fee awards in the principal amounts awarded to each respective plaintiff or plaintiffs in the judgment, which amounts total $1,748,215. We have recalculated the principal as $1,326,604 by excluding the attorney fee and mental anguish awards from consideration, and by subtracting the duplicate awards for land and tree loss and the awards for "adverse servitude," which awards we reverse. We now consider whether an award of one-third of the principal, or $442,201, is a reasonable fee, realizing, of course, that no court is bound to, but may, award a reasonable amount that plaintiffs have contracted to pay their attorneys. Pillow, supra.

    The three plaintiff attorneys represent 33 individuals who own 17 residential lots, about which the attorneys were required to familiarize themselves in detail. Both the number of plaintiffs involved and the nature and variance of the damage to their property distinguish this case from the typical inverse condemnation case.

    Plaintiffs' attorneys filed suit in 1987 and were involved in extensive discovery before the six-day trial began in May 1992. They reviewed and introduced in evidence several hundred documents, including excerpts from the State's daily and weekly project diaries, the technical engineering reports of the Soil Conservation Service and of Demopulos & Ferguson, and the engineering and appraisal reports of plaintiffs' experts, Cagle and Russell. The attorneys' thorough and orderly presentation of much highly technical evidence reveals that the attorneys had extensively prepared for trial on the issue of liability as well as quantum. Defendants did not stipulate to liability until the eve of trial. The Board stipulated only to strict liability under CC Art. 667, leaving the issue of its liability on constitutional grounds for the *958 court to decide. Some of the liability evidence that plaintiffs' attorneys had amassed was introduced to show that the Board had "taken or damaged" plaintiffs' property in the constitutional sense, and to support the mental anguish claims.

    The two attorneys representing the Simmons plaintiffs filed affidavits in the trial court itemizing about 1550 hours of work before trial began. One attorney filed a supplemental affidavit in this court listing an additional 408 hours from the beginning of trial through preparation of the appellate brief, which includes filing and arguing the motion for new trial but not the preparation and presentation of oral argument in this court. Defendants do not suggest any inaccuracy or unreasonableness in these hourly charges.

    The attorney for the Grosjean plaintiffs has handled that case from its inception but did not itemize his time in affidavit form or otherwise. The record indicates he obviously spent comparable time on the case and that he did not simply "ride the coattails" of the other plaintiff attorneys who maintained time records.

    On this record, it is reasonable to conclude that the three plaintiff attorneys have collectively spent well over 2,000 hours working on this case. Their assertion that their representation of plaintiffs sometimes precluded other employment appears credible. They were quite successful, as evidenced by the trial court judgment for over $1.5 million in principal awards, which, in most respects, have been upheld on appeal. We also consider their appellate briefs and arguments. By answering the appeal, they have substantially increased the awards for their attorney and expert witness fees and for legal interest (see discussion of expert witness fees and legal interest below).

    The factors above mentioned bear on the reasonableness of the "attorney fees actually incurred" as LRS 13:5111 contemplates. Pillow, supra. See also and compare the "reasonableness" analysis in State, DOTD v. Williamson, 597 So.2d 439 (La.1992) and Mini Togs Products, Inc. v. Wallace, 513 So.2d 867 (La.App. 2d Cir.1987), writ denied.

    In the light of these factors, we conclude that an attorney fee of $442,201, or one-third of the principal awards for constitutional compensation, is reasonable. We shall amend the judgment accordingly.

    Plaintiffs also contend the trial court should have awarded them about $15,000 in expenses for paralegal services, for the cost of reproducing evidence and for taking depositions that were not introduced in evidence. In the light of our significant increase in the attorney fee award, we decline to make a separate award for these expenses.

    EXPERT WITNESS FEES

    The trial court awarded plaintiffs $25,000 in expert witness fees for Cagle and $12,000 for Russell, and taxed the awards as costs under LRS 13:3666. Defendants contend the awards are excessive, while plaintiffs ask that they be increased to the full amount of the experts' charges, or $40,456 for Cagle and $18,612 for Russell.

    Cagle studied the bank failures in detail and made his projections of future soil movement, which he diagrammed for the court. He also inspected each plaintiff's home and other structures for evidence of structural damage, and made a written report of his findings. Russell separately appraised each of the 17 homes and calculated severance damages for each in the manner we have explained. Both witnesses testified at trial. The trial court obviously found their testimony and reports of great value, accepting Cagle's opinion about the likelihood of further soil movement and structural damage to the property, and applying Russell's values to calculate plaintiffs' losses in most respects.

    The trial court's expert witness fee awards of about two-thirds of each expert's total charges are clearly not excessive, as defendants argue. On this record, and considering both the amount of work performed by each expert and its obvious value to the trial court, we find that the court abused its discretion in failing to award plaintiffs the full amount of their loss with respect to charges by their expert witnesses. See and compare State v. Centuries Memorial Park Ass'n, 391 So.2d 489 (La.App. 2d Cir.1980); State, Department of Highways v. Allen, 332 So.2d *959 922 (La.App. 2d Cir.1976); and State Dept. of Highways v. Gormley, 357 So.2d 859 (La. App. 3d Cir.1978).

    We shall amend the judgment to increase the awards.

    LEGAL INTEREST ISSUES

    Interest From Date of Taking?

    Plaintiffs contend the trial court should have awarded legal interest from the date of the taking rather than from date of judicial demand. Defendants again argue that no taking occurred. We have found to the contrary. If legal interest is awarded from the date of the taking, defendants ask that the interest award be offset by some unspecified amount to reflect that plaintiffs have continued to live in their homes since the time of the taking.

    We agree that plaintiffs are entitled to legal interest on the principal amounts awarded as constitutional compensation from the date of the taking. See A.K. Roy, Inc. v. Board of Com'rs for Pontchartrain L.D., 238 La. 926, 117 So.2d 60 (1960); Reddel v. State through Dept. of Highways, 340 So.2d 1010 (La.App. 4th Cir.1976); and Mathis v. City of DeRidder, supra, at p. 392. Mental anguish damages under CC Art. 667 bear interest, not from the taking, but from the date of judicial demand, as in other conventional damage actions. LRS 13:4203.

    Defendants cite no authority for their contention that the interest award should be offset to reflect that plaintiffs still live in their homes. The awards that bear legal interest from the date of the taking do not purport to compensate plaintiffs for the loss of physical possession of their homes, but rather for the loss of land and trees from the rear of their lots and for severance damages to the remaining property, including a decline in its market value. Severance damage exists whether or not plaintiffs occupy the property. We find no factual or legal basis to offset the interest award in the manner that defendants suggest.

    Legal Interest on Court Costs

    Legal interest is owed on all court costs, including the expert witness and attorney fees, from the date of the judgment fixing the fees and costs. Cajun Elec. v. Owens-Corning Fiberglass, cited supra, Packard's Western Store v. State, DOTD, 618 So.2d 1166 (La.App. 2d Cir.1993). These cases, rather than our earlier contrary decision in Pillow v. Board of Com'rs, supra, control this issue.

    The trial court cast defendants with "all court costs in these proceedings." The only dollar amounts fixed as costs in the September 28, 1992, judgment are expert witness fees for Cagle, Russell and other of plaintiffs' witnesses, which, as amended, total $60,996. That amount shall bear legal interest from the date of the trial court judgment.

    DECREE

    We affirm the awards to each respective plaintiff or plaintiffs for the value of the land and trees actually lost. We reduce the severance damage awards to each respective plaintiff or plaintiffs to delete the duplicate awards for land lost and the cost of replacing some trees on the remainder of the property. We also delete the award for adverse servitude damage in favor of those plaintiffs who were awarded that damage. We increase the trial court's attorney fee award to each respective plaintiff or plaintiffs to allow one-third of the total constitutional compensation as reasonable attorney fees under LRS 13:5111.

    We affirm the awards for mental anguish damage for each respective plaintiff or plaintiffs.

    On the awards of constitutional compensation for the taking and severance damage, which awards total $1,326,604, legal interest is awarded from the date of the taking. On the one-third attorney fees, which total $442,201, legal interest is awarded from date of the trial court judgment.

    We amend to increase the awards for all expert witness fees to $60,996, with legal interest thereon from the date of the trial court judgment.

    The respective awards for mental anguish damages, totaling $111,500, shall bear legal interest from date of judicial demand.

    *960 The date of the taking by defendants obviously differs for each respective plaintiff or plaintiffs. The litigants may stipulate or resolve by rule in the trial court the applicable date of taking for each respective plaintiff or plaintiffs.

    Costs of the appeal are assessed to defendants to the extent allowed by law.

    The judgment, as thus amended, is AFFIRMED.