Loading...
EIR: Appendix C.1. Draft Habitat Valuation for Proposed Alternatives • APPENDIX Cl DRAFT HABITAT VALUATION FOR PROPOSED ALTERNATIVES 1 & 2 RANCHO PALOS VERDES FEASIBILITY STUDY Prepared by Daniel J. Pondella, II and John S. Stephens, Jr. Vantuna Research Group Occidental College Developed for U.S.Army Corps of Engineers January 1998 • • Executive Summary The following analysis was conducted in order to assess the environmental restoration which would accompany the U.S. Army Corps of Engineers proposed dykes in Portuguese Bend. The purpose of Alternative 1,a dyke 200'offshore,and Alternative 2,a dyke 400'offshore,is to stop the sedimentation and turbidity associated with the erosion from the Portuguese Bend landslide. Since 1956,this turbidity plume(Please see Figure 2 of DEIS from Pondella et al. 1996) has negatively impacted the nearshore rocky reef environment from Pt.Fermin to Abalone Cove,an area that represents nearly half of the Palos Verdes Peninsula,one of the most prominent rocky headlands in the Southern California Bight. Currently,there is no other known source of turbidity in the area,other than natural erosion from the palisades. And,no other negative environmental impacts on the nearshore fauna of the Palos Verdes Peninsula have been demonstrated. Using the Marine Habitat Valuation Method(Bond et al. in review)developed by the Vantuna Research Group,we found that this restoration effort would be the equivalent of restoring 386 acres of kelp bed habitat for Alternative 1 and 379 acres of kelp bed habitat for Alternative 2. These values represent the single largest restoration,mitigation or enhancement effort for the nearshore rocky reef environment in the Southern California Bight to date. Characterization of the Study Sites Based upon previous surveys of the study area(Envirosphere 1989, Stephens 1990, Pondella et al. 1996, Sadd and Davis 1997) five areas of impact have been identified: Portuguese Bend, Abalone Cove, Bunker Point, Bunker Point-Whites Point, and Point Fermin. The area of primary impact is Portuguese Bend. In Portuguese Bend the sedimentation has covered up the natural reefs resulting in a transformation from a historic kelp bed (John Stephens personal observation) to a soft bottom community. The current status of the environment of Portuguese Bend consists of fine sediments which are uncharacteristic in the shallow reaches of the Palos Verdes Peninsula. This homogenous soft bottom habitat is broken up by sporadic boulders which protrude through the mud as a reminder of the rocky habitat buried below. Beyond the approximately the 10' isobath the habitat is uniform without any observable changes at the greater depths of the bend. At the base of the slide there are large boulders which have been deposited from the land movement. There is also rocky reef habitat on the leeward side of Inspiration Point. The rocky reef and soft bottom fauna of these areas has been characterized and turbidity is extremely high (Pondella et al. 1996). Immediately downcoast the soft bottom is replaced by rocky reefs, both low and high relief; which support a kelp bed on Bunker Point to a depth of at least 60'. The turbidity plume is normally present here, except when the longshore current runs upcoast. This area receives a heavy influx of sediment as it is transported towards Point Fermin. The area is reminiscent of other points on the peninsula with a noticeable reduction in benthic macroalgae and invertebrates. The ichthyofauna was surveyed by SCUBA divers for conspicuous fishes in November 1997. This methodology was consistent with previous fish surveys on the peninsula(Terry and Stephens 1976, Stephens et al. 1984 and Pondella et al. 1996). In November, the area between Bunker Point and Whites Point was also surveyed by SCUBA for conspicuous fishes and general observations. The level of impact in this area was striking. The benthic community of invertebrates and algae was absent from the 30' isobath to the low intertidal zone. In the quiet waters between the points on the peninsula, kelp may or may not be present, C1-1 but in its absence an urchin barren community, consisting primarily of purple urchins, Strongylocentrotus purpuratus, dominates the substrate. As such, this area is atypical for the • Palos Verdes Peninsula. In deeper water(45 to 60'), there was a relatively healthy kelp bed. The Point Fermin area has not been surveyed, but the turbidity is still present in this area(personal observation and please see Figure 2 of DEIS from Pondella et al. 1996). Abalone Cove has also been included in this analysis for two reasons. First, the longshore current does run upcoast extending the turbidity plume into this area. Second, comparing the area as surveyed in the 1988 study (Envirosphere 1989)with the 1995-96 surveys there appears to have been an increase in deposited sediments on the low lying reef within the cove. There appears to be approximately 6" of new sediment in Abalone Cove. In other words, the rocky reef habitat has been degraded over the last decade. Abalone Cove and the accompanying area of Pirates Cove have sheer reefs proximate to the palisades which support many rocky reef species on the points. At the base of these reefs along the 30' isobath the rock is replaced by a sand bottom. The fine sedimentation associated with the slide can be seen on these reefs, but as on Bunker Point, it gets transported by the longshore currents and deposited elsewhere. Unfortunately, it appears to be smothering the low lying reefs within Abalone Cove. Palos Verdes Point was used as a control for this analysis. It represents a healthy rocky reef area on the Palos Verdes Peninsula which normally supports a kelp head. The density of giant kelp fluctuates normally on this point and was present during the surveys for this study. In our long term surveys of this region kelp has been absent. This point was ideal as a control because it is •. outside of the proposed areas impacted by the turbidity plume and relatively close to the study area(also see Stephens et al 1996:1). For this analysis the question of the ability of the proposed dykes to stop turbidity is not addressed. The assumption is that the turbidity will he storprperd_ Excessive turbidy ithag been � � noted to cause significant problems on nearby rocky reefs(Pondella and Stephens 1994). Yet, reefs affected by turbidity do recover when the sediment input has stopped. The proposal to put a dyke on what will become a rocky reef is unique in Southern California for a couple of reasons. It will insure that the breakwater does not sink because there is hard substrate beneath it. And, as the sediment erodes away from its base, it will become contiguous with the natural reef essentially becoming an extension of the natural reef Analyses For this analysis the Marine Habitat Valuation method (Bond et al. in review) was utilized. This is a data driven three parameter model based upon fish lengths, fidelity and abundance (for further discussion see Stephens et al. 1996). The methodology was specifically designed to be objective; it does not weight the importance of specific organisms or habitats and diversity is not included as a component. With the realization that most organisms are not shared between habitats, it utilizes fish guilds to categorize the sampled fishes. The results are Habitat Values which can be directly C1-2 • compared between habitats. For this analysis, the three parameter valuation method was used in conjunction with acreage of the areas discussed above. The existing (without project) habitat value of areas within the Study Area was calculated (Table 1). This existing Habitat Value was multiplied by the acreage which results in a term referred to as Habitat Value Units (HVU) (Table 2). A restored HVU was also calculated using a Palos Verdes Point as a control (i.e., with the assumption that all areas would be restored to the quality of habitat at Palos Verdes Point) . Palos Verdes Point represents a historically healthy kelp bed/rocky reef community which is outside of the area of impact. The difference between the restored HVU and the current HVU can be converted in a similar manner to the acreage any habitat of interest. For this analysis, the restoration effort has been presented as acres of kelp habitat. In all areas, except Bunker Point Kelp, the acreage has been limited to within the 30' isobath. This is a conservative estimate because the turbidity could be impacting deeper environments. Fish data was used from Pondella et al. 1996 and surveys conducted in the fall of 1997. The calculated parameters for Palos Verdes Point (control reef), Portuguese Bend, and the cryptic fishes and beach seine data of Abalone Cove were from Stephens et al. 1996 (Table 1). The Habitat Value for Palos Verdes Point (5719) was used from the 1995-96 surveys (reported in Stephens et al. 1996) because it temporally comparable to the other surveys reported in Stephens et al. 1996. It is important to note that this value is very close to the Habitat Value (5754) calculated for Palos Verdes Point when the entire data base from 1974-1994 was included. Using the lower Habitat Value is also a conservative approach in this analysis because it gives a slightly lower value to a healthy kelp bed community. The Abalone Cove diver transects were recalculated to include only the transects within the 30' isobath (Appendix I). This increased the Habitat Value from 1073 to 1124 which also has the conservative effect of reducing the restoration total (because it increases the without project habitat value). For Bunker Point Kelp and the area between Bunker Point and White's Point, the diver transects were conducted in the Fall of 1997 (Appendix II and III, respectively). The value of the Portuguese Bend beach seines was used to increase the relative worth of these two areas in the comparison with the canopy of a kelp bed. The beach seine values from Abalone Cove as the comparison to the canopy transects at Palos Verdes Point could have been used, but we felt that the higher value of the Portuguese Bend beach seines to be representative of this fauna (personal observations) and a more conservative number for the overall analysis. There is no cryptic fish data available for Bunker Point Kelp, Bunker Point to White's Point or Point Fermin. Due to the heavy impact on the macroalgae and invertebrates in these regions, the numbers of cryptic fishes is likely to be extremely low. The value for Abalone Cove's cryptic fish fauna was used in spite of the fact that this data was collected in 1988 in Abalone Cove, and that currently Abalone Cove, Bunker Point Kelp and the Bunker Point to White's Point areas are probably significantly below that figure. The values for Point Fermin reflect those of the Bunker Point to White's Point area C1-3 and are the highest impacted area Habitat Values reported. • The calculation of Habitat Value Units (HVU) is in Table 2. For reference all impacted areas are incorporated into the analysis. The acreage behind the dykes has been calculated as the sum of the quiet area behind the footprint of the breakwaters plus half of the acreage of the breakwaters (Figures supplied by US Army Corps of Engineers). The rationale for this calculation is that the ocean side of the breakwaters will become rocky reef/intertidal habitat and everything behind the breakwater will be lost. An acreage figure for Portuguese Bend was estimated as all nearshore habitat within the 30' isobath. This is the depth to which the Corps assumes normal wave action will remove the deposited sediment to uncover the hard bottom. The other four acreage estimates are taken from the hydro surveys of the area and with the exception of Bunker Point Kelp all estimates are from the MLLW line to the 30' isobath(see Fig 5.1 and 5.2 of the DEIS). The acreage value for Bunker Point Kelp extends to the 60' isobath. The acreage for Alternatives 1 and 2 were calculated as the difference between Portuguese Bend and the lost acreage behind the two dykes previously calculated. All values were converted into hectares and the existing Habitat Values are reported from Table 1. Multiplying these two parameters (existing Habitat Value x Hectares)results in the Existing HVU; while multiplying hectares by the control (i.e., Palos Verdes Point)Habitat Value (5719) gives the Restored HVU. The units for current HVU and restored HVU are directly comparable. The adjusted difference between these two values has been divided by the value for the kelp habitat and converted to acres for easy reference. Kelp habitat restored acreage for Alternatives 1 and 2 equal the sum of the restored acreage for Bunker Point Kelp +Bunker Point to Whites Point+Abalone Cove+ • Point Fermin+Alternative 1 or 2. The values for the dykes and Portuguese Bend are only used to calculate the acreage for Alternatives 1 and 2. The Kelp Habitat Restored equals 386 acres for Alternative 1 and 379 acres for Alternative 2. Discussion In all areas of this analysis whenever a decision needed to be made about the relative value of particular habitats, the more conservative value was chosen. This analysis, therefore, provides a conservative estimate of the potential restoration habitat values. As more data is collected about currently impacted areas, the estimate of the restoration effort will most likely increase to over 400 acres of kelp habitat. The acreage lost behind the perspective dykes would easily be offset by the kelp acreage gained in Portuguese Bend alone(Table 2). While the marine habitat valuation method was originally developed to cross compare habitats in which there is little species overlap, in this analysis it becomes an even greater tool. With the exception of Portuguese Bend all of the habitats compared are currently rocky reef/kelp bed areas. Thus the species composition which was examined does not turn over between areas resulting in the difference in the abundance of fishes between habitats driving the restoration model. The strength of the analysis also is apparent when comparing the values for the various areas. For instance, even though on Bunker Point there is a persistent kelp head, this was found C1-4 to be less valuable than any other rocky reef area in the analysis. While kelp can persist in disturbed areas, primarily due the loss of benthic competitors, its associated community of invertebrates, fishes and benthic algae are degraded due to the turbidity. As we move away from the effects of the Portuguese Bend slide, the relative abundance of fishes increases in spite of a decrease in kelp cover. This can be seen in the nearly identical values of the Abalone Cove and Bunker Point to White's Point diver transect data (1124 and 1156 respectively). Thus,while kelp is normally thought to be an indicator of a healthy marine environment, in this situation it is not the case. The scope and magnitude of such a restoration effort cannot be over emphasized. The amount of acreage being discussed would easily surpass the value of all nearshore artificial reefs or other mitigative programs in the Southern California Bight. Commercially important species such as abalone, lobster, urchins, white seabass, black seabass and rockfish (to name only a few) all are found on the peninsula. While rocky reefs encompass only 10% of the nearshore habitat in the bight, a restoration effort of the Palos Verdes Peninsula represents not only a potential fisheries increase, but also represents a greater potential to enhance stocks in other areas through natural recruitment and emigration processes. For example, black, green and pink abalone can still be found at Palos Verdes Point. Since we know that their larval/dispersal period is only a few days, the degradation of the benthic environment which we see will inhibit the recruitment of abalone to a large portion of the peninsula and slow the return of these stocks. Certainly we can also associate an economic increase for the sport fishing and sport diving industries with this project. Increasing the water clarity of this nearshore environment will immediately enhance the SCUBA diving opportunities from Abalone Point to Point Fermin providing a close to home option for the SCUBA diving interests based in Los Angeles Harbor. Thus, if we look at Palos Verdes in a pluralistic manner the importance of this habitat will go beyond any specific anthropogenic need. Literature Cited Bond, A., J. S. Stephens, D. J. Pondella II, P. Morris and M. J. Allen. (In review) A method for the estimation of neritic marine habitat values in the southern California bight based on fish guilds. In press, Bulletin of Marine Science. Envirosphere Company. 1989. Draft Environmental Impact Report for Abalone Cove Landslide Stabilization Project. County of Los Angeles Department of Public Works. Sadd, James L. and N. Davis. 1997. Final Report of the Sediment Surveys,Portuguese Bend Area, Rancho Palos Verdes, CA. U.S. Army Corps of Engineers. Pondella, D. J., II and J. S. Stephens Jr. 1994. Factors affecting the abundance of juvenile fish species on a temperate artificial reef. Bulletin of Marine Science, 55 (2-3):1216-1223. Pondella,D.J., II, Morris P., Stephens J.S., Jr. and Davis N. 1996. Marine biological surveys of • the coastal zone off the city of Rancho Palos Verdes. U.S. Corps of Engineers. C1-5 Stephens, J.S., Jr., Morris, P.A., Zerba, K. and Love, M. 1984. Factors affecting fish diversity on a temperate reef: the fish assemblage of Palos Verdes Point, 1974-1981. Env. Biol. O., Fish. :25 - 11 4 ( ). 9 275. Stephens, J.S., Jr. 1990. The effect of the Portuguese Bend Landslide upon the nearshore biota of Palos Verdes. Prepared for the U.S. Army Corps of Engineers. Stephens, J.S., Jr., D. J. Pondella, II, and P. Morris. 1996. Habitat Value Determination of the Coastal Zone off the City of Rancho Palos Verdes Based on Habitat-Specific Assemblage Data. U.S. Corps of Engineers. Terry, C.B. and Stephens, J.S., Jr. 1976. A study of the orientation of selected environmental fishes to depth and shifting seasonal vertical temperature gradients. Bull. S. Calif. Acad. Sci. 57(2):170-183 List of Figures Table 1. The square-root standardized habitat values (hectares) were calculated the five impacted areas and the control (Palos Verdes Point) by each sampling method based upon the methods of Bond et al. in review and incorporated values from Stephens et al. 1996. Table 2. The restoration effort for Alternatives 1 and 2 were calculated in acres of restored kelp • habitat. These values were based upon the current habitat values, acreage of impacted areas and the value of the control, Palos Verdes Point, which is not impacted. Appendix I. The habitat value for the diver transects at Abalone Cove was calculated from the data collected in 1995 —96 surveys (Pondella et 1. 1996). Only transects within the 30' isobath were used (n=18). Appendix II. The habitat value for the diver transects counting conspicuous fishes at Bunker Point Kelp was calculated from the diver transects conducted in November 1997 (n=8). Appendix III. The habitat value for the diver transects counting conspicuous fishes was calculated from the diver transects conducted in November 1997 (n=10). 4110 C1-6 • List of Figures Table 1. The square-root standardized habitat values (hectares) were calculated the five impacted areasand the control (Palos Verdes Point) by each sampling method based upon the methods of Bond el al. in review and incorporated values from Stephens el al. 1996. Table 2. The restoration effort for Alternatives 1 and 2 were calculated in acres of restored kelp habitat. These values were based upon the current habitat values, acreage of impacted areas and the value of the control, Palos Verdes Point, which is not impacted. Appendix I. The habitat value for the diver transects at Abalone Cove was calculated from the data collected in 1995 — 96 surveys (Pondella el al. 1996). Only transects within the 30' isobath were used (n=18). Appendix II. The habitat value for the diver transects counting conspicuous fishes at Bunker Point Kelp was calculated from the diver transects conducted in November 1997 (n=8). Appendix Ill. The habitat value for the diver transects counting conspicuous fishes was calculated from the diver transects conducted in November 1997 (n=10). • • Table 1. Calculation of Habitat Values by Location. The square-root standardized habitat values(hectares) were calculated the five impacted areas and the control (Palos Verdes Point) by each sampling method based upon the methods of Bond ei al. in review and incorporated values from Stephens et a1. 1996. Palos Verdes Abalone Portuguese Bunker Point Bunker Point to ' Method Point Cove Bend Kelp White's Point Point Fermin Diver Transects 2171 1124 37 963 1156 1156 Canopy/Beach Seines 2342 674 798 798 798 798 Cryptic Fish 1206 253 () 253 253 253 Totals: 5719 2051 834 2013 2206 2206 • • • • • Table 2. Calculation of Restoration Effort. The restoration effort for Alternatives 1 and 2 were calculated in acres of restored kelp habitat. These values were based upon the current habitat values, acreage of impacted areas and the value of the control, Palos Verdes Point, which is not impacted. Restored Existing Restored Restored - Kelp Existing Habitat Value Habitat Value Existing Habitat Units (Acres) Impacted Areas Acres Hectares Habitat Value Units (HVL (H�7 �� 200' dyke 14 5 834 4558 31247 26689 12 400' dyke 22 9 334 7259 49763 42504 18 Portuguese Bend 80 32 834 27010 185165 158154 68 Bunker Point Kelp 125 51 2013 101837 289320 187483 81 Bunker Point to White's Point 150 61 2206 133920 347 184 213264 952 Alternative 1 67 27 834 22452 153918 131466 Altcriv,Itivc 2 59 24 834 19751 135402 115650 50 Abalone Cove 100 40 2051 82999 231456 148457 64 Point Fermin 150 61 2206 133920 347184 213264 92 Alternative: 1 2 Kelp Habitat Restored: 386 379 • Appendix I. Abalone Cove. The habitat value for the diver transects at Abalone Cove was calculated from the data collected in 1995-96 surveys(Pondclla et al. 1996). Only transects within the 30' isobath were used(n=18). Guild Square Age Density Guild Mean Root Species Class Length Guild Abundance Fidelity (Hectares) Density Length Product Chromis pw;ctipinnis A 120 5 126 1.0 292 847 99 290 Chromis punchpnris J SO 5 100 1.0 23 I Chromis putctipiruus S 120 5 132 1.0 306 0.i)julis calif-arnica 1 48 5 8 _ 1.0 19 JJrachyissiusfrenatus A 97 8 1 0.5 2 113 148 92 0Ayiulis californica A 149 8 48 1.0 111 Paralabraz clathrates A 277 9 42 1.0 97 123 286 187 Parolabrax clathratus S 277 9 2 1.0 5 Paralabrca nebulifer A 329 _ 9 9 1.0 21 Embiotoca jacksoni A 152 11 11 1.0 72 192 167 179 Embiotoca jacksoni S 152 11 10 1.0 23 Ilypsurus caryi A 159 11 2 1.0 5 Hypsypops rubicundus A 186 11 36 1.0 83 flypsypops rubicundus 3 40 11 I 1.0 2 JJj psypops rubicundus S 186 11 2 1.0 5 1 Phanerodon Arcot= A 160 1 1 I 0.5 2 l-alichoeressemicinclus A 237 12 20 1.0 46 109 213 152 Rhacochilus racca A 195 12 16 1.0 37 Rhacochilus vacca .1 90 12 1 1.0 2 Rhacochilus vacca S 195 12 6 1.0 14 Semicossyphus pulcher A 223 12 3 1.0 7 Semicossyphus pulcher S 223 12 1 1.0 2 Girella nigricans A 173 13 58 1.0 134 160 174 167 Girella nigricans S 173 13 8 1.0 19 Ilermosilla a:urca A 218 13 1 0.5 2 1,1edialuna californiensis A 196 13 _ 3 0.5 5 Rhacochilus lozotes A 246 14 _ 1 0.5 2 2 246 17 Paralichthys californicus 3 90 16 1 0.5 2 5 234 23 Paralichthys californicus S 378 16 1 0.5 2 • Plalyrhinoidis friseri ata S 250 2.2 1 0.5 2 2 250 17 -___ Total: 1124 • I 41k. S. 0 • Appendix LI. Bunker Point Kelp. The habitat value for the diver transects counting conspicuous fishes at Bunker Point Kelp was calculated from the diver transects conducted in November 1997 (n=8). • Guild Square Age Density Guild Mean Root Species Class Length Guild Abundance Fidelity (Hectares) Density Length Product Chromic punctipinnis J 50 5 34 1 177 365 50 135 Chromic punctipinnis S 120 5 1 1 5 Oxyjulis californica J 48 5 35 1 182 Oxyjulis californica A 149 8 23 1 120 219 149 180 Oxyjulis californica S 149 8 19 1 99 Paralabrax clathratus A 277 9 9 1 47 120 295 188 Paralabrav clathratus S 277 9 6 1 31 Paralabrax nebulifer A 329 9 4 1 21 Paralabrav nebulifer S 329 9 4 1 21 ,Embiotoca jacksoni A 152 11 10 1 52 120 151 134 ,Embiotocajackconi S 152 11 1 1 5 Hypsvpops rubicundus A 186 11 9 1 47 Hvpsypops rubicundus L J 40 11 3 1 16 _ Halichoeres senticinctus A 237 12 5 1 26 57 227 114 Rhacochilus yucca S 195 12 1 1 5 Semicossvpbus pulcher A 223 12 3 1 16 Semico ssyphus pulcher S 223 12 2 1 10 • Girella nigricans A 173 13 15 1 78 99 177 132 Girella nigricans S 173 13 2 1 10 Hermosilla azurea A 218 13 2 1 10 Rhacochilrts toxotes A 246 14 2 1 10 10 246 51 Cortiphoplerus nicholsii A 49 19 2 1 10 16 49 28 Cor}phopterus nicholsii S 49 19 1 1 5 Total: 963 • r'Appendix.III. Bunker Point to White's Point. The habitat value for the diver transects counting conspicuous fishes was calculated from the diver transects conducted in November 1997 (n=10). Guild Square Density Guild Mean Root Species Age Class Length Guild_Abundance Fidelity (Hectares) Density Length Product Cluomis pwtctipinnis 1 50 5 9 1 38 38 50 43 Oxyjulis californica A 149 8 22 1 92 117 149 132 Oxyjulis cali/ornica S 149 8 _ 6 _ 1 25 Paralabrax clathratus A 277 9 55 I 229 425 290 351 Paralabrax clathranus S 277 9 17 1 71 Purulabrax nebulifer A 329 9 14 1 58 Paralabrax ncbulifer J 90 9 1 1 4 Parulabrax nebullfer S 329 9 _ 15 1 63 Sebastes atrovirens A 200 10 2 I 8 8 200 41 • Embiolocajacksoni A 152 11 12 1 50 163 150 156 Embiotoca jacksoni J 90 11 2 1 • 8 Embiotoca jacks oni S 152 11 2 I S Girella nigricans J 63 11 5 I 21 Hypsurus caryi A 159 1 I 4 1 17 Hypsypops rub icundus A 186 1114 14 1 58 . Halichoeres semicinctus A 237 12 16 1 67 188 167 177 Hulichoeres sem icinctus J 50 12 16 1 67 Halichocres semicinctus S 237 12 4 1 17 Rhacochilus vacca A 195 12 1 1 4 Semicossyphus puicher A 223 12 _ 8 1 33 Girella nigricans A 173 13 26 1 108 183 176 180 Girella nigricans S 173 13 12 1 50 Medialuna californie.uis A 196 13 4 I 17 Medialuna californiensis S 196 13 _ 2 1 8 Anisotretnus davidsonii 1 65 14 6 1 25 33 113 61 Anisotremus davidsonii S 257 14 2 1 8 Coryphopterus nicholsii A 49 19 I 1 4 4 49 14 To tal: 1156 • I 4,EN •F • .i' United States Department of the Interior H 3 U.S. GEOLOGICAL SURVEY Midcontinent Ecological Silence Center 4512 McMurry Avenue Fort Collins,CO 80525-3400 March 18, 1998 In Reply Refer To: 82020 OCD:703.05 Mr.John R. Hanlon Chief, Branch of Federal Projects U. S.Fish and Wildlife Service Carlsbad Field Office-Ecological Services • 2730 Loker Avenue West Carlsbad, California 92008 Dear Mr. Hanlon: In response to your memorandum of 02/11/98,I asked Dr.Richard Stiehl to coordinate a review of the Habitat Valuation Model by several Center staff members. Overall, I believe the model is a reasonable approach to modeling the marine habitats in your area,but all of the • reviewers had some questions about it. One thread that seems to cross our comments is the objectives for the model. The model may be a useful tool if a measure of biodiversity(which the model derives from density, diversity and mean size)is useful for decision making. Even if it is a meaningful measure for decision making, •¢ there are different scales at which biodiversity can be measured.The model attempts to characterize biodiversity on a small,patch basis,and does not deal with biodiversity on a larger,cross-habitat scale. Of course, measures of diversity treat all species as being of equal importance for decision making,but this may not be the case in your area. For example,if management priorities in your area focus on rare species, then a measure of biodiversity may not be meaningful. Another thread in our comments is the argument that the model is totally objective,and is based on no arbitrary decisions. However,all models include such a bias,especially where there is an attempt to put a numerical score on dissimilar habitats. Our reviewers identified several arbitrary and subjective decisions that the authors make in developing the model. AlI models have some subjective choices involved with development,but one should strive to keep these choices to a minimum and provide logical support for the choices that are made. The following pages have specific comments from each reviewer. If Dr. Stiehl can provide any additional assistance,please contact me. Sincerely, • • Rey C. Stendell 11 Director Enclosures • _ . _ Comments of Richard Stiehl-HEP specialist and Avian Ecologist Pg. 3) I The authors must justify why the"fish assemblage is a good indicator of the health of the entire community". What aspect of the fish assemblage do the authors consider"good"? Why do the authors consider this aspect (or these aspects)"good". Could other aspects also be indicators of the health of the entire community? What is"health" and how is it measured? The limits of"the entire community" needs to be established. Pg. 12) The desire to have the three parameters contribute equally to the final estimate of habitat value assumes that the three parameters are biologically equal in value. The authors must defend their desire, otherwise one might combine the three parameters with the first parameter having a weighting of ten times the second parameter having ten times the third parameter because 'we wished the parameters to contribute this way'. Why should these three parameters be considered any more than any other group of parameters? The authors must defend their choices and respective contributions based on biology and a measure of how close the choices and respective contributions reflect truth. Pg. 19) How is this new method of habitat valuation consistent with earlier approaches? I think the authors mean that the values of their approach are similar to the BEST technique. The authors claim that their measure"does track the changing trends in fish abundance",yet Figure 11 displays `Total Fish Density",not abundance. Further,figure 11 shows that their measure has correlated tracking in five out of nine years. Where is it stated that fish abundance(or density), regardless of species, is an accurate measure of'value'? Pg. 22) • The authors state that the use of different survey techniques make it difficult to obtain a precise estimate of the magnitudeothe Aimerence,but earlier(Figure 11)suggest that their method iscorrelatedlatCu with trawl data and diver transects. I do not understand how an underestimation of 50%to 85%is"roughly the same order of magnitude as is 30% to 511%tinrirrr stimatinn Pg. 24) The authors state that since the density of fish from open coast shallow sand(54 fish per 1000m2)is identical to cobble lacking kelp (52 fish per 1000m2), the nature of the substrate has a minimal influence on habitat quality. If the species comprising each community were identical,and all other environmental parameters were identical,and the onlyecological difference between the two communities was the type of substrate,then such a conclusion could be made. I think it more probable that several environmental factors are different, and that the relatively similar densities of fish is a coincidence. To assert any type of relationship, Iet alone a cause and effect relationship is not advisable when all other parameters are not held constant. • Comments of Brian Cade-Biostatistician • Pg. 2) It is untrue that HSI models"are not amenable to cross-habitat comparisons". Pg. 13) The authors state that the distribution product of the three parameters has the"undesirable consequence of emphasizing abundance at the expense of ecological diversity". For whom is this `undesirable'? Why was a density greater than five fish per hectare chosen as a cut off? What measure is obtained when the number of guilds with a density of greater than five fish per hectare is determined? Why do the authors want a single measure based on density,fidelity,and size of fish guilds to correlate with the number of fish guilds? Pg. 15) I do not see a confidence limit spanning 18%of the mean in Fig 11. Table I states that 1000 replications are used for jackknife estimates of CI. This appears to be bootstrapping, as jackknifing only uses n replicates. Pg. 16) The authors conclude that the"two comparisons provide a substantial assurance of the legitimacy for inference based on the jackknife analysis". The comparisons do not provide any assurance. If some assurance that the CI's are reasonable is desired,the authors need to do simulation studies where truth is controlled. Simulations with multiple truths provide the best method of assessing whether CI will provide correct coverage given a real data set with unknown truth. Is the analysis a jackknife analysis or a bootstrap analysis? Pg. 18) The authors claim that"the results of the resampling suggest that the guild-based method of habitat valuation is robust and efficient". The results do not suggest that. The authors claim that the method is"relatively objective",but the authors have made very subjective decisions about how the three measures are combined into one measure. To be objective,the authors must either use commonly accepted combination methods or justify the combination approach they used. Comments of Adrian Farmer-Avian Ecologist and Modeler The habitat quality index(based on density,diversity and mean size)may be dependent on two spatial factors that are not considered in the model. These are:juxtaposition to other habitat types;and scale. Of the two spatial factors,patch size is problematic if the model is intended to provide the quality(HSI)index in HEP to produce HUs. The computation of HUs as a product of quality and area is not meaningful if quality is dependent on area. This model would not be useful for performing impact assessments unless the list of habitat types addressed is exhaustive and can be used to characterize all conditions one is likely to encounter in the future,with and • without the project(i.e., every point in space belongs to one habitat type or another, and there is no variation within a habitat type). The paper does not state this,but this is an assumption that'should be stated clearly, and early in the paper. Comments of Jim Terrell - Aquatic Ecologist and Modeler The basic idea is sound: use data on fish species composition and relative abundance to develop a community 114 metric for individual habitat types. Whether or not this(or any other)population based metric represents"habitat value" depends solely on the point of view of the person asking the question. The authors claim they have a robust,objective technique. "Objective" is a misnomer. Numerous, very subjective, decisions had to be made to select the various data manipulation and aggregation techniques used to develop the index. The method is replicable;that does not make it"objective." There is nothing objective about saying(p 13) that it is "undesirable" to emphasize abundance at the expense of diversity. The "desirability" of such characteristics are human values that depend on the proposed use of the index. It has very little to do with biology and could just as easily be replaced by a goal to emphasize abundance of a species championed by a very powerful political lobby. Providing an index with (or without)such characteristics is thriven by the intended application. The method is not"fairly free of subjective bias" (p. 27), rather it is a mathematical formulation of an index designed to meet subjective goals.Equitably rating"diversity"and"abundance" is a subjective goal, accomplishing it through a square root transformation, (or a cube root.transformation for that matter)does not yield an objective index,there is no universal truth against which the results can be compared. The transformation simply provides a replicable description of the bias. There is no need to waste time selling the method as"objective." If potential users of the method are willing to trade off habitats at the ratios suggested by the habitat type scores, then fine. However,based on my past experiences with environmental planners looking for HEP and HSI models (which the authors erroneously claim arc incapable of making cross habitat comparisons) to get them off the hook in the inherently subjective problem of cross habitat tradeoffs, objections will be raised by certain interest groups that the values provided by this rating system are"biased"because they do not meet their needs. The best way to insure that the system will be used in the decision making process is to be up front about what it does not address. One rating system does not have to provide all of the answers. • VANTUNA RESEARCH GROUP OCCIDENTAL COLLEGE MOORE LABORATORY OF ZOOLOGY • LOS ANGELES, CALIFORNIA 90041 John S. Stephens, Jr. Daniel J. Pondella, II Executive Director Director (805)546-0443 (213)259-2955 or 2891 Fax: (213)259-2887 August 20, 1998 Rey Farve CESPL-PD-RQ P.O. Box 2711 Los Angeles, CA 90053-2325 Rey, In reviewing the U.S. Geological Survey's memo to John Hanlon, I do not find any of the objections particularly ground shaking nor problematic. In fact, most of them have already been dealt with based upon the anonymous review which I have received post distribution of this manuscript to you from the NMFS. Also, I think it is important to note none of the reviewers claim that the technique/model is poor in fact it is considered a `reasonable approach'. All of the reviews are generally constructive pointing out • semantic problems, questions of objectivity, and references to cross-habitat comparisons in HEP. Constructive reviews of manuscripts are always helpful and provide the valuable feedback that is the cornerstone of the scientific review process. I will approach this memo by discussing some of the major points and then go through the specific points one at a time. As we have seen with the multiple meetings and reviews by all of the agencies involved with this project and the seemingly endless reviews of this model, there has yet to be an objection to this approach in evaluating marine habitat quality. The most pertinent point is that there are other factors, which are important in management decisions concerning the marine environment. Certainly this is the case. However, it is not appropriate for us to include multiple constituencies various concerns and ideas about what is valuable in the marine environment for this model. The intention of the model is to provide a valuable tool, which will enable scientists, resource managers and inevitably politicians, a means for answering various habitat questions. Other concerns about habitat quality including water or sediment quality, endangered, commercial or sport species and other taxonomic groups to name a few are perfectly valid concerns for evaluating relative marine habitat values. However, weighting habitat values according to a particular constituencies interest, we feel is undesirable because these weighted values may change over time and we cannot predict that change. For instance, an endangered species may not always be endangered. If a species becomes delisted, does that change the inherent value of the habitat? Or, a species which was not of commercial interest in the past,may • become valuable in a future fishery. Do we change the habitat value? It is not feasible to account for these qualitative changes based upon subjective interests. On a more practical note adding more parameters to the model will increase the variability and reduce its applicability. Its value is in the fact that we can go out and sample the marine environment and then be able to set the data we collect in the appropriate context. The next step would be to add various constituencies concerns when deciding upon habitat quality. That step is the subjectivity, which we are avoiding. A common thread throughout the reviews, is the question of objectivity in the approach. While certain `subjective' decisions are made in the development of any model, it is not possible to get around this fact. Without making decisions and trying new things science does not go forward. Thus all new models are slightly subjective. Taken literally it appears that we should become so objective that we should not try anything new or use `commonly accepted combination methods' (please see Brian Cole's last comment). I do not think that this is the intent of the reviewers. What needs to be done is a clarification of what we mean by objective, and the decisions, which were made in setting up the model. The model is still objective, in that, the decisions to run the model are set out ahead of time and the results stand for themselves. The model and the assumptions that were made in its implementation were not modified based upon the results. The question about objectivity appears to be more of a semantic concern, which is easily resolved and really does not change the results of the paper. The other common thread is the reference to HEP as a tool for cross-habitat analyses. Not being a HEP expert myself, this just an introductory comment and it also does not weigh on the results of this paper or the efficacy of the model. I think that this reference to HEP can easily be edited out of the manuscript without any deleterious effects. • Since the submission of this,manuscript for publication Dr. Stephens has added additional databases to the analysis including more wetland data and offshore oil platforms. This information has broadened, strengthened and enhanced the applicability of this method for what are undoubtedly very complex marine habitat questions. Take care, Dan Pondella Director Vantuna Research Group Occidental College The memo from the USGS is dealt with in the order in which the comments are presented. First off, the cover letter apparently perceives our model as being some type of biodiversity measure derived from 'density, diversity and mean size'. Diversity is not a measure included in the analysis. We used density, fidelity and mean size. This misrepresentation appears to stem from Adrian Farmer's first comment. We specifically designed the model to not be a measure of biodiversity. We knew, a priori, that specific marine habitats have varying degrees of diversity. With this in mind, we did not want to weight diversity as the indicator of environmental quality. We go to great lengths to keep diversity from influencing the results of the analysis. The positive regression between the number of guilds (a proxy of diversity) and mean product (Figure 9) is a demonstration of the model not being driven by a particular parameter. Not weighting habitats by diversity, in my view, is one of the real strengths of our analysis. Also mentioned in the cover letter is the question of objectivity, which I have already discussed. Richard Stiehl's comments: Pg. 3-An unhealthy environment, for instance a heavily polluted environment, will not • have any fish in it or anything else for that matter. As environmental quality increases so does the health of the community. This is basic ecology and going into a discussion of why it is important to have fishes in marine environments is not appropriate for a publishable manuscript of this scope. There are numerous reasons for using fish assemblage data as an indicator for environmental quality. Fishes occupy numerous trophic levels and ecological niches, they relatively easy to sample and taxonomically straight forward and the data are relatively available. Also, measurement of biomass (directly correlated with lengths) is easily determined. Further, there is an incredible wealth of this type of fisheries data in the literature. In order to use organisms as a measure of habitat quality in the marine environment there are basically five groups which can be examined: fishes, macroinvertebrates, macroalgae, plankton and infaunal organisms. Only fish data crosses all the habitats that we are studying and is reasonably useful for this type of comparison. Certainly other aspects of the community could be used. Pg. 12-We did not want to weight any parameter over another because we did not have any reason to think that one parameter is more important than another is. Weighting would be more subjective than the assumptions we already used. I do not know of any criteria that differentiate between density, fidelity and mean standard length (biomass)for marine habitats. All are important so we just weighted them equally. Pg. 19-these are both semantic/typo questions which are addressed in the final • manuscript. Density has been confused with abundance. The point of Figure 11 is that the habitat value is tractable throughout this 10-year study 0. period for the King Harbor data set. It is important to note that it does not mirror just the fluctuations in density because of the other two parameters in the model. We know that there are great fluctuations in density both seasonal and annual in the marine environment. Demonstrating that the model is robust enough to overcome these fluctuations in one parameter strengthens its applicability. We never state that fish abundance or density is an accurate measure of`value'-that is one of the important points of the paper. Pg. 22-Figure 11 is only from diver transects and it does not suggest that the method is correlated with diver transects or trawl data. Please see comments on Figure 11 above. Our interpretation of`orders of magnitude' is based upon factors of ten. For example, 100% is an order of magnitude greater than 10% and 1000% is an order of magnitude greater than 100%. 30-85% are all the same order of magnitude. I've never heard of the term used any other way. Pg. 24-Yes, this is correct. We may need to clarify this a bit. Brian Cade's comments: Pg. 2) please see above. Pg. 13) We did not want one parameter to drive the model, perhaps we need to be more clear on this point. We go to lengths in the analysis to see that this does not happen. Both abundance and diversity are important parameters in marine systems and we did not want one parameter to swamp out the other. Certain habitats can have higher diversity and low abundance while others have low diversity and high abundance. To weight one over the other is a subjective choice (which we did not want to make) concerning the relative value of those parameters. A density of five fish per hectare was used as a cut off because guilds with such low densities were so rare that their inclusion would not affect the analysis. The measure would be the same. The habitat value should correlate with the number of fish guilds (as a proxy for diversity) so that it accurately represents habitats of various guild numbers. In this manner,the habitat value does not weight preferentially habitats that have more or less diversity. We wanted to stay away from this (see above response pg. 13) This refers to Figure 10, confidence limits are not given in Figure 11. I queried Alan Bond directly about this point, he had written bootstrapping in the original manuscript, but what he did is technically a resampling technique without replacement. This is jackknifing and we hadn't caught all of the changes before the manuscript went • out. Pg. 16 This is an interesting idea, but not the approach we chose for this manuscript. Please see Pg. 15 response. Pg. 18-this sentence should probably be rewritten. The method has proved to be robust and efficient for a variety of reasons. The resampling technique is merely a tool for setting confidence limits and other more traditional statistical parameters on our habitat values. This is important for numerous reasons including cross-habitat comparisons and the evaluation of our comparison sites, which were sampled for only a year or two in most cases. Without the confidence limits, it becomes very difficult to interpret the results. How can we do something `novel' and 'use commonly accepted combination methods'? Please see the previous comments on objectivity. Adrian Farmer's comments: The habitat quality index is not based on diversity. There always are other factors, which may not be included in a model; otherwise it would not be a model. These two spatial factors:juxtaposition and scale are more applicable to terrestrial habitats. However,the question of juxtaposition of habitat types is solved by the guilds because the guilds cross habitats. The scale of different habitats is in part resolved by the standardization of the density measures. Further to weight habitats variably by their scale in the southern California bight is adding a subjective parameter to the model, which we did not have any reason to do. We have used an exhaustive list of habitat types (past, present and foreseeable future) in the southern California bight in this paper. We certainly can be clearer about this in the paper(good point). Jim Terrell's comments: As I have stated more than once in the presentation of this model and have added to the final manuscript, the model is simply another tool and not the final word on habitat valuation. Different constituencies have various points of view on what is important in the marine environment. Obviously all of these ideas are important but do not want to weight the model towards a particular point of view. Especially when we know that ideas of what people view as important in environmental valuation may change over time. • Please refer to the comments on objectivity. Absolutely correct (please see first Terrell response) however we do need a tool for S-. valuating marine habitats and this is an effective method for that purpose. There are other things that can be considered in environmental quality, but they all cannot be put in one model or analysis. .- _ G �.