Regressions for Estimating Gastropod Biomass with Multiple Shell Metrics Author(S): Adam Obaza & Clifton B
Total Page:16
File Type:pdf, Size:1020Kb
Regressions for Estimating Gastropod Biomass with Multiple Shell Metrics Author(s): Adam Obaza & Clifton B. Ruehl Source: Malacologia, 56(1–2):343-349. 2013. Published By: Institute of Malacology URL: http://www.bioone.org/doi/full/10.4002/040.056.0224 BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, and environmental sciences. BioOne provides a sustainable online platform for over 170 journals and books published by nonprofit societies, associations, museums, institutions, and presses. Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of BioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use. Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries or rights and permissions requests should be directed to the individual publisher as copyright holder. BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research libraries, and research funders in the common goal of maximizing access to critical research. 0$/$&2/2*,$ 2): 343349 REGRESSIONS FOR ESTIMATING GASTROPOD BIOMASS WITH MULTIPLE SHELL METRICS Adam Obaza1,2* & Clifton B. Ruehl1,3* INTRODUCTION power and provide more accurate biomass es- timates for snails because they exhibit a diver- Snails are among the most important primary sity of shapes. Comparisons among possible consumers in freshwater ecosystems. They eat regression models may reveal that different large amounts of periphyton, facilitate nutrient metrics or combinations of metrics are better regeneration for the resources they consume, predictors of a particular mass measure for a and in ecosystems with greater than two trophic given group (e.g., AFDM vs. wet mass). OHYHOVVQDLOVDUHDYLWDOFRQGXLWIRUHQHUJ\ÀRZ In this study, we consider three common, to upper trophic levels by serving as prey to nu- structurally diverse freshwater snail genera merous invertebrate and vertebrate predators Planorbella, Pomacea and Haitia to examine (Dillon, 2000). Estimates of individual biomass the utility of using multiple length measures to (mass/individual) enable calculations of stand- predict multiple soft-tissue mass measures. ing stock (mass/area) and production (mass/ Planorbella (Planorbidae) are pulmonate snails area/time). These measures are superior to with planispiral shells that grow sinistrally (left GHQVLW\ QRDUHD IRUDVVHVVLQJWKHLQÀXHQFH handed). Haitia (Physidae) are small pulmo- of animals on communities and ecosystems nates with a sinistrally growing conispiral shell. because they incorporate size-structured Pomacea (Ampulariidae) are large caenogas- DVSHFWVRIHQHUJ\ÀRZ 5HJHVWHUHWDO tropods that have globose, conispiral shells Rosas-Luis et al., 2008). that are generally dextral (right handed) with Length-mass regressions enable quick DQRSHUFXOXP$GGLWLRQDOO\ZHUHSRUWWKH¿UVW estimation of individual biomass to assist length-to-mass regression for 3RPDFHDÀDJHO- estimating standing stock and production. lata Say, 1827. Regressions provide mass estimates without VDFUL¿FLQJDQLPDOVZKLFKPD\QRWEHIHDVLEOH for protected taxa or desirable because repeat- METHODS edly sampling the same population can cause bias (Cahill et al., 2002). Obtaining biomass We collected P. duryi Wetherby, 1879 data from preserved specimens is facilitated ±PPVKHOOOHQJWKn = 89) and H. by regressions because length is a relatively cubensis 3IHLIIHU ± PP VKHOO stable measure compared to mass (Wetzel et length, n = 30) from populations maintained al., 2005). However, historically most studies of at the Daniel Beard Research Center in Ev- freshwater snails report density without mea- erglades National Park. These populations VXUHVRIELRPDVV *DUQHU +DJJHUW\ originated from collections made throughout Ruehl & Trexler, 2011). Among those studies the Everglades and were supplemented an- that report biomass data, few report regres- nually with wild-caught individuals. Pomacea sions that include more than one measure of paludosa 6D\ ±PPn = 19) length (e.g., shell length or width) or mass (e.g., were obtained from live-caught animals and wet mass or ash-free dry mass, a measure of their offspring. Pomacea paludosa sample size organic content) and rarely combine multiple was comparatively small, but we believe it is measures of length to improve mass estimates VXI¿FLHQWIRURXUSXUSRVHEHFDXVHSUHOLPLQDU\ HJ(FNEODG%RHUJHU&ROOLQV UHJUHVVLRQVLQGLFDWHGWKDWPRGHO¿WZDVYHU\ %HQNHHWDO*RQ]iOH]HWDO good. 3RPDFHD ÀDJHOODWD ± PP n = Deichmann et al., 2008). Combining multiple 49) were collected from Mexico (Sian Kaan, length measures should improve predictive Quitana Roo, 19.20°N, 87.89°W) and Belize 1Department of Biological Science, Florida International University, 3000 NE 151st Street, North Miami, Florida 33181, U.S.A. 2Ocean Associates, Inc., 501 W. Ocean Blvd, Suite 4200, Long Beach, California 90802, U.S.A. 3Department of Biology, Columbus State University, Columbus, Georgia 31907, U.S.A. *Corresponding author: [email protected] 343 344 OBAZA & RUEHL FIG. 1. Graphical description of how aperture width and length, and shell length were mea- sured P. duryi. Shell length is the lateral distance from anterior to posterior, aperture length is the distance across the aperture from anterior to posterior, aperture width is the length across the aperture at its widest point using calipers to 0.01 FIG. 3. Graphical description of how aperture and PP6FDOHEDU PP operculum width, length, and shell length were measured for P. paludosa and 3ÀDJHOODWD. Shell length is the lateral distance from anterior to pos- terior, aperture length is the distance across the &UDE&DWFKHUODJRRQ2UDQJH:DON1 aperture from anterior to posterior, aperture width : )DFLOLWLHVZHUHQRWDYDLODEOHWRJHW is the length across the aperture at its widest point dry and ash mass of 3 ÀDJHOODWD. All snails and opercula length and width that correspond with aperture length and width measurements ZHUHVDFUL¿FHGE\IUHH]LQJIRUK 1LHOVHQ using calipers to 0.01 mm. Scale bar = 10 mm. Gosselin, 2011). Shell measurements were made to the near- est 0.01 mm with calipers and weighed to the furnace and reweighed. We subtracted the nearest 0.0001 g. Shell length and aperture ashed mass from the dried mass to calculate width and length were measured on Planorbella AFDM. (Fig. 1) and Haitia (Fig. 2). We measured shell We quantified the capacity for multiple length, aperture width and length as well as length measures to predict mass compared operculum width and length on P. paludosa to single measures using a model selection (Fig. 3). After measurements, the soft tissue procedure with regression. Regressions were was separated from the shell with forceps. Soft used to determine relationships between mass tissue and shell were weighed and then dried (wet, dry, and AFDM) and shell size (length, DW&WRDFRQVWDQWPDVV'ULHGVRIWWLVVXH width, operculum length, operculum width). shell, and opercula were weighed and then :HWHVWHGDOOPHWULFFRPELQDWLRQVWR¿QGWKH FRPEXVWHGDW&IRUIRXUKRXUVLQDPXIÀH best model for each measure of mass for each species. Models were compared using the information-theoretic approach proposed by Burnham & Anderson (2002). Each model was ranked relative to the others using second order Akaike’s Information Criterion for small sample sizes (AICc) where the lowest value LQGLFDWHG WKH EHVW ¿W7KH$,&c calculation adds another parameter to the AIC formula WKDWIXUWKHUSHQDOL]HVPRGHO¿WIRULQFOXGLQJ more independent variables. Models were considered equal if the difference between FIG. 2. Graphical description of how aperture the calculated AICc value and the lowest AICc width and length, and shell length were measured YDOXHIRUHDFKGDWDVHWZDV &RPSWRQHW for H. cubensis. Shell length is the lateral distance al., 2002). Model weights facilitated better from anterior to posterior, aperture length is the distance across the aperture from anterior to UHVROXWLRQRIPRGHOVZLWK$,&FYDOXHV'DWD posterior, and aperture width is the length across were natural-log (ln) transformed because a the aperture at its widest point using calipers to plot of mass by length suggested a power 0.01 mm. Scale bar = 3 mm. relationship. BIOMASS FROM MULTIPLE LENGTH METRICS 345 RESULTS AND DISCUSSION measurements changed depending on the mass measurement, but always included The information-theoretic approach revealed shell length. Wet mass was predicted best by that in many cases, multiple measures of shell shell length and aperture length or width for morphology predicted tissue mass better than P. duryi (Table 1). Similarly, a combination of single shell measurements. Model weights, the shell length and aperture width, or these and relative probability of model accuracy, indicated aperture length, predicted wet mass best in that inclusion of a length and width parameter H. cubensis. However, models that included provided more robust models (Table 1). Even only shell length and aperture width predicted ZKHQWKHǻ$,&FYDOXHVIRUPRGHOVZLWKRQO\ dry mass best for P. duryi, while shell length one parameter were low, the model weight with aperture length best predicted dry mass was correspondingly low meaning that it was in H. cubensis. Finally, a combination of shell unlikely that single parameter models provided length and aperture width best predicted AFDM WKHEHVW¿W,QFOXVLRQRIH[WUDSDUDPHWHUVDOVR for both P. duryi and H cubensis. Therefore, increased the R2 value (Table 2). For the two regardless of