Printer Friendly

Modeling common trends of zooplankton, atherinopsid, and barred splitfin time series in a shallow tropical lake.

Silversides and whitefish of the genus Chirostoma (Atherinopsidae) comprise the most-important native fishery resource in Lake Chapala, Jalisco-Michoacan, Mexico. These fish species are mainly endemic to the Mexican Central Plateau (Barbour, 1973a, 1973b, 2002). The atherinopsids represent a species flock due to high levels of sympatry, morphological evolution, and species diversity in a relatively small geographic area (Barbour and Chernoff, 1984; Echelle and Echelle, 1984; Greenwood, 1984). In Lake Chapala, the atherinopsid flock is composed of six zooplanktivorous species of silversides: charal (Chirostoma jordani); smallmouth silverside (Chirostoma chapalae); ranch silverside (Chirostoma consocium); sharpnose silverside (Chirostoma labarcae); scowling silverside (Chirostoma aculeatum); and Ajijic silverside (Chirostoma contrerasi). There are three whitefish that ontogenetically shift from zooplanktivores to predatory species: longjaw silverside (Chirostoma lucius); bigmouth silverside (Chirostoma sphyraena); and blacknose silverside (Chirostoma promelas). The scowling silverside is currently a threatened species due to habitat degradation (mouth of the polluted Lerma River; Bloom et al., 2008). Chirostoma species flocks have been of great economic and ethnic importance for centuries; nonetheless, little is known about their basic biology. Further knowledge is critical for their conservation and for developing rational fisheries management programs. Atherinopsids have been the focus of distribution and abundance studies (Becerra-Mufioz et al., 2003; Ruiz-Gomez et al., 2008; Arce et al., 2011).

Modeling efforts of Lake Chapala fisheries have included time series analysis of fish density and biomass time series of atherinopsids and barred splitfin (Becerra-Munoz et al., 2003). However, time series modeling did not provide any information on common trends among several fish density and biomass time series in Lake Chapala. Conversely, there is a statistical technique that detects common trends among multiple concurrent time series. Dynamic factor analysis (DFA) is a statistical technique for the analysis of relatively short (>15-25 points) multivariate time series; it identifies trends for a set of time series with a smaller number of common trends (Zuur et al., 2003a, 2007). For example, DFA has been applied to estimate underlying common trends among multiple time series of catch per unit effort of Norwegian lobster Nephrops norvegicus (Zuur et al., 2003b) and squid Loligo forbesi (Zuur and Pierce, 2004) with sea surface temperature and North Atlantic Oscillation index as the explanatory variables.

The analysis of common temporal trends of atherinopsid densities and biomass, along with physicochemical variables, may shed light on the main environmental variables influencing their population dynamics throughout Lake Chapala. The interrelationships between spatiotemporal distribution patterns of physicochemical variables and fish densities and biomass are not known in Lake Chapala. It is important to know whether or not common trends are spatially coherent throughout Lake Chapala because this information could provide a basis for improved atherinopsid stock assessment and fisheries management. The goals of the present study were to (1) identify common trends in zooplankton and fish density and biomass time series in Lake Chapala, and (2) elucidate the importance of water temperature, electrical conductivity, dissolved oxygen, pH, and alkalinity on these common trends.


METHODS--Study Site and Data Collection--From June 1997--February 1999, we implemented a survey for zooplankton, barred splitfin, and atherinopsids throughout Lake Chapala. Zooplankton and fish were collected monthly at 15 spatially systematic, equidistant sampling stations (except station [ST] 3) in Lake Chapala (Fig. 1). We used a zooplankton net with a 0.188-[m.sup.2] mouth area and a mesh size of 240 pm. The net-zooplankton samples were fixed in the field with a 70% ethanol solution. For the fish, we utilized a midwater trawl with a 6.6-[m.sup.2] mouth area and a 1.2-cm stretched mesh from the mouth to the cod end. A General Oceanics flowmeter (General Oceanics, Miami, Florida), suspended inside the zooplankton net, was set at the start of zooplankton collection and net trawling, and then read when the zooplankton net and trawl net reached the boat to estimate water flow through the zooplankton net and trawl net. Volume of water filtered was calculated as the product of estimated tow velocity and mouth areas of the zooplankton and trawl nets, respectively. Collected zooplankton was thoroughly washed and preserved with 70% ethanol in the field. Fish were placed in bags inside ice coolers. In the laboratory, fish were counted and their biomass was estimated as wet weight. Because thousands of atherinopsids were caught, the Chirostoma flock was analyzed as a group because their proper identification to species had to be done under a stereoscopic microscope. Zooplankton samples were measured for the total settled volume of all zooplankton collected by pouring each sample into an Imhoff cone and measuring the volume of zooplankton which had settled to the bottom of the cone overnight. Water temperature and dissolved oxygen were measured at the surface and bottom of each of the 15 stations with a YSI oxygen meter (Xylem, Inc., Yellow Springs, Ohio) with a long sensor cable calibrated in meters. Conductivity was recorded with a YSI conductivity meter (Xylem, Inc.); pH was measured at approximately 25 cm below the surface with a Corning pH-meter (Corning Life Sciences, Tewksbury, Massachusetts); and total alkalinity was estimated in the lab (American Public Health Association, 1985) from water samples collected approximately 25 cm below the surface that had been stored in ice coolers.

Statistical Analysis--Multiple time series were modeled as a function of a constant, a linear combination of 'M' common trends, explanatory variables, and noise. Instead of multiple 'N' time series, a linear combination of M common trends was estimated where M < N. The M trends represent tendencies (joint signals) across all the time series (Zuur et al., 2007). Multiple time series were selected and each group of selected time series was standardized (i.e., zero mean and standard deviation of 1). Different models containing one or two common trends were applied to standardized time series as well as models with and without explanatory variables. DFA models with minimum Akaike's Information Criterion were selected as best representing common trends (Zuur et al., 2007). Analyses were carried out in Brodgar software (DFA version 2.7.2, Highland Statistics, Inc.,

RESULTS--We identified very similar common trends for the zooplankton and atherinopsid time series, which peaked during the winter. The common trends for density and biomass for atherinopsids were different temporally and peaked during the winter and summer, respectively. There were two common trends among the barred splitfin time series. The first common trend for barred splitfin density and biomass were remarkably similar, peaking during the summer, but the second trend exhibited an opposite behavior through time (Fig. 2).

The best DFA models per station had one common trend and one or two explanatory physicochemical variables, except for one near Mismaloya (ST 7), where all studied physicochemical variables were selected as explanatory. The order of importance of explanatory physicochemical variables to biological time series was defined by the selected DFA models per sampling station. We found that DFA methodology was useful to assess the overall importance of measured physicochemical variables on zooplankton and fish time series throughout Lake Chapala. Conductivity and pH were selected for stations near the cities of Jocotepec (ST 1) and Tizapan (ST 9). Alkalinity was the best explanatory variable for biological series near San Juan Cosala (ST 2). Water temperature was selected as the best explanatory variable near the cities of San Luis Soyatlan (ST 3), Chapala (ST 6), Emiliano Zapata (ST 11), Agua Caliente (ST 12), and Petatan (ST 15); conductivity and dissolved oxygen were best explanatory variables for Ajijic (ST 4) and west of Mezcala Island (ST 8); and conductivity was the best explanatory variable for the station near Palo Alto (ST 13). Alkalinity and conductivity were selected for DFA models near San Pedro Itzican (ST 10) and alkalinity and dissolved oxygen were selected for DFA models at San Nicolas (ST 5) and Ocotlan (ST 14). All measured physicochemical variables were selected for the DFA model for Mismaloya (ST 7).


DISCUSSION--Multiple time series can be considered strongly coherent, weakly coherent, or noncoherent. Lack of coherence is characterized by more than one common trend among time series, with only a few time series associated with each trend. Under strongly coherent survey time series, DFA usually identifies only one common trend to describe all series (Nye et al., 2010). Zooplankton density and atherinopsid density and biomass time series were strongly coherent in Lake Chapala, and barred splitfin density and biomass time series were weakly coherent. Zooplankton and atherinop sid series had only one distinctive common trend, which was both temporally similar and strongly coherent throughout Lake Chapala. Consequently, environmental policy for improved stock assessment and atherinopsid fisheries management can be done knowing that these biological series behave homogeneously temporally and spatially. The common trends in zooplankton and atherinopsid density were synchronized. The common trend of atherinopsid biomass time series peaked months after the common trend of zooplankton time series. Hence, other food resources consumed by atherinopsids, such as suspended clay-organic matter-bacteria aggregates (Lind and Davalos-Lind, 1991; Davalos-Lind and Lind, 2005), could be responsible for an additional fish biomass increase when zooplankton density exhibited a downward tendency. Hence, additional studies are necessary to investigate those other variables which might influence zooplankton and fish density dynamics in Lake Chapala. We identified patterns in the zooplankton and fish density time series that could not have been revealed by other, more-traditional approaches.

The Lake Chapala fisheries research survey was supported through grant IPN-CEGEPY 970468 from the Mexican National Polytechnic Institute. We are thankful to F. Villalpando Barragan for providing the map of Lake Chapala.


AMERICAN PUBLIC HEALTH ASSOCIATION. 1985. Standard methods for the examination of water and wastewater. Sixteenth edition. APHA, Washington, D.C.

ARCE, E., J. ALCOCER, X. CHIAPRA-CARRARA, AND L. ZAMBRANO. 2011. Hydroacoustic study of spatial and temporal distribution of Poblana alchichica (De Buen, 1945) in Lake Alchichica, Mexico. Canadian Journal of Fisheries and Aquatic Sciences 6:624-632.

BARBOUR, C. D. 1973a. The systematics and evolution of the genus Chirostoma Swainson (Pisces, Atherinidae). Tulane Studies of Zoology and Botany 18:97-141.

BARBOUR, C. D. 1973b. A biogeographical history of Chirostoma (Pisces, Atherinidae): a species flock from the Mexican plateau. Copeia 3:533-556.

BARBOUR, C. D. 2002. Chirostoma contrerasi (Atherinopsidae, Menidiinae), a new species from Lago de Chapala, Mexico. Pages 23-33 in Libro Jubilar en honor al Dr. Salvador Contreras Balderas (M. L. Lozano-Vilano, editor). Universidad Autonoma de Nuevo Leon, Monterrey, Mexico.

BARBOUR, C. D., and B. Chernoff. 1984. Comparative morphology and morphometrics of the pescados blancos (genus Chirostoma) from Lake Chapala, Mexico. Pages 111-127 in Evolution of fish species flocks (A. A. Echelle and I. Kornfield, editors). University of Maine, Orono, Maine.

BECERRA-MUNOZ, S., H. R. BUELNA-OSBEN, AND J. M. CATALAN-ROMERO. 2003. Spatial patterns of ARIMA modeled rates of change of atherinids (Chirostoma spp.) and goodeid Chapalichthys encaustus from Lake Chapala, Mexico. Ecological Modelling 165:237-250.

BLOOM, D. D., K. R. PILLER, J. LYONS, AND N. MERCADO-SILVA. 2008. Threatened fishes of the world: Chirostoma aculeatum Barbour, 1973 (Atherinopsidae). Environmental Biology of Fishes 82:221-222.

DAVALOS-LIND, L., AND O. T. LIND. 2005. Bacteriozooplankton grazing by fish and atherinopsid in clay-rich and clay-free water. Verhandlungen des internationalen Verein Limnologie 29:140-142.

ECHELLE, A. A., AND A. F. ECHELLE. 1984. Evolutionary genetics of a "species flock:" atherinid fishes on the Mesa Central of Mexico. Pages 93-110 in Evolution of fish species flocks (A. A. Echelle and I. Kornfield, editors). University of Maine, Orono, Maine.

GREENWOOD, P. H. 1984. What is a species flock? Pages 13-19 in Evolution of fish species flocks (A. A. Echelle and I. Kornfield, editors). University of Maine, Orono, Maine.

LIND, O. T., AND L. DAVALOS-LIND. 1991. Association of turbidity and organic carbon with bacterial abundance and cell size in a large turbid tropical lake. Limnology and oceanography 36:1200-1208.

NYE, J. A., A. BUNDY, N. L. SHACKELL, K. D. FRIEDLAND, AND J. S. LINK. 2010. Coherent trends in contiguous survey time series of major ecological and commercial fish species in the Gulf of Maine ecosystem. ICES Journal of Marine Science 67:26-40.

RUIZ-GOMEZ, M. D.,J. F. MENDEZ-SANCHEZ, F. D. RODRIGUEZ-ROMERO, AND C. M. TAYLOR. 2008. Spatiotemporal changes in fish assemblages of Los Terreros Creek, an isolated stream system in headwaters of the Lerma River, Central Mexico. Southwestern Naturalist 53:224-229.

ZUUR, A. F., R. J. FRYER, I. T. JOLLIFFE, R. DEKKER, ANDJ. J. BEUKEMA. 2003a. Estimating common trends in multivariate time series using dynamic factor analysis. Environmetrics 14:665-685.

ZUUR, A. F., E. N. IENO, AND G. M. SMITH. 2007. Analyzing ecological data. Volume 1. Springer Science Business Media, New York.

ZUUR, A. F., AND G. J. PIERCE. 2004. Common trends in Northeast Atlantic squid time series Journal of Sea Research 52:57-72.

ZUUR, A. F., I. D. TUCK, AND N. BAILEY. 2003b. Dynamic factor analysis to estimate common trends in fisheries time series. Canadian Journal of Fisheries and Aquatic Sciences 60:542552.

Submitted 25 March 2015.

Acceptance recommended by Associate Editor, Mark Pyron, 21 January 2016.

Salvador Becerra-Munoz, * Hector R. Buelna-Osben, Juan M. Catalan-Romero

Pacific States Marine Fisheries Commission, P.O. Box 10, La Grange, CA 95329 (SBM) IPN-CIIDIR Michoacan, Justo Sierra 28 Centro, Jiquilpan, Michoacan, 59510, Mexico (HRBO, JMCR)

* Correspondent:
COPYRIGHT 2016 Southwestern Association of Naturalists
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2016 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Becerra-Munoz, Salvador; Buelna-Osben, Hector R.; Catalan-Romero, Juan M.
Publication:Southwestern Naturalist
Article Type:Report
Geographic Code:1MEX
Date:Mar 1, 2016
Previous Article:Small mammals of Pyramid Lake and Anaho Island (Nevada).
Next Article:Distribution and density of zebra mussels in four Kansas reservoirs.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters