LIFE HISTORY VARIATION AND ARTISANAL FISHERY ASSESSMENT OF THE GIANT CICHLID (Petenia splendida Gunther 1862) IN GUATEMALAN LAKES

By

CHRISTIAN ALBERTO BARRIENTOS CONTRERAS

A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2015

© 2015 Christian Alberto Barrientos Contreras

To my parents for their strong belief in me and to my wife and daughters who go together with me in my life journey

ACKNOWLEDGMENTS

I thank my advisor Jeff Hill for his understanding, encouragement, and perseverance with me in my journey to develop my scientific skills and the research needs of Guatemala. I appreciate his time and effort in the peaks and troughs of my writing.

I am extremely grateful to all the members of my dissertation committee: Dr. Debra

Murie, Dr. Mike Allen, Dr. Mark Brenner, and Dr. Collete St. Mary, for their generous contributions in the form of advice, expertise, and good-natured support to better my work.

My deep appreciation goes to my Guatemala colleagues, especially Yasmin Quintana and

Diego Elias, because enthusiasm and dedication to the work as well as the humorous stories made the data collection such an enjoyable experience. Several other people participated in my data collection including Diego Juarez, Byron Cruz, Obet Gonzalez, Don Julio Chan, and

Gustavo Orellana. I received help with permits and logistics from Francisco Castaneda Moya,

Mercedes Barrios, Julio Madrid, Emilio Mattus, Luis Guerra, Erick Villagran, Julio Morales, and

David, Kelsey and Rosita Kuhn. My parents provided lodging and transportation in Guatemala

City, which was greatly needed and I cannot ever thank them enough. My brother and sister also contributed their time and ran errands on several occasions, and their efforts are also recognized.

I also thank the funding sources without which my Ph.D. program would not have been possible. My stay in the US was supported by the University of Florida through an Alumni

Award Scholarship from the Program of Fisheries and Aquatic Sciences and from a Teaching

Assistantship from the Biology Department. Funding for my field research was from the

National Council of Science and Technology in Guatemala and from the Lake Petén Itzá

National Management Authority. The National Council of Protected Areas facilitated personnel and equipment inside Yaxhá National Park, and the University of San Carlos, through the

Conservationist Studies Center, provided personnel and facilities in Petén. La Casa de Don

4

David provided a base for operations in El Remate, Petén and the National Conservation and

Environment Organization provided facilities in Guatemala City. I am grateful to all of them for their trust and support of this project. I also thank Dr. Jose Miguel Ponciano for several discussions about fish, numbers, Guatemala, swimming and life.

Finally, I owe thanks to the many people who helped in many ways, but who I am not able to list here. To them, I promise I will “pay it forward.”

5

TABLE OF CONTENTS

page

ACKNOWLEDGMENTS ...... 4

LIST OF TABLES ...... 8

LIST OF FIGURES ...... 9

ABSTRACT ...... 11

CHAPTER

1 THE GIANT CICHLID, LIFE HISTORY AND ARTISANAL FISHERY ...... 13

2 SEASONAL CHANGES IN THE FISH COMMUNITY THAT INHABITS THE LITTORAL SUBMERSED AQUATIC VEGETATION OF LAKE PETEN ITZÁ, GUATEMALA ...... 16

Introduction ...... 16 Methods ...... 19 Habitat Characterization and Plant Sampling ...... 19 Fish Sampling ...... 20 Analysis ...... 22 Results...... 22 Habitat Characterization ...... 22 Fish Communities ...... 23 Temporal Analysis ...... 23 Discussion ...... 24

3 REPRODUCTIVE LIFE HISTORY TRAITS OF THE GIANT CICHLID (Petenia splendida Gunther 1862) IN GUATEMALAN LAKES ...... 35

Introduction ...... 35 Study Site ...... 37 Environmental Factors ...... 38 Methods ...... 39 Fish Sampling ...... 39 Reproduction ...... 39 Batch Fecundity ...... 40 Maturity ...... 40 Results...... 42 Spawning Season ...... 42 Fecundity ...... 43 Maturity ...... 43 Discussion ...... 43 Sex Ratio and Size Distribution ...... 44

6

Spawning Season ...... 45 Fecundity ...... 46 Maturity Schedule ...... 47

4 AGE, GROWTH AND MORTALITY RATES OF THE GIANT CICHLID (Petenia splendida Gunther 1862) IN GUATEMALA ...... 58

Introduction ...... 58 Methods ...... 61 Fish Sampling ...... 61 Length Frequency ...... 62 Otolith Processing ...... 62 Age Estimates and Validation ...... 63 Growth ...... 64 Estimates of Total Mortality ...... 65 Results...... 66 Length Frequency ...... 66 Length-Weight Relationships ...... 66 Age Estimates ...... 67 Growth ...... 68 Mortality Estimates ...... 68 Discussion ...... 69 Length Frequency ...... 69 Age Estimates ...... 70 Growth ...... 71 Mortality ...... 73

5 ASSESSMENT OF ARTISANAL FISHERY POLICY FOR THE GIANT CICHLID (Petenia splendida Gunther 1862) IN GUATEMALAN LAKES ...... 88

Introduction ...... 88 Methods ...... 89 Results...... 92 Discussion ...... 93

EPILOGUE ...... 102

REFERENCES ...... 104

BIOGRAPHICAL SKETCH ...... 116

7

LIST OF TABLES

Table page

2-1 Mean and standard error of habitat characteristics from blocknet samples collected at Lake Peten Itzá in years 2009-2010...... 32

2-2 Mean ± standard error of biomass (g 100 m-2) of fish collected in blocknets in different months during 2009-2010 at Lake Peten Itzá...... 33

2-3 Mean and standard error of fish metrics from blocknet samples collected at Lake Peten Itzá in 2009-2010...... 34

3-1 Developmental stages of gonads in female giant cichlids from Lakes Yaxhá and Petén Itzá (adapted from Perez-Vega et al. 2006 and Nuñez and Duponchelle 2009)...... 49

3-2 Candidate models fitted to maturity schedule for Petenia splendida in Lakes Petén Itzá and Yaxhá. (*) indicate the used model...... 52

4-1 Age-Length key for giant cichlid from Lake Petén Itzá, Guatemala. Percentage of fish in 20-mm length categories as a function of age...... 84

4-2 Age-Length key for giant cichlid from Lake Yaxhá, Guatemala. Percentage of fish in 20-mm length categories as a function of age...... 85

4-3 Candidate models for different sex fitted to growth parameters estimated for giant cichlid in Lake Yaxhá. (*) indicate the used model...... 86

4-4 The von Bertlantffy parameters estimated for Petenia splendida by lake and sex...... 86

4-5 Average Total Length at observed age (±S.E) of giant cichlid from Lakes Petén Itzá and Yaxhá, Guatemala...... 87

4-6 Total mortality (Z), natural mortality (M) and annual survival rate (S) of giant cichlid captured in Lakes Petén Itzá and Yaxhá...... 87

5-1 Parameters for giant cichlid populations from two Guatemalan lakes, that were used in a yield-per-recruit model. Lake Petén Itzá is a lake with an artisanal fishery and Lake Yaxhá is unfished...... 97

8

LIST OF FIGURES

Figure page

2-1 Study area in Lake Petén Itzá, Guatemala, including sites sampled with blocknets (•)...... 29

2-2 Average monthly rainfall (mm) in Petén, Guatemala, in 2009 (solid line) and 2010 (dotted line)...... 30

2-3 Plant biomass and percentage of area covered (PAC) measured inside blocknets at Lake Peten Itzá, in five months 2009-2010...... 31

3-1 Map of the study area, showing geographic location of Lake Petén Itzá and Lake Yaxhá within the Maya Biosphere Reserve...... 50

3-2 Monthly rainfall (mm) and lake level (m) in Lake Petén Itzá, Guatemala in 2010...... 51

3-3 Length–frequency distribution for Petenia splendida immature (I), male (M) female (F) from Lake Petén Itzá (n=339) and Lake Yaxhá (n=327), Guatemala...... 53

3-4 Average gonadosomatic indices for female Petenia splendida in Lakes Petén Itzá and Yaxhá captured during July and October 2010 and March-November 2011...... 54

3-5 Batch fecundity for Petenia splendida females in Lakes Yaxhá (top) and Petén Itzá (middle), and for all lakes combined (bottom) from Guatemala ...... 55

3-6 Proportion of mature male and female Petenia splendida in Lake Petén Itzá (PI) and Yaxhá (Y), Guatemala, as a function of total length (TL)...... 56

3-7 Proportion of mature males and females of Petenia splendida in Lake Petén Itzá (PI) and Yaxhá (Y), Guatemala, as a function of age (years)...... 57

4-1 The Mayan Biosphere Reserve and the aquatic systems studied in Guatemala. Study lakes are in different management zones, with difference in fishing pressure...... 75

4-2 A whole sagittal otolith from giant cichlid captured at Lake Petén Itzá, showing the nucleus (N), an opaque (O) zone, and a translucent (T) zone...... 76

4-3 Otolith section of giant cichlid from Lake Petén Itzá. Opaque zones (OZ) were counted for age estimation. Translucent zones (TZ) occur between OZ. T ...... 76

4-4 Length–frequency distribution for immature (I), male (M) and female (F) giant cichlid from Lake Petén Itzá and Lake Yaxhá, Guatemala...... 77

4-5 Total body weight as a function of total length by month of giant cichlid from Lake Petén Itzá and Lake Yaxhá, Guatemala...... 78

4-6 Giant cichlid age frequency distribution collected in Lakes Petén Itzá (n=339) and Yaxhá (n=327) with electro fishing...... 79

9

4-7 Bias plot for age estimates from giant cichlid otoliths (n=103) from Lakes Yaxhá and Petén Itzá. Numbers above the read 1 (x-axis) are numbers of individuals (n). T ...... 80

4-8 Monthly percent occurrence of the opaque zone at the edge and the percent index of completion of the translucent zone of otoliths of giant cichlid ...... 81

4-9 Length at age observed for giant cichlid fitted to the von Bertalanffy growth model by sex from fish sampled in Lake Yaxhá and Lake Petén Itzá Guatemala...... 82

4-10 Catch curve analysis for giant cichlid from Lake Yaxhá and Lake Petén Itzá, Guatemala. Lake Yaxhá ages 0 -2 were not included in Z estimation ...... 83

5-1 Spawning potential ratio (SPR) on the harvest exploitation rate (U) and minimum length limits. Values of SPR below 0.3 are usually indicative of recruitment overfishing ...... 98

5-2 Spawning potential ratio (SPR) on the harvest exploitation rate (U) and minimum length limits. Values of SPR below 0.3 are usually indicative of recruitment overfishing...... 99

5-3 Yield (kg) isopleths plotted on harvest exploitation rate (U) and minimum length limits. Parameters for the model from Lake Petén Itzá ...... 100

5-4 Yield (kg) isopleths plotted on harvest exploitation rate (U) and minimum length limits. Parameters for the model from Lake Yaxhá...... 101

10

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

LIFE HISTORY VARIATION AND ARTISANAL FISHERY ASSESSMENT OF THE GIANT CICHLID (Petenia splendida Gunther 1862) IN GUATEMALAN LAKES

By

Christian Alberto Barrientos Contreras

May 2015

Chair: Jeffrey Hill Major: Fisheries and Aquatic Sciences

Life history traits are described in terms of factors such as reproduction, survival, growth, age and size at maturity, and are influenced by environmental variation. It is important to consider inter-population variations in fish life history characters for effective fisheries management. I examined the life history variations and juvenile ecology of the giant cichlid

(Petenia splendida) in Mesoamerican lakes that are critical in the overall distribution of the species. I studied the giant cichlid because it is the most important fishery species in the area, has many life history characteristics common to other cichlids, and exists across a wide range of ecological conditions, all of which make it a good model to investigate variations in life history traits. Despite the importance of life history traits and ecology to fisheries management, there have been few attempts to assess them in Mesoamerican freshwater systems. Giant cichlid populations are subject to a large range of exploitation and are considered overexploited in some systems. Nonetheless, there are some systems with minimal habitat transformation and no artisanal fishery, though they are few. I found that giant cichlid reproduces year-round. Thus, juveniles also use the habitat throughout the year. Interdemic life history traits that best differentiated populations from fished Lake Petén Itzá and unfished Lake Yaxhá were sex ratio, egg size, gonadosomatic index (GSI), growth, size and age frequency distributions. I explored

11

spawning potential ratio (SPR) and yield-per-recruit (YPR) models to assess the status of the fishery in its present condition. Lake Yaxhá is more vulnerable to recruitment overfishing if exploitation occurs. Lake Petén Itzá is less vulnerable to recruitment overfishing under the exploitation scenarios that were explored. Higher growth rates of giant cichlid in Lake Yaxhá would yield higher biomass-per-recruit under the same exploitation level compared to Lake

Petén Itzá. These results provide useful baseline data for management of exploited and unexploited giant cichlid populations in Mesoamerica. Lake Yaxhá should be open to well- planned fishing, but should be monitored, to assess the fishery sustainability. Quantitative fisheries studies will benefit all stakeholders, including fishers, scientists and governmental authorities.

12 CHAPTER 1 THE GIANT CICHLID, LIFE HISTORY AND ARTISANAL FISHERY

Introduction: The giant cichlid (Petenia splendida Gunther 1862) is distributed across a mosaic of freshwater systems in Mesoamerica (Greenfield and Thomerson 1997; Miller 2005;

Perez et al. 2010), hence populations occur under different conditions. Moreover, because the giant cichlid is the main target of the artisanal fishery, populations are subject to a large range of exploitation and are considered overexploited in some systems (Noiset and Micha 1996).

Nonetheless, there are systems with minimal habitat transformation and no artisanal fishery, but such systems are exceptionally rare (Granados-Dieseldorff et al. 2012; Barrientos et al. in press).

The giant cichlid has many life history characteristics common to cichlids and survives under a broad range of ecological conditions, which makes it a good model to investigate variations in reproductive traits.

In recent years there has been a worldwide decline in fish stocks attributable to over- exploitation. Rapid human population growth in Petén, northern Guatemala, occurred as part of a government development plan (Schwartz 1990). The giant cichlid is the main target of the gillnet fishery throughout Petén, including in Lake Petén Itzá. One of the most conspicuous effects of fishing is high fish mortality. Fishing gear has a size bias, preferentially capturing larger and hence older or faster growing individuals. Fishing mortality can reach the point where the stock can be overfished and in some cases, depleted (Meyers and Worm 2003; Orensanz et al. 2005). Most freshwater fisheries in Guatemala are considered artisanal, with some exceptions.

Because artisanal fisheries are usually small-scale and spatially structured (Orensanz et al. 2005), fishing is largely carried out by the communities around the lake. Similar to other artisanal fisheries, data scarcity is a problem in Guatemala (Orenzans et al. 2005), in part because it is expensive to collect and because of the poor spatial resolution.

13

Life history factors are fundamental to make generalizations about species responses or tradeoffs in different environments (Winemiller 1992; Winemiller and Rose 1992; Vila-Gispert et al. 2002). Consequently, differences in reproduction, survival, age and size at maturity and growth are common because of environmental variation (Stearns 1992; Winemiller and Rose

1992; Vila-Gispert and Moreno-Amich 2002). In fishes, tradeoffs between growth and reproduction are critical, because fish typically divert energy from somatic growth to reproduction at sexual maturity (Stearns 1992; Glazier 2000; Schultz and Rounton 2001). A premature halt in somatic growth can result in reduced body size, fecundity, mobility, longevity, predatory ability and increased predation risk (Freedman and Noakes 2003). Consequently, life history characters are essential to fish population performance and are a central subject in ecology and resource management.

Despite the importance of life history traits and ecology for fisheries management, there have been few attempts to assess them in Mesoamerican freshwater systems. Lakes Petén Itzá and Yaxhá are located in the northern lowlands of Guatemala and share the same tectonic/karstic origin, climate (Brezonik and Fox 1974; Perez et al. 2010) and ichthyofauna (Barrientos et al. in press), but differ in area, depth (Perez et al. 2010), watershed development (Rosenmeier et al.

2004) and protection status (Barrientos et al. in press). Because the lakes differ in the levels of development and protection, i.e. Lake Petén Itzá is highly exploited relative to Lake Yaxhá, they offered an opportunity to explore life history variation.

The overall goal for my dissertation research was to examine life history variations and juvenile ecology of the giant cichlid in northern Guatemala lakes. These lakes are critical to the giant cichlid, because they are the only protected lentic aquatic ecosystems in Mesoamerica. The specific objectives were to: 1) assess habitat characteristics of giant cichlids, including

14

distribution, abundance, and temporal variability of fish assemblages, 2) to examine aspects of the reproductive biology of the giant cichlid, 3) to assess population dynamics including age, growth, and mortality, and 4) to evaluate suitable fisheries models to assess management policies, including length limits.

15

CHAPTER 2 SEASONAL CHANGES IN THE FISH COMMUNITY THAT INHABITS THE LITTORAL SUBMERSED AQUATIC VEGETATION OF LAKE PETEN ITZÁ, GUATEMALA

Introduction

The association between fish and aquatic vegetation in freshwater systems has been widely studied in temperate zones. The vegetated littoral zone is important habitat for juvenile and small-bodied adult fishes (Dibble et al. 1996). Aquatic macrophytes provide food (Keast

1984; Petty and Grossman 1996; Vono and Barbosa 2001; Bickel and Closs 2008) and protection from predation (Savino and Stein 1982; Gotceitas and Colgan 1987; Dione and Folt 1991) and therefore influence fish abundance and distribution. Fish abundance is usually higher in areas where plants are present (Killgore et al.1989; Lubbers et al.1990; Vono and Barbosa 2001;

Rogers and Allen 2008) thus spatial differences in plant distribution affect fish distribution

(Killgore et al.1989; Bettoli et al.1993; Rogers and Allen 2008). Aquatic plants also contribute to primary production, stabilize sediments, maintain water clarity, and provide habitat for zooplankton and macroinvertebrates (Carpenter and Lodge 1986; Dibble et al. 1996).

Fish abundance and distribution within the littoral zone is influenced by seasonal factors in both tropical and temperate aquatic ecosystems. Temperature often drives seasonal changes in fish community attributes in temperate systems (Killgore et al.1989) with higher species richness, density, and biomass of fishes in littoral zones during warmer months (Fischer and

Eckmann 1997; Bickel and Closs 2008). Less is known about temperature effects in most tropical systems where diurnal and annual temperature swings are not large (Lewis 1996).

Hydrology plays a major role in freshwater fish community dynamics in the tropics (Winemiller

1990; Jepsen et al. 1997; Winemiller and Montoya 2006). The concentration of fishes that occurs in many systems during the dry season is due to a reduction in space and habitat, leading to increased density (Dibble and Pelicice 2010). Conversely, rising water levels associated with

16

wet seasons increase the availability of habitat and decrease fish density through dilution and habitat shifts (Jepsen et al. 1997). The timing of reproduction in relation to hydrological changes also alters fish community metrics due to pulses of recruits (Faunce and Lorenz 2000).

Environmental factors thus strongly influence richness, species associations and abundance of small fishes in the littoral zone and consequently have implications for habitat use and fisheries management of larger, predatory species.

The rich and diverse freshwater fish fauna of Mesoamerica (Greenfield and Thomerson

1997; Miller et al. 2005) coupled with a variety of lentic systems, many of which exhibit seasonal fluctuations in water levels (Perez et al. 2011), makes this tropical region an ideal place to investigate seasonal changes in littoral fish communities. Despite the acknowledged importance of vegetated littoral habitat for fish, little work has been done in Mesoamerica.

Barrientos and Allen (2008) worked on fish of the littoral zone in Lake Izabal, Guatemala. They studied the influence of different plant species, including non-native Hydrilla verticillata (L.F.)

Royle, and found that fish assemblages were different between vegetated and non-vegetated sites, with lower fish density and biomass in areas without macrophytes. Unlike many

Mesoamerican lakes, Lake Izabal has relatively little seasonal fluctuation in water levels and therefore seasonal aspects were not an important part of their study. Conversely, Lake Petén Itza within the Petén region of northern Guatemala has extensive vegetated littoral zones in the southern and western portions and has seasonal water level fluctuations of 60-90 cm annually

(AMPI; Monitoring Program). Research in this lake could provide key insights into seasonal effects on littoral fish communities. The relation between plant and fish abundance is an important management consideration when assessing juvenile habitat for species important in artisanal fishery like giant cichlid.

17

To my knowledge, seasonal changes in the fish community of the littoral zone have not been assessed in any Mesoamerican lake. I chose Lake Petén Itzá to: a) characterize the submersed vegetation in the littoral zone, b) assess the distribution, richness, abundance, and other fish metrics across seasons in the vegetated littoral zone, and c) examine key environmental and habitat variables in relation to fish assemblages. This investigation was designed to obtain information that is required to manage littoral habitats in lentic systems, in particular because many such environments in Mesoamerican have recently been subjected to increasing human pressures.

Study area: The northern Petén part of Guatemala lies northeast of the Sierra Madre

Mountains and constitutes the country’s lowland region (Figure 2-1). The Petén is the southern extent of the Yucatan Peninsula, a large limestone platform of marine origin. Petén possesses a lake district comprised of 14 large waterbodies, of which Lake Petén Itzá is the largest (~100 km2). Most of the Petén region lies within the greater Usumacinta River basin, which includes all the major rivers and the border zone with Mexico. Annual rainfall can vary from ~1200 to ~2000 mm (Figure 2-2), and such inter-annual precipitation fluctuations drive shifts in regional lake levels (Deevey et al. 1980). Lake Petén Itzá owes its origin to a combination of limestone dissolution and tectonism. It is about 32 km long and 3 km wide, with a maximum water depth of

~165 m. It is thought be the deepest lake in lowland Central America. The lake water chemical

+ + - - composition is dominated by cations Ca and Mg and anions SO4 and HCO3 (Perez et al.

2011).

The lake has two elongate basins that run east-west (Fig. 2-1). The larger north basin supports at least four towns along its shoreline, including San Andres, San Jose, Jobompiche and

El Remate. The smaller south basin supports the largest population (~160,000, National Statistics

18

Institute 2008) in the towns of Flores, Santa Elena and San Benito. The lake is a closed hydrologic basin with no visible surface outlet and water level fluctuates, with a slight lag, in response to precipitation amount. The littoral zone is most extensive in the shallow west part of the south basin, with a variety of plant species in the zone. Leon and Morales (2000) reported eight species of submersed vegetation in the northwestern region of Petén, at Laguna del Tigre

National Park, with Vallisneria americana Michx present only in the San Pedro River. Reyes

(2008) evaluated aquatic vegetation in lakes and lagoons in the central Tikal-Yaxhá region and reported V. americana and Potamogeton illinoensis Morong as dominant species in cases where human activities have increased the nutrient status of lakes.

Ichthyological studies of Lake Petén Itzá, Guatemala, began with collections by

European researchers in the early 1800s. These early sampling expeditions were followed by those of North American researchers in the 1900s (Kihn et al. 2006). The most recent fish species list for Petén was compiled by Valdez-Moreno et al. (2005) with cichlids and poecilids as the most speciose families. Previous studies of the ichthyofauna in this area emphasized taxonomy, with no effort to evaluate population or community variables. The artisanal fishery has a long history in northern Guatemala. The giant cichlid, Petenia splendida Günther, also called the “bay snook,” is the main target of the gillnet fishery throughout Petén, including in

Lake Petén Itzá. Other cichlid species like the Mayan cichlid, Cichlasoma urophthalmum

(Günther) and Paraneetroplus melanurus (Günther), are fished with gillnets as well, but lower prices in the local market make them less of a target species for fishermen.

Methods

Habitat Characterization and Plant Sampling

The entire lake shoreline was explored (~60 km) for specific habitats in which to set blocknets. The habitats chosen for sampling were those possessing the most common submersed

19

aquatic vegetation in the littoral zone, which included the dominant and widely distributed V. americana as well as P. illinoensis and the macroalgae Chara sp. (Reyes 2008; Barrientos and

Quintana 2012). I sampled the littoral zone between 0.7 and 1.6 m depth (Figure 2-1) in April

2009 and March, May, July and October 2010 (Table 1). Percent area coverage by vegetation

(PAC) was visually estimated in the field. Total vegetation biomass was estimated using three randomly placed 0.25-m2 quadrats inside each block net. In each quadrat, above-substrate plant material was removed, placed into a nylon mesh bag, spun to remove excess water and then weighed to the nearest 0.1 kg (Canfield et al. 1990). Within each block net, I measured water depth at three points, close to shore, in the middle and far from shore. I also measured the thickness of sediments in the middle of the block net by pushing a measuring stick into the substrate by hand. Dissolved oxygen (DO), temperature (T), total dissolved solids (TDS) and pH were measured within 5 m of every habitat sampled, using a multi-parameter portable meter

(Hanna HI 9828). The distance from the closest part of the block net to the shore was measured.

I used data from the National Institute of Seismology, Volcanology, Meteorology and

Hydrology (INSIVUMEH, Spanish acronym) for air temperature, precipitation and lake level for

2009 and 2010. The Petén region has two marked seasons, a dry season from November to April, usually with <150 mm of rain per month, and a rainy season from May to October, typified by

>150 mm of rain per month (Figure 2-2). Water level fluctuation in the lake ranged ~0.6 m in

2010 (Figure 2-2), but has fluctuated over a range of ~2.5 m since the 1990s.

Fish Sampling

Lake Petén Itzá was sampled using block nets, seines and electrofishing to evaluate the fish community. A total of 32 block nets were set in littoral habitats of Lake Petén Itzá in March,

April, May, July and October. Fish were sampled by applying rotenone inside a 100-m2 (10 m x

10 m) block net. Each block net was set from a boat, with a 5-kg weight at each corner.

20

Personnel walked the perimeter of the block net to insure that the lead line made contact with the substrate. After net placement, powdered rotenone was mixed with water to produce slurry.

Using a small pump, the rotenone slurry was sprayed onto the water surface within each block net to achieve a concentration in the water of 3 mg/L of active ingredient (Finlayson et al. 2000).

Fish began to surface after the slurry application and were collected with dip nets by three personnel for 1 hr (Bettoli and Maceina 1996). When a species was abundant, we measured TL

(mm) and weighed (g) 30 individual specimens and the rest were counted and weighed as a batch.

Fish were also sampled with seines (6-mm mesh, 4-m length, and 1.5 m high) along the shoreline. The seine was used to sweep areas of ~12 m2. Sites were chosen in zones were there was not enough space to set a 100 m2 block net or because of the shallowness of the water (<0.6 m). Seined areas were usually between the submersed vegetation and the shoreline. These zones were typically un-vegetated and common in areas close to beaches and docks, and offshore of

Typha sp. beds. It was difficult to predict how many sweeps I could get in each area, but I strived to have at least three in each zone. A total of 30 electrofishing transects were conducted in the littoral zone. The electrical output ranged from 6 to 9 A of pulsed DC (KVA 1.5 Smith and Root). The littoral zones were sampled for at least 1 hr of pedal time split across 5-6 transects parallel to shore, with a total length traveled of 800-1000 m, once a month. Only block net data were used for statistical analysis on total; data from seining and electrofishing were used to estimate overall fish community richness.

Collected fish were placed on ice for further analysis in the laboratory. Specimens were identified to the lowest taxonomic level using the key for continental waters of Belize

(Greenfield and Thomerson 1997) and original descriptions for some taxa. The taxonomic list

21

was compiled using the nomenclature of Eschmeyer and Fong (2012). Voucher specimens were deposited in the fish collection of the Museo de Historia Natural (MUSHNAT), Universidad de

San Carlos de Guatemala (USAC), Guatemala City.

Analysis

Fish community metrics, vegetation and water quality parameters were assessed: vegetation biomass (kg m2), DO, TDS, temperature, pH, total fish biomass (kg 100 m2), total fish density (fish 100 m2), fish species richness and fish diversity for each block net in each month. I only used fish <200 mm TL, because small block nets are ineffective at capture larger fish

(Rogers and Allen 2008; Barrientos and Allen 2008). The null hypothesis was that there was no difference in these variables across months. Diversity was estimated using the Shannon-Wiener index (H’) for fish species collected in each block net (Krebs 1999). Petenia splendida Günther biomass was assessed separately because it is the main target of the artisanal fishery in the lake.

Thorychthys affinis (Günther 1862) was assessed separately because it accounted for >50% of the fish biomass and numbers across all months and thus likely an important driver of observed patterns. The cichlid family was considered separately because it is the most speciose and was found to be the most important family in artisanal fisheries. Vegetation biomass, as well as fish density and biomass were log10 (x+1) transformed to improve normality. One-way ANOVA, followed by Tukey’s test for significance was used in all analysis. Differences were considered significant at P ≤ 0.05 for all analyses. The analyses were performed in R, version 2.13.1.

Results

Habitat Characterization

The dominant aquatic vegetation in the littoral zone was V. Americana (63%), with other plant taxa, including P. illinoiensis (28%) and Chara sp. (6%), present in lesser amounts. Percent

22

area coverage (PAC) was positively related to plant biomass (r2=0.42, Figure 2-3). There was no difference in plant biomass among months (Table 2-1) (F 4, 27 = 1.14, P= 0.35). Temperature, dissolved oxygen, conductivity and pH were within expected ranges for this tropical lake, and none showed significant differences across months (F 4,27=1.27-1.70, P =0.18-0.30). Water temperature ranged from 25 to 32 ºC, pH ranged from 6.7 to 8.1, conductivity ranged from 480 to 730 µS/cm and total dissolved solids ranged from 235 to 630. Distance to shore, soft sediment thickness and depth of block nets were not significantly different among months (F 4,27=0.65-

1.53, P =0.22-0.49) (Table 2-1).

Fish Communities

I found 20 fish species from 8 families including 18 native (7 families) and 2 non-native

(2 families). Among gears, block nets (19 spp.) and electrofishing (20 spp.) were the most effective for sampling fish richness in the lake and seines were less effective (16 spp). Only one species was captured exclusively by electrofishing, the non-native Pterygoplichthys pardalis

(Castelnau 1855), thus is not included in the block net Table 2-2. Nine species were present in every month sampled and five species were found in three months or less (Table 2-2). For block nets, average species captured by month ranged from 8 to 11. Some species were uncommon and found only in the dry season, like Belonesox belizanus Kner, Gambusia yucatana Regan and

Poecilia petenensis Günther, all in the family Poeciliidae. Other species were found only in the rainy season, such as Rhamdia quelen (Quoy and Gaimard) and the non-native Oreochromis niloticus (Linnaeus) (Table 2-2).

Temporal Analysis

Total mean fish biomass based on block nets differed among months (Table 2-3) (F

4,27=2.71, P=0.04). In March, total fish biomass was different than July (lowest). Three species,

T. affinis, Cichlasoma urophthalmum (Günther) and Poecilia mexicana Steindachner comprised

23

75% of the total biomass and T. affinis represented >50% of the total biomass across all months.

T. affinis showed the highest biomass in March and the lowest in July (F4,27 =6.05, P=0.001).

The average biomass or density for P. splendida did not differ among months sampled (F4, 27=1.

98, P=0.12) (Table 2-3).

The total average density among months ranged from 278 to 732 individuals per 100 m2.

Total fish density did not differ (F4,27 =1.47, P=0.23) among months sampled and only T. affinis

(F4,27 = 6.51, P<0.001) showed differences in density among months, with the highest density in

March and the lowest in July.

Fish species richness ranged from 8 to 11 species. The diversity index (H’) showed no difference in any month sampled (F4, 27=1.69, P=0.18), and ranged from 1.15 in March to 1.52 in

October.

Discussion

In tropical Lake Petén Itzá, I found considerable year-round use of vegetated littoral habitat by small fishes (< 200mm TL), including juveniles and adults of important fishery species, but also found strong evidence of seasonal trends related to lake level. Seasonal increases in water levels created flooded zones which provided new habitat and corresponded with the lowest fish density and biomass of the sampling. Seasonal differences in fish biomass were driven by dispersal of the most common fish, T. affinis, into newly flooded habitat, away from littoral vegetation stands. The lower biomass in July is consistent with the comments of local artisanal fisherman who report lower CPUE in the rainy season due to dispersal to new habitats (C.A. Barrientos, unpublished data). Moreover, higher total fish biomass in the dry season and lower total fish biomass in the rainy season is consistent with findings in other tropical areas (Winemiller 1990; Jepsen et al. 1997; Winemiller and Montoya 2006; Lopez-

Lopez et al. 2009; Dibble and Pelicice 2010).

24

Species richness did not change with season but fish assemblages’ composition was different. Cichlid taxa are present year-round, whereas poeciliid richness was highest in the dry season and low in the rainy season. Two species, Rhamdia quelen and non-native Oreochromis niloticus, were found only in the rainy season, but in low abundance. All fish species reported in the lake by Valdez et al. (2005) were captured in the littoral zone in at least one sampling event, and the only taxa I added to the list were the non-natives O. niloticus and Pterygoplichthys pardalis. Consistent with previous findings in the Usumacinta Province (Miller 1976; Willink et al. 2000; Valdez et al. 2005), cichlids and poeciliids accounted for 65% or more of the species richness in Lake Petén Itzá.

Fish richness and cichlid diversity have been associated with resource partitioning of different water column zones (Lowe-McConnell 1987; Jepsen et al. 1997; Cochran and

Winemiller 2010). It seems, however, that the vegetated littoral habitat provides sufficient complexity and ample structure to accommodate the cichlids of the Lake Petén Itzá littoral habitat year-round. Moreover, only one cichlid, T. affinis, displays significant differences in density and biomass across seasons. This is probably because T. affinis is the main substrate sifter in the littoral zone (Cochran and Winemiller 2010) and with the onset of the rainy season there are newly flooded benthic areas with high numbers of aquatic invertebrates (Garcia 2009), found away from the littoral vegetation.

Juvenile fishes, mostly cichlids, were common in the samples consistent with previous studies that found that the vegetated littoral zone usually has juveniles that find refuge and food there (Lubbers et al.1990; Savino and Stein1982; Chick and McIvor 1994; Barrientos and Allen

2008; Rogers and Allen 2008). In contrast, the small-bodied poeciliids were found mainly as adults. Most poecilid species occupied the habitat between the vegetation and the shoreline

25

(Bettoli and Morris 1991). They thus can be captured with blocknets when this strip of water dries or becomes too shallow in the peak of the dry season and they move to deeper vegetated zones. I found that pelagic planktivorous fishes Dorosoma petenense (Günther) and Atherinella alvarezi (Diaz-Pardo) had low density and biomass in the sampled habitat, probably because they prefer areas without dense vegetation (Killgore et al.1989; Gelwick and Matthews 1990; Bettoli and Morris 1991; Bettoli et al.1993; Barrientos and Allen 2008). Only 20% of the habitat I sampled had <50% PAC and I did not sample habitat without vegetation.

Non-native species O. niloticus and Pterygoplichthys pardalis were rare in my samples and only collected in the western part of Lake Petén Itzá. Both species have been reported in the greater Usumacinta Basin (Willink et al.2000; Valdez-Moreno et al.2005; Wakida-Kusunoki et al.2007) but these are the first recorded specimens from Lake Petén Itzá, a closed-basin lake, which suggests that the non-natives were probably introduced directly into the lake. Tilapias are among the most widely introduced freshwater fishes in tropical regions of the world

(http://www.issg.org/) and are known from several lake systems in Mesoamerica as well as rivers in the eastern portion of Petén and Belize (Esselman et al. 2013). Fishermen mention escapes of

O. niloticus, in the early 1990s from ponds and cages at El Remate, a town on the east shore of the lake (National Authority for Lake Petén Itzá Management; AMPI executive director personal communication). Pterygoplichthys pardalis has no documented point of introduction. This introduction is somewhat surprising because P. pardalis, unlike O. niloticus, is not a commonly cultured fish in the area. P. pardalis, however, is a popular fish in the aquarium trade (Gertzen et al. 2008), a potential pathway of introduction (Duggan et al. 2006). At least two local hotels and the local zoo (Petencito) in central Santa Elena have outdoor ponds that contain P. pardalis, other potential sources. Although introduced tilapias and loricariid catfishes are thought to

26

negatively affect native species and alter habitat in some regions, the low abundance and limited distribution of these non-natives in Lake Petén Itza suggests little current impact. Monitoring to detect increased abundance and impacts of both species should be incorporated into lake management planning.

Biomass estimates in the littoral zone of Lake Petén Itzá was relatively low compared with another Guatemalan lake, Lake Izabal, during June-July (Barrientos and Allen 2008). This was expected because Izabal is a highly productive lake and receives large inputs of materials/nutrients through the Polochic River (Perez et al. 2010). In contrast, Lake Petén Itzá is oligotrophic and has no major river input (Perez et al. 2010). The positive relationship between trophic state and fish biomass is known and well documented in subtropical Florida lakes

(Bachmann et al. 1996). There is, however, recent evidence of cultural eutrophication in the southwest basin of Lake Petén Itzá (Rosenmeier et al. 2004; Oliva 2005), associated with the urban centers where the human population is concentrated, which probably has led to an increase in fish biomass in Lake Petén Itzá. Olive et al. (2005) found higher centrarchid biomass associated with an increase in nutrients. Centrachids are considered to be the North American ecological equivalents of cichlids in Central America (Miller 1976; Norton and Brainerd 1993) and thus cichlid responses to increased nutrients might be similar. Fish biomass could be also affected by the intra-annual variability in lake levels (Soria-Barreto et al. 2008), which was evident in Lake Petén Itzá.

Lake Petén Itzá lies in a region that has sustained large anthropogenic changes over the last 40 years (Schwartz 1990), including development along the shoreline and cultural eutrophication (Rosenmeier et al. 2004; Oliva 2005). This is of concern because all species reported in the lake are present in the vegetated littoral zone at some time of the year.

27

Submersed littoral aquatic vegetation provides important fish habitat in Lake Petén Itzá, which is large and deep, and has no vegetation at water depths >6 m. The vegetated littoral zone should be a priority habitat to conserve in terms of fish species richness and abundance. It is a critical area for the P. splendida fishery, because juveniles of the species occupy this habitat year-round similar to largemouth bass in temperate systems (Hoyer and Canfield 1996). Moreover, aquatic vegetation provides other ecosystem services such as sediment stabilization, water clarity maintenance and zooplankton habitat (Carpenter and Lodge 1986). The vegetated littoral zone is critical not only for fish, but other wildlife as well. As in many parts of Mesoamerica, lakes in

Petén, Guatemala, serves several functions for humans, as sites for recreation, protein production, transportation, and water supply. They must be managed with a holistic approach.

28

Figure 2-1. Study area in Lake Petén Itzá, Guatemala, including sites sampled with blocknets (•). Towns are in red: a) San Andres, b) San Jose, c) Jobompiche, d) El Remate, e) Flores, f) Santa Elena and g) San Benito.

29

500 1.4 2010 450 2009 1.2

400 2010 L

350 1

300 0.8 250 0.6 200

Precipitation (mm) Precipitation 150 0.4 Water level (m) level Water 100 0.2 50 0 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month

Figure 2-2. Average monthly rainfall (mm) in Petén, Guatemala, in 2009 (solid line) and 2010 (dotted line). Note the second axis; Lake Petén Itzá water level (m) in 2010 (dashed line). Data from INSIVUMEH, Guatemala.

30

4 October March

April

2

- 3 May July

2

1 Log10 Log10 plant biomassm kg

0 0 20 40 60 80 100 Percentage of area covered

Figure 2-3. Plant biomass and percentage of area covered (PAC) measured inside blocknets at Lake Peten Itzá, in five months 2009-2010.

31

Table 2-1. Mean and standard error of habitat characteristics from blocknet samples collected at Lake Peten Itzá in years 2009-2010. Blocknets (n), water depth (cm), sediment thickness (cm), distance to shore (m), vegetation biomass (g) and range of percent area covered (%PAC), Temperature (C), Conductivity and pH. Mar Apr May Jul Oct Blocknets (n) 6 7 6 6 7 Water depth (cm) 138.3±6.7 141.4±3.4 140±9.3 126.7±8 122±8.4 Sediment thickness 13.3±1 12.1±2.1 19.2±2.4 16.7±5.1 14.5±2.8 (cm) Distance to shore (m) 7.7±0.9 7.8±1.6 12±6.5 5.9±1.4 10.7±4.3 Vegetation biomass (g 780.9±425.8 803.4±189.7 905.5±302.5 880.7±279.8 1870.6±478.5 0.25 m2) PAC (%) 25-95 40-85 50-90 50-80 40-85 Temperature 27.3±3.1 29.2±2.4 29.9±1.9 30.9±1.9 28.9±3.1 Conductivity 548±45 521±38 535±21 519±32 553±42 pH 7.1±1.1 7.4±0.9 6.9±1.3 8.1±1 7.7±1.1

32 Table 2-2. Mean ± standard error of biomass (g 100 m-2) of fish collected in blocknets in different months during 2009-2010 at Lake Peten Itzá. Total is the sum of biomass of all individuals of a species collected during all months (g). * indicates a non- native species. March 2010 April 2009 May 2010 October 2010 Total July 2010 (n=6) Species (n=7) (n=6) (n=6) (n=7) (all) Family Clupeidae 1.8 ± 0.8 26.6 ± 3.8 2.2 ± 1.8 14.2 ± 3 284.4 Dorosoma petenense (Günther) Family Characidae Astyanax aeneus (Günther) 43.1 ± 14 21 ± 2 18 ± 3.2 6.1 ± 1.7 22.9 ± 5 730.3 2.1 ±1.2 6.8 ± 1.4 1.4 ± 1 134.1 ± 22 78.9 Hyphessobrycon compressus (Meek) Family Pimelodidae Rhamdia quelen (Quoy & Gaimard). 6.5 ± 4.6 3.8 ± 2 6.5 ± 3.1 106.6 Family Poeciliidae 2 ± 0.3 12.1 Belonesox belizanus Kner 2.1 ± 0.5 12.7 Gambusia yucatana Regan 1.4 ± 1 0.9 ± 0.7 1.4 ± 0.2 0.2 ± 0.1 25.1 Gambusia sexradiata Hubbs Poecilia mexicana Steindachner 47.2 ± 9.1 256.3 ± 29.3 206.9 ± 31.2 202.9 ± 14.5 127.8 ± 11.1 5,215.5 1 ± 0.7 1.1 0.5 13.2 Poecilia petenensis Günther Family Atherinidae Atherinella alvarezi (Díaz-Pardo) 1.5 ± 0.9 85.5 ± 14.8 2.1 ± 1.1 3 ± 2.1 6.3 ± 2.1 579.7 Family Synbranchidae Ophisternon aenigmaticum Rosen & 77.2 8.9 ± 6.2 2 ± 0.2 Greenwood Family Cichlidae 1.8 ± 3.8 6.3 ± 3.2 0.6 ± 0.9 52.6 Amphilophus robertsoni (Regan) Cichlasoma salvini (Günther) 76. ± 32 30.1 ± 5.3 51 ± 9.5 14.9 ± 3 44.8 ± 5 1,425.3 Cichlasoma urophthalmum (Günther) 93.3 ± 21 184.9 ± 39.1 289.1 ± 24.3 50.5 ± 7 207 ± 17.4 5,248.9 5.6 ± 1.1 39.2 *Oreochromis niloticus (Linnaeus) Parachromis friedrichsthalii (Heckel) 13.3 ± 3 29.3 ± 12.2 11.7 ± 7.7 2.1 ± .4 461.7 Paraneetroplus melanurus (Günther) 144.1 ± 28 144.5 ± 10.7 6 5 ± 27.4 49.7 ± 32.1 112.5 ± 44.3 3 Petenia splendida Günther 92.9 ± 37.5 234 ± 59.7 81.7 ± 31.6 42.1 ± 25.9 123.9 ± 58.4 3,655.8 Thorichthys affinis (Günther 1862) 1722.7 ± 531.3 1271 ± 615.8 604.2 ± 432.1 94.6 ± 32.6 460.3 ± 156.1 25,823.6 *Pterygoplichthys pardalis was only collected with electrofishing and is not included in the table.

33

Table 2-3. Mean and standard error of fish metrics from blocknet samples collected at Lake Peten Itzá in 2009-2010. Fish species richness is the cumulative richness across all samples (n = 6-7) for each month. Values in a column without a letter in common are significantly different (P < 0.05) by ANOVA and Tukey’s test. Columns without letters are not significantly different. Shannon- Fish density (fish 100 m-2) Fish biomass (g 100 m-2) Total Wiener Petenia Thorichthys affinis Petenia Thorichthys Month Species diversity Total splendida Total splendida affinis Richness index March 14 1.15 ± 0.14 732 ± 168 21.7 ± 6.1 533 ± 144.6 a 2573 ± 658 a 92.9 ± 37.5 1723 ± 531 a (n=6) April 15 1.26 ± 0.13 454 ± 154 8.6 ± 5.9 222 ± 122.4 ab 1470 ± 743 ab 234 ± 59.7 1271 ± 615 ab (n=7) May 12 1.30 ± 0.12 414 ± 123 14.5 ± 3.6 146 ± 118.7 bc 1257 ± 464 ab 81.7 ± 31.6 604 ± 432 bc (n=6) July 14 1.52 ± 0.10 278 ± 44 23.8 ± 8 27 ± 14 c 480 ± 93 b 42.1 ± 25.9 94 ± 32 c (n=6) October 15 1.53 ± 0.13 503 ± 127 26.9 ± 9.4 207 ± 69.8 ab 1412 ± 442 ab 123.9 ± 58.4 460 ± 156 abc (n=7)

34

CHAPTER 3 REPRODUCTIVE LIFE HISTORY TRAITS OF THE GIANT CICHLID (Petenia splendida Gunther 1862) IN GUATEMALAN LAKES

Introduction

Life history theory provides a fundamental framework to predict the responses of species in different environments and conditions (Winemiller 1992; Winemiller and Rose 1992; Vila-

Gispert et al. 2002). Variation in key traits provides alternative strategies that relate to increasing lifetime fitness. Consequently, differences in reproduction, survival, age and size at maturity and growth are common, given environmental variation (Stearns 1992; Winemiller and Rose 1992;

Vila-Gispert and Moreno-Amich 2002). Alternative strategies frequently involve tradeoffs such as often observed between growth and reproduction in fishes. Fish, divert energy from somatic growth to reproduction at sexual maturity (Stearns 1992; Glazier 2000; Schultz and Rounton,

2001) but may trade reduced mean or maximum body size, fecundity, foraging ability, and survival for an early reproductive advantage while in most of the cases increased predation risk

(Freedman and Noakes 2003). Variations in life history traits are essential to fish population performance and provide a rich context for fisheries and ecosystem management (Walters and

Martell 2004).

Fishing is size-selective and targets in most cases larger, presumably faster growing or older individuals. The selective removal of large adults by a fishery may select for a more precocial life history style (Reznik et al 1990; Olsen et al. 2004; Walsh et al. 2006). A more precocial life history style can include a reduction in length at maturity (Faunce and Lorenz

2000; Haugen and Vollestad 2001), with associated changes in fecundity (Olsen and Rulifson

1992; Duponchelle et al.2000), smaller egg size (Duponchelle et al.2000), and behavior (Walsh et al. 2006; Biro and Post 2008). Comparing exploited and unexploited populations can provide key data to improve the predictive ability of fisheries models.

35

The importance of life history variation in fishes is mostly studied in temperate regions

(Winemiller and Rose 1992; Vila-Gispert et al. 2002; Marchetti et al. 2004; Mims et al. 2010) or aquaculture species (Duponchelle et al 2000; Efitre et al. 2009; Ishikawa et al. 2012) with far fewer studies in tropical systems (Winemiller 1989). Mesoamerica has a well-known ichthyofauna and is a center of diversity for cichlids (Greenfield and Thomerson 1997; Willink et al. 2000; Hulsey et al. 2004; Miller et al.2005; Schmitter-Soto 2007; Matamoros et al.2012;

McMahan et al.2013). Despite the importance of cichlids in fisheries and aquaculture, cichlid life history variation in this region is poorly studied at species or interdemic scales across different environments.

The giant cichlid (Petenia splendida Gunther 1862) is a good species to investigate variation in life history. It is distributed across a mosaic of freshwater systems in Mesoamerica

(Greenfield and Thomerson 1997; Miller 2005; Perez et al. 2010), hence populations occur under different environmental conditions. Moreover, the giant cichlid is the main fisheries target, and populations are subject to a large range of exploitation from no fishing to intense pressure leading to overexploitation (Noiset and Micha 1996; Ixquiac 2007). Human population growth and concurrent environmental change such as cultural eutrophication impact some aquatic systems harboring giant cichlid but there are systems with minimal habitat transformation and no fishery (Valdez-Moreno et al. 2005; Granados-Dieseldorff et al. 2012; Barrientos et al. in press).

Lakes Petén Itzá and Yaxhá are located in the northern lowlands of Guatemala and share the same tectonic/karstic origin climate (Brezonik and Fox 1974; Perez et al.2010) and ichthyofauna (Barrientos et al. in press; Chapter 1), but differ in area, depth (Perez et al.2010), watershed development (Rosenmeier et al. 2004) and protection status (Barrientos et al. in

36

press). Giant cichlid in Lake Petén Itzá is subject to high exploitation whereas this species is not fished in Lake Yaxhá, thus providing a contrast to explore life history variation.

Reproduction is a fundamental biological process that frequently interacts intimately with a suite of important life history traits (Reznik et al 1990; Duponchelle et al. 2000). Some of the most dramatic and important changes in life history traits documented in other fish species occur in reproductive traits (Reznik et al. 1990; Olsen and Rulifson 1992). Consequently the aim was to examine variation in the reproductive biology of the giant cichlid in Lake Petén Itzá and Lake

Yaxhá. To that purpose, I quantified life history reproductive traits such as sex ratio, size at maturity, gonadosomatic index, batch fecundity and egg size and then compared reproductive traits between Lakes Petén Itzá and Yaxhá.

Study Site

Petén, northern Guatemala, is the southern extent of the Yucatan Peninsula, a large marine limestone platform. Petén has a lake district comprised of about 14 large water bodies, of which Petén Itzá is the largest (~100 km2) and Yaxhá is the second largest (~7.4 km2; Figure 3-

1). Most of Petén lies within the greater Usumacinta River Basin, which includes all the major rivers and the border zone with Mexico. Annual rainfall can vary from ~1200 to ~2000 mm and inter-annual precipitation fluctuations drive shifts in regional lake levels (Deevey et al. 1980).

Lake Petén Itzá owes its origin to a combination of limestone dissolution and tectonism.

It is about 32 km long and 3 km wide, with a maximum water depth of ~165 m. It is thought to be the deepest lake in lowland Central America. The lake is oligotrophic and water chemical composition is dominated by cations Ca and Mg and anions SO4 and HCO3 (Perez et al. 2011).

The lake has two elongate basins that run east-west. The larger north basin supports at least four towns along its shoreline, including San Andres, San Jose, Jobompiche and El Remate. The smaller south basin supports the largest population in the towns of Flores, Santa Elena and San

37

Benito (~160,000, National Statistics Institute 2008). The littoral zone is most extensive in the shallow west part of the south basin, with a variety of plant species in the zone. The south basin is influenced by recent (~50 years) cultural eutrophication and is a eutrophic basin in an oligotrophic lake (Brezonik and Fox 1974; Rosenmeier et al. 2004). Although fishing is not legal, there is an artisanal fishery with gillnets targeting giant cichlid and other cichlids in the lake (Barrientos et al. 2012).

Lake Yaxhá is part of the Mopan River Basin that runs from east to west in the south part of the Yaxhá-Nakum-Naranjo National Park. The lake origin is tectonic/solution, similar to Lake

Petén Itzá, but is of lower productivity being mesotrophic (Deevey et al. 1980; Perez et al.

2008). Within the lake, there are only three small boats used for transportation to the archaeological site of Topoxte. Fishing and hunting are prohibited. The National Council of

Protected Areas (Consejo Nacional de Areas Protegidas; CONAP) and the Institute of

Anthropology and History (Instituto de Antropologia e Historia; IDAEH) share park administration. Cultural eutrophication does not play a large role in Yaxhá because of its location within the Maya Biosphere Reserve (MBR) and its relative lack of human development in the watershed. Only a few farms are located along the south basin of Lake Yaxhá. Fish communities are similar in both lakes (Barrientos et al. in press) despite the differences between the two lake watersheds.

Environmental Factors

The National Institute of Hydrology (INSIVUMEH, Spanish acronym) provided temperature and precipitation data for the region, with the station located at Lake Petén Itzá.

This region has two seasons, a “dry” seasonally (December to April), usually characterized by

<100 mm of rain per month, and a “rainy” season (May to November), with >100 mm of rain per month (Figure 3-2). There was a difference of about 60 cm in water level between the end of the

38

dry season and the end of the rainy season during the span of the study in Lakes Petén Itzá

(Figure 3-2) and Yaxhá (Barrientos et al. in press). The lowest average monthly temperature was

21°C in January and the highest was 30°C in May.

Methods

Fish Sampling

I collected giant cichlid from both lakes using a boat-mounted electrofisher (1.5 KVA;

Smith-Root, Vancouver, Washington) with an electrical output of 5 to 8 A of pulsed DC, along the shoreline at depths of 1-2.5 m. Samples were taken from the lakes in July and October 2010 and monthly from March to October 2011. Fish were sampled until I captured at least 30 fish

>150 mm or for two days, but I kept all Petenia splendida, of any size. Fish were measured for total length (TL) to the nearest 1 mm, blotted dry and weighed to the nearest 0.1 g. I removed gonads and determined sex of individual fish using visual examination in the field.

Reproduction

The sex of each individual was identified macroscopically and classified into three major categories—males, females and immature. To test the theoretical sex ratio (i.e. 1:1 females:males), I used a Chi-square test on the number of males and females in each lake.

Females were assigned an ovary developmental stage based upon macroscopic examination following the criteria of Perez-Vega et al. (2006) for giant cichlid and used the criteria for mature and immature from Nuñez and Duponchelle (2009; Table 3-1). Gonads from both sexes were preserved on ice and transferred to 10% formalin within 12 h of capture and processed for further analysis. I subsampled gonads to make histological preparations of segments of the gonads to compare them to field assignments to test whether the correct gonadal stage was assigned in the field. Histological preparations were made using a small cube of the right gonad

(5 mm), taken near the center of the gonad. After field fixation with formalin, the gonad was

39

transferred to 70% ethanol later. Samples were then dehydrated with a series of alcohol strengths (70%, 80%, 90% and 100%), embedded in paraffin, and sectioned to 5-µm thickness.

The gonad sections were stained with hematoxylin and eosin and mounted onto slides

(Laboratorio Soluciones Analiticas; Guatemala).

The gonadosomatic index (GSI) was calculated as: GSI = (WG ⁄ (WT-WG))*100, where

WG is the weight of the gonad and WT the weight of the fish. Average GSI was plotted for all monthly samples to determine seasonal cycles. I only used fish in stages III and IV for the GSI because using other stages can lead to an underestimation of the breeding season (Nuñez and

Duponchelle 2009).

Batch Fecundity

Batch fecundity was estimated using female with ovaries in stages III and IV (Table 3-1).

These stages were used to prevent overestimating (i.e. stage II) or underestimating batch fecundity (Nuñez and Duponchelle 2009). I used the gravimetric method to estimate batch fecundity. To do this, I subsampled 10% of the ovary and counted the individual hydrated eggs, then adjusted for the total weight of the gonad (West 1990). Linear regression was used to analyze the data (r2) using batch fecundity as a function of TL or body weight of the gonad. Egg counts were loge transformed to improve normality. Analysis of Covariance (ANCOVA) was used to detect differences in fecundity between lakes. Differences were significant at P < 0.05 for all analyses. I also estimated the average weight of an oocyte by dividing eggs counted by the subsampled gonad weight.

Maturity

To estimate the maturity schedule, male and female fish were assigned into a category of either immature or mature as a function of TL. Females defined as being sexually mature had

40

ovaries in stage II (corresponding to vitellogenic oocytes in the ovaries) through stage V (Nuñez and Duponchelle 2009; Table 3-1). The only female stage that was considered immature in this study was stage I. To ensure that fish from stage II were not immature, I used histological cuts of the gonads to confirm oocyte development, according to the procedure of Perez-Vega et al.

(2006). Because fish were collected at the peak of the reproductive season, I was able to stage fish as mature or immature throughout the sampling period using macroscopic ovary examination because resting females were thought to be rare (Nuñez and Duponchelle 2009). For males, I used the criteria developed by Nuñez and Duponchelle (2009) to define males as immature if their testes were silvery or translucent filaments, and thinner and usually longer than the ovaries of immature females. Males were considered mature if the testes were whitish to pinkish, relatively long, and swollen, with a triangular-circular cross section. I used logistic regression to model the proportion of mature fish as a function of TL and age. Length at 50%

(TL50) maturity was estimated using the size for which 50% of the sampled fish were sexually mature, i.e. in stage II or above.

1 (3-1) P = 1−(TL−TL50)/σ

Where TL50 and sigma (σ) are the two model parameters estimated and P is the proportion of mature fish.

To compare sexual maturity among sexes and lakes, I used the Akaike Information

Criteria (AIC) on different models (Strong 1999). The models have two parameters that can potentially vary; I used the most probable combinations. The TL50 and sigma were allowed to be independent for each sex and lake, only by sex or only by lake (Table 3-2). The difference in

AIC is dependent on the maximized likelihoods and the number of parameters between models.

41

Difference in AIC >2 was considered to be significant (Taper 2004). If models had similar AIC values, then I used the most parsimonious model.

Results

I collected a total of 667 giant cichlid in both lakes, 403 from Lake Petén Itzá and 264 from Lake Yaxhá. Total lengths of all fish collected ranged from 55-440 mm with 544 fish larger than 150 mm. Lake Yaxhá showed a shifted TL distribution to larger sizes regardless of sex. The catches displayed a bimodal structure in Lake Yaxhá, with males larger than females, but a unimodal structure in Petén Itzá, with both sexes distributed in the same size classes

(Figure 3-3). The ratio of females to males deviated from 1:1 in Lake Petén Itzá (2.5:1;

X2=19.01, P<0.05), but not in Lake Yaxhá (X2=1.63, P=0.21).

Spawning Season

Only 57% of all fish captured were mature and used in the GSI analysis in both lakes.

Giant cichlids presented mature gonads throughout the duration of the study (March to October).

Weight of females ranged from 158 to 692 g and ovaries ranged from 0.56 to 17.5 g. Female

GSI ranged from 0.46 to 4.52 % (Figure 3-4). Weight in males ranged from 88 to 1024 g and testes from 0.1 to 2.1g. In general, male GSI increased from 0.05% in March to 0.7% in

October, within a similar time frame as the females. The gonadal stage in males was difficult to determine, because of low change in gonad size, causing low staging accuracy in male cichlids.

Thus, males are not included in Figure 3-4.

The pattern of female GSI was similar in both lakes, with a peak in May-June, less activity in July-August, and with another peak in September-October. The GSI pattern seemed to be correlated with the rainfall pattern in the area. Overall, the GSI was lower in Lake Petén Itzá compared to Lake Yaxhá (t-test, df=8, P<0.05). Only May and November showed similar GSI in both lakes (Figure 3-4).

42

Fecundity

Females from Lake Petén Itzá were generally smaller in size and weight than the ones from Lake Yaxhá. Despite the difference in female size and gonad weights between lakes, I found no significant difference in batch fecundity as a function of female body weight

(ANCOVA: homogeneity of slopes by lake, P=0.22; intercepts P=0.51) (Figure 3-5). I therefore pooled females from both lakes, resulting in the following relation between egg numbers and body weight: #eggs=7.14+0.0032 (female weight) (Figure 5c; r2 =0.63). Although fecundity for

Lake Petén Itzá was similar to Lake Yaxhá as a function of female weight, I found a difference in average egg weight between lakes, with Lake Yaxhá having 3.01 mg/egg and Petén Itzá having 1.66 mg/egg (t-test df 39=-3.76, P<0.05).

Maturity

The smallest mature female in Lake Petén Itzá was 175 mm TL, whereas in Lake Yaxhá, it was 180 mm TL. Males in Lakes Petén Itzá showed earliest maturity at 195 mm and 198 mm

TL in LakeYaxhá. Average size at maturity was similar for females in both lakes (L50 = 21-22 cm) and was slightly smaller than for males (L50 = 23-24 cm; Figure 3-6). There were four models with similar AIC, among them the one with the most parameters (8), which had different parameter for each sex in each lake, and the most parsimonious (2) that was the same for all giant cichlid in both lakes (Table 3-2). The most parsimonious model; had one set of parameters for both lakes and sexes. I found that females, regardless of the lake, matured between 0.9 and

1.1 years, whereas males mature later, at 1.20 and 1.25 years in both lakes (Figure 3-7).

Discussion

Variation was found in reproductive traits between populations of giant cichlid consistent with differences in exploitation. The life history traits that best differentiated populations from

43

fished Lake Petén Itzá and unfished Lake Yaxhá were sex ratio, egg size and GSI. Lake Yaxhá had significantly greater GSI and average egg weight and a sex ratio of 1:1. Variation in reproductive traits might be expected in the mosaic of natural conditions and separate stocks given different historical conditions. However, significant differences at the giant cichlid population level can be the result of fishing mortality targeting large individuals (Walsh et al.

2006; Edeline et al. 2007), in Lake Petén Itza mostly males, causing skew sex ratio with more females and large females that can invest more in gonads.

Sex Ratio and Size Distribution

Lake Yaxhá showed a bimodal size distribution, with small females (mean = 27 cm) and large males (mean = 34 cm). This is not uncommon in cichlids with larger males having advantages in reproduction including defense of nesting territory (Gomiero and Braga 2004;

Bwainka et al. 2007). The male size advantage was, however, non-existent in Lake Petén Itzá; males and females were commonly < 25 cm and rarely exceeded 30 cm, probably because the artisanal gillnet fishery targets fish >25 cm. This was reflected in the truncated giant cichlid TL distribution in Lake Petén Itzá that was skewed toward smaller and probably younger individuals

(Chapter 4). This may result in a strong directional selection toward those individuals, which may lead to smaller size at age (Conover and Munch 2002; Walsh et al. 2006).

In Lake Petén Itzá, the sex ratio differed from 1:1 and was dominated by females (2.5:1).

Again, size-selective mortality could be the reason for the skewed sex ratio, because males are larger in giant cichlid populations. De Souza et al. (2008) found that cichlid populations with small TL have female predominance, similar to what I found in Petén Itzá. Unbalanced sex ratios could be a consequence of a seasonality event or bias created by the sampling gear. The reproductive season and sampling sites were similar in both lakes, so these factors were unlikely

44

to be causal of the skewed sex ratio. Thus, selective male mortality, arising from sexual dimorphism and size-selective fishing probably was the cause. Unbalanced sex ratios in cichlids with the smaller sex overrepresented can lead to reduced mean size compared to balanced populations (Toguyeni et al. 2002). Lake Yaxhá did not show a sex ratio different from 1:1, which is common in cichlids (Chellappa et al. 2003; Gomez-Marquez et al. 2003) and was consistent with the assumption that large individuals are taken by artisanal fishing in Lake Petén

Itzá. Fishing mortality in Lake Yaxhá is not significant (Chapter 3). The sex ratio could have implications because of the fewer males to pair up with the overabundant females, probably reducing the reproduction success in the population.

Spawning Season

The GSI value showed a peak reproductive period that correlates with the rain pattern.

Reproduction related to precipitation is seen in other cichlids in the Americas (Jensen et al. 1999;

Chellappa et al. 2003; Gomiero and Braga 2004) and Africa (Bwainka et al. 2007). Another cue could be water levels, which is common for cichlids that inhabit rivers (Jepsen et al. 1999). The

GSI in Lake Petén Itzá peaked when the lake levels were the lowest, at the beginning of the rainy season. This correlation between spawning behavior and water level, which are usually correlated with rain patterns, is found in other Mesoamerican cichlids (Martinez-Palacios and

Ross 1992). The GSI values were lower in Petén Itzá because of the smaller egg size, but showed similar patterns of peaks during the duration of the study. Despite the peaks detected in the study,

I found juveniles in the littoral zone year-round, with no change in the TL mode in the littoral habitats (Chapter 1), which could indicate constant reproductive output by adults or stunted populations of age-0 fish, but the latter explanation is unlikely.

I did not detect a multiple stages of oocytes in the gonads using macroscopic examination. Moreover when I examined histological preparations I found only two different

45

oocyte stages represented, but one stage always dominated, typical of a synchronous ovulatory teleosts (Tyler and Sampter 1996). This is consistent with Perez-Vega et al. (2006) who found mature gonads in giant cichlids, mainly in stage IV oocytes (i.e. mature), but also in other oocytes stages, in small amounts. Perez-Vega et al. (2006) hypothesized that giant cichlid individuals may experience more than one spawning event during a single reproductive period.

However, repeated spawning in tropical cichlids like Parachromis managuensis and Cichla spp.

(Chellappa et al. 2003; Gomiero and Braga 2004; Normando et al. 2009) showed oocytes in distinct groups, with different maturation stages, differences in the gonad macroscopic appearance, and differences in egg size. Repeated spawning during a year is possible for giant cichlids given a sufficient food supply and favorable environmental conditions (Arredondo-

Figueroa et al. 2013), however, high individual investment of energy in reproduction and brood defense for 6-8 weeks would probably result in a single annual spawning event for most individuals.

Fecundity

This study is the first to quantify fecundity in wild giant cichlids. Fecundity was correlated to female size, similar to most fishes (Olsen and Rulifson 1992; Tyler and Sampter

1996; Alejo et al. 2011), including other cichlids (Duponchelle et al. 2000; Faunce and Lorenz

2000; Normando et al. 2009). Despite the larger gonad size in fish from Lake Yaxhá, I found no fecundity difference between lakes in females of the same size range. Although fecundity is similar in both lakes, egg size is larger in Lake Yaxhá. According to life history theory, a decrease in fecundity with an increase in egg size occurs when juvenile mortality increases.

Thus, unfavorable conditions for juveniles may occur in Lake Yaxhá, probably including cannibalism (unpublished diet data), because of the proportion of large fish in the population.

Conversely, Walsh et al. (2006) found declines in fecundity when size-selective fishing occurred

46

over five generations, which is probably a low value for Lake Petén Itzá, where there has been more than four decades of strong human colonization in the area (Scharwtz 1990). Although fecundity did not show any difference between lakes, egg size was smaller in Lake Petén Itzá.

Fecundity and egg size are highly variable traits that can be affected by environmental factors. Smaller eggs usually hatch smaller larval fish with lower survival (Walsh et al. 2006) and survival can be much lower with fish that mature early and are inexperienced in parental care. Usually, larger eggs are associated with old-larger parents (Trippel 1995; Duponchelle et al. 2000) and in stable conditions (Winemiller and Rose 1992). On the one hand, I have larger eggs but the same fecundity, which is a mixed result according to life history theory. My results both support and contradict life history theory predictions, similar to Edeline et al. (2007) who found alternating pulses between natural selection and fishing selection,. However, as in any study of natural populations, I cannot rule out the possibility that the effects attributed to factors included in my models instead reflect effects of other, untested factors.

Maturity Schedule

My models suggest no difference between lakes in size at maturity of giant cichlids. This is surprising because other studies suggest that cichlids have the ability to switch to a more opportunistic life history strategy (Dupocehlle et al. 1990; Faunce and Lorenz 2000) with a higher adult mortality, as in Lake Petén Itzá (Chapter 4). Trends in maturity have been associated with changes in growth and mortality (Trippel 1995; Olsen et al. 2004; Walsh et al. 2006). Lake

Yaxhá showed higher size at age in both males and females (Chapter 4) than Petén Itzá, but size at maturity was the same. Despite the difference in length frequency distribution structure size difference and the lack of large individuals in Lake Petén Itzá, the maturity schedule was best explained by a single model. The only other model that can explain the data as well was when males in LakeYaxhá mature at larger size. On the other hand smaller males in Lake Petén Itzá

47

could be explained by males maturing early. Because life history traits examined in this study are dynamic and giant cichlids can show dynamic plasticity, it is necessary to follow the maturity schedule over time to understand interactions between resource limitation, changes in allocation of energy and increases in large adult mortality.

Finally, some life history traits differ at the population level, like sex ratio, egg size and

GSI, but others show great similarity. I cannot completely rule out differences between lakes or explain all differences in terms of higher mortality of large individuals. Trying to explain reproductive trait differences using data from only a single annual cycle emphasizes the importance and inherent difficulties of this theme in fish ecology and management. A wider monitoring program will be desirable to test not only annual differences, but to identify any possible bias in our approach. The age of first maturation that can be affected by fishing and should be included in the monitoring program. Additionally parameters of larvae and juvenile survival and predation ability still needed to evaluate the impacts in giant cichlid life history parameters.

48

Table 3-1. Developmental stages of gonads in female giant cichlids from Lakes Yaxhá and Petén Itzá (adapted from Perez-Vega et al. 2006 and Nuñez and Duponchelle 2009). Ovary Stage Macroscopic description Maturity stage I. Immature Ovaries are formed but appear as tiny tubes. Partly Immature pinkish-yellowish and opaque. Oocytes are not readily visible to the naked eye. Ovary weight 0.1–0.5 g II. Maturing Soft pink/yellowish ovary is easily recognized as Mature an ovary by the appearance of small (~200 μm) oocytes throughout. Ovary weight is generally 0.5– 2 g II. Advanced Ovaries highly vascularized and increased volume Mature maturation occupying a significant part of the abdominal cavity. Ovary weights generally 2–5 g. Large (~600 μm) yellow/salmon oocytes and more homogenous in size. IV. Ripe The largest size of the ovaries. Salmon or bright Mature orange gonad is highly distended and occupies almost double the area in the gut. Gonad weight very large (5–15 g). Fish can expel oocytes with a gentle abdominal oppression. V. Spent/Resting Gonad size similar to stages 1 and 2. Oocytes seen Mature only in portions of the gonad. High degree of vascularization. Thicker ovary wall compared to immature fish.

49

Figure 3-1. Map of the study area, showing geographic location of Lake Petén Itzá and Lake Yaxhá within the Maya Biosphere Reserve.

50

500 1.4 2010 450 2010 L 1.2

400

350 1

300 0.8 250 0.6

200 Levels (m) Levels

Precipitation (mm) Precipitation 150 0.4 100 0.2 50 0 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month

Figure 3-2. Monthly rainfall (mm) and lake level (m) in Lake Petén Itzá, Guatemala in 2010.

51

Table 3-2. Candidate models fitted to maturity schedule for Petenia splendida in Lakes Petén Itzá and Yaxhá. (*) indicates the selected model. Model Parameters Log AIC likelihood differences *TL50a σa 2 -154.72 0 TL50Y.m σY.m TL50a σa 4 -152.87 0.30 TL50PI.f TL50Y.f TL50PI.m TL50Y.m σPI.f σY.f 8 -149.48 1.51 σPI.m σY.m TL50PI σPI TL50Y.f σY.f TL50Y.m σY.m 6 -151.65 1.86 TL50Y σY TL50PI.f σPI.f TL50PI.m σPI.m 6 -154.41 3.03 TL50PI TL50Y σPI σY 4 -152.24 3.39 Note: Total length at which 50% are mature (TL50); Petén Itzá (PI); Yaxhá (Y); all (a); female (f); male (m); σ Sigma

52

Figure 3-3. Length–frequency distribution for Petenia splendida immature (I), male (M) female (F) from Lake Petén Itzá (n=339) and Lake Yaxhá (n=327), Guatemala.

53

Figure 3-4. Average gonadosomatic indices for female Petenia splendida in Lakes Petén Itzá and Yaxhá captured during July and October 2010 and March-November 2011. The bars represent the standard error and the table contains sample size in each month sampled (n).

54

Figure 3-5. Batch fecundity for Petenia splendida females in Lakes Yaxhá (top) and Petén Itzá (middle), and for all lakes combined (bottom) from Guatemala, as a function of female weight.

55

Figure 3-6. Proportion of mature male and female Petenia splendida in Lake Petén Itzá (PI) and Yaxhá (Y), Guatemala, as a function of total length (TL). Size at L50 ranged between 210 and 240 mm. (+) proportion of mature in a given size category.

56

Figure 3-7. Proportion of mature males and females of Petenia splendida in Lake Petén Itzá (PI) and Yaxhá (Y), Guatemala, as a function of age (years). Age at L50 ranged between 0.90-1.25 years. (+) proportion of mature in a given age.

57

CHAPTER 4 AGE, GROWTH AND MORTALITY RATES OF THE GIANT CICHLID (Petenia splendida Gunther 1862) IN GUATEMALA

Introduction

Age estimation is critical in determination of growth, survival and mortality of fishes and can be linked to different environmental factors that affect those life history variables. Fisheries science relies on age data more than other non-fish fields (Campana 2001). Today, fish stocks are often managed with age-structured models, and growth models based on age are used routinely (Walters and Martell 2004). Calcified structures in fish, such as vertebrae, scales, rays, opercules and otoliths that display sub-annual (Bwainka et al. 2007) or annual resolution

(Beckman and Wilson 1995) are key in aging fishes.

Using otoliths to determine age of fish is widely accepted and used mostly in temperate zones (Campana 2001) because strong intra-annual temperature differences, which affect growth patterns in the otoliths, making them act as “natural data loggers” (Beckman and Wilson 1995).

The physiological origin of the growth zones in calcified structures is not clear (Ferreire and

Russ 1994), but both abiotic and biotic factors probably play a role in creating “growth zones.”

Aging tropical fishes, however, is more difficult because it is hard to discern “macro-zones” that correspond to seasonal patterns, probably because of less extreme differences in temperature throughout the year. Growth is probably more consistent year-round, thereby reducing differences across otolith zones or in any calcified structures. Although seasonal fluctuations in temperature are not as extreme as in temperate latitudes, other seasonal cycles are important, such as wet and dry season (Bwanika et al. 2007). In any case, otoliths in fish from tropical zones can be used to age fish successfully (Beckman and Wilson 1995; Bwanika et al. 2007).

Growth provides a general assessment of habitat suitability, prey availability or the influence of management activities. Because fishing reduces intraspecific competition,

58

populations can be released from density-dependent processes, which results in faster growth

(Law 2000). Faster growth provides larger fish for fishers. Growth rate can influence size and age at maturity, and variation in size and age at maturity of individuals may also affect growth.

Thus, assessing growth is one of the most important steps to manage a fishery (Walters and

Martell 2004).

One of the most conspicuous effects of fishing is higher fish mortality. Usually fishing gear is selective for fish size, selecting for larger sizes and indirectly older or faster-growth individuals. High fishing mortality can result in overfished and depleted stocks (Meyers and

Worm 2003; Orensanz et al. 2005). Empirical studies show that strong size-selective fishing mortality over time leads to a downward shift in size at age (Olsen et al. 2004). Mortality can be estimated by different methods, but the most used by far depends on age.

Only a few age and growth studies in tropical areas have been undertaken, mainly for

African cichlids. Egger et al. (2004) found contrasting otolith growth zones in cichlids that were a consequence of seasonal rainfall differences in Lake Tanganyika. Bwanika et al. (2007) found definable “bi-annuli” (two opaque/translucent zones per year) in Nile tilapia, related to wet and dry seasons in two lakes in Uganda. In tropical America, there are even fewer examples of comprehensive readings of macrozone formation than in Africa. Noiset and Micha (1996) compared aging with scales, otoliths and length frequency in giant cichlids from the San Pedro

River, Mexico, and reported they formed a single annulus per year. Lack of older age groups,

due to high fishing pressure, made it impossible to determine L∞ (asymptotic length) in the growth model, and consequently mortality was not estimated in that study.

The giant cichlid, Petenia splendida, is the target of the most important freshwater artisanal fishery in north Guatemala and south Mexico. Aquatic ecosystems like Lake Yaxhá, in

59

Guatemala which have minimal habitat transformation and no artisanal fishery, are exceptionally difficult to find (Granados-Dieseldorff et al. 2012; Barrientos et al. in press). Thus, Lake Yaxhá provides a rare, unfished situation to compare to Lake Petén Itzá, where an artisanal fishery has long been established. I examined the age, growth, and mortality of the giant cichlid (Petenia splendida) in Lake Petén Itzá and Lake Yaxhá. The specific objectives were to a) develop an aging method using otoliths, b) create an age-length key to determine population age structure, and c) estimate total mortality and survival rates.

Study site: Petén, Guatemala is the southern extent of the Yucatan Peninsula, a large marine limestone platform in Mesoamerica. Petén encompasses a lake district comprised of about 14 large water bodies, of which Lake Petén Itzá is the largest (~100 km2) and Yaxhá is the second largest (~7.4 km2) (Figure 4-1). Most of Petén lies within the greater Usumacinta River

Basin, which includes all the major rivers and the border zone with Mexico. Annual rainfall can vary from ~1200 to ~2000 mm and such inter-annual precipitation fluctuations drive shifts in regional lake levels (Deevey et al. 1980).

Lake Petén Itzá owes its origin to a combination of limestone dissolution and tectonism.

It is about 32 km long and 3 km wide, with a maximum water depth of ~165 m. It is thought be the deepest lake in lowland Central America. The lake water chemical composition is dominated by cations Ca and Mg and anions SO4 and HCO3 (Perez et al. 2011). The lake has two elongate basins that run east-west. The larger north basin supports at least four towns along its shoreline, including San Andres, San Jose, Jobompiche and El Remate. The smaller south basin supports the largest population (~160,000, National Statistics Institute 2008) in the towns of Flores, Santa

Elena and San Benito. The lake is a closed hydrologic basin with no visible outlet and water level fluctuates, with a slight lag, in response to the amount of precipitation. The littoral zone is

60

most extensive in the shallow west part of the south basin, with a variety of plant species in the zone. Although fishing is not legal, there is an artisanal fishery with gillnets based on Petenia splendida and other cichlids in the lake (Barrientos et al. 2012).

Lake Yaxhá is part of the Mopan River Basin that goes from east to west in the southern part of the Yaxhá-Nakum-Naranjo National Park. The lake origin is attributed to tectonic/solution processes, similar to Lake Petén Itzá. It is about 4 km long and 1 km wide, with a maximum depth of ~26 m. Access to the lake is limited to ~15 km of non-paved road, which restricts vehicle visits in the rainy season. Within the lake, there are only a few small boats that are used for transport to the archaeological site of Topoxte. Fishing is prohibited. The

National Council of Protected Areas (Consejo Nacional de Areas Protegidas; CONAP) and

Institute of Anthropology and History (Instituto de Antropologia e Historia; IDAEH) share park administration. Differing from Lake Petén Itzá, cultural eutrophication does not play a large role in this lake, because of its location within the Maya Biosphere Reserve (MBR) and the lack of communities along the shores. Only a few farms are located along the south basin of Lake

Yaxhá. Despite the difference in size and being located in different basins, fish communities are similar between the lakes, with the exception of the fact that there are introduced species in Lake

Petén Itzá (Barrientos et al. in press).

Methods

Fish Sampling

Fish were collected using a boat-mounted electrofisher (1.5 KVA; Smith-Root,

Vancouver, Washington) with an electrical output of 5 to 8 A of pulsed DC, in shoreline areas where water depth usually ranged from 1-2.5 m. Sampling of the lakes was completed monthly from March to October 2011 in both lakes on the same dates. Sampling usually lasted 2 days and an attempt was made to capture 30 fishes. All fish were storage on ice to be transported into

61

the laboratory. Fish were measured for TL to the nearest 1 mm, blotted dry and weighed to the nearest 0.1 g. Both sagittal otoliths were cleaned with water and dried in the field and stored dry in plastic vials with a field collection number until processed for aging. Gonads were removed and sex of individual fish was determined using visual examination of the gonads. Small fish

(<200 mm TL) that could not be sexed using visual examination and were assign a immature stage.

Length-Weight Relationships

Fisheries management usually requires body weight for catches and biomass regulations, but often only has length. Weight can be predicted from length with the help of length-weight relationship. A power function was fitted to length and weight data,

푊 = 푎(푇퐿푏) (4-1)

where W is the weight in grams without gonads; a is the scale factor; TL is the total length of the fish in centimeters and b is the power exponent. The length-weight relationship

(LWR) was fitted separately for males and females in each lake. I transformed the data (log10) and used ANCOVA to determine if the parameters of the length-weight relationship were different between lakes, sexes or sampled month. All tests of significance were at p≤0.05.

Length Frequency

I plotted length frequencies for males and females of the giant cichlid to detect potential differences between sexes and between lakes. Length frequencies were compared using

Kolmogorov-Smirnov (KS) tests by lake and sex.

Otolith Processing

The left otolith was used to estimate giant cichlid age, but in case of loss or damage to the left otolith, I used the right otolith. First I used whole otoliths that were immersed in water and

62

examined under a stereoscope (Figure 4-2). Because of the difficulty of seeing more than one annulus in the whole otoliths, which might result in an underestimate of age, I opted to thin- section any otolith with at least one ring present (Ferreira and Russ 1994; Newman et al. 2000).

Each otolith was secured to a frosted glass slide with crystal bond and sectioned into two different thicknesses (0.2 and 0.4 mm) through the core region using a low speed saw (350 rpm).

The frosted glass with the otolith already sectioned was reheated to make removal of sections easier and to reduce damage to sections. Sections were mounted on a clear slide using

Histomount (National Diagnostics)..

Age Estimates and Validation

For age estimation I started by using whole otoliths. However, because cross-sections of otoliths usually yield better readability, I began to use cross-sections to check and identify the opaque and translucent zones (Figure 4-3). The sections were still difficult to read because of problems like split opaque zones or indistinct differences in the opaque zones. I followed the opaque zones close to the sulcus when possible. Most nuclei were difficult to interpret because of differences in size among them. I read each otolith section two times, using a Leica stereoscope

(4-10x) under transmitted light to count opaque growth zones, because the translucent zones were broad and more diffuse. There was a 6-week interval between the first and the second reading. If there was a discrepancy between the first and the second reading then a third reading was done 6 weeks later. All age estimations were made without knowledge of sampling date, TL or any other information. The periodicity of annulus formation was validated using edge analysis. For edge analysis, the percentage of fish having otoliths with an opaque zone versus a translucent zone on the edge was plotted monthly. For edge type, the following codes were used: i) the opaque zone was at the edge (edge = 0); ii) narrow translucent zone at the edge (width less than about 30% of the previous increment = 1); iii) medium translucent zone at the edge (width

63

about 30-60% of the previous increment = 2), and iii) wide translucent zone at the edge (width more than 60% of the previous increment = 3).

The narrow opaque zones were counted to estimate the number of annuli. The annulus was followed right to the sulcus in the sectioned otoliths when possible. Then number of annuli and edge type and collection date were used to assign ages to individual fish through advancing or not advancing counts of annuli (i.e. fish with 2 annuli and edge = 3 in March will be advanced to age class 3).

To assess the accuracy of readings, I used Hoenig’s test of symmetry (Hoenig et al.

1995), the coefficient of variation (CV) (Chang 1982) and the average percent error (APE)

(Beamish and Fournier 1981).

Growth

To assign ages to giant cichlids that were only measured and not aged directly, age-length keys were constructed for fish in both lakes. Age-length keys were based on pooling of the sexes because in the fishery they do not distinguish sexes. Age data were grouped into 20-mm length classes.

To model age and growth of giant cichlids, I used data from male and female fish from the two lakes independently to fit von Bertalanffy growth curves to observed TL at observed age data (Ricker 1975). The form of the von Bertalanffy growth curve for modeling fish length as a function of its age was:

(−푘(푡−푡표)) (4-2) 퐿푡 = 퐿∞(1 − 푒 )

where Lt is the length at age t, L∞ is the average maximum length that the species will reach if it lived indefinitely, k is a growth coefficient that measures the rate at which the average maximum size is reached ( known as the Brody growth coefficient) and t0 is the theoretical age at zero

64

length. To test for homoscedasticity and that error were normally distributed; I used a plot with residuals versus fitted values and a histogram of the residuals. I used a two-way ANOVA to test for differences in length at age class between lakes by sex. The main effects were lakes and sex, and length at age was the response variable. I used Akaike Information Criteria (AIC) to test for differences among reduced or complete models. I considered a ∆AIC larger than 2 significant

(Taper 2004). For complete models, I used different L∞, k, and t0 for males and females in the same lake. Because I found differences between males and females (AIC >2), I compared data between lakes using only the same sex.

Estimates of Total Mortality

Total mortality (Z) was determined using catch curve analysis (Ricker 1975), with the

Chapman and Robson (1960) mortality estimator. Because the method of fish collection was consistent between lakes, I assumed that age distribution from both lakes was represented in a similar manner for comparison. I only used fish fully recruited to the gear in the initial years (i.e. only descending right data from the catch curve). For Petén Itzá this is age 0 and for Lake Yaxhá it was age 2. I used an age-length relation to assign age to a number of fish that were not aged directly. Because of different growth rates between the two lakes, I used different age-length keys for Petén Itzá and Yaxhá. Chapman and Robson (1960) suggest that age classes above the age when the catch falls below five individuals should be excluded. I, however, followed the recommendation of Dunn et al. (2002) who suggested one individual as the cutoff, and showed that the regression performed better using all available age classes.

Apparent survival (S) was then estimated from:

S= e-z (4-3)

where Z is the instantaneous total annual mortality (Ricker 1975) .

65

I also estimated the total annual survival using the relationship of A=1-S where A is annual total mortality. I estimated natural mortality (M) for Lake Petén Itzá using the equation proposed by Hewitt and Hoenig (2005):

4.22 (4-4) 푀 = 푡푚푎푥 where tmax is the maximum age observed. Considering that Lake Petén Itzá is size truncated with probably older fish absent from my sampling, I used Lake Yaxhá non-fishing situation as my natural mortality (Z=M) for both lakes and used Hoenig only as a reference point.

Results

Length Frequency

Length frequencies showed differences between the two lakes that were consistent with expectations based on differential fishing pressure. A total of 667 Petenia splendida were collected in 2011, 339 from Petén Itzá and 327 from Lake Yaxhá. Fish ranged from 44 to 436 mm TL in Lake Yaxhá and 40 to 393 mm TL in Lake Petén Itzá. Males were larger than females in Lake Yaxhá (KS test; P<0.0001). Males were slightly larger than females in Lake Petén Itzá

(KS test, P<0.0005). Therefore males and females were separated in subsequent analyses. The observed trend was that fish in Lake Yaxhá were larger, regardless of the sex (KS test,

P<0.0001; Figure 4-4).

Length-Weight Relationships

No difference in length-weight relations between sexes was found for Lake Yaxhá

(ANCOVA; slopes, P=0.64; intercepts, P=0.67) thus I pooled all data from the lake. Lake Yaxhá showed a curvilinear relation W= 6×10-6 TL3.1245 (r2= 0.984). Conversely, there was a difference in Lake Petén Itzá between sexes (ANCOVA slopes, P =0.0146; intercepts, P =0.0143) so I used different models for males (W=2×10-6 TL3.2787; r2= 0.948) and females (W=3×10-7TL3.6384; r2=

66

0.906). However, it is usually difficult for fisherman or enforcement agency personnel to distinguish sex, thus I also calculated an equation using pooled data for Petén Itzá (W=5×10-6

TL3.1489; r2= 0.972). The common equation for both lakes combined data (size range 40-436 mm) was: W=5×10-6 TL3.1311 (r2= 0.982). March had lower weight at length compared to all other months, thus it was excluded from the common equation for both lakes (ANCOVA slopes,

P=0.0.0007; Figure-4-5).

Age Estimates

Despite difficulties in reading and interpreting otolith zones and nuclei (see methods), high agreement was found between readings. The giant cichlid collected in Lake Yaxhá ranged from 0-8 years old and Lake Petén Itzá from 0-5 years old. Fish sampled from Lake Yaxhá showed a unimodal age distribution dominated by fish of age 3 (30%), whereas the sample from

Lake Petén Itzá was dominated by fish of age 1 (35%) and age 0 (33%) (Figure 4-6). No deviation from the 1:1 relation was found in ages assigned between the first and the second read

(Hoenig’s test, X2= 6.61; df=5; P= 0.25). An age bias plot with 95% confidence intervals did not reveal non-linear biases caused by under or over estimation of age (Figure 4-7). The CV was

4.41% and the APE was 3.12%, with 85% agreement between reads. All aging precision tests indicated that sagittal otolith sections were precise for aging giant cichlids. In total 254 fish were aged from the Lake Yaxhá collections and 225 from Lake Petén Itzá. The age-length keys for each lake showed a range of ages for a given size. For Lake Petén Itzá (Table 4-1), a fish in the size range of 280 mm could be from 1 to 3 years old, with approximately half (65%) of them 2- year-old. In addition, the largest (i.e. > 340 mm) fish in Petén Itzá were nearly the oldest fish.

The largest fish from Lake Yaxhá, however, were not always the oldest (Table 4-2).

67

I used giant cichlid over the whole range of ages in the edge analysis. Based on edge analysis for the ages assessed, the majority of giant cichlid deposited an opaque zone between

May and September, represented by the unimodal plot (Figure 4-8). The period of the opaque zone deposition coincides with the peaks of the rainy season. Conversely, the translucent zones were the widest at the beginning and the end of the rainy season. Based on the validation of aging method for giant cichlid the age of the fish was estimated as the number of opaque annuli in each bone, which was reasonable despite not having samples in the dry season (i.e. Nov, Dec,

Jan and Feb). Because of the unimodal plot I assumed that fish in March, April and May with a

3 edge code was going to form an opaque zone in the later months in the same season, thus advanced them one year to the age class of the cohort.

Growth

Growth curves for males and females in both lakes had different L∞ and k (Table 4-3; 4-

4). Separate growth curves were therefore retained for males and females in Lake Yaxhá and

Lake Petén Itzá (Figure 4-9). Total length at age was larger for males than females in Yaxhá (F

3,1= 107 ; P< 0.0001) and Petén Itzá (F3,1=31; P<0.0001)(Table 4-5). On average, males from

Lake Yaxhá had greater total length at age than males in Lake Petén Itzá (F 3,1= 15.32 ; P<

0.0001). On average, females from Yaxhá were not significantly different in total length at age

>2 (P>0.05) whereas males had no difference in total length in age >3 (P>0.05), indicating reduced growth rate beyond age two in females and age three in males (Table 4-6)

Mortality Estimates

Instantaneous mortality did not differ among females (ages 2-5), and males (ages 2-5;

ANCOVA: homogeneity of slopes, P=0.14) in Lake Yaxhá; similarly there was no mortality difference in Lake Petén Itzá females (ages 2-5) and males (ages 2-5; ANCOVA: homogeneity of slopes, P=0.51). I therefore pooled the loge catch as a function of age for both sexes, from

68

Lake Yaxhá (r2= 0.76) and Lake Petén Itzá (r2=0.70) (Figure 4-10). Instantaneous total mortality estimated for Lake Petén Itzá (ages 2-5) was higher than Lake Yaxhá (ages 2-5; ANCOVA: homogeneity of slopes, P=0.06). Lake Yaxhá had a lower total instantaneous mortality than

Petén Itzá, i.e. survival was lower in Lake Petén Itzá. Natural mortality for Lake Petén Itzá

(Table 4-6) was M=Z from Lake Yaxhá.

Discussion

Contrasting giant cichlid age, growth, size/age distribution in lakes subject to different fishing pressure scenarios adds to the understanding of fish population response to induced high mortality of large individuals. Sex-specific differences in growth, size and age structure were evident between giant cichlids from fished Lake Petén Itzá and unfished Lake Yaxhá. In this discussion, I summarize the major trends I found and interpret the variation in life history traits observed based on lakes differences in environment and fishing.

Length Frequency

Fishing pressure is probably the most important factor to explain the different size and age structures between the two lakes, however I considered other possible factors such as habitat differences and fish behavior. Larger fish (>15 cm) that dwell in littoral habitats move considerably more than small fish that live mainly in this zone (Chapter 1; Barrientos et al.

2008). Large fish move in and out of the littoral zone, depending on their feeding behavior.

Because I sampled similar habitats in both lakes, it is not likely that large fish from only Lake

Petén Itzá moved out of the littoral zone during sampling. Moreover, fish were obtained on the same dates and with the same gear, which probably led to sampling fish with comparable nesting patterns in the littoral zone. If behavior and habitat are less likely factors to explain the smaller

69

size and age structure in Lake Petén Itzá, then higher fishing mortality is the more likely explanation (Berkeley et al. 2004).

Age Estimates

The present study on giant cichlid advances the use of calcified structures to estimate age and growth parameters for tropical fish species. I used sagittal otoliths, which were successfully used before in giant cichlid (Noiset and Micha 1996) and Nile tilapia (Egger et al. 2004;

Bwanika et al. 2007; Efitre 2007). Agreement between readers in fish sampled from temperate zones is usually >90%, but can be lower in tropical species (Campana 2001). Tilapias and reef fish are the tropical taxa with more age studies, probably because of their importance in fisheries and aquaculture. Bwanika et al. (2007), using Nile tilapias in tropical Africa, found an APE of

5.6 and CV of 7.9, which was within the range of most values for APE and CV in a review by

Campana (2001). My results are in the same range, probably because of better than expected readability of the cross-sections of otoliths embedded in Histomount, which improves light penetration (Figure 4-3). Although, it was apparent that the opaque zones become less distinct with increasing size of the otoliths and added some variability between reads.

The timing of deposition of the opaque zone in giant cichlid was validated with edge analysis and coincides with the noticeable rainy season common for these latitudes. Similar to other tropical species, that had opaque zone deposition related to rainfall, that potentially could increase growth with more food availability (Egger et al. 2004; Bwanika et al. 2007; Efitre

2007). Conversely, Noiset and Micha (1996) hypothesized that the annuli formation was at the end of the dry season for giant cichlid; however they did not validate the periodicity of the annuli deposition (Campana 2001). Moreover, Noiset and Micha (1996) only sampled 5 months and skipped the crucial months that showed the tendency of the completion of the translucent zone.

70

This study underscore the need of validate the deposition of the macrozones of a fish in systems specific basis (Bwanika et al. 2007).

Growth

The giant cichlid shows sex- and lake-specific growth. Males were consistently larger than females in Lake Yaxhá. Although sexual dimorphism it is not uncommon for tropical fish with parental care (Magurran and Macias 2000; Egger et al. 2004), it was previously unknown for giant cichlid. Females tend to be larger in other fish families in Mesoamerica (e.g.,

Poeciliidae and Atherinidae), probably because fecundity is related to female weight. However because the cichlid mating system in Mesoamerica is based on nest site protection, larger males have a reproductive advantage if large (Magurran and Macias 2000). In cichlids from Africa,

Bwanika et al. (2007) found males had greater growth rates and attributed this to a size advantage conferred upon males defending a territory. Egger et al. (2004) found differential growth between sexes, with size dimorphism increasing with age. Similarly, size differences in older giant cichlid were evident in Yaxhá where the population age structure was older and larger individuals were common in the sample. The male size advantage was less evident in Lake

Petén Itzá, possibly because the advantage attributed to sexual selection for larger males is countered by fishing pressure on large, potentially faster-growing individuals (Berkeley et al.

2004; Carlson et al. 2007; Allendorf and Hard 2009).

It is counterintuitive that giant cichlid from Lake Petén Itzá grew slower than those from

Lake Yaxhá. Growth can be influenced by a variety of factors, including environment, fishing, and genetics. Fishing pressure in Lake Petén Itzá should reduce intraspecific competition, which in turn should result in higher growth rates (Ward et al. 2006), but I found exactly the opposite.

Conversely, low fishing pressure may result in high population density, which increases competition for resources, leading to density dependent growth (i.e. slower growth), but I found

71

the opposite in Lake Yaxhá. One plausible explanation could be that productivity in the two lakes differs; based on Secchi disk measurements (Canfield et al. 1983; Perez et al. 2010), Yaxhá is mesotrophic and Petén Itzá is oligotrophic. Recently, however, Lake Petén Itzá’s trophic state was locally driven to mesotrophy/eutrophy by cultural eutrophication (Rosenmeier et al. 2004), mainly in the south basin. Although our sample was from the entire littoral zone in Lake Petén

Itzá, most of the fish (~70%) were captured in the south basin, which has seen the largest impact from cultural eutrophication, as reflected by Secchi disk readings that are comparable to those made in Yaxhá (AMPI 2010; Barrientos unpublished data). Another explanation is that fishing is typically size-selective, targeting large individuals, which are presumably the faster-growing and older individuals in a population (Law 2000). This may favor slow growers and a smaller size at age alive to reproduce (Olsen et al. 2004; Walsh et al. 2006). In laboratory experiments, Atlantic

Silverside, Menidia menidia, Conover and Munch (2002) found selection for slower somatic growth following selective removal of the larger individuals in the population. My data suggest a genetic response to a robust directional selection for smaller-slow-grow individuals in Lake

Petén Itzá. On the other hand, natural selection probably selects against slower-growing individuals in Lake Yaxhá because they vulnerable to predation for a longer time (Carlson et al.

2007; Edeline et al. 2007; Allendorf and Hard 2009)

Growth studies in Petenia splendida are rare. Noiset and Micha (1996) fitted data from fish sampled in the San Pedro River using three different methods (ELEFAN, scale reading, and

otoliths) to obtain parameters for the von Bertalanffy growth curve. They found L∞ ranging from

431-452 mm and a growth coefficient (k) ranging from 0.30-0.38. Lack of older individuals was a problem for fitting the growth curve from otoliths because it did not reach an asymptote. In my study, the von Bertalanffy model provided a good fit for Petenia splendida for the different sexes

72

in Lake Yaxhá. Fit, however, was not as good in Lake Petén Itzá, probably because the age class truncation due to the effects of fishing. My growth coefficients (k) for both lakes were larger

(range of 0.51-0.97) than that obtained by Noiset and Micha (1996) and showed that the giant cichlid is a short-lived, fast-growing fish.

Mortality

Total mortality estimates for giant cichlid were low for Lake Yaxhá and high for Lake

Petén Itzá, consistent with expectations based on the low fishing pressure in Lake Yaxhá. To my knowledge there is no estimate of mortality rates based on age for Petenia splendida in any other part of their distribution. An estimate of total instantaneous mortality of 0.874, using Length

Frequency Distribution Analysis (LFDA) in Lake Petén Itzá from a national report on the fishery

(Ixquiac 2007), is clearly lower than my estimates. Ixquiac (2007) reported an annual total mortality of 58% in Lake Petén Itzá, whereas my estimate of total annual mortality in Lake Petén

Itzá of 65% was higher and lower (54%) in Lake Yaxhá. Comparison of giant cichlid age structure between lakes suggests that it will be impossible to have such low annual mortality in

Lake Petén Itzá. Age data provide better estimates of mortality and are preferred to LFDA for stock assessment (Walters and Martell 2004). Lake Yaxhá, which is an unfished lake, presented an age structure based on 2- and 3-year-old fish, not 0- and 1-year-old fish, as in Petén Itzá.

Ixquiac (2007) is a report from the Guatemala Science and Technology Office (SENACYT), and could be used to make fishery management decisions. Underestimating mortality, however, could be misleading, and inappropriately enable fishing effort to be increased in management plans and eventually result in stock overfishing.

These results represent the first estimates of age, growth and mortality for Petenia splendida in Guatemala, and the first in lentic systems for Mesoamerica, and provide an important first step toward appropriate fishery management in the region. They also provide

73

needed data to inform fish population dynamics models used to test management actions. My sampling design included two large areas in the Mayan Biosphere Reserve under different management regimes, enabling study of baseline conditions in an unfished lake, with better longevity and distinctive size structure. Growth of Petenia splendida can be modulated from lake to lake and selection possible by favoring small, slow-growing fish, as in Lake Petén Itzá and larger, faster-growing individuals in Lake Yaxhá (Edeline et al. 2007; Carlson et al. 2007;

Allendorf and Hard 2009). This is important to consider in light of the recent interest to use giant cichlid as an aquaculture species (Arredondo-Figueroa et al. 2013). Guatemala is now stocking giant cichlid in Lake Petén Itzá, and it will be important to understand the growth patterns and choose a parent generation of fish that show high growth rates. Further exploration of growth in regional giant cichlid populations will be needed to manage the species in different ecosystems

(e.g. lakes, rivers) across its wide distribution range. Like other artisanal fisheries, the giant cichlid fishery is spatially structured, small-scale and targets sedentary stocks (Orensanz et al.

2005), but supports the livelihoods of thousands of families, so it is necessary to manage such fisheries at the local level, with explicit regulations. I recommend a fishing and natural mortality study to better understand the sources of mortality. I also recommend a fish movement study to understand the dynamics of fish migration in and out of the littoral zone.

74

Figure 4-1. The Maya Biosphere Reserve and the aquatic systems studied in Guatemala. Study lakes are in different management zones, with difference in fishing pressure. (Sampling occurred in March-October 2011)

75

Figure 4-2. A whole sagittal otolith from giant cichlid captured at Lake Petén Itzá, showing the nucleus (N), an opaque (O) zone, and a translucent (T) zone.

Figure 4-3. Otolith section of giant cichlid from Lake Petén Itzá. Opaque zones (OZ) were counted for age estimation. Translucent zones (TZ) occur between OZ. This is an individual with 3 annuli.

76

Figure 4-4. Length–frequency distribution for immature (I), male (M) and female (F) giant cichlid from Lake Petén Itzá and Lake Yaxhá, Guatemala.

77

Figure 4-5. Total body weight as a function of total length by month of giant cichlid from Lake Petén Itzá and Lake Yaxhá, Guatemala.

78

Figure 4-6. Giant cichlid age frequency distribution collected in Lakes Petén Itzá (n=339) and Yaxhá (n=327) with electro fishing.

79

Figure 4-7. Bias plot for age estimates from giant cichlid otoliths (n=103) from Lakes Yaxhá and Petén Itzá. Numbers above the read 1 (x-axis) are numbers of individuals (n). The dashed line is the 1:1 relation. Bars represent the 95% confidence interval.

80

100 500 Rainfall opaque zone on edge Index of Completion

80 400

60 300

% % Opaque 40 200

PRECIPITATION (mm) PRECIPITATION % Index of Completion or Completion of % Index 20 100

0 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec MONTH

Figure 4-8. Monthly percent occurrence of the opaque zone at the edge and the percent index of completion of the translucent zone of otoliths of giant cichlid from in Lake Yaxhá and Lake Petén Itzá, Guatemala, relative to monthly precipitation in Lake Petén Itzá.

81

Figure 4-9. Length at age observed for giant cichlid fitted to the von Bertalanffy growth model by sex from fish sampled in Lake Yaxhá and Lake Petén Itzá Guatemala.

82

Figure 4-10. Catch curve analysis for giant cichlid from Lake Yaxhá and Lake Petén Itzá, Guatemala. Lake Yaxhá ages 0 -2 were not included in Z estimation, and only included in the figure for illustrative purposes. Lake Petén Itzá age 0 was not included in the Z estimation for and only included in the figure for illustrative purposes.

83

Table 4-1. Age-Length key for giant cichlid from Lake Petén Itzá, Guatemala. Percentage of fish in 20-mm length categories as a function of age. TL (mm) N Age 0 1 2 3 4 5 50-69 6 100 70-89 5 100 90-109 15 100 110-129 18 100 130-149 19 100 150-169 12 92 8 170-189 4 50 25 25 190-209 9 67 22 11 210-229 30 90 10 230-249 34 3 59 35 3 250-269 21 38 48 9 5 270-289 17 29 65 6 290-309 14 14 58 21 7 310-329 7 43 43 14 330-349 8 50 25 25 350-369 2 50 50 370-389 2 50 50 >390 2 50 50

84

Table 4-2. Age-Length key for giant cichlid from Lake Yaxhá, Guatemala. Percentage of fish in 20-mm length categories as a function of age. TL (mm) N Age 0 1 2 3 4 5 6 7 8 50-69 2 100 70-89 3 100 90-109 10 100 110-129 9 100 130-149 4 100 150-169 4 100 170-189 0 190-209 5 80 20 210-229 10 10 60 20 10 230-249 16 50 32 18 250-269 28 32 43 19 3 3 270-289 30 7 33 43 17 290-309 22 4 37 37 18 4 310-329 32 25 56 16 3 330-349 14 29 35 29 7 350-369 11 9 36 36 19 370-389 34 50 35 9 3 3 390-409 20 5 20 55 20 410-429 5 40 40 20 >430 2 50 50

85

Table 4-3. Candidate models for different sex fitted to growth parameters estimated for giant cichlid in Lake Yaxhá. (*) indicate the used model. Model df AIC AIC differences

2 L∞, 2k, 2t0 7 2588.6 2.1

1 L∞, 2k, 2t0 6 2626.9 40.5

2 L∞, 1k, 2t0 6 2598.8 12.2

*2 L∞, 2k, 1t0 6 2586.7 0

1 L∞, 1k, 2t0 5 2689.4 102.8

1 L∞, 2k, 1t0 5 2660.2 73.6

2 L∞, 1k, 1t0 5 2620.6 34

L∞, k, t0 4 2689.3 102.7

Table 4-4. Candidate models for different sex fitted to growth parameters estimated for giant cichlid in Lake Petén Itzá (*) indicate the used model. Model df AIC AIC differences

2 L∞, 2k, 2t0 7 2197.4 0

1 L∞, 2k, 2t0 6 2218.6 21.2

2 L∞, 1k, 2t0 6 2205.9 8.5

*2 L∞, 2k, 1t0 6 2197.8 0.4

1 L∞, 1k, 2t0 5 2230.9 33.5

1 L∞, 2k, 1t0 5 2226.3 28.9

2 L∞, 1k, 1t0 5 2213 15.6

L∞, k, t0 4 2229 31.6

Table 4-5. The von Bertalanffy parameters estimated for Petenia splendida by lake and sex. Sex L∞ k to Petén Itzá Male 362 0.59 -0.58 Female 277 0.94 -0.58

Yaxhá Male 406 0.51 -0.49 Female 297 0.97 -0.49

86

Table 4-6. Average Total Length at observed age (±S.E) of giant cichlid from Lakes Petén Itzá and Yaxhá, Guatemala. Yaxhá Petén Itzá Male Female Male Female Age Mean S.E. Mean S.E. Mean S.E. Mean S.E. (mm) (mm) (mm) (mm) 1 232 14.29 230.3 6.34 233.8 6.83 223.1 3.67 2 292.7 11.87 271.5 4.86 278 6.71 239.9 5.38 3 349.8 7.32 288.5 4.36 337.3 8.96 270.6 12.49 4 368.9 5.03 294.3 6.21 349 12.92 301.8 18.84 5 374.1 12.25 295 349 6 392.7 6.28 8 398

Table 4-7. Total mortality (Z), natural mortality (M) and annual survival rate (S) of giant cichlid captured in Lakes Petén Itzá and Yaxhá. Petén Itzá Yaxhá tmax 5 8 Age range used 0-5 3-8 Z 1.05 0.81 S 0.35 0.45 M 0.81 0.81 A 65% 55%

87

CHAPTER 5 ASSESSMENT OF ARTISANAL FISHERY POLICY FOR THE GIANT CICHLID (Petenia splendida Gunther 1862) IN GUATEMALAN LAKES

Introduction

Rapid human population growth in the Petén region of Guatemala occurred since 1960 as a consequence of government development plans (Schwartz 1990). This population growth has led to increasing demands on freshwaters and other ecosystems of northern Guatemala. Artisanal fishing is a traditional activity within lakes in Petén, but has increased in effort in some water bodies because of human population expansion (Madrid J. personal communication; National

Council of Protected Areas, CONAP). Gillnets and to a lesser extent hook-and-line are the main fishing gears. Artisanal fisheries are largely defined as being small-scale and spatially structured

(Orenzans et al. 2005), which means fishing effort is largely restricted to communities around the lake. A common management problem associated with artisanal fisheries is scarce data

(Orenzans et al. 2005); this situation is difficult to remedy because available data may not be collected at the appropriate spatial resolution and needed data are expensive to collect. Thus, the effects of fishing on fish populations are usually unknown.

The giant cichlid (Petenia splendida Gunther 1862) is distributed across a mosaic of freshwater systems in Mesoamerica, including the Petén region of Guatemala (Greenfield and

Thomerson 1997; Miller 2005; Perez et al. 2010), hence populations occur across different environmental conditions. In addition to differences in environment, populations are subject to a large range of exploitation levels. Although fishing harvest is not quantified by any agency, the

Guatemalan government considers the giant cichlid overexploited in some systems and a threatened species nationally (CONAP 2001). It is clear that fishing pressure has already resulted in population changes in some lakes (Chapter 4), potentially influencing the results of fisheries investigations. Nonetheless, there are lakes with minimal habitat degradation and no artisanal

88

fishery (Granados-Dieseldorff et al. 2012; Barrientos et al. in press). Such systems can provide key insights into fisheries management by allowing estimation of fisheries model parameters from lightly or unexploited populations (Chapters 3,4).

The artisanal fishery for the giant cichlid is vital to the livelihood of hundreds of fishers and their families in the Petén region and thus there is a need to assess the impact of fishing and policy (e.g. fishing regulations) on sustainability of the fishery. To meet this need, I compiled information on life-history traits such as growth, mortality, and reproduction and developed a yield-per-recruit and spawning potential ratio model to evaluate management policies, including length limits under contrasting harvest strategies. Data were obtained from life history studies of the giant cichlid in two lakes within the Petén region that differ markedly in fishing pressure

(Chapters 3,4). Lake Petén Itzá has regulations prohibiting gill nets, but has a large human population in cities along its shores and experiences considerable illegal fishing for giant cichlids

(J. Madrid personal communication; CONAP). Conversely, Lake Yaxhá is located within the

Maya Biosphere Reserve, has limited access, and is protected from fishing (Barrientos et al. in press). These two lakes provide a contrast in intraspecific life history parameters and population characteristics to compare fisheries management modeling scenarios.

Methods

I constructed a yield-per-recruit model similar to the model proposed by Allen and

Hightower (2004). The model employed survival schedules, growth and a Botsford and

Wickman incidence function, summarized by Walters and Martel (2004), which estimates biomass per recruit for a fish population. Length-weight relationships for giant cichlids were estimated separately for each lake (Chapter 4).

The form of the von Bertalanffy growth curve for length at age was:

89

(−푘(푡−푡표)) (5-1) 퐿푡 = 퐿∞(1 − 푒 )

where Lt is the length at age t, L∞ is the average maximum length that the species will reach if it lived indefinitely, k is a growth coefficient that measures the rate at which maximum size is reached ( known as the Brody growth coefficient) and t0 is the theoretical age at zero length. The von Bertalanffy growth curve parameters used for all simulations were obtained from Barrientos

(Chapter 4).

Total mortality (Z) was determined using catch curve analysis (Ricker 1975) with the

Chapman and Robson (1960) mortality estimator (Chapter 4). Apparent survival was then estimated from:

S= e-z (5-2)

where Z is the instantaneous total annual mortality.

I estimated natural mortality (M) for Lake Petén Itzá and Lake Yaxhá using Z from

Yaxhá, because in the absence of fishing F=0 (Chapter 4):

푍 = 푀 + 퐹 (5-3)

Finally, the vulnerability was set to the size reported by the artisanal gillnet fishery

(Table 5-1). I simulated a range of exploitation and harvest size to explore probable outcomes of length limit regulations.

Model: I used the function of unfished (ɸ0) and fished egg production per recruit (ɸf) to account for the effects of fishing on the reproductive capacity of the population (Walters and

Martell 2004). These incidence functions were calculated as:

(5-4) ɸ0 = ∑ 푓푎푙푎 푎

90

(5-5) ɸf = ∑ 푓푎푙푓푎 푎

where 푓푎 represents age-specific fecundity, and 푙푎 and 푙푓푎 are the survivorship schedules of the unfished and fished state, respectively. The value of (푓푎) was set to zero if age was less than age at maturity (Chapter 2), resulting in a fecundity-age relationship.

The model used survivorship curves to calculate the survivors per recruit to each age.

Survivorship to age a, in the absence of fishing, was found as:

푙푎 = 푆푎푙푎−1 (5-6)

where 푙푎 is the age-specific finite annual natural survival rate.

The survivorship schedules in the fished condition incorporate natural mortality and harvest as:

푙푓푎 = 푙푓푎−1푆푎(1 − 푈푉푎−1) (5-7)

where 푙푓푎 is the survivorship in the fished condition, U is the finite annual exploitation rate and V is the vulnerability to harvest.

Mean total length at age, 퐿푎, was calculated from the von Bertalanffy growth model

(Chapter 4).

To assess the performance of various harvest regulations, I simulated a range of exploitation (0.05-0.9) and size limits (200-390 mm total length [TL]). Model output included yield (kg) and spawning potential ratio (SPR). I used the static spawning potential ratio to evaluate the extent to which fishing mortality can reduce reproductive output for the giant cichlid:

91

ɸ (5-8) SPR = f ɸ0

The SPR measures the lifetime fecundity per recruit for a given level of fishing mortality and is commonly used to assess fisheries sustainability (Goodyear 1993). Recruitment overfishing is prevented by maintaining an SPR >0.3 (Mace 1994).

Parameter estimates used in the model simulation are shown in Table 1. Giant cichlid size at 50% maturity was below 226 mm TL (chapter 3), which is the TL age-1 in Lake Petén

Itzá, so I considered all fish mature at age 1.

The von Bertalanffy growth curve shows different asymptotic lengths and metabolic parameters for Lakes Petén Itzá and Yaxhá (Chapter 4). Walters and Martell (2004) observed that the value of K provides similar estimates of M calculated from maximum age. I favored this model because it incorporates information on age, growth and mortality, which I estimated

(Chapter 4) and I did not have effort/catch data.

Results

The model predicted that SPR for Lake Petén Itzá was larger than the threshold of 0.3 at the current exploitation level, with the average TL size captured by the fishery (Figure 5-1). At the current fishing “default” size (Table 5-1), only fishing mortalities higher than 0.9 would drive

SPR lower than 0.3. However, if the fishery changes gear to capture smaller fish (e.g. 20 cm)

SPR will be considered overfished at a range of lower exploitation levels, from 0.4-0.9 (Figure 5-

1). According to Walter and Martell (2004), sustainable rates of exploitation can be obtained by multiplying 0.8 by the natural mortality (M). That will be 0.64 for Lake Petén Itzá.

If a fishery develops in Lake Yaxhá and targets the same fish size as in Lake Petén Itzá

(i.e., 25 cm), the model predicts an SPR above the overfishing threshold at exploitation of 0.75 or higher (Figure 5-2). If fishers target giant cichlids <24 cm, the SPR will be below 0.3 at any

92

exploitation level higher than 0.4. Assuming that 0.8xM represents sustainable exploitation, the exploitation rate for Lake Yaxhá would be 0.48.

Isopleths for yield-per-recruit show that giant cichlids harvested at an average TL of 25 cm produces maximum yield at Lake Petén Itzá (Figure 5-3). In Lake Petén Itzá, the growth rate coefficient (k) was lower than natural mortality (M) obtained by maximum age. If fishers wait for fish to grow larger, the fishery is not productive, given the fact that M is high.

Isopleths for yield-per-recruit in Lake Yaxhá differ from Lake Petén Itzá in the yield quantity that can be obtained per 1000 recruits (Figure 5-4). The Lake Yaxhá maximum yield was ~60% more than the maximum yield in Lake Petén Itzá. Thus, if a fishery develops in Lake

Yaxhá and targets the same fish size as in Lake Petén Itzá, at the same exploitation level, the yield would greater. Both lakes show that SPR can be sustainable at levels where maximum yield is achieved, however length at harvest would be different between the lakes.

Discussion

Giant cichlid SPR and yield-per-recruit models differed between my study lakes. Lake

Yaxhá had lower natural mortality (Chapter 4), making it unlikely that it could sustain harvest rates similar to Lake Petén Itzá and a SPR below 0.3. Natural mortality in Lake Petén Itzá was higher, which makes the population there more likely to be able to sustain higher harvest rates and still have SPR above the threshold of 0.3 (Mace 1994). Lake Yaxhá will yield higher biomass per recruit than Lake Petén Itzá, because growth is slower in Lake Petén Itzá (Chapter

4).

There are no fishing mortality estimates for any freshwater fishery in Guatemala, so I used a realistic range in my model. Nevertheless, more accurate information is required for models to obtain meaningful results. Recently, a tagging and reward experiment in Lake Petén

Itzá indicated that annual exploitation could be larger than 40% (Barrientos; preliminary data).

93

Even with this exploitation rate, according to my model the SPR would not be lower than 0.4.

Future research should include estimates for fishing mortality, combined with telemetry methods to obtain a better understanding of natural mortality (Pine et al. 2003). Estimates of fish harvest are also necessary, but are difficult to obtain in a spatially expansive artisanal fishery (Orensanz et al. 2005), like the one in Petén Itzá. Moreover, because younger individuals can be similar in size to older fish, catch data do not reflect changes in age structure.

The natural mortality used in the model to estimate SPR and YPR was obtained from

Lake Yaxhá; however I did not find the same age structure in both lakes, with Lake Petén Itzá, having an age truncated structure. One probable explanation for the low numbers of large fish in

Lake Petén Itzá is that vulnerability to electrofishing was low. Under these circumstances, size/age structure reflects gear vulnerability more than fishing mortality in the system, causing underestimation of age, which affects the growth parameters and mortality rates I used in my model. Deriso (1987) claimed a theoretical basis for using an F value similar to M. Alternatively,

Walter and Martell (2004) suggested a more conservative approach and estimate F as 80% of M.

Using Deriso approach will yield an F that likely lead to recruit overfishing in Lake Petén Itzá, instead I recommend the more conservative approach.

The model was tested for different exploitation scenarios, with the assumption of constant recruitment. Recruitment, however, is not likely to remain the same. One fisherman from Lake Petén Itzá recalled that when the lake level rose in the early 1990s, giant cichlid abundance increased strongly over the next couple of years. Several fishermen claimed that when rain was poor, giant cichlids were scarce during the following year. This anecdotal information could indicate temporal differences in recruitment, which are not uncommon in

94

freshwater systems (Allen et al. 2009). It is unlikely that population parameters from only two geographic locations will be useful for managing all giant cichlid stocks.

Fishing effort is a concern for managers because fishers display different behaviors with regulations, season and in different areas. Although the only regulation in Lake Petén Itzá is not enforced, Lake Yaxhá is inside the Maya Biosphere Reserve, thus no fishing is allowed and this rule is enforced. The age/size structure in Lake Yaxhá reflects the no-fishing status of the system. Any change in fishing regulations in the zone will result in behavior changes among fishermen. For example, in Lake Petén Itzá, size “regulation” comes from the market. Tourism in the area creates a demand for local fish from the lake, which they like to be “plate size” (i.e. 25-

28 cm). Thus, there is a strong market for fish that size or a little larger, but not for smaller individuals (personal observation). Effort tends to be “regulated” by catch and gasoline price.

Low catch usually discourages more trips in the same week, because fishermen cannot afford them. Fisher behaviors are constrained by management regulations and economics (Hilborn et al.

2004a).

Exploitation is a concern for the government and the giant cichlid is on Guatemala’s imperiled species list, the “Lista Roja.” My model of the central lakes of Petén found populations in a sustainable status. SPR was sustainable, probably because the market demands specific sizes, which leads to a safeguard for reproduction. My study, however, was limited to a computer simulation, with associated error. Models are useful for detecting population trends, addressing uncertainty in management decisions and making quantitative predictions (Allen and Pine 2000).

Management implications

Simulation indicates that the current status of the fishery is sustainable in Lake Petén Itzá.

Giant cichlid should be managed as different stocks, but one species. Orensanz et al. (2005)

95

showed that unrestricted access or application of inadequate management options can result in a failure of the artisanal fishery. Moreover, only the right incentives for fishers can make sustainability possible. The best incentive for responsible behavior in fishing is probably long- term guarantee of fishing rights. This involves participation of fishers in survey and monitoring activities, which is needed in a Guatemala. Fishers are willing to help with data collection if the right incentive is provided (i.e. rewards given in the mortality study; Barrientos unpublished data). Lake Yaxhá should be open to fishing in the future, though the system must be monitored and hypotheses about the fishery must be tested. Lake Yaxhá, under its current no-fishing status, provides an excellent “control” system for study of the giant cichlid in a “no-take zone” (Hilborn et al. 2004b) that has not only preserved a potential future fishery, but allowed the government to test co-management strategies, with a guarantee of long-term fishing rights in the lake, a groundbreaking approach in Mesoamerica (Orensanz et al. 2005). Well-planned, hypothesis- based studies that are designed to evaluate a management tool, provide opportunities that potentially benefit all stakeholders involved, including fishers, scientists, government agencies and the giant cichlid.

96

Table 5-1. Parameters for giant cichlid populations from two Guatemalan lakes, that were used in a yield-per-recruit model. Lake Petén Itzá is a lake with an artisanal fishery and Lake Yaxhá is unfished. All parameters are from Chapters 3 and 4, except for exploitation (range) and regulatory information. Parameter Value Petén Yaxhá Itzá Natural Mortality M Instantaneous natural mortality (year -1) 0.81 0.81 S Annual natural survival 0.44 0.44 Exploitation U Annual harvest exploitation 0.05-0.9 Growth

L∞ Asymptotic length (mm) 331 383 k Metabolic coefficient (year-1) 0.68 0.47 t0 Time at zero length (year) -0.63 -0.71 Length-Weight a Length-weight coefficient (mm to grams) 0.000005 0.000006 b Length-weight exponent 3.15 3.124 Recruits 1000 1000 Regulatory Average size reported in fishery (mm) *250 - * average size in the gillnet fishery at lake Petén Itzá

97

Figure 5-1. Spawning potential ratio (SPR) on the harvest exploitation rate (U) and minimum length limits. Values of SPR below 0.3 are usually indicative of recruitment overfishing. Model parameters from Lake Petén Itzá.

98

Figure 5-2. Spawning potential ratio (SPR) on the harvest exploitation rate (U) and minimum length limits. Values of SPR below 0.3 are usually indicative of recruitment overfishing. Parameters for the model from Lake Yaxhá.

99

Figure 5-3. Yield (kg) isopleths plotted on harvest exploitation rate (U) and minimum length limits. Parameters for the model from Lake Petén Itzá

100

Figure 5-4. Yield (kg) isopleths plotted on harvest exploitation rate (U) and minimum length limits. Parameters for the model from Lake Yaxhá.

101

CHAPTER 6 EPILOGUE

The giant cichlid is an apex aquatic predator in most of the freshwater systems where it is found. Moreover, it is an important species to the livelihood of hundreds of fishermen. Previous management efforts in Guatemala have been ineffective, probably because of the perceived threatened status of the species and lack of life-history parameters for the taxon in fished and unfished conditions. This has resulted in the giant cichlid being assigned “red list” status in

Guatemala. Fishing is therefore prohibited in several systems, with hundreds of fishers marginalized economically, but continuing to fish illegally. Change in life history traits, like growth, age at maturity, can be triggered by high mortality and eventually can be catastrophic to stocks at the local and regional level. It is believed that giant cichlid is overexploited in several systems in Mexico (Noiset and Micha 1996; Perez-Vega et al. 2006; Arredondo-Figueroa et al.

2013), with devastating economic and social consequences. The giant cichlid fishery could be characterized as an S-fishery (Orensanz et al. 2005), which are small-scale, spatially distributed and target sedentary stocks. Effects of such fisheries are typically localized. At the same time, S- fisheries support the livelihoods of thousands of fishers and their families. A traditional enforcement-dependent management option is probably not viable in Guatemala, and instead, an approach that fully involves fishers and provides them with incentives is probably a better strategy for managing the giant cichlid fishery.

The goal of this study was to examine life history variations and juvenile ecology of the giant cichlid in Mesoamerican lakes, and based on that information, explore suitable fisheries models to evaluate management policies. This study showed that giant cichlid juveniles are found in the vegetated area of the littoral zone year-round (Chapter 2). The study also provided size at maturity, sex ratio, a fecundity estimate for wild populations and updated information on

102

the spawning season in the southern area of the giant cichlid distribution, in both fished and unfished lakes (Chapter 3). Whereas other studies speculated about several aspects of giant cichlid life history, this study provides specific data on the timing and characteristics of giant cichlid reproductive traits. This study found that males in an unfished lake had the largest size and age in the population, but this was not the case in a fished lake.

The age at maturity, coupled with natural mortality (Chapter 4) illustrates how this stock can support high fishing mortality, as long as the size demanded by the market remains >25cm

(Chapter 5). Lake Yaxhá, in the current no-fishing status, is an excellent example of an unfished giant cichlid population in a lentic system; the lake is equivalent to a “no-take zone” in marine systems (Hillborn et al. 2004b), which not only preserve the fishery, but also biodiversity. My recommendation is to explore innovative, alternative management schemes for S-fisheries, like territory use rights, which can provide a platform for fishers and scientists to work together in future management of the fishery. Future research should include direct estimation of fishing and natural mortality, and quantify fished volume to elucidate the economic impact of the lake on the livelihood of fishers. This dissertation pioneered study of the life history of Mesoamerican cichlids and represents the most up-to-date analyses of wild populations of giant cichlids. The information should be incorporated into future stock assessment models to improve management policies, and can be treated as useful baseline information on giant cichlids in Guatemala and elsewhere in Mesoamerica.

103

REFERENCES

Alejo-Plata C., Diaz-Jaimes P., and Salgado-Ugarte I.H. (2011). Sex ratios, size at sexual maturity, and spawning seasonality of dolphinfish (Coryphaena hippurus) captured in the Gulf of Tehuantepec, Mexico. Fisheries Research 110:207-216.

Allen M.S. and W.E. Pine. (2000). Detecting fish population responses to a minimum length limit: effects of variable recruitment and duration of evaluation. North American Journal of Fisheries Management 20: 672–682.

Allen M.S., Walters C.J. and R.M. Myers (2008). Temporal trends in largemouth bass mortality, with fishery implications. North American Journal of Fisheries Management 28: 418–427

Allen M.S., Brown P., Douglas J., Fulton W. and Catalano M. (2009). An assessment of recreational fishery harvest policies for Murray cod in southeast Australia. Fisheries Research 95:260-267

Allendorf F.W. and J.J. Hard (2009). Human-induced evolution caused by unnatural selection through harvest of wild animals. Proceedings of the National Academy of Sciences of the United States of America. 106: 9987-9994

Arredondo-Figueroa J.L., Nunez-Garcia L.G., Perez-Vega M.H., Ponce-Palafox J.T. and I.A. Barriga-Sosa. (2013) Culture of founders and breeding of the bay snook Petenia splendida (Pisces:Cichlidae) in a recirculation aquaculture system. European International Journal of Science and Tehcnology 2:123-132

Bachmann R.W., Jones B.L., Fox D.D., Hoyer M., Bull L.A. and D.E. Canfield (1996). Relations between trophic state indicators and fish in Florida (USA) lakes. Canadian Journal of Fisheries and Aquatic Sciences 53:842–855.

Barrientos C. and M.S. Allen (2008). Fish abundance and community composition in native and non-native plants following hydrilla colonization at Lake Izabal, Guatemala. Fisheries Management and Ecology 15:99-106.

Barrientos C. and Y. Quintana (2012) Evaluación del impacto de especies no nativas en los lagos Atitlán, Izabal y Peten Itzá; y caracterización del hábitat de especies nativas y no nativas de peces. Informe Final. Fondo Nacional de Ciencia y Tecnología, pp 1-75.

Barrientos C., Elias D. and Y. Quintana (in press) Fishes from Lake Yaxhá, Mayan Biosphere Reserve, Petén, Guatemala. Check List

Bayley P. B. and D. J. Austen. (2002). Capture efficiency of a boat electrofisher. Transactions of the American Fisheries Society 131:435–451.

Beamish R. J. and D. A. Fournier. (1981). A method for comparing the precision of a set of age determinations. Canadian Journal of Fisheries and Aquatic Sciences 38:982-983.

104

Beckman D. W. and Wilson C. A. (1995). Seasonal timing of opaque zone formation in fish otoliths. Pages 27-44 In: Secor, D. H., J. M. Dean, and S. E. Campana (editors). Recent Developments in Fish Otolith Research. University of South Carolina Press, Columbia.

Berkeley S.A., Hixon M.A., Larson R.J. and M.S. Love (2011). Fisheries sustainability via protection of age structure and spatial distribution of fish populations. Fisheries 29: 8, 23-32.

Bettoli P.W. and M.J Maceina. (1996). Sampling with toxicants. In: BR Murphy, DW Willis (eds) Fisheries Techniques, 2nd edn. Bethesda, MD: American Fisheries Society, pp 303–333.

Bettoli P.W. and J.E. Morris. (1991). Changes in the abundance of two atherinid species after aquatic vegetation removal. Transactions of the American Fisheries Society 120:90-97.

Bettoli P.W., Maceina M.J., Noble R.L. and R.K. Betsill (1993). Response of a reservoir fish community to aquatic vegetation removal. North American Journal of Fisheries Management 13:110-124.

Bickel T.O. and G.P. Closs (2008). Fish distribution and diet in relation to the invasive macrophyte Lagarosiphon major in the littoral zone of Lake Dunstan, New Zealand. Ecology of freshwater 17:10-19.

Biro P.A. and J.R. Post (2008) Rapid depletion of genotypes with fast growth and bold personality traits from harvested fish populations. Proceeds from National Academy of Sciences 105:2919-2922

Brezonik P. L. and J. L. Fox (1974). The limnology of selected Guatemalan lakes. Hydrobiologia 45: 467–487.

Bwanika G.N., Murie D.J. and L.J. Chapman (2007). Comparative age and growth of Nile tilapia (Oreochromis niloticus L.) in lakes Nabugado and Wamala, Uganda. Hydrobiologia. 589:287- 301.

Campana S.E. (2001) Accuracy, precision and quality control in age determination, including a review of the use and abuse of age validation methods. Journal of Fish Biology 59:197-242

Canfield D.E., Hoyer M.V. and C.M. Duarte (1990). An empirical method for characterizing standing crops of aquatic vegetation. Journal of Aquatic Plant Management 28:64–69.

Carlson S.M., Edeline E., Vollestad L.A. Haugen T.O., Winfield I.J., Fletcher J.M., James J.M. and N.C. Stenseth (2007). Four decades of opposing natural and human-induced artificial selection acting on Windermere pike (Esox Lucius). Ecology Letters 10:512-521

Carpenter S.R. and D.M. Lodge (1986). Effects of submersed macrophytes on ecosystem processes. Aquatic Botany 26:341--370.

Chang W. Y. B. (1982). A statistical method for evaluating the reproducibility of age determination. Canadian Journal of Fisheries and Aquatic Sciences 39:1208-1210.

105

Chapman D. G. and D. S. Robson. (1960). The analysis of a catch curve. Biometrics 16:354-368.

Chellappa S., Camara M.R., Chellappa N.T., Beveridge M.C.M. and F.A. Huntingford (2003). Reproductive ecology of a neotropical cichlid fish, Cichla monoculus (Osteichthyes: Cichlidae). Brazilian Journal of Biology 63:17-26.

Chick J.H. and C.C. Mclvor (1994). Patterns in the abundance and composition of fishes among beds of different macrophytes: viewing a littoral zone as a landscape. Canadian Journal of Fisheries and Aquatic Sciences 51:2873–2883.

Chick J.H. and C.C. Mclvor (1997). Habitat selection by three littoral zone fishes: effects of predation pressure, plant density and macrophyte type. Ecology of Freshwater Fish 6:27-35.

Cochran-Biederman J.L. and K.O. Winemiller (2010). Relations among habitat, ecomorphology and diets of cichlids in the Bladen River, Belize. Environmental Biology of Fishes 88:143-152.

Coggins Jr. L.C., Catalano M.J., Allen M.S., Pine III W.E. and C.J. Walters (2007). Effects of cryptic mortality and the hidden costs of length limits in fishery management. Fish and Fisheries 8: 196–210.

Concheiro Pérez G.A., Rícan O., Ortí G., Bermingham E., Doadrio I. and R. Zardoya (2007). Phylogeny and biogeography of 91 species of heroine cichlids (Teleostei:Cichlidae) based on sequences of the cytochrome b gene. Molecular Phylogenetic Evolution 43:91–110.

Conover D. O. and S. B. Munch (2002). Sustaining fisheries yields over evolutionary time scales. Science 297: 94-96.

Deevey E.S., Brenner M., Flannery M.S. and G.H. Yezdani (1980). Lakes Yaxhá and Sacnab, Petén, Guatemala: limnology and hydrology. Archives of Hydrobiology 57: 419–460.

DeVries D.R. and Frie R.V. (1996). Determination of age and growth. pp. 483–512. In: B.R. Murphy and D.W. Willis (ed.) Fisheries Techniques, 2nd edition, American Fisheries Society, Bethesda

Deriso R.B. (1987). Optimal F0.1 criteria and their relationship to maximum sustainable yield. Canadian Journal of Fisheries and Aquatic Sciences 44: 339-348.

Dibble E.D., Killgore K.J. and S.L. Harrel (1996). Assessment of fish-plant interactions. American Fisheries Society Symposium 16:357–372.

Dibble E.D. and F.M. Pelicice (2010). Influence of aquatic plant-specific habitat on an assemblage of small neotropical floodplain fishes. Ecology of Freshwater Fish 19:381–389.

Dionne M. and C.L. Folt (1991). An experimental analysis of macrophyte growth forms as fish foraging habitat. Canadian Journal of Fisheries and Aquatic Sciences 48:123-131.

106

Dunn A., Francis R.I.C.C. and I. J. Doonan. (2002). Comparison of the Chapman-Robson and regression estimators of Z from catch-curve data when non-sampling stochastic error is present. Fisheries Research 59:149-159.

Duponchelle F. and J.Panfili (1998). Variation in age and size at maturity of female Nile Tilapia, Oreochromis niloticus, populatios from man-made lakes of Cote d’Ivoire. Environmental Biology of Fishes 52:453-465.

Duponchelle F., Cecchi P., Corbin D., Nuñez J., and M.Legendre (2000). Variation in fecundity and egg size of female Nile tilapia Oreochromis niloticus, from man-made lakes of Cote d’Ivoire. Environmental Biology of Fishes 57:155-170

Edeline E., Carlson S.M., Stige L.C., Winfield I.J., Fletcher J.M., James J.B., Haugen T.O., Vollestad L.A. and N.C. Stenseth (2007). Trait changes in a harvested population are driven by a dynamic tug-of-war between natural and harvest selection. Proceeds from National Academy of Sciences 40:15799-15804

Efitre J. (2007) Life history variation in tilapia populations within the crater lakes of western Uganda: the role of size-selective predation. Ph.D. dissertation, University of Florida, Gainesville, USA.

Efitre J., Chapman L.J. and D.J. Murie (2009). Fish condition in introduced tilapias of Ugandan crater lakes in relation to deforestation and fishing pressure. Environmental Biology of Fishes 85:63-75.

Egger B., Meekan M., Salzburger W., Mwape L., Shapola R. and C. Sturmbauer (2004). Validation of the periodicity of increment formation in the otoliths of a cichlid fish from Lake Tanganyika, East Africa. Journal of Fish Biology 64: 1272-1284.

Eschmeyer WN (ed) (2012) Catalog of fishes electronic. California Academy of Sciences. Electronic version accessed (JAN 2012) at: http://research.calacademy.org/research/ichthyology/catalog/ fishcatmain.asp

Faunce C.H. and J.J. Lorenz (2000) Reproductive biology of the introduced Mayan cichlid, Cichlasoma urophthalmus, within an estuarine mangrove habitat of southern Flroida. Environmental Biology of Fishes 58:215-225

Ferreira B. P. and G. R. Russ (1994). Age validation and estimation of growth rate of the coral trout, Plectropomus leopardus, (Lacepede 1802) from Lizard Island, Northern Great Barrier Reef. Fishery Bulletin 92: 46-57.

Finlayson B.J., Schnick R.A., Cailteux R.L., DeMong L., Horton W.D., McClay W., Thompson C.W. and G.J. Tichacek (2000). Rotenone use in fisheries management: administrative and technical guidelines manual. Bethesda, MD: American Fisheries Society.

Ficke A.D., Myrick C.A. and L.A. Hansen. (2007). Potential impacts of global climate change on freshwater fisheries. Review in Fish Biology and Fisheries 17 581-613.

107

Fischer P. and R. Eckmann (1997) Seasonal changes in fish abundance, biomass and species richness in the littoral zone of a large European lake, Lake Constance, Germany. Archiv für Hydrobiologie 139:433-448.

Freedman J.A. and L.G. Noakes (2002). Why are there no really big bony fishes? A point-of- view on maximum body size in teleost and elasmobranchs. Reviews in Fish Biology and Fisheries 12:403-416.

Froese R., Thorson J.T. and R.B. Reyes Jr. (2013). A Bayesian approach for estimating length- Weight relationships in fishes. Journal of Applied Ichthyology. doi: 10.1111/jai.12299

Garcia P. (2009). Distribución de los insectos acuáticos en cuerpos lenticos de la Reserva de Biosfera Maya: indicadores biológicos de la calidad del agua. Dirección general de investigación. Universidad de San Carlos de Guatemala, pp1-66.

Garcia-Lizarraga M.A., Soto-Franco F.E., Jesus J.M., Velazco-Arce R., Velasquez-Abunader J.I. Ramirez-Perez J.S. and E.Pena-Messina (2011). Population structure and reproductive behavior of Sinaloa cichlid Cichlasoma beani (Jordan, 1889) in a tropical reservoir. Neotropical Ichthyology 9: 593-599.

Gelwick F.P. and W.J. Matthews (1990) Temporal and spatial patterns in littoral-zone fish assemblages of a reservoir (Lake Texoma, Oklahoma-Texas, U.S.A.). Environmental Biology of Fishes 27:107-120.

Glazier D.S. (2000) Trade-offs between reproductive and somatic (storage) investments in animals: a comparative test of the Van Noordwijk and De Jong Model. Evolutionary Ecology 13: 539–555.

Gómez-Marquez J. L., Peña-Mendoza B., Salgado-Ugarte I. H. and M. Guzmán-Arroyo (2003). Reproductive aspects of Oreochromis niloticus (Perciformes: Cichlidae) at Coatetelco lake, Morelos, Mexico. Revista de Biología Tropical 51: 221-228.

Gomiero L.M. and F.M.S. Braga (2004). Reproduction of species of the genus Cichla in a reservoir in southeastern Brazil. Brazilian Journal of Biology 63: 613-624

Goodyear C.P. (1993). Spawning stock biomass per recruit in fisheries management: foundation and current use. In: Risk Evaluation and Biological Reference Points for Fisheries Management. Canadian Special Publication of Fisheries and Aquatic Sciences 120 (eds S.J. Smith, J.J. Hunt and D. Rivard), NRC Press, Ottawa, Canada.

Gotceitas V. and P. Colgan (1987). Selection between densities of artificial vegetation by young bluegills avoiding predation. Transactions of the American Fisheries Society 116:40–49.

Granados-Dieseldorff P., Christensen M.F. and P.H. Kihn-Pineda. (2012). Fishes from Lachua Lake, Upper Usumacinta Basin, Guatemala. Check List 8:095-101

Greenfield D.W. and J.E. Thomerson (1997). Fishes of the continental waters of Belize. Gainesville, FL: University Press of Florida.

108

Haugen T. O. and L. A. Vollestad (2001). A century of life-history evolution in grayling. Genetica 112: 475-491.

Hewitt D.A. and J.M. Hoenig (2005). Comparison of two approaches for estimating natural mortality based on longevity. Fishery Bulletin 103:433-437

Hightower J.E., Jackson J.R. and K.H. Pollock (2001). Use of telemetrymethods to estimate natural and fishing mortality of striped bass in Lake Gaston, North Carolina. Transactions of American Fisheries Society 130: 557–567.

Hilborn R., Punt A. E. and J. M. Orensanz. (2004). Beyond band-aids in fisheries management: fixing world fisheries. Bulletin of Marine Sciences 74: 493–507.

Hilborn R., Stokes J K., Maguire J., Smith T., Botsford L. W., Mangel M., Orensanz J. M., Parma A., Rice J., Bell J., Cochrane K. L., Garcia S., Hall S. J., Kirkwood G. P., Sainsbury K., Stefansson G. and C. Walters. (2004). When can marine reserves areas improve fisheries management? Ocean Coastal Management 47: 197–205.

Hoenig J. M., Morgan M. J. and C. A. Brown. (1995). Analysing differences between two age determination methods by tests of symmetry. Canadian Journal of Fisheries and Aquatic Systems 52:364-368.

Hoyer M.V. and D.E. Canfield (1996) Largemouth bass abundance and aquatic vegetation in Florida Lakes: An empirical analysis. Journal of Aquatic Plant Management 34:23-32.

Hulsey C.D., García de León F.J., Johnson Y.S., Hendrickson D.A. and T.J. Near (2004). Temporal diversification of Mesoamerican cichlid fishes across a major biogeographic boundary. Molecular Phylogenetic. Evolution 31:754–764.

Ishikawa T., Shimose T. and K. Tachihara (2012). Life history of an invasive and unexploited population of Nile tilapia (Oreochromis niloticus) and geographical variation across its native and non-native ranges. Environmental Biology of Fishes. DOI 10.1007/s10641-012-0050-1

Ixquiac M. 2010. Crecimiento del pez blanco (Petenia splendida) en tres hábitats: cultivo, lago (Petén Itzá) y en Río San Pedro por medio de marcaje y recaptura”. Centro de Estudios del Mar y Acuicultura : CONCYT.

Jepsen D.B., Winemiller K.O. and D.C. Taphorn (1997). Temporal patterns of resource partitioning among Cichla species in a Venezuelan black water river. Journal of Fish Biology 51:1085-1108.

Jepsen D.B., Winemiller K.O., Taphorn D.C. and Rodriguez-Olarte D. (1999). Age structure and growth of peacock cichlids from rivers and reservoirs of Venezuela. Journal of Fish Biology 55:433-450.

Keast A. (1984). The introduced aquatic macrophyte, Myriophyllum spicatum, as habitat for fish and their invertebrate prey. Canadian Journal of Zoology 62:1289-1303.

109

Killgore K.J., Morgan R.P. II and N.B. Rybicki (1989). Distribution and abundance of fishes associated with submersed aquatic plants in the Potomac River. North American Journal of Fisheries Management 9:101–111.

Kinh H., Cano E. and A. Morales (2006). Peces de las aguas interiores de Guatemala. En: Biodiversidad de Guatemala. En: Cano E (ed) Biodiversidad en Guatemala. Guatemala: Universidad del Valle de Guatemala, pp. 456-485.

Krebs C.J. (1999) Ecological methodology. 2nd edition. Addison-Welsey Educational Publishers Inc. New York.

Law R. (2000) Fishing, selection, and phenotypic evolution. ICES Journal of Marine Science 57: 659-669.

Leon B. and J. Morales (2000). The aquatic macrophyte communities of Laguna del Tigre national park, Peten, Guatemala. In: Bestelmeyer B, Alonso L (eds) A biological Assessment of Laguna del Tigre national park, Peten, Guatemala. RAP Bulletin of Biological Assessment 16, Conservation International, Washington, DC, pp 35-40.

Lewis Jr. W. M. (1996) Tropical lakes: how latitude makes a difference. Perspectives in Tropical Limnology 43-64

López-López E., Sedeño-Díaz J.E., López-Romero F. and P. Trujillo- Jiménez (2009). Spatial and seasonal distribution patterns of fish assemblages in the Rio Champotón, southeastern Mexico. Review in Fish Biology and Fisheries 19:127-142.

Lowe-McConnell R.H. (1987) Ecological Studies in Tropical Fish Communities. Cambridge: Cambridge University Press.

Lubber L., Boyton W.R. and W.M. Kemp (1990). Variations in structure of estuarine fish communities in relation to abundance of submersed vascular plants. Marine Ecology Progress Series 65:1-14.

Mace P.M. (1994). Relationships between common biological reference points used as thresholds and targets of fisheries management strategies. Canadian Journal of Fisheries and Aquatic 51:110–122.

Magurran A. E. and C. Macias Garcia (2000). Sex differences in behavior as an indirect consequence of mating system–a review. Journal of Fish Biology 57: 839–875.

Marchetti M.P., Moyle P.B. and R. Levine (2004). Alien fishes in California watershed: characteristics of successful and failed invaders. Ecological applications 14:587-596

Martinez-Palacios C.A., Chavez-Soza J.C., Santoyo-Guzman V.O., Campos Mendoza A., Martinez-Chavez C.C. and L.G. Ross (2007).The effect of photoperiod on the reproduction of Chirostoma estor Jordan 1879 from Lago de Patzcuaro, Mexico. Journal of Applied Ichthyology 23:612-623

110

Matamoros, W.A., Schaefer, S. and B. Kreiser (2009). Annotated checklist of the freshwater fishes of continental and insular Honduras. Zootaxa 2307: 1–38.

McMahan C.D., Matamoros W.A., Alvarez-Calderon F.S., Henriquez W.Y., Recino H.M., Chakrabarty P., Barraza E. and N. Herrera (2013). Checklist of inland fishes of El Salvador. Zootaxa 3608:440-456.

Meyers R. and B.Worm (2003). Rapid worldwide depletion of predatory fish communities. Nature 423:280-283

Miller R. R. (1966). Geographical distribution of CentralAmerican fishes. Copeia 1966: 773-802.

Miller R.R. (1976). Geographical distribution of Central American freshwater fishes. In Thorson, T.B. (Ed.), Investigations of the Ichthyofauna of Nicaraguan Lakes. University of Nebraska, Lincoln, pp. 125–156. [16]

Miller R.R., Minckley W.L. and S.M. Norris (2005). Freshwater Fishes of México. University of Chicago Press, USA, 490pp.

Mims M.C., Olden .D., Shattuck Z.R. and N.L. Poff (2010). Life hisotry trait diversity of native freshwater fishes in North America. Ecology of Freshwater Fish 19:390-400

Murua H., Kraus G., Saborido-Rey F., Witthames P.R., Thorsen A. and S. Junquera (2003). Procedures to estimate fecundity of marine fish species in relation to their reproductive strategy. Journal of Northwest Atlantic Fishery Science 33:33-54.

Newman S. J., Cappo M. and D. M. B. Williams (2000). Age, growth, mortality rates and corresponding yield estimates using otoliths of the tropical red snappers, Lutjanus erythropterus, L. malabaricus and L. sebae, from the central Great Barrier Reef. Fisheries Research 48: 1-14.

Noiset J.L. and J.Cl. Micha (1996). Age and growth of Petenia splendida Gunther in the san Pedro River (Southern Mexico). In: Cows G (ed) Stock assessment in inland fisheries I. Fishing News Books, London, UK, pp 455–462.

Normando F.T., Arantes R.P., Luz R.K.,Thome R.G., Rizzo E., Sato Y. and N. Bazzoli (2009). Reproduction and fecundity of tucunare, Cichal kelberi (Perciformes: Cichlidae), an exotic species in Tres Marias Reservoir, Southeastern Brazil. Journal of Applied Ichthyology. 25:299- 305

Norton S.F. and E.L. Brainerd (1993). Convergence in the feeding mechanics of the ecomorphological similar species in the Centrarchidae and Cichlidae. Journal of experimental Biology 176:11-29

Nuñez J., and F. Duponchelle (2009). Towards a universal scale to assess sexual maturation and related life history traits in oviparous teleost fishes. Fish Physiology and Biochemistry 35:167– 180.

111

Oliva B.E. (2005). Contaminación en el Lago Peten Itzá. Consejo Nacional de Ciencia y Tecnología. pp. 1-64

Olive J.A., Miranda L.E. and W.D. Hubbard (2005). Centrarchid assemblages in Mississippi State-operated Fishing Lakes. North American Journal of Fisheries Management 25:7-15.

Olsen E.M., Heino M., Lilly G.R., Morgan M.J., Bratley J., Ernande B. and U. Dieckmann (2004). Maturation trends indicative of rapid evolution preceded the collapse of northern cod. Nature 428: 932-935

Olsen E.J. and R.A. Rulifson (1992). Maturation and fecundity of Roanoke River-Albermarle sound striped Bass. Transactions of the American Fisheries Society 121: 524-537

Orensanz J.M., Parma A.M., Jerez G., Barahona N., Montecinos M. and I. Elías (2005). What are the key elements for the sustainability of “S-fisheries”? Insights from South America. Bulletin of Marine Science 76:527-556.

Ozen O. and R. L. Noble (2002). Relationship between water level fluctuation and largemouth bass spawning in a Puerto Rico reservoir. Pp 213–220: in D. P. Philipp and M. S. Ridgway, editors. Black bass: ecology, conservation, and management. American Fisheries Society, Symposium 31, Bethesda, Maryland.

Pauly D. (1990). Length-converted catch curves and the seasonal growth of fishes. ICLARM Fishbyte 8: 33-38.

Perez L, Bugja R, Lorenschat J, Brenner M, Curtis J, Hoelzmann P, Islebe G, Scharf B. and A. Schwalb (2011) Aquatic ecosystems of the Yucatan Peninsula (Mexico), Belize and Guatemala. Hydrobiologia 661:407-433.

Perez-Vega M.H., Uribe-Aranzabal M.C., Garcia-Lorenzana M., Romero-Ramirez M.C. and J.L. Arredondo-Figueroa (2006). Description of the ovarian follicle growth of the neotropical cichlids Petenia splendida and Parachromis managuensis (Perciformes: Cichlidae). Journal of Applied Ichthyiology 22:515-520.

Petty J.T. and G. D. Grossman (1996). Patch selection by mottled sculpin (Pisces:Cottidae) in a southern Appalachian stream. Freshwater Biology 35:261-276.

Pine W.E., Pollock K.H., Hightower J.E., Kwak, T.J. and J.A Rice (2003). A review of tagging methods for estimating fish population size and components of mortality. Fisheries 20: 10–23

Poot-Lopez G.R., Arce-Ibarra A.M., Elias-Gutierrez M. and A.Cervantes-Martinez (2009). Difference in reproductive seasonality of the Central American cichlid Cichlasoma urophthalmus from three ‘cenotes’(sinkholes). Journal of Applied Ichthyology 25:89-90

Quintana Y. and C. Barrientos C (2012). La pesca en Río Dulce. Documento técnico de Proyecto Especies pesqueras de importancia comercial en el Parque Nacional Río Dulce: Valoración económica y estrategias para su manejo. Reporte Técnico en Fondo Nacional para la Conservación.

112

Reardon E. E and L. J. Chapman (2008). Reproductive seasonality in a swamp-locked African cichlid. Ecology of Freshwater Fish 17:20-29

Reyes F. (2008). Los cuerpos de agua de la región Maya Tikal-Yaxhá: importancia de la vegetación acuática asociada, su conservación y el valor desde el uso humano. Reporte Técnico en Fondo Nacional de Ciencia y Tecnología 25:1-87.

Reznik D. A. and H.Bryga (1987). Life history evolution in guppies: phenotypic and genotypic changes in an introduction experiment. Evolution 41:1370-1385.

Reznik D. A., Bryga H. and J. A. Endler (1990). Experimentally induced life-history evolution in a natural population. Nature 346: 357-359.

Ricker W. E. (1975). Computation and interpretation of biological statistics of fish parameters in the von Bertalanffy equation. Canadian Journal of Fisheries and Aquatic Sciences 47: 1416- 1426.

Rogers M.W., Allen M.S. and W.F. Porak (2006). Separating genetic and environmental influences on temporal spawning distributions of largemouth bass (Micropterus salmoides). Canadian Journal of Fisheries and Aquatic Sciences 63:2391-2399

Rogers M.W. and M.S. Allen (2008). Hurricane impacts to Lake Okeechobee: Altered hydrology creates difficult management tradeoffs. Fisheries 33:11-17.

Rosenmeier M.F., Brenner M., Kenney W.F., Whitmore T.J. and C.M. Taylor (2004). Recent eutrophication in the Southern Basin of Lake Petén Itzá, Guatemala: Human impact on a large tropical lake. Hydrobiologia 511:161-172.

Savino J.F. and R.A. Stein (1982). Predator-prey interaction between largemouth bass and bluegills as influenced by simulated submersed vegetation. Transactions of the American Fisheries Society 111:255–266.

Schmitter-Soto J.J. (2007). A systematic revision of the genus Archocentrus (Perciformes: Cichlidae), with the description of two new genera and six new species. Zootaxa 1603: 78 pp.

Schultz E.T. and P.C. Rounton (2001). Analysis of daily growth patterns in young-of-year male dwarf surfperch (Embiotocidae: Micrometrus minimus) suggests alternative tactics: breed or grow after birth. Copeia 2001: 14–24.

Schwartz N. (1990).Forest society: a social history of Petén, Guatemala. The University of Pennsylvania Press.

Soria-Barreto M. and R. Rodiles-Hernandez (2008). Spatial distribution of cichlids in Tzendales River Biosphere Reserve Montes Azules, Chiapas, Mexico. Environmental Biology of Fishes 83:459-469.

113

de Souza J.E., Fragoso-Moura E.N., Fenerich-Verani N., Rocha O and J.R. Verani (2008). Population structure and reproductive biology of Cichla kelberi (Perciformes, Cichlidae) in Lobo Reservoir, Brazil. Neotropical Ichthyology 6:201:210.

Stearns S. C. (1992). The evolution of life histories. Oxford University Press, Oxford, UK.

Strong D.R., Whipple A.V., Child A.L. and B. Dennis (1999). Model Selection for a subterranean trophic cascade: root-feeding caterpillars and entomopathogenic nematodes. Ecology 80: 2750-2761

Taper M. L. (2004). Model identification from many candidates. Pages 448–524 in M. L. Taper and S. R. Lele, editors. The nature of scientific evidence: statistical, philosophical and empirical considerations. The University of Chicago Press, Chicago, Illinois, USA

Toguyeni A., Fauconneau B., Fostier A., Abucay J., Mair G. and J. Baroiller (2002). Influence of sexual phenotype and genotype, and sex ratio on growth performances in tilapia,Oreochromis niloticus. Aquaculture 207: 249-261

Trippel E. A. (1995) Age at maturity as a stress indicator in fisheries. BioScience 45: 759-771.

Tyler C. R. and J. P. Sumpter (1996). Oocyte growth and development in teleosts. Reviews in Fish Biology and Fisheries 6:287–318.

Valdez-Moreno M.E., Pool-Canul J. and S. Contreras-Balderas (2005). A checklist of the freshwater ichthyofauna from El Petén and Alta Verapaz, Guatemala, with notes for its conservation and management. Zootaxa 1072:43-60.

Vila-Gispert A. and R. Moreno-Amich (2002). Life-history patterns of 25 species from European freshwater fish communities. Environmental Biology of Fishes 65:387-400

Vila-Gispert A., Moreno-Amich R. and E. Garcia-Berthou (2002). Gradients of life-history variation: an intercontinental comparison of fishes. Reviews in Fish Biology and Fisheries 12:417-427

Vono V. and F.A.R. Barbosa (2001). Habitats and littoral zone fish community structure of two natural lakes in southeast Brazil. Environmental Biology of Fishes 61:371-379.

Wakida-Kusunoki A.T., Ruiz-Carus R. and E. Amador-Del-Angel (2007). Amazon salifin catfish Pterygoplichthys pardalis (Castelnau 1855) (Loricariidae), another exotic species established in southeastern Mexico. The Southern Naturalist 52:141-144.

Wallace R.A. and K. Selman (1981). Cellular and dynamic aspects of oocyte size in teleost. American Zoology 21:325-343

Walsh M. R., Munch S. B., Chiba S. and D. O. Conover (2006). Maladaptive changes in multiple traits caused by fishing: impediments to population recovery. Ecology Letters 9:142-148.

114

Walters C.J. and S.J.D. Martell (2004). Fisheries Ecology and Management. Princeton University Press, Princeton, NJ.

Werner E.E. and D.J. Hall (1988). Ontogenetic habitat shifts in bluegill: the foraging rate- predation risk trade-off. Ecology 69:1352–1366.

West G. (1990) Methods of assessing ovarian development in fishes: a review. Australian Journal of Marine and Freshwater Research 41:199-222.

Willink P., Barrientos C., Kihn H.A. and B. Chernnoff (2000). An ichthyological survey of the Laguna del Tigre National Park, Peten, In: Bestelmeyer B, Alonso L (eds) A biological Assessment of Laguna del Tigre national park, Peten, Guatemala. RAP Bulletin of Biological Assessment 16, Conservation International, Washington, DC, pp 41-48.

Winemiller K.O. (1990). Spatial and temporal variation in tropical fish trophic networks. Ecological Monograph 60:331–367.

Winemiller K.O. (1992). Life-history strategies and the effective- ness of sexual selection. Oikos 63:318–327.

Winemiller K.O. and D.C. Taphorn (1992). a evolucion de las estrategias de la vida en los peces de los llanos occidentales de Venezuela. Biollania 6: 79-122.

Winemiller K.O. and K.A. Rose (1992). Patterns of life-history diversification in North American fishes: implications for population regulation. Canadian Journal of Fisheries and Aquatic Sciences 49: 2196–2219.

Winemiller K.O., Montoya J.V., Roelke D.L., Layman C.A. and J.B. Cotner (2006). Seasonally varying impact of detritivorous fishes on the benthic ecology of a tropical river. Journal of North American Benthological Society 25:250-262.

115

BIOGRAPHICAL SKETCH

Christian Barrientos is from Guatemala, Central America. He grew up in Guatemala City, where he obtained a biology degree from the University of San Carlos, Guatemala. He married another biologist and moved to Petén, home of the Maya Biosphere Reserve. He worked at the

Scarlet Macaw Biological Station located in Laguna del Tigre National Park, spending most of his time on the beautiful San Pedro River. During that time, he worked in conservation of terrestrial systems. But his heart was always loyal to aquatic systems and his research started to move toward investigation of water. He always enjoyed fishing trips on the San Pedro River.

During his time as coordinator of a Global Environment Facility/World Bank Project for Laguna

Del Tigre, he improved his English writing and communication abilities. Following the advice of several friends and his parents, he decided to pursue further education at the University of

Florida. During this new student phase, and as a father, he realized how valuable aquatic systems are for this and future generations, and he decided to continue work on the conservation and management of aquatic systems. He originally went to the University of Florida with the idea of working on giant cichlids, but as so often happens in life, he took another path and worked on a non-native fish species in Lake Izabal, Guatemala. That research led his Master of Science

(M.Sc.) degree. After completing his M.Sc., he was awarded a prestigious Knauss fellowship to work at the National Oceanic and Atmospheric Administration (NOAA), in the office of Science and Technology in the National Fishery Service. He learned to be a liaison between science and policy. After finishing his stint at NOAA, he decided to return to the University of Florida and was accepted in the doctoral degree program in Fisheries and Aquatic Sciences. He obtained a prestigious Alumni Award and a teaching assistantship from the Biology Department. At the same time, he was awarded three different projects by the Guatemalan government to evaluate

116

non-native fishes, life history and fishery dynamics of giant cichlid. In addition to his professional interests, he enjoys sports, family time and food.

117