Amphibian-Human Coexistence in Urban Areas

Urban areas are growing rapidly throughout the world, fragmenting and replacing preexisting

habitats. Not all habitats are being equally overrun by urbanization: human settlements are often

situated in flat, agriculturally productive, lowland areas where wetlands formerly existed (Mitsch

and Gosselink 2007). Organisms that depend on such habitats are therefore likely to be

disproportionately impacted by urbanization. However, most cities contain a variety of

freshwater ponds that are potential breeding habitats for some of the species that historically

inhabited the area. Potential habitats in urban areas include remnant ponds that have persisted in

some form throughout the development of cities, degraded ponds that have been restored, and

ponds that were newly created for aesthetic, environmental, or water management reasons, and

that may or may not resemble ponds in the original landscape.

The magnitude of harm that urbanization causes to resident organisms can vary greatly

depending on habitat availability and quality within cities. Although urban habitats are often

substantially different from their natural counterparts, they have the potential to provide habitat

for numerous species (Hobbs et al. 2009; Magle et al. 2012). It is increasingly apparent that

while ecological reservation and restoration strategies are necessary, they are not sufficient for

the conservation of Earth’s species under current human growth and land use trends

(Rosenzweig 2003; Hobbs et al. 2011). Reserves are often too small for long-term persistence

and can be isolated by human development; for these reasons a hospitable matrix within

humandominated areas may greatly benefit their effectiveness (Ricketts 2001; Kupfer et al.

2006). Additionally, the focus on reserves has historically led to cities being disregarded as

possible settings for conservation (Rosenzweig 2003). Documenting the importance of

humanconstructed habitats such as urban stormwater ponds to native biota can initiate efforts to

provide or improve habitat in areas that otherwise would be overlooked.

Amphibians that depend on freshwater ponds are of particular conservation concern.

Many amphibians are experiencing striking global declines (Collins and Storfer 2003; Stuart et

al. 2004; Beebee and Griffiths 2005), and urbanization has been identified as greatly contributing

to their downward trend (Cushman 2006; Hamer and McDonnell 2008). A recent review of urban

wildlife research identified urban amphibian studies as one of the most critical gaps in the field

(Magle et al. 2012). Pond-breeding amphibians are important links within and between aquatic

and terrestrial food webs, and are often a major food source for fish, birds, snakes, and aquatic

invertebrates (Stebbins and Cohen 1995 and references therein). Amphibians are particularly

well suited to providing city residents with a personal connection with nature because they are

often conspicuous and many produce calls that remind city residents that they live within a larger

system.

Amphibians have been found breeding in a variety of urban water bodies including ponds

and lakes in parks or private yards, stormwater facilities, and golf course ponds, thus using

habitats that are substantially different from their former pristine breeding habitats (e.g. Husté et

al. 2006; Colding et al. 2009; Hamer et al. 2012; Scheffers and Paszkowski 2013). Native

wildlife can often adapt to novel and altered habitats if the right conditions exist, and an

understanding of what factors influence amphibians in urban areas can lead to management that

is better informed to mitigate the impacts of urbanization and promote persistence.

Many habitat characteristics have been shown to affect urban pond-breeding amphibians

in other systems (Hamer and McDonnell 2008). The biological community of urban ponds

affects native amphibians by setting the conditions for habitat structure, food, competition, and

predation (Srdedl and Collins 1992; Stebbins and Cohen 1995; Lawler et al. 1999). Nutrient and

chemical pollution can have lethal and negative sublethal effects on pond-breeding amphibians

(Rouse et al. 1999; Hatch and Blaustein 2003; Otto et al. 2007). Physical attributes such as depth

and seasonality influence breeding success through effects on metabolism, activity, and survival

of amphibians (Stebbins and Cohen 1995; Babbitt 2005). Landscape-level factors such as

surrounding road density and forest cover have all also been shown to impact amphibian use of

urban ponds by influencing upland habitat use and movement among patches (Hamer and

McDonnell 2008; McCarthy and Lathrop 2011). Here I examine patterns of amphibian use of

wetlands in a large urban area in the Willamette Valley of Oregon where wetlands loss has been

rampant, but where constructed urban wetlands are recently becoming abundant. In this study I

aim to: (1) determine which regionally occurring amphibian species use remnant and constructed

wetlands in the city of Portland, Oregon, (2) investigate habitat characteristics associated with

amphibian species richness in this city, and (3) develop management recommendations to benefit

amphibians in urban areas in this region. I investigated amphibian use of 62 remnant and

constructed wetlands in relation to various biological, chemical, physical, and landscape features.

I predicted that the city would harbor a subset of the regionally occurring species, that

constructed wetlands would provide some habitat for native amphibians, but not to the extent that

remnant wetlands do, and that vegetation cover would be strongly positively associated with

species richness.

Methods

Study Sites

This study was conducted in Portland, Oregon which is situated at the confluence of the

Willamette and Columbia Rivers in the Willamette Valley ecoregion. The area was formerly

dominated by interspersed wetlands and forests (Habeck 1961; Taft and Haig 2003). An

estimated 87% of wetlands have been lost in this area since 1850 (Oregon Biodiversity Project

1998). The Portland area was inhabited by Upper Chinook people until the 1850s when European

settlement initiated rapid urban growth (Marschner 2008). Portland is the nation’s 29th largest

city with ~2.3 million people in the metropolitan area (United States Census Bureau 2010). It

experiences damp, mild winters and warm, dry summers. Portland has an urban growth boundary

which, by law, separates areas of dense development from rural areas, reducing urban sprawl. All

of the ponds in this study were within the urban growth boundary. Portland is an ideal city in

which to study amphibian urban habitat use because of the formerly extensive wetlands on which

it was built and the large amount and variety of ponds that have been retained, restored, or

created. Additionally, this region was identified in a recent review of urban amphibian studies in

North America as relatively understudied (Scheffers and Paszkowski 2012).

I surveyed 62 ponds in 18 sites within 4 watersheds (Fig 1.1). A site was defined as the

contiguous area surrounding a pond before reaching impervious surfaces in all directions. The

average distance between sites was 10.3km (range: 0.5-26.4km) and the average distance

between ponds within a site was 137m (range: 5-668m). Although all ponds are within the urban

growth boundary of the

Fig. 1.1 shows the location of the survey sites, each of which contained between one and seven

ponds.

city, most ponds in Portland are distributed around the edges of the city and are generally absent

from the urban core. These ponds represent a variety of sizes, uses, and management authorities.

Approximately half of the ponds were remnants from areas that had historically been wetlands

while the other half were constructed; the majority of the constructed ponds were excavated,

unlined, and built for stormwater purposes. Ponds ranged in size from 0.25m2 to 3.2 hectares.

Measurement of Amphibians

I surveyed ponds during peak tadpole/larvae season for three years: May-August of 2008, 2009,

and 2010. I assessed the presence of amphibian species using dip-net surveys. I walked the

perimeter of each pond at a depth of ~25cm, swept a handheld net every three steps for a length

of ~75cm, and counted and identified all tadpoles and larvae to the species level before returning

them to the water. Each pond was surveyed three times during a season. This type of dip-net

survey has been shown to be useful for determining presence/absence of amphibian species in

ponds and is appropriate for use in water bodies with spatial structure and vegetation, such as

many of those in this study (Shaffer et al. 1994; Kline 1998). Detection probabilities were

calculated using the program PRESENCE (Hines 2006).

Selection and Measurement of Habitat Characteristics

The habitat characteristics measured for each pond were chosen through a literature review of

factors influencing pond-breeding amphibians and collaborative meetings with managers and

ecologists from the various governing bodies of the sites: Metro Regional Government, private

land owners, and the City of Portland’s Bureau of Environmental Services, Water Bureau, and

Department of Parks and Recreation. Selected factors were perceived by managers to have

potential impacts on amphibians and were regarded as feasible to manage in the future, either for

the improvement of existing ponds or for the creation of new ponds. The following factors were

measured for each pond: depth, nitrate level, amount of aquatic refugia, percent cover of aquatic

vegetation, surrounding vegetation cover, water permanence, fish presence, introduced American

bullfrogs presence—Rana (Lithobates) catesbeiana, whether the pond was remnant natural or

constructed, surrounding road density, and surrounding forest cover (Table 1.1). The majority of

these factors are at the local scale because the managers indicated that information concerning

Habitat characteristic Mean Range

Water depth at deepest point during we season (average cm) 70 7- >150

Nitrate level (average mg/L with sample integrated through water column) 1.6 0-20

Aquatic refugia (average % cover of submerged material) 58 0-100

Aquatic vegetation (average % cover of all aquatic plants) 38 0-100

Surrounding vegetation (average % cover at 1m height in 10m buffer) 69 0-100

Water permanence (temporary/permanent) NA 27/35

Formation (constructed/natural) NA 39/23

Fish presence (absent/present) NA 46/16

Bullfrog presence (absent/present) NA 40/22

Road density (km of road length/km2 in 1km radius) 15 1-43

Forest cover (% in 1km radius) 22 8-47

Table 1.1 Definitions and descriptive statistics for the habitat characteristics measured in the 62 ponds

this scale would be most useful to them because their management decisions and actions are

generally focused at the site or pond scale.

Depth and nitrate level were measured three times per season at the deepest point of the

pond up to 1.5m. Nitrate level was assessed from a 30mL water sample that was integrated

throughout the water column and measured with test strips from Industrial Test Systems, Inc.

The amount of aquatic refugia in a pond was visually estimated by the same observer

three times per season to the nearest 20%. It was defined as the amount of submerged material,

including: total aquatic vegetation (assessed as a stand-alone predictor as well), branches, and

human-constructed objects. I quantified surrounding vegetative cover three times per season as

the amount of a 10m buffer surrounding the pond to the nearest 20% where vegetation obstructed

the bare ground from a vantage point of 1m.

I assessed fish as present in a pond if they were found during any amphibian dip-net

survey or observed at any time during data collection. I determined presence of bullfrogs by

conducting visual encounter surveys during the day by walking the entire edge of each pond

three times per season. Juveniles and adults generally sit on the edges and dive into the water

upon approach giving off a distinctive squeak. Any bullfrog tadpoles or eggs encountered during

native amphibian surveys also constituted presence.

I visited each pond during the driest time of each season to observe whether or not it

dried out. For ponds that are routinely drained, the managers were contacted to determine if a

pond was drained that year. Each pond was scored as either temporary or permanent for each

year. I contacted the managers to determine whether each pond was a natural remnant or if it was

constructed. I measured road density (km of road length per km2) and forest cover (percent) in a

1km radius of each site using satellite imagery analyses.

Data Analyses

To determine which factors best explained species richness I ran mixed-effects multiple

regression analyses with the habitat characteristics as fixed factors. I examined: 1) top models as

determined by corrected Akaike Information Criterion (AICc; Hurvich and Tsai 1995), and 2)

factor AICc weights, described below. Analyses included only a single value of each variable for

each pond in each year even though several variables were measured three times in a year. The

following data were averaged among the three points for a given season: depth, nitrate level,

amount of aquatic refugia, percent aquatic vegetation, and surrounding vegetative cover. A data

point consisted of the compiled data for each pond for each year surveyed. There were a total of

128 data points as not all 62 ponds were surveyed each year. Due to spatial and temporal

autocorrelation, year and site were included in models as random factors.

All factors were tested for collinearity (Appendix 1.A). In final models I did not include

pairs of factors that had collinearity absolute values of ≥0.5. There were three pairs of factors

with collinearity that exceeded this level: 1) the amount of aquatic vegetation and the amount of

refugia in the pond, 2) the depth of the pond and whether it contained fish, and 3) whether the

pond dried in the summer and whether it contained fish. In preliminary analyses the amount of

aquatic vegetation explained more variation in amphibian richness than the amount of aquatic

refugia, and depth and whether a pond dried in the summer explained more variation than the

presence of fish; therefore, aquatic refugia and fish presence were not included in the remaining

model analyses.

All possible multiple regression models were run with all combinations of predictor

variables for a total of 512 candidate models from which I identified the top models (those for

which ∆AIC≤2; Appendix 1.B). Models were run in R (R Core Development Team 2008) using

the glmer function of the lme4 package (Bates et al. 2012). In addition to identifying top AICc

models I calculated AICc weights for each factor. AICc weights are calculated by summing the

delta AICc values for all models that include a given factor and then scaling them to be between

0 and 1. A weight close to zero indicates that the factor was not included in many models that

had relatively high explanatory power while a weight close to 1 indicates the factor was included

in many high-ranking models (Burnham and Anderson 2002). Relative AICc weights of variables

were examined in addition to top models rather than hypothesis testing because the relative

importance of each variable was of more interest in this study than identifying exactly which

variables to include for a predictive model (Wagenmakers and Farrell 2004).

Results

Species presence

Five of the six native pond-breeding amphibian species that currently occur in the Willamette

Valley were found to be breeding in the city of Portland, Oregon during these surveys: Pacific

chorus frog (Pseudacris regilla), northern red-legged frog (Rana aurora), long-toed salamander

(Ambystoma macrodactylum), northwestern salamander (Ambystoma gracile), and rough-skinned

newt (Taricha granulosa). P. regilla and A. macrodactylum were the most common (present in

58% and 54% of ponds surveyed, respectively) while T. granulosa was relatively rare (present in

6% of ponds). R. aurora and A. gracile were intermediate, being found in 30% and 20% of

surveyed ponds, respectively. The western toad (Bufo boreas) was not found during this survey

and is currently rare or absent throughout the Willamette Valley (Pearl et al. 2005). Detection

probabilities of individual species were as follows: P. regilla—0.65, R. aurora—0.42, A.

macrodactylum—0.81, A. gracile—0.67, T. granulosa—0.50, and bullfrogs 0.85.

Habitat Characteristics

Models showed that surrounding forest cover, amount of aquatic vegetation, and low nitrate

levels were most associated with high species richness (Table 1.2 and Appendix 1.B). The R2

value for the top model was 0.67 for a simple regression of predicted vs. observed values.

The amount of surrounding forest cover in a 1km radius was most influential in models and

was positively correlated with species richness. The percent of aquatic vegetation in a pond was

also highly positively correlated with species richness. The plants species that were abundant in

ponds with amphibians present were (in alphabetical order): Alisma plantago, Carex obnupta,

Eleocharis spp., Juncas spp., Ludwigia spp., Phalaris arundinacea, Polygonum amphibium,

Scirpus spp., Typha latifolia, and Veronica spp. Nitrate level was highly influential in models

and was negatively associated with richness.

The amount of surrounding vegetative cover, water permanence, whether the pond was

natural or constructed, surrounding road density, and pond depth had low influence on models

(Table 1.2). The direction of relationships is shown in Table 1.2. Bullfrog presence had no

influence on models and was not associated with species richness. There was a trend for higher

species presence in constructed vs. natural ponds that was consistent across species (Figure 1.2).

Habitat characteristic AICc weight Direction of relationship; higher richness in ponds:

Forest cover 0.87 with more surrounding forest cover

Aquatic vegetation 0.77 with more aquatic vegetation

Nitrate level 0.63 with lower nitrate levels

Surrounding vegetation 0.21 with more surrounding vegetation

Water permanence 0.19 that dry in the summer

Formation 0.10 that were constructed rather than natural remnants

Road density 0.06 with lower surrounding road densities

Depth 0.06 that were shallower

Bullfrog presence 0.00 NA

Table 1.2 AICc weights of predictor variables for models of species richness. Higher AICc weight

indicates that a given factor is more influential in describing species richness. Variables with high

influence are bolded

Fig. 1.2 Proportion of constructed and natural ponds occupied by amphibian species Discussion

This study of pond-breeding amphibians in the large city of Portland, Oregon showed that all of

the regional species except Bufo boreas were present in remnant natural as well as constructed

ponds within the city boundary. Models of native amphibian species richness identified the

relative influence of a set of habitat characteristics which can be affected by management

actions. Urbanization greatly alters existing landscapes, but this process often leaves some

remnant habitats and creates novel habitats that have potential to support some biodiversity.

Wildlife abundance and diversity can be fairly high in urban areas even though biotic

communities are often relatively homogenized among cities (Chace and Walsh 2006; McKinney

2006). Even artificial features that are entirely novel for a species can have the potential to be

used as habitat (Rosenzweig 2003).

Species Presence

The presence of five of the six regionally occurring native pond-breeding amphibians inside the

urban growth boundary of the city indicates that most of these species have been able to tolerate

urbanization when appropriate habitat was provided. A seventh species, the Oregon spotted frog

(Rana pretiosa), was at one time recorded in the region, but is presumed to be extirpated from

the Willamette Valley (Nussbaum et al. 1983; Hayes 1994; Pearl et al. 2005). The cool and moist

climate, relative abundance of terrestrial vegetation, and numerous streams in Portland may

contribute to amphibian persistence. Similar results were found in Portland for other types of

amphibians: surveys in the city detected four of the five terrestrially breeding species and the

only stream-breeding species native to the region (K Holzer, unpublished data).

Urban tolerance was not limited to common species: Rana aurora is absent from ~70%

of historical sites in the Willamette Valley (Kiesecker unpublished data referenced in Kiesecker

et al. 2001) and is a Species of Concern in Oregon, yet it bred in a variety of ponds in the city. A

direct comparison of species prevalence in and outside of the city is needed to better understand

the effects of urbanization. Additionally, ponds are lacking from Portland’s urban core, indicating

that although many species are breeding within the city limits, they are generally confined to the

edges.

Detection of a species during surveys does not necessarily indicate that the species will

persist in the city in the long term. Urban ponds may be experiencing an extinction debt scenario

where some species that are currently present are awaiting inevitable local extinction (Tilman et

al. 1994, Gagne and Fahrig 2010). However, informed management could conceivably stave off

this

fate.

Habitat Characteristics

The examination of habitat characteristics associated with urban amphibian species richness

produced results that can guide management. Ponds with high native amphibian species richness

tended to have high surrounding forest cover, some aquatic vegetation, and low nitrate levels.

There are two main caveats for this study. First, the relationships between factors and species

richness are correlational and may not be causal, although known mechanisms do exist for the

influence of each factor on amphibians. Second, there are likely habitat characteristics that were

not measured in this study that are influential.

The amount of forest cover surrounding a pond was the most influential factor in

explaining species richness. This is consistent with other studies of urban amphibians where 14

out of 17 studies showed forest cover to be positively associated with species presence, richness,

or abundance (Hamer and McDonnell 2008). This is likely because many pond-breeding

amphibian species require terrestrial upland habitat in addition to aquatic breeding habitat and

use vegetation corridors for movement (Semlitsch and Bodie 2003, McCarthy and Lathrop

2011). Studies examining the movement of amphibians in urban areas are rare and would

improve understanding of the effect of landscape connectivity on promoting persistence in these

systems. While forest cover was highly associated with richness at the landscape level, road

density was not. This indicates that areas of dense urban development may be able to sustain

amphibian populations if upland habitat is present.

The amount of aquatic vegetation in ponds was highly positively associated with species

richness. Aquatic vegetation has been found to be favorable to amphibian populations for a

variety of reasons including refuge from predators, oviposition sites, shade, periphyton substrate

for grazing, and oxygen production (Sredl and Collins 1992; Stebbins and Cohen 1995; Tarr and

Babbitt 2002). These findings are consistent with most urban amphibian studies: twelve out of

sixteen studies found aquatic vegetation cover to be positively associated with native amphibian

presence, abundance, richness, and/or diversity (Pearl et al. 2005; reviewed in Hamer and

McDonnell 2008; Hamer and Organ 2008; Hamer and Parris 2011; Hamer et al. 2012). During

this study several ponds with sparse vegetation were observed to have amphibian egg masses on

over 90% of plant stems. In some ponds that entirely lacked vegetation I observed egg masses on

materials such as plastic fencing and barbed wire that had fallen into the water. This indicates

that, for at least some sparsely vegetated ponds, adult amphibians are present and attempting to

breed but may be limited by oviposition sites provided by vegetation. Therefore, planting and

maintaining aquatic vegetation in urban ponds likely benefits native amphibians in this area.

Further studies of specific amphibian-plant relationships with a focus on plant types and species

are needed.

The only water pollutant measured during this study was nitrate (NO3-) which showed a

high negative influence on models. It is a naturally occurring compound that is often elevated in

urban and agricultural areas; at low concentrations it can stimulate primary productivity, but has

been shown to have detrimental effects on amphibian larvae at high levels through direct toxicity

and eutrophication (Hatch and Blaustein 2003; Rouse et al. 1999). Nutrient load in urban ponds

has been shown to be negatively associated with amphibian species richness (Ensabella et al.

2003; Houlahan and Findlay 2003), although Scheffers and Paszkowski (2013) found total

nitrogen to be positively associated with boreal chorus frog (Pseudacris maculata) occurrence.

Elevated urban nitrate may originate from fertilizers, industrial effluents, waste-water treatment

discharge, animal waste, and deposition from motor vehicles and industrial exhaust (Rouse et al.

1999; Camargo et al. 2005). Nitrate can be managed through regulating sources and by

constructing series of runoff ponds that remove contaminants as the water moves towards

downstream ponds. It is possible that nitrate itself is responsible for the negative trend found

in this study, but it is also likely that nitrate levels are correlated with other pollutants which were

not measured. Ponds with high nitrate levels often receive large amounts of runoff from nearby

yards, gardens, farms, and roads, which likely contains additional pollutants. These correlated

factors may underlie the trend observed because some of these factors, such as water

conductivity and heavy metals, have been shown to be detrimental to amphibians in urban ponds

(Paul and Meyer 2001; Wilson and Dorcas 2003; Simon et al. 2009). Although it is difficult to

study synergistic effects of multiple pollutants, recent studies addressing this issue for aquatic

organisms indicate that pollutant synergisms are likely important (e.g. Sullivan and Spence 2003;

Boone 2008; Relyea 2009). Sites in Portland that were designed as a series of ponds for runoff

filtration showed a trend where amphibians were often absent in the first pond but were generally

present in subsequent ponds (pers. obs.). This indicates that runoff treated in these ponds may

contain concentrations of pollutants high enough to limit amphibian use, but that pollutant levels

may be sufficiently reduced through the filtration process.

The remaining factors measured had little influence on models of species richness even

though these factors are known to be important in some circumstances. Hydroperiod, amount of

surrounding vegetative cover, and depth were likely within tolerable ranges for these species in

this area. Hydroperiod has been shown to be a strong predictor of urban amphibians and may

have failed to emerge as influential for overall richness in this study due to differences in life

histories: most of the species in this study generally do better in temporary ponds while

Ambystoma gracile benefits from permanent ponds (Rubbo and Kiesecker 2005; Shaffer 2005).

The lack of influence on species richness of bullfrog presence and whether a pond was natural or

constructed was more surprising.

Bullfrogs have been introduced to the west coast from the southeastern United States

(Nussbaum et al. 1983), and have been shown to be detrimental to native amphibians in a

number of ways including competition, predation, and disease transmission (summarized by

Casper and Hendricks 2005). In theory bullfrogs can be removed, but in practice, a large amount

of effort has not lead to substantial reduction in this region (pers. obs.; Adams and Pearl 2007).

In this study, bullfrog presence had no influence on models of native species richness. This is

particularly surprising because bullfrogs have been documented to negatively impact several

native amphibians, especially R. aurora and their sister species the California red-legged frog, R.

draytonii (Kiesecker and Blaustein 1998; Lawler et al. 1999; Kiesecker et al. 2001). It is

plausible that bullfrogs are in fact negatively impacting amphibians in this area but are not

completely excluding them. It is also possible that while bullfrogs are detrimental to native

amphibians in other systems, they have less of an impact in this region. Other studies conducted

in the Cascadia bioregion that either included urban areas (Pearl et al. 2005; Richter and Azous

1995; Johnson et al. 2011) or did not include urban areas (Adams et al. 1998; Adams 1999;

Adams et al. 2011) failed to find evidence of bullfrogs negatively impacting native amphibians.

The lessened impact of bullfrogs in this region compared to others may be due to the cooler

climate: this region is at the northern edge of current bullfrog range and possibly near its

physiological limits (Casper and Hendricks 2005). This indicates that money and effort currently

spent on bullfrog eradication for the sake of native amphibians in this region may be better spent

in other ways. However, due to changing climate, monitoring the effects of bullfrogs should

continue because conditions may shift in their favor (Rahel and Olden 2008).

All amphibian species surveyed were found breeding in constructed ponds. Contrary to

my hypothesis of constructed ponds sustaining lower amphibian use, there was a trend for all

species to be found in a higher proportion of constructed ponds than natural ponds. Results of

prior studies were mixed: Scheffers and Paszkowski (2013) found a variety of amphibians using

constructed urban ponds less than natural remnant ponds while Brand and Snodgrass (2010)

found the opposite. Creating new ponds could be an important part of a conservation strategy for

improving amphibian populations in cities, provided the new ponds possess characteristics

favored by native amphibians. The age of constructed ponds was not examined explicitly during

this study due to lack of accurate information. Through time ponds may become unsuitable as

they fill in with sediment and organic matter, especially those designed to collect suspended

solids. Dredging can reduce this effect, but this possibility should be explored with caution

because it can also be detrimental to amphibians (Aresco and Gunzburger 2004).

Management Implications

These findings indicate that amphibians in urban areas use both natural and constructed ponds as

long as there is appropriate vegetation and low levels of pollutants. This study supports the

following recommendations for managing ponds in this region to improve their value as

amphibian habitat: maintain and improve forest cover and aquatic vegetation, reduce runoff that

potentially contains pollutants, construct new ponds (with appropriate characteristics), and

deprioritize bullfrog removal while continuing to monitor for changes in their effects on native

amphibians. The specific influences of factors cannot necessarily be extrapolated to other species

or areas. However, these findings were generally in agreement with the existing studies of habitat

variables in urban ponds (Hamer and McDonnell 2008), and therefore may represent general

trends beyond this study site.

Conclusion

Pond-breeding amphibians are often assumed to be particularly vulnerable in human-dominated

landscapes because of their requirements for sufficient aquatic and terrestrial habitats and the

movement corridors between them. This study shows that many pond-breeding amphibian

species are able to breed in altered and completely constructed ponds in a large city, given

suitable habitat conditions. The species in this study did not avoid novel, constructed habitats—

the results indicate similar or even higher presence and richness in constructed ponds than in

natural remnant ponds. Many species were surprisingly adaptable, and managers can consider

urban ponds as potential amphibian breeding habitats that are not to be discounted due to

structural differences from wetlands that existed there before urbanization. Studies examining

which aspects are most influential for focal species in remnant and novel habitats are

increasingly important conservation tools to allow for species coexistence as land continues to be

converted into human-dominated landscapes.

Acknowledgements

I would like to thank Sue Thomas, Meghan Young, Robert P. Bayers, and Ian Pearce for help in

planning, data collection, and analysis of this study. I would also like to thank Sharon P. Lawler,

Peter B. Moyle, H. Bradley Shaffer, and two anonymous reviewers for improving earlier versions

of this manuscript. This study was supported by funding from the City or Portland, OR, the

University of California, Davis, and the National Science Foundation Graduate Research

Fellowship Program (grant number 1148897).

References

Adams MJ (1999) Correlated factors in amphibian decline: Exotic species and habitat change in

western Washington. J Wildl Manage 63:1162-1171

Adams MJ, Bury RB, Swarts SA (1998) Amphibians of the Fort Lewis Military Reservation,

Washington: Sampling techniques and community patterns. Northwest Nat 79:12-18

Adams MJ, Pearl CA (2007) Problems and opportunities managing invasive Bullfrogs: is there

any hope? In: Gherardi J (ed) Biological invaders in inland water: Profiles, distribution,

and threats. Springer, New York. pp 679-693

Adams MJ, Pearl CA, Galvan S, McCreary B (2011) Non-native species impacts on pond

occupancy by an anuran. J Wildl Manage 75(1):30-35

Aresco MJ, Gunzburger MS (2004) Effects of large-scale sediment removal on herpetofauna in

Florida wetlands. J Herpetol 38:275-279.

Babbitt KJ (2005) The relative importance of wetland size and hydroperiod for amphibians in

southern New Hampshire, USA. Wetl Ecol Manag 13(3):269-279

Bates D. Maechler M, Bolker B (2012) Linear mixed-effects models using S4 classes.

<[email protected]>

Beebee TJC., Griffiths RA (2005) The amphibian decline crisis: A watershed for conservation

biology? Biol Conserv 125:271-285

Boone MD (2008) Examining the single and interactive effects of three insecticides on

amphibian metamorphosis. Environ Toxicol Chem 27(7):1561-1567

Brand AG, Snodgrass JW (2010) Value of artificial habitats for amphibian reproduction in altered

landscapes. Conserv Biol 24(1):295-301

Burnham KP, Anderson DR (2002) Model Selection and Multi-model Inference: A Practical

Information-theoretic Approach, 2nd edn. Springer, New York.

Camargo JA, Alonso A, Salamanca A (2005) Nitrate toxicity to aquatic animals: A review with

new data for freshwater invertebrates. Chemosphere 58(9):1255-1267

Casper GS, Hendricks R (2005) Rana catesbeiana. In: Lanoo M (ed) Amphibian declines: The

conservation status of United States species. University of California Press

Chace JF, Walsh JJ (2006) Urban effects on native avifauna: a review. Landscape Urban Plann

74:46-69

Colding J, Lundberg J, Lundberg S, Andersson E (2009) Golf courses and wetland fauna. Ecol

Appl 19(6):1481-1491

Collins JP, Storfer A (2003) Global amphibian declines: Sorting the hypotheses. Divers Distrib

9:89-98

Cushman SA (2006) Effects of habitat loss and fragmentation on amphibians: A review and

prospectus. Biol Conserv 128:231-240

Ensabella F, Loriga S, Formichetti P, Isotii R, Sorace A (2003) Breeding site selection of Bufo

viridis in the city of Rome (Italy). Amphibia-Reptilia 24:396-400

Gagne SA, Fahrig L (2010) Effects of time since urbanization on anuran community composition

in remnant urban ponds. Environ Conserv 37(2):128-135

Habeck JR (1961) The original vegetation of the Mid-Willamette Valley of Western Oregon.

Bull Torrey Bot Club 65(4):262-280

Hamer AJ, McDonnell MJ (2008) Amphibian ecology and conservation in the urbanizing world:

A review. Biol Conserv 141(10):2432-2449

Hamer AJ, Organ AK (2008) Aspects of the ecology and conservation of the Growling Grass

Frog Litoria raniformis in an urban-fringe environment, southern Victoria. Austral Zool

34(3):393-407

Hamer AJ, Parris KM (2011) Local and landscape determinants of amphibian communities in

urban ponds. Ecol Appl 21(2):378-390

Hamer AJ, Smith PJ, McDonnell MJ (2012) The importance of habitat design and aquatic

connectivity in amphibian use of urban stormwater retention ponds. Urban Ecosyst

15:451-471

Hatch AC, Blaustein AR (2003) Combined effects of UV-B radiation and nitrate fertilizer on

larval amphibians. Ecol Appl 13(4):1083-1093

Hayes MP (1994) Current status of the spotted frog (Rana pretiosa) in western Oregon. Oregon

Department of Fish and Wildlife, Portland, OR, USA. Technical Report #94-1-01.

Hines JE (2006) PRESENCE—Software to estimate patch occupancy and related parameters.

Version 6.4. USGS-PWRC. http://www.mbr-pwrc.usgs.gov/software/presence.html.

Hobbs RJ, Higgs E, Harris JA (2009) Novel ecosystems: Implications for conservation and

restoration. Trends Ecol Evol 24(11):599-605

Hobbs RJ, Hallett LM, Ehrlich PR, Mooney HA (2011) Intervention ecology: Applying

ecological science in the twenty-first century. BioScience 61(6):442-450

Houlahan JE, Findlay CS (2003) The effects of adjacent land use on wetland amphibian species

richness and community composition. Can J Fish Aquat Sci 60:1078-1094

Hurvich CM, Tsai C-L (1995) Model selection for extended quasi-likelihood models in small

samples. Biometrics 51(3):1077-1084

Husté A, Clobert J, Miaud C (2006) The movements and breeding site fidelity of the natterjack

toad (Bufo calamita) in an urban park near Paris (France) with management

recommendations. Amphibia-Reptilia 27:561-568

Johnson PT, McKensize, VJ, Peterson AC, Kerby JL, Brown J, Blaustein AR, Jackson T (2011)

Regional decline of an iconic amphibian associated with elevation, land-use change, and

invasive species. Conserv Biol 25(3):556-566

Kiesecker JM, Blaustein AR (1998) Effects of introduced bullfrogs and smallmouth bass on

microhabitat use, growth, and survival of native red-legged frogs (Rana aurora). Conserv

Biol 12(4):776-787

Kiesecker JM, Blaustein AR, Miller CL (2001) Potential mechanisms underlying the

displacement of native red-legged frogs by introduced bullfrogs. Ecology 82(7):1964-

1970

Kline J (1998) Monitoring amphibians in created and restored wetlands. In: Lanoo MJ (ed)

Status and Conservation of Midwestern Amphibians. University of Iowa Press, Iowa, pp

360-368

Kupfer JA, Malanson GP, Franklin SB (2006) Not seeing the ocean for the islands: The

mediating influence of matrix-based processes on forest fragmentation effects. Glob Ecol

Biogeogr 15(1):8-20

Lawler SP, Dritz D, Strange T, Holyoak M (1999) Effects of introduced mosquitofish and

bullfrogs on the threatened California red-legged frog. Conserv Biol 13(3):613-622

Magle SB, Hunt VM, Vernon M, Crooks KR (2012) Urban wildlife research: Past, present, and

future. Biol Conserv 155:23-32

Marschner J (2008) Oregon 1859: A snapshot in time. Timber Press, Inc., Oregon

McCarthy K, Lathrop KG (2011) Stormwater basins of New Jersey coastal plain: Subsidies or

sinks for frogs and toads? Urban Ecosyst 14:395-413

McKinney ML (2006) Urbanization as a major cause of biotic homogenization. Biol Conserv

127:247-260

Mitsch WJ, Gosselink JG (2007) Wetlands. Van Nostrand Reinhold, New York

Nussbaum RA, Brodie ED Jr, Storm RM (1983) Amphibians and reptiles of the Pacific

Northwest. University Press of Idaho, Idaho

Oregon Biodiversity Project (1998) Oregon’s Living Landscape: Strategies and opportunities to

conserve biodiversity. Defenders of Wildlife

Otto CR, Forester DC, Snodgrass JW (2007) Influences of wetland and landscape characteristics

on the distribution of carpenter frogs. Wetlands 27(2):261-269

Paul MJ, Meyer JL (2001) Streams in the urban landscape. Annu Rev Ecol 32:333-365

Pearl CA, Adams MJ, Leuthold N, Bury B (2005) Amphibian occurrence and aquatic invaders in

a changing landscape: Implications for wetland mitigation in the Willamette Valley,

Oregon, USA. Wetlands 25(1):76-88

R Core Development Team (2008) R, Version 2.8.1. The R Foundation, Vienna, Austria

Rahel FJ, Olden JD (2008) Assessing the effects of climate change on aquatic invasive species.

Conserv Biol 22(3):521-533

Relyea RA (2009) A cocktail of contaminants: How mixtures of pesticides at low concentrations

affect aquatic communities. Oecologia 159(2):363-376

Richter KO, Azous AL (1995) Amphibian occurrence and wetland characteristics in the Puget

Sound Basin. Wetlands 15(3):305-312

Ricketts TH (2001) The matrix matters: Effective isolation in fragmented landscapes. Am Nat

158(1):87-99

Rosenzweig ML (2003) Win-win ecology—How Earth’s species can survive in the midst of

human enterprise. Oxford University Press, NewYork

Rouse JD, Bishop CA, Struger J (1999) Nitrogen pollution: An assessment of its threat to

amphibian survival. Environ Health Perspect 107(10):799-803

Rubbo MJ, Kiesecker JM (2005) Amphibian breeding distribution in an urbanized landscape.

Conserv Biol 19:504-511.

Scheffers BR, Paszkowski CA (2012) The effects of urbanization on North American amphibian

species: Identifying new directions for urban conservation. Urban Ecosyst 15:133-147

Scheffers BR, Paszkowski CA (2013) Amphibian use of urban stormwater wetlands: The role of

natural habitat features. Landscape Urban Plann 113:139-149

Semlitsch RD, Bodie JR (2003) Biological criteria for buffer zones around wetlands and riparian

habitats for amphibians and reptiles. Conserv Biol 17(5):1219-1228

Shaffer HB, Alford RA, Woodward BD, Richards SJ, Altig RG, Gascon C (1994) Quantitative

sampling of amphibian larvae, In: Heyer WR, Donnelly MA, McDiarmid RW, Hayek

LAC, Foster MS (eds.) Measuring and Monitoring Biological Diversity: Standard

Methods for Amphibians, Smithsonian Institution Press, Washington and London

Shaffer HB (2005) Ambystoma gracile. In: Lanoo M (ed) Amphibian declines: The conservation

status of United States species. University of California Press

Simon JA, Snodgrass JW, Casey RE, Sparling DW (2009) Spatial correlates of amphibian use of

constructed wetlands in an urban landscape. Landscape Ecol 24:361-373.

Sredl MJ, Collins JP (1992) The interaction of predation, competition, and habitat complexity in

structuring an amphibian community. Copeia 3:607-614

Stebbins RC, Cohen NW (1995) A natural history of amphibians. Princeton University Press,

New Jersey

Stuart SN, Chanson JS, Cox NA, Young BE, Rodrigues ASL, Fischman DL, Waller RW (2004)

Status and trends of amphibian declines and extinctions. Science 306:1783-1786

Sullivan KB, Spence KM (2003) Effects of sublethal concentrations of atrazine and nitrate on

metamorphosis of the African clawed frog. Environ Toxicol 22(3):627-635

Taft OW, Haig SM (2003) Historical wetlands in Oregon’s Willamette Valley: Implications for

restoration of winter waterbird habitat. Wetlands 23(1):51-63

Tarr TL, Babbitt, KJ (2002) Effects of habitat complexity and predator identity on predation of

Rana clamitans larvae. Amphibia-Reptilia 23:13-20

Tilman D, May RM, Lehman CL, Nowak MA (1994) Habitat destruction and the extinction debt.

Nature 371(6492):65-66

United States Census Bureau (2010) 2010 Census Redistricting Data, Public Law 94-171

Summary File

Wagenmakers EJ, Farrell S (2004) AIC model selection using Akaike weights. Psychon Bull

Rev 11(1):192-196

Willson JD, Dorcas ME (2003) Effects of habitat disturbance on stream salamanders:

implications for buffer zones and watershed management. Conserv Biol 17:763-771

YEAR SITE DPTH NITR REF AVEG SVEG DRY NATR FISH BLFG ROD FORT

YEAR - 0 -0.4 0 -0.3 0 0 0.2 0 0 0.1 0 0.2

SITE - 0.3 0 0 0 0.1 -0.4 -0.2 0.4 0 0 -0.2

DPTH - 0 -0.1 -0.2 0 -0.4 0 0.5 0.1 -0.1 0

NITR - 0 -0.2 -0.2 -0.4 0 0 -0.2 0.4 -0.1

REF - 0.8 0 0.3 0 0 -0.2 -0.1 -0.4

AVEG - 0.3 0.3 0 0 -0.1 -0.2 0.2

SVEG - 0 0 0. 0.1 -0.3 0.2

DRY - 0 -0.5 0 -0.3 0.2

NATR - 0 0 -0.1 0

FISH - 0.2 -0.1 0

BLFG - -0.4 0

ROD - -0.1

FORT -

Appendix 1.A Collinearity values for predictor variables. Variable pairs with collinearity of absolute values ≥0.5

are bolded and were assessed for the variable with more predictive power; the other variable in the pair was

excluded from analyses. The predictor variable abbreviations are as follows: YEAR-year of the survey, SITE-site of

the survey, DPTH-depth at the deepest point, NITR-nitrate level, REF-aquatic refugia, AVEG-aquatic vegetation,

SVEG-surrounding vegetative cover, DRY-if the pond dried in the summer, NATR-if the pond was a natural

remnant, FISH-presence of fish, BLFG-presence of introduced bullfrogs, ROD-road density in 1km radius,

FORTforest cover in 1km radius

FORT AVEG NITR SVEG DRY NATR ROD DPTH BLFG Delta AICc

1 1 -1 0 0 0 0 0 0 0

1 1 -1 1 0 0 0 0 0 0.86

1 1 -1 0 1 0 0 0 0 1.39

1 1 -1 0 0 -1 0 0 0 1.49

1 1 -1 1 0 -1 0 0 0 1.66

1 1 -1 0 0 0 1 0 0 1.80

1 1 -1 0 0 0 0 -1 0 1.83

1 1 -1 1 1 0 0 0 0 1.86

Appendix 1.B Top Akaike Information Criterion models for species richness, corrected for small sample sizes

(AICc). All models are shown where the difference in AICc was <2 from the top AICc model. A “0” indicates that a

factor was not included in a given model, “1” indicates that it was included and positively associated with richness,

and “-1” indicates that it was included and negatively associated. The predictor variable abbreviations are as

follows: FORT-forest cover in 1km radius, AVEG-aquatic vegetation, NITR-nitrate level, SVEG-surrounding

vegetative cover, DRY-if the pond dried in the summer, NATR-if the pond was a natural remnant, ROD-road density

in 1km radius, DPTH-depth at the deepest point, BLFG-presence of introduced bullfrogs

Chapter Two

In Chapter One I found that many native amphibians were breeding in Portland, and that they did

not avoid constructed wetlands. Associated habitat features corroborate what other studies have

shown by demonstrating that water both upland and aquatic vegetation were influential for native

amphibians, and that water quality likely also plays an important role. These results are fairly

typical for studies of amphibian use of urban landscapes, most of which have been conducted in

North America, Europe, or Australia. It is less clear if other areas of the world with relative high

biodiversity and high human population growth experience similar trends. To gain a perspective

on amphibian use of rapidly urbanizing areas with high biodiversity, this chapter is comprised of

surveys of amphibian breeding and associated habitat characteristics in and around the three

largest urban centers of Vietnam.

Habitat Value of Cities and Rice Paddies for Amphibians in Rapidly Urbanizing Vietnam

Katie A Holzerab*; Robert P Bayersb; Nguyen Thien Taoc

Department of Entomology and Nematology, University of California, Davis, One

Shields Ave, Davis, CA 95616

Vietnam National University of Ho Chi Minh City, 227 Nguyen Van Cu St., Dist. 5, Ho Chi

Minh City, Vietnam

Vietnam National Museum of Nature, Vietnam Academy of Science and Technology, 18 Hoang

Quoc Viet Road, Hanoi, Vietnam

*corresponding author: holzer[email protected]; 1-507-581-1327

Abstract

Urban and agricultural areas are growing rapidly throughout the world and may represent the

majority of potential habitat for many lowland species. Understanding species’ use of these

landscapes is integral to conservation management efforts worldwide, and especially in

understudied areas rich in biodiversity such as Southeast Asia. We quantified amphibian use of

rice cultivation fields, sprawling urban fringes, and dense urban cores in this region. We

surveyed amphibian breeding in 180 water bodies in these three land use types in and around the

three largest cities in Vietnam: Hanoi, Danang, and Ho Chi Minh City. We also assessed habitat

characteristics to examine associations with species richness. Surprisingly, we found that all

twelve species detected during our surveys were breeding in all three land use types. However,

we did observe decreased abundance and occupancy with increasing urbanization. In cities,

species richness was positively associated with the amount of surrounding vegetated upland

habitat and the presence of shallows, natural banks, and aquatic vegetation, even when

vegetation was simply potted ornamental plants. Our findings suggest important modifications to

general urban ecology theories in an understudied system. Theory predicts that urban areas have

reduced richness, increased abundance of common taxa, and homogenized wildlife communities,

whereas we found sustained richness, decreased prevalence of all taxa, and lack of biotic

homogenization. Future planning and management can incorporate habitat features that

associated with native species richness in areas of intense human use.

Keywords: human-dominated landscape; reconciliation ecology; anuran; novel habitat;

countryside biogeography

Highlights

-Amphibian species richness was remarkably comparable in urban and rice water bodies

-Anuran prevalence was lower in urban cores than suburban or rural areas

-Urban planning incorporating aquatic plants and upland access may favor amphibians

-Communities were region-specific, challenging the paradigm of homogenized urban taxa

1 Introduction

The majority of the world’s pristine ecosystems are declining in spatial extent while

humandominated landscapes, such as agriculture and urban areas, are growing rapidly. The

continued persistence of many species depends on the habitability of these altered landscapes

(Rosenzweig, 2003). The impacts of urbanization on wildlife have recently been studied

extensively but within a narrow range of study systems, generally focusing on birds and

mammals in urban vs.

“natural” areas in North America, Europe, and Australia (Aitkenhead-Peterson and Volder, 2010;

Magle et al., 2012). Additional taxa need to be studied in urban areas to gain a comprehensive

understanding of how biotic communities respond to urban environments. Especially few urban

biodiversity studies are located in South America, Africa, or Asia which contain rapidly

expanding cities along with relatively high levels of biodiversity (Aitkenhead-Peterson and

Volder, 2010; Magle et al., 2012). In order to direct management for the conservation of global

biodiversity it is important to research biodiversity in human-dominated landscapes for

lessstudies regions and taxa. Here, we present the first study conducted in Southeast Asia on the

effects of urbanization on amphibians, looking both at long-standing urban centers and the rapid

expansion of suburbs into surrounding agricultural areas.

Urbanization often influences areas that are already dominated by human activities

(McCleery, 2010; Shochat et al., 2010). Agricultural is particularly common surrounding cities

because humans tend to build settlements in flat, productive areas with access to fresh water, and

which are also attractive for growing food (Huston, 1993; Mitsch and Gosselink, 2007). Despite

the frequency with which agriculture surrounds cities, few studies address the further impact of

urban growth into agricultural areas. Rice cultivation fields are now the most abundant wetland

type in many tropical and subtropical countries (Aselmann and Crutzen, 1989), yet, apart from

the notable exceptions of Brazilian and Japanese rice paddies (e.g. Machado and Maltchik, 2010;

Naito et al., 2013), their use by amphibians is poorly understood, as is the effect of urban

encroachment. Amphibians face many threats globally (Stuart et al., 2004), and urbanization has

been identified as a leading cause of decline (Cushman, 2006). Documenting amphibian use of

these cities and rice paddies, and understanding associated habitat characteristics, can assist

planners and managers seeking to promote coexistence of these species in areas of intense human

use.

Existing research on urban wildlife has produced three general patterns. First, only a

subset of species from surrounding areas are found living in cities; those that do not are termed

urban avoiders and likely cannot tolerate some physical or biological aspect of urban living

(McCleery, 2010; McKinney, 2002; Shochat et al., 2010). Second, those species that do persist in

cities are often found at high densities and are termed urban exploiters as they are often able to

benefit from reduced predation and/or increased food from anthropogenic sources (McCleery,

2010; McKinney, 2002; Shochat et al., 2010). Lastly, wildlife communities are often

homogenized among cities because urban habitats are often more similar to each other than to the

surrounding areas (McKinney, 2006).

We anticipated some deviations from these patterns in amphibian communities of urban

and nearby agricultural areas of Vietnam. First, we expected to find some urban avoiders as well

as some species that avoided agricultural areas. The vast, connected rice cultivation fields likely

benefit species that can utilize the specific vegetation structure and hydrology, but they are

relatively homogeneous. Other species may benefit from the greater variety of water bodies in

cities if they can cope with the fragmentation created by human structures and roads. Second, we

did not expect to find any urban exploiters because most major amphibian predators—including

fish, aquatic invertebrates, and birds—were expected to be present in urban water bodies, and

because amphibians generally do not consume anthropogenic food. Lastly, we did not expect

homogenization of amphibian communities among cities because the metropolitan areas

possessed a substantially different climates and amphibians generally have smaller ranges and do

not migrate or disperse as far as many birds and mammals (Stebbins and Cohen, 1995);

therefore, a given species is less likely to occur in multiple distant cities.

If observed, these deviations from general urban wildlife patterns would highlight the

need to incorporate more diverse study systems and identify directions for developing

comprehensive theories of wildlife use of growing urban habitats. Recent results from

understudied systems demonstrate some of these possible deviations. For example, parasitoid

wasp communities in individual Californian cities were more reflective of nearby reference

communities than of communities in other cities in the region (Herrmann et al., 2012), riparian-

associated reptile communities in South Carolina maintained species richness in cities (Hunt et

al., 2013), and urban amphibian communities in Oregon contained nearly all regionally occurring

species

(Holzer, 2014).

Where wildlife is able to persist in cities, understanding patterns of use in relation to

habitat characteristics can allow for informed management decisions to maintain and improve

coexistence with humans (Rosenzweig, 2003). Amphibians are affected by a variety of local and

landscape-level factors in urban areas (Hamer and McDonnell, 2008). Physical attributes such as

water body size, slope of edge, and presence of shallows can affect ability of amphibians to gain

resources, maintain beneficial metabolic rates, and move in and out of water bodies (Babbitt,

2005; Stebbins and Cohen, 1995). Biological aspects of water bodies, such as water-pervious

banks and aquatic vegetation, benefit amphibians by providing oviposition sites and refuge from

predators (Hamer and McDonnell, 2008; Stebbins and Cohen, 1995; Tarr and Babbitt, 2002).

Landscape-level factors, such as amount of accessible and vegetated upland habitat, can

influence persistence by providing non-breeding habitat and connection to other breeding

populations (Hamer and McDonnell, 2008; McCarthy and Lathrop, 2011). We predicted that

amphibian species richness would be strongly affected by aquatic vegetation and upland habitat

availability because these features were expected to be limiting in dense urban systems. The

goal of this study was to measure the effects of urban encroachment into an agricultural

landscape on native amphibians. We aimed to determine how amphibian species richness,

diversity, occupancy, and abundance differ between urban areas and surrounding rice fields, and

to investigate which habitat features are most associated with amphibian use of water bodies in

urban areas. We documented the occurrence and abundance of amphibian species in urban water

bodies and surrounding rice fields for the three largest cities in Vietnam and assessed association

of local and landscape habitat characteristics with amphibian use in these landscapes.

2 Materials and Methods

2.1 Study Sites

Vietnam is a narrow country spanning 16 degrees of latitude with marked climatic

differences from north to south, and with major cities occurring in the lowlands. To capture the

breadth of diversity of Vietnamese lowland climates we surveyed the three largest cities in

Vietnam: Hanoi in the north, Danang on the central coast, and Ho Chi Minh City (formerly

Saigon) in the south. Each of these cities is ~600 km from the next. Hanoi exhibits a humid

subtropical climate with hot, humid summers and relatively cool, dry winters; Danang has a

tropical monsoon climate with warm, dry summers and warm, wet winters; Ho Chi Minh City

exhibits a tropical savanna climate with hot, rainy summers and warm, dry winters (Nguyen et

al., 2000). All three of these cities contain dense urban centers and are set in a surrounding

matrix of rice cultivation fields. The edges of all three cities are experiencing rapid development

and expansion, converting rural areas into neighborhoods and high-rises.

We selected study sites within each metropolitan area which represented the dominant

types of lowland land use. We characterized land use into three broad categories: dense urban

cores (hereafter: urban), developing boundary areas with mixed land cover types at the edges of

the cities (hereafter: suburban), and active rice cultivation fields (here after: rice paddies). We

used the proportion of water-impervious surface in a 500 m radius to distinguish between urban

and suburban land use types: those with >80% were considered urban (mean=95%) while those

with <70% were considered suburban (mean=48%). We selected 20 representative water bodies

for each of the three land use types in each of the three regions for a total of 180 water bodies.

We first identified potential water bodies with satellite imagery, then visited the water bodies

during the day to characterize the habitat features as described in section 2.3, and finally

randomly selected 20 in each land use type in each region that maintained the relative

distribution of habitat characters measured. All urban and suburban water bodies were separated

from each other by at least one road, and all rice paddies were >1 km apart. All rice paddies

surveyed were within 50 km of the urban core.

2.2 Amphibian surveys

We used call surveys to determine presence of amphibian species at each site. This

reflects the majority of amphibians in these areas because all frog species known to the region

produce identifiable calls, there are no known salamanders in these areas, and caecilians are

especially rare and difficult to detect. We used call surveys instead of egg, tadpole, metamorph,

or adult surveys because a many urban and suburban water bodies were contained within

restricted areas such as airports, construction areas, or military sites. Although we were allowed

to observe from the edge of the water, we were not given permission to enter the water in such

sites. Anuran call surveys have been shown to be a reliable estimate of breeding effort at a site

when there are limits to studying other life stages (Dorcas et al., 2009; Weir and Mossman,

2005), and were at least as informative as line transects and traps in a study of a similar

amphibian community in Taiwan (Hsu et al., 2005). Through direct observation, we used Nguyen

et al. 2009 to identify species.

We conducted call surveys on rainy nights in 2011-2012 during the breeding season for

each region: September-November 2011 in Danang, March-May 2012 in Hanoi, and May-June

2012 in Ho Chi Minh City. We surveyed each water body three times during this period. A

survey consisted of standing at the edge of the water body for five minutes and noting the

presence and relative abundance of all frog species calling. Preliminary surveys indicated that

five minutes was long enough to capture >80% of the species calling that evening, and surveys

of this length have been shown to be sufficient for many species and are considered standard

protocol for monitoring amphibians in North America (Gooch et al., 2006; Weir and Mossman,

2005). We noted abundance of each species as low (1-2 individuals), medium (3-8 individuals),

or high (>8 individuals), and conducted all surveys between half an hour after sunset and 1:00am

(Weir and Mossman, 2005). To reduce observer bias, K.A.H. and R.P.B. were present for all

surveys and reached consensus for the presence and abundance of species at each site during

each survey (Lotz and Allen, 2007).

2.3 Habitat Characteristics

We characterized five local habitat features and one landscape feature for all suburban

and urban water bodies: size, presence of aquatic vegetation, type of bank, slope of bank,

presence of shallows, and amount of surrounding vegetated upland habitat. The size of water

bodies changed throughout the season, and we therefore classified water bodies into log-scaled

size categories: <100m2, 100-1000m2, 1000-10 000m2, 10 000-100 000m2, or >100 000m2. We

assessed rooted aquatic vegetation as being present or absent. We classified the bank type as

either soil or waterimpervious (concrete, stone, or brick), and described the slope of the bank as

being either vertical or sloped. We defined the presence of shallows as at least 10m2 that was <20

cm deep. We used satellite images to quantify surrounding upland habitat by measuring the

amount of waterpervious surface before encountering a water-impervious surface in all

directions. This measure was log-transformed to meet normality assumptions. We did not include

rice paddies in this analysis because paddy habitat was relatively homogenous with respect to

these features and thus did not provide enough variation for informative results.

2.4 Statistical Analyses

We used permutational multivariate analysis of variance with simple linear regressions to

determine if amphibian occupancy and alpha, beta, and gamma diversity were influenced by

region or land-use type. We used Tukey’s Honestly Significant Difference test for multiple

comparisons to determine which land-uses and regions differed for these measures. We ran

nonmetric multi-dimensional scaling analyses on amphibian community composition (presence

and occurrence) to examine patterns of community similarity within and among regions and land

use types. We conducted generalized linear mixed model analyses to determine the relative

influence of habitat characteristics on amphibian richness and used a stepwise AIC approach for

variable selection. We avoided pseudoreplication by treating sites within one land use type for

one metropolitan area as subsamples. For land use analyses we considered nine data points (three

land use types for the three metropolitan areas), each with 20 subsamples; for habitat

characteristic analyses we included metropolitan area as a random factor during modeling. All

analyses were conducted in R Version 3.0.1 with the lme4, MASS, and Vegan packages (R Core

Development Team 2008).

3 Results

3.1 Land use

During our surveys we identified twelve species of amphibians, all of which were found

in each of the three land use types: urban, suburban, and rice paddies (Table 2.1). However, any

given metropolitan area contained only seven to nine species (Table 2.1). All of these species are

native to the region and belong to eight genera and five families. The Asian painted frog

(Kaloula pulchra) was only detected on a handful of very rainy nights and was therefore not

included in the remaining analyses. Although we considered the common treefrog (Polypedates

leucomystax) as one species, it is now recognized as a species complex, and individuals surveyed

likely consisted of two species—Polypedates megacephalus and Polypedates mutus (Kuraishi et

al., 2013).

Gamma diversity (total species richness) did not vary among land use types, (Fig. 2.1;

F2,4=2.91, P=0.1660), but was lower in Hanoi than the other metropolitan areas (F2,4=13.27,

P=0.0175; Hanoi-Danang: Padj=0.0202; Hanoi-Ho Chi Mihn City: Padj=0.0300; Danang-Ho Chi

Minh City: Padj=0.5015). Alpha diversity (average species richness per site) did not differ among

regions (F2,4=1.73, P=0.2876), but there was a trend of decreasing alpha diversity with increasing

urbanization, and urban areas had significantly lower alpha diversity than suburban areas or rice

paddies (F2,4=31.11, p=0.0036; urban-suburban: Padj=0.0139; urban-rice: Padj=0.0033;

suburbanrice: Padj=0.1365). Beta diversity (species turnover) therefore increased with

urbanization and was higher in urban areas than in suburban areas or rice paddies (F2,4=25.38,

P=0.0053; urbansuburban: Padj=0.0123; urban-rice: Padj=0.0056; suburban-rice: Padj=0.4684).

Family

Species

Presence

Land use where species

was most abundant

Bufonidae

Duttaphrynus melanostictus

(Schneider, 1799)

Mycrohylidae

Rice

Kaloula pulchra

(Gray, 1831)

suburban

Microhyla butleri

(Boulenger 1900)

suburban

Microhyla fissipies

(Boulenger, 1884)

Dicroglossidae

suburban

Fejervarya limnocharis

(Gravenhorst, 1829)

Rice

Hoplobatrachus rugulosus

(Wiegmann, 1834)

Rice

Occidozyga lima

(Gravenhorst, 1829)

Rice

Occidozyga martensii

(Peter, 1867)

Rice

Occidozyga vittata

(Andersson, 1942)

Ranidae

Rice

Hylarana guentheri

(Boulenger, 1882)

Rice

Hylarana taipehensis

(Van Denburgh, 1909)

Rhacophoridae

Rice

Polypedates leucomystax complex

(Gravenhorst, 1829)

suburban

Location of species detection: darkened quadrants indicate presence while white quadrants represent absence.

Columns are metropolitan areas: H-Hanoi, D-Danang, C-Ho Chi Minh City; rows are land-use types: R-rice,

SSuburban, U-urban.

Table 2.1 Location and most common land-use types for species recorded during surveys.

Fig. 2.1 Effects of land-use type on amphibian diversity. Columns and error bars represent the means +/- standard

deviations of values for the three metropolitan areas surveyed. Lower case letters indicate statistical significance

with Tukey’s Honestly Significant Difference Tests at alpha=0.05.

The occupancy of water bodies by any calling amphibian was lower in urban areas than

in the other two land use types (Fig. 2.1; F2,4=130.87, P<0.001; urban-suburban: Padj<0.001;

urbanrice: padj<0.001; suburban-rice: padj=0.2169; Fig. 2.1), but did not differ among metropolitan

areas (F2,4=1.526, P=0.3217). Across all cities, we detected at least one calling amphibian in all

surveyed rice paddies, in 91% of suburban water bodies, and in 62% of urban water bodies.

Abundance of each species peaked either in rice paddies (eight species; Table 2.1) or in suburban

water bodies (four species). No species was most abundant in urban areas.

Species composition clustered both by metropolitan region and by land-use type,

indicating that geographic location and immediate surrounding land use both contribute to

composition at a site (Fig. 2.2). Our measure of community composition takes into account both

species identity and occurrence, and was explained by both region (F2,4=4.01, P<0.001) and land

use (F2,4=3.60, P<0.001). Species identity was generally explained by region while relative

occupancy of species was influenced more by land-use type (Table 2.1 and Fig. 2.2).

Fig. 2.2 Non-metric multi-dimensional scaling analysis of amphibian community composition for three land-uses in

and around three Vietnamese cities. Community composition incorporated both presence and occurrence of species.

Each two-letter data point represents the amphibian community for one land-use in one city. The first letter

represents the metropolitan area: H-Hanoi, D-Danang, C-Ho Chi Minh City, while the second letter represents the

land-use: R-rice paddy, S-suburban, U-urban. The dashed lines group sites by region and the solid lines group sites

by land-use.

3.2 Habitat Characteristics

Species richness in urban and suburban water bodies was best explained by amount of

surrounding vegetated upland habitat and the presence of shallows, as well as either natural

banks or presence of aquatic vegetation (Table 2.2). The two top models contained these

variables and had R2’s of 0.38 and 0.41. Bank type and aquatic vegetation were highly correlated,

and therefore either one or the other was included in any given model. The size of water bodies

and slope of the sides were not influential in explaining species richness. Richness was on

average about twice as high in water bodies with shallows than in those without, three times

higher in water bodies with natural banks than those with constructed edges, four times higher in

those with aquatic vegetation than those without, and four times higher in those with >1 hectare

of surrounding vegetated upland habitat than those with less (Table 2.2). Rice paddies, which

were not included in these analyses due to their relative homogeneity with respect to the

measured features, were generally >100 000m2, with rooted aquatic vegetation, with natural

banks of slopes <90°, with shallow areas, and surrounded by at least 1 hectare of upland habitat.

Habitat characteristic Description Distributiona Directionb

Effect sizec

Natural banks

sides of water body

are natural (dirt

substrate) vs.

constructed (stone,

brick, or concrete)

natural: 75 constructed:

3.2

Shallows

at least 10 m2 is

>20 cm in depth

present: 99 absent:

+ 3.8

Surrounding

vegetated upland

habitat

amount of pervious

surface until an

impervious surface is

reached in all

directions (hectares)

mean: 29.5

median: 5.0 range:

0-319

+ 2.3

(at 1 hectare

cutoff)

Aquatic vegetation

rooted vegetation in

the water body

present: 91 absent:

+ 3.9

Slope of edge

edge slope

<90°: 96

90°: 24

ns NA

Size

average size during

breeding season

Mean: 61,945m2

Median: 5,395m2

Range: 0.6-120,785 m2

For categorical features the number of water bodies (out of the total 120) in each category are listed. For

continuous feature, the mean, median, and range values are given.

b The direction of association is for generalized linear mixed model averaged effects where “ns”

indicates non-significance.

The effect size is given as proportional change in species richness.

Table 2.2 Habitat characteristics measured for water bodies in urban and suburban areas and their associations with

amphibian richness.

4 Discussion

Results indicate that many current Vietnamese lowland amphibian species can persist in a

variety of human land uses when the right local and landscape habitat characteristics are present.

Similar results from prior work in temperate urbanized landscapes suggest that this may be a

general pattern (Hamer and McDonnell, 2008; Holzer, 2014).

4.1 Land Use

The most surprising result of this study was that all twelve species of amphibians

detected during surveys were present and calling in all three land-use types examined: in rice

cultivation fields, in growing suburban areas, and in dense urban cores. We expected to find

some species avoiding urban areas and others avoiding agricultural areas, but we found that none

of these species avoided either of these potential habitats. This result is unlikely to be a fleeting

artifact of extinction debt (sensu Tilman et al., 1994) because two of the three cities examined are

large (>7 million people) and hundreds to thousands of years old, and rice cultivation has

dominated the surrounding landscape for hundreds to thousands of years (Hanks, 1972).

However, extinction debt in the urban core cannot be entirely ruled out because suburban areas

have been expanding dramatically in recent years. If rice fields are important source populations

for urban cores, the cores might be increasingly isolated. The presence of a variety of amphibian

species across these landscapes is likely related to these species being relatively tolerant of

altered landscapes and may also be related to land use practices that make the areas relatively

hospitable for amphibians. Rice paddies throughout the world are able to support diverse

native aquatic wildlife communities under some management regimes, especially those without

rapid hydrologic fluctuations and those with low or no pesticide use (Lawler, 2001). Because rice

fields can

mimic seasonal wetlands, their habitat value may be higher in areas that have been converted

from seasonal wetlands than from other land cover, such as forest (Lawler, 2001). Many

Vietnamese lowland rice paddies replace former seasonal wetlands and do not experience rapid

fluctuations in hydrology, which likely benefits the native amphibians. Pesticide use in rice fields

in Vietnam rose rapidly in the early 1990s and, although use has been drastically reduced through

recent campaigns, pesticide use is currently practiced to some extent by the majority of rice

farmers (Berg and Tam, 2012; Huan et al., 1999; Nguyen et al., 1999). The effects of pesticides

on amphibians in this region are little known, but likely to be detrimental for many species (Liu

et al., 2011), and have been shown to have negative impacts on amphibians in other regions

(Egea-Serrano et al., 2012).

Urban development in this region has generally incorporated a variety of water bodies

during growth (Mitsch and Gosselink, 2007). This is opposed to general practices in the global

West of frequently filling wetlands during urban growth and sometimes reconstructing them in

some form later (Mitsch and Gosselink, 2007). All three cities in this study contained over one

hundred water bodies including urban lakes surrounded by cafes, large vegetated wet fields to

capture storm water, ornamental fountains in urban parks, large ponds for fish rearing and

harvest, and spiritual pools in temples; amphibians were found breeding in each of these types

when appropriate habitat characteristics were present.

We could not provide direct comparison of these human-dominated landscapes to pristine

lowland landscapes similar to what was present before development due to the rarity of the latter.

However, it is likely that amphibian species richness and diversity was much higher when

pristine areas were more common. Amphibian communities in primary and secondary forests and

palm plantations in Malaysia have been shown to have similar richness, but differing community

composition: the forests contained rare, forest-dependent species while the plantations contained

common species generally considered tolerant of human disturbance (Faruk et al., 2013).

Amphibian communities in pastures and coffee plantations in Costa Rica also had similar

richness to a nearby forest reserve but contained different species (Mendenhall et al., 2014).

Most of the species identified in the current study are considered common and tolerant of human

disturbance (Nguyen et al., 2009). Contemporary studies of relatively pristine areas in other

lowland areas in Southeast Asia as well as historical surveys in this region have indicated that

16-42 amphibian species may have formerly been present (Das and Dijk, 2013; Nguyen et al.,

2009). The twelve species found during this survey have apparently been able to adapt and

persist in the agriculture and urban areas which now constitute much of the limited habitat

options for many lowland species.

Although all species detected were found breeding in urban cores, none were most

successful there, that is, we did not find evidence for ‘urban exploiters’. Two species—

Hoplobatrachus rugulosus and Hylarana taipehensis—were very rare in urban water bodies.

Occupancy was lower in urban areas than in the other two land uses, and abundances for each

species peaked in either rice or suburban areas. This was expected because amphibians often

require vegetated uplands (Harper et al., 2008; Trenham and Shaffer, 2005), and are not known to

take advantage of anthropogenic food. Major amphibian predators—such as fish, aquatic

invertebrates, and birds— were observed frequently during our surveys in or near the urban

water bodies. This is consistent with past findings where only one out of the 25 studies examined

in a recent review found urbanization to increase amphibian presence or abundance (Hamer and

McDonnell, 2008).

Although urban exploiter species are common for birds and mammals (McCleery, 2010;

McKinney, 2002; Shochat et al., 2010), they are likely not common in amphibians.

Many plant and animal assemblages have been found to be relatively homogenous among

urban areas across regions (McKinney, 2006); however, we did not find this for urban

amphibians throughout Vietnam. We had considered homogenization unlikely in these

communities due to the relatively low vagilities and small species ranges of amphibians. Results

showed that the geographic location of each metropolitan area generally set the bounds on which

species were found during surveys, but land use also played a role in structuring amphibian

communities by affecting relative occupancy of species.

Species turnover among sites was higher in urban areas than in other land use types, and

fewer species on average were found at a given site in urban areas (1 or 2 species per water

body) than in suburban areas or rice paddies (3 or 4 species per water body). The typical rice

paddy possessed the characteristics associated with increased species richness per site: soil banks

with <90° slope, rooted aquatic vegetation, and >1 hectare of surrounding vegetated upland

habitat. Rice paddies were generally much larger than the urban or suburban water bodies and,

therefore; a species-area relationship may explain the higher number of species in rice paddies

(Gleason, 1922). However, this cannot explain the pattern in suburban vs. urban water bodies

where there was no difference in water body size or density. In addition, water body size was a

poor predictor of species richness in these land uses.

The reduced occupancy and number of species at a given site in cities are likely related to

landscape isolation by roads and buildings in urban areas (Hale et al., 2012; Hamer and

McDonnell, 2008; Van Buskirk, 2012). Due to the dense infrastructure and age of the urban

centers of two of the cities, it was surprising to find any amphibians persisting at sites in the

urban core. Local populations either must have persisted for decades or centuries, or some

individuals were able to traverse the built landscapes of downtown Hanoi and Ho Chi Minh City.

Studies examining movement of amphibians among sites in urban areas are needed to better

understand the degree of isolation.

Our findings of amphibian use of urban and surrounding agricultural rice paddies in

Vietnam do not follow general patterns in the relatively new field of urban ecology which predict

that urban areas have reduced richness, increased abundance, and homogenized wildlife

communities (Aitkenhead-Peterson and Volder, 2010). Instead, in urban areas we found sustained

richness, decreased abundance, and lack of biotic homogenization. Our pattern is similar to that

found by Herrmann et al., 2012 with parasitoid wasp species. Therefore urban ecology theories

may require additional principles for low-vagility taxa that do not consume human-associated

foods. Some such communities may have substantial presence in urban environments when

suitable conditions exist, providing opportunities for their conservation.

4.2 Habitat Characteristics

We were interested in determining which habitat characteristics best corresponded with

species richness to better understand the conditions that allow for persistence in urbanized areas.

We found that both local and landscape-scale factors contributed to explaining amphibian

distribution in the urban and suburban sites.

Banks made of soil (rather than water-impervious material) were highly correlated with

the presence of rooted aquatic vegetation, and these correlated factors together were associated

with higher species richness. This is consistent with other studies showing that aquatic vegetation

and natural banks are often important for urban amphibians (Cornelis and Hermy, 2004; Hamer

and

McDonnell, 2008). Even potted ornamental plants seem to provide substantial benefits for many

species. For example Tao Dan Park in the center of Ho Chi Minh City’s urban core contains

several bodies of water, most of which are small, vertically sided, concrete fountains; those

without any aquatic vegetation had no breeding activity while those with potted ornamental lilies

contained up to six breeding species (Fig. 2.3a). There is a general construction trend in

Southeast Asia of converting the banks of water bodies from soil to impervious surfaces (Fig.

2.3b). When this occurred during our surveys we noticed a decline in amphibians in the affected

water bodies. Studies of amphibian use of drainage ditches in Japan and rice paddy banks in

Taiwan showed a similar detriments to amphibians with concrete vs. soil banks (Chang et al.,

2011; Fujioka and Lane, 1997; Hou et al., 2010; Naito et al., 2012). Ideally, this practice would

be halted and reversed, but this is unlikely due to perception that it promotes clean and modern

city landscapes. However, it is possible that damage to amphibian communities could be

lessened if rooted aquatic vegetation is provided after alteration, even if it is in the form of potted

ornamentals.

Many urban and suburban water bodies in Vietnam are deep and steep-sided such that

there are no shallow areas, and we found that these rarely supported breeding amphibians.

Amphibian richness was associated with water bodies that provided some shallow areas, likely

because tadpoles benefit from the ability to bask (Stebbins and Cohen, 1995). Even artificial

shallows provided by garbage and broken walls were observed being used by tadpoles for

basking during this study. Due to limitations in access to water bodies, this study did not examine

the role of pollution, which is likely a limiting factor at many sites and should be examined in the

future. The amount of vegetated upland area surrounding a water body was positively

associated with amphibian richness. This is expected because many of the species found

generally only come to water bodies to breed, spending the majority of their adult lives in the

surrounding terrestrial environment. The strong positive association between upland habitat and

species richness is consistent with many other urban amphibian studies (Hamer and McDonnell,

2008). What was surprising was that the prominent threshold occurred at only one hectare of

surrounding vegetated upland habitat, and that many parks and storm water sites around this size

contained several amphibian species. Although more upland habitat is generally better, it is

encouraging that setting aside a relatively small amount of water-pervious land may have a large

positive impact on the amphibian community.

To benefit amphibians in the vast and rapidly expanding urban areas of Southeast Asia,

we recommend maintaining vegetated upland habitat surrounding water bodies where possible,

allowing water-pervious banks in water bodies, ensuring presence of shallows, and promoting

the presence of aquatic vegetation.

Fig. 2.3 Examples of urban and suburban water bodies with and without major habitat features that

contributed to amphibian species richness. On the left are urban fountains with (a) and without (b) potted

ornamental vegetation from Tao Dan Park in Ho Chi Minh City. On the right is a suburban water body

before (c) and during (d) conversion from a natural bank to an impervious bank in northern Danang.

Ponds in (a) and (c) supported the most amphibians.

5 Conclusion

The continued decline of pristine landscapes emphasizes that urban and agricultural areas

cannot be discounted from conservation planning. Our results indicate that areas of intense

human use, including cities and rice paddies, can provide habitat for many amphibian species.

However, the specific planning and management of these landscapes can have large impacts on

whether or not they are used by native wildlife species.

The lowlands of Vietnam are dominated by rice cultivation and rapidly growing cities,

and have likely already lost many species that were not able to cope with these land use changes.

Currently one dozen amphibian species breed in rice cultivation, suburban sprawl, and dense

urban cores of the country’s largest cities. Understanding how amphibians use these

humandominated areas can guide conservation and management to ensure their continued

coexistence. Past land management practices, such as allowing diverse rice ecosystems and

incorporating water bodies into urban growth, have likely contributed to amphibian use of these

areas. However, some worrying trends—such as widespread pesticide use, rapid urban growth

incorporating fewer water bodies and green spaces, and conversion of soil to water-impervious

banks—will decrease the suitability of these habitats. Understanding the drivers of biodiversity

in cities and rice fields will allow planners and managers to develop practices that promote

human-wildlife coexistence in areas of intense human use.

Acknowledgements

We would like to thank the Vietnam National University and the Vietnam National Museum of

Nature for allowing and facilitating this research. We would also like to thank Dr. S.P. Lawler,

Dr. P.B. Moyle, and Dr. H.B. Shaffer for improving earlier versions of this manuscript. Funding

was provided by University of California, Davis, the Explorers’ Club, and National Science

Foundation Graduate Research Fellowship Program grant to K.A.H (#1148897).

Literature Cited

Aitkenhead-Peterson, J., Volder, A. (Eds.), 2010. Urban Ecosystem Ecology. American Society of

Agronomy.

Aselmann, I., Crutzen, P.J., 1989. Global Distribution of Natural Freshwater Wetlands and Rice

Paddies , their Net Primary Productivity , Seasonality and Possible Methane Emissions. J.

Atmos. Chem. 8, 307–358.

Babbitt, K.J., 2005. The relative importance of wetland size and hydroperiod for amphibians in

southern New Hampshire, USA. Wetl. Ecol. Manag. 13, 269–279. doi:10.1007/s11273-004-

7521-x

Berg, H., Tam, N.T., 2012. Use of pesticides and attitude to pest management strategies among

rice and rice-fish farmers intheMekong Delta, Vietnam. Int. J. Pest Manag. 58, 153–164.

doi:10.1080/09670874.2012.672776

Chang, Y.-H., Wang, H.-W., Hou, W.-S., 2011. Effects of construction materials and design of

lake and stream banks on climbing ability of frogs and salamanders. Ecol. Eng. 37, 1726–

1733. doi:10.1016/j.ecoleng.2011.07.005

Cornelis, J., Hermy, M., 2004. Biodiversity relationships in urban and suburban parks in

Flanders. Landsc. Urban Plan. 69, 385–401. doi:10.1016/j.landurbplan.2003.10.038

Cushman, S. a., 2006. Effects of habitat loss and fragmentation on amphibians: A review and

prospectus. Biol. Conserv. 128, 231–240. doi:10.1016/j.biocon.2005.09.031

Das, I., Dijk, P.P. van, 2013. Species richness and endemicity of the herpetofauna of South and

Southeast Asia. Raffles Bull. Zool. Supp. 29, 269–277.

Dorcas, M.E., Price, S.J., Walls, S.C., Barichivich, W.J., 2009. Auditory monitoring of anuran

populations., in: Dodd, Jr., C.K. (Ed.), Amphibian Ecology and Conservation. Oxford

University Press, New York, New York, pp. 281–298.

Egea-Serrano, A., Relyea, R. a, Tejedo, M., Torralva, M., 2012. Understanding of the impact of

chemicals on amphibians: a meta-analytic review. Ecol. Evol. 2, 1382–97.

doi:10.1002/ece3.249

Faruk, A., Belabut, D., Ahmad, N., Knell, R.J., Garner, T.W.J., 2013. Effects of Oil-Palm

Plantations on Diversity of Tropical Anurans 27, 615–624. doi:10.1111/cobi.12062

Fujioka, M., Lane, S.J., 1997. The impact of changing irrigation practices in rice fields on frog

populations of the Kanto Plain, central Japan. Ecol. Res. 12, 101–108.

doi:10.1007/BF02523615

Gleason, H.A., 1922. On the Relation Between Species and Area. Ecology 3, 158–162.

Gooch, M., Dorcas, M., Price, S., Heupel, A., 2006. The effects of survey protocol on detection

probabilities and site occupancy estimates of summer breeding anurans. Appl. Herpetol. 3,

129–142. doi:10.1163/157075406776984211

Hale, J.M., Heard, G.W., Smith, K.L., Parris, K.M., Austin, J.J., Kearney, M., Melville, J., 2012.

Structure and fragmentation of growling grass frog metapopulations. Conserv. Genet. 14,

313–322. doi:10.1007/s10592-012-0428-9

Hamer, A.J., McDonnell, M.J., 2008. Amphibian ecology and conservation in the urbanising

world: A review. Biol. Conserv. 141, 2432–2449. doi:10.1016/j.biocon.2008.07.020

Hanks, L.M., 1972. Rice and Man: Agricultural Ecology in Southeast Asia. Aldine Publishing

Company, Chicago.

Harper, E.B., Rittenhouse, T.A.G., Semlitsch, R.D., 2008. Demographic consequences of

terrestrial habitat loss for pool-breeding amphibians: predicting extinction risks associated

with inadequate size of buffer zones. Conserv. Biol. 22, 1205–15. doi:10.1111/j.1523-

1739.2008.01015.x

Herrmann, D.L., Pearse, I.S., Baty, J.H., 2012. Landscape and Urban Planning Drivers of

specialist herbivore diversity across 10 cities. Landsc. Urban Plan. 108, 123–130.

doi:10.1016/j.landurbplan.2012.08.007

Holzer, K., 2014. Amphibian use of constructed and remnant wetlands in an urban landscape.

Urban Ecosyst. doi:10.1007/s11252-014-0373-0

Hou, W., Chang, Y., Wang, H., Tan, Y., 2010. Using the behavior of seven amphibian species for

the design of banks of irrigation and drainage systems in Taiwan. Irrig. Drain. 59, 493–

505. doi:10.1002/ird

Hsu, M., Kam, Y.-C., Fellers, G.M., 2005. Effectiveness of amphibian monitoring techniques in a

Taiwanese subtropical forest. Herpetol. J. 15, 73–79.

Huan, N.H., Mai, V., Escalada, M.M., Heong, K.L., 1999. Changes in rice farmers’ pest

management in the Mekong Delta , Vietnam. Crop Prod. 18.

Hunt, S.D., Guzy, J.C., Price, S.J., Halstead, B.J., Eskew, E. a., Dorcas, M.E., 2013. Responses of

riparian reptile communities to damming and urbanization. Biol. Conserv. 157, 277–284.

doi:10.1016/j.biocon.2012.08.035

Huston, M., 1993. Biological diversity, soils, and economics. Science (80-. ). 262, 1676–1680.

Kuraishi, N., Matsui, M., Hamidy, A., Belabut, D.M., Ahmad, N., Panha, S., Sudin, A., Yong,

H.S., Jiang, J.-P., Ota, H., Thong, H.T., Nishikawa, K., 2013. Phylogenetic and taxonomic

relationships of the Polypedates leucomystax complex (Amphibia). Zool. Scr. 42, 54–70.

doi:10.1111/j.1463-6409.2012.00562.x

Lawler, S.P., 2001. Rice fields as temporary wetlands: A review. Isr. J. Zool. 47, 513–528.

Liu, W.-Y., Wang, C.-Y., Wang, T.-S., Fellers, G.M., Lai, B.-C., Kam, Y.-C., 2011. Impacts of the

herbicide butachlor on the larvae of a paddy field breeding frog (Fejervarya limnocharis) in

subtropical Taiwan. Ecotoxicology 20, 377–84. doi:10.1007/s10646-010-

0589-6

Lotz, A., Allen, C.R., 2007. Observer Bias in Anuran Call Surveys. J. Wildl. Manage. 71, 675–

679. doi:10.2193/2005-759

Machado, I.F., Maltchik, L., 2010. Can management practices in rice fields contribute to

amphibian conservation in southern Brazilian wetlands? Aquat. Conserv. Mar. Freshw.

Ecosyst. 20, 39–46. doi:10.1002/aqc

Magle, S.B., Hunt, V.M., Vernon, M., Crooks, K.R., 2012. Urban wildlife research: Past, present,

and future. Biol. Conserv. 155, 23–32. doi:10.1016/j.biocon.2012.06.018

McCarthy, K., Lathrop, R.G., 2011. Stormwater basins of the New Jersey coastal plain:

Subsidies or sinks for frogs and toads? Urban Ecosyst. 14, 395–413.

doi:10.1007/s11252011-0161-z

McCleery, R., 2010. Urban Mammals, in: Aitkenhead-Peterson, J., Volder, A. (Eds.), Urban

Ecosystem Ecology. American Society of Agronomy, pp. 87–102.

McKinney, M.L., 2002. Urbanization, biodiversity, and conservation. Bioscience 52, 883–390.

McKinney, M.L., 2006. Urbanization as a major cause of biotic homogenization. Biol. Conserv.

127, 247–260. doi:10.1016/j.biocon.2005.09.005

Mendenhall, C.D., Frishkoff, L.O., Santos-Barrera, G., Pacheco, J., Mesfun, E., Quijano, F.M.,

Ehrlich, P.R., Ceballos, G., Daily, G.C., Pringle, R.M., 2014. Countryside biogeography of

Neotropical reptiles and amphibians. Ecology 95, 856–870.

Mitsch, W.J., Gosselink, J.G., 2007. Wetlands, 4th ed. John Wiley & Sons Inc.

Naito, R., Sakai, M., Morimoto, Y., 2012. Negative effects of deep roadside ditches on

Pelophylax porosa brevipoda dispersal and migration in comparison with Hyla japonica in a

rice paddy area in Japan. Zoolog. Sci. 29, 599–603. doi:10.2108/zsj.29.599

Naito, R., Sakai, M., Natuhara, Y., Morimoto, Y., Shibata, S., 2013. Microhabitat use by Hyla

japonica and Pelophylax porosa brevipoda at Levees in rice paddy areas of Japan. Zoolog.

Sci. 30, 386–91. doi:10.2108/zsj.30.386

Nguyen, H.D., Thien, T.C., Hong, N. Van, Loc, N.T., Minh, D. Van, Thau, T.D., Thi, H., Nguyen,

L., Phong, N.T., 1999. Impact of Agro-Chemical Use on Productivity and Health in

Vietnam. Econ. Environ. Progr. South East Asia Res. Rep. Ser. 1–69.

Nguyen, K. V., Nguyen, T.H., Phan, K.L., 2000. Bioclimatic diagrams of Vietnam. Vietnam

National University Publishing House, Hanoi, Vietnam.

Nguyen, V.S., Ho, T.C., Nguyen, Q.T., 2009. Herpetofauna of Vietnam. Edition Chimaira.

Rosenzweig, M.L., 2003. Reconciliation ecology and the future of species diversity. Oryx 37,

194–205. doi:10.1017/S0030605303000371

Shochat, E., Lerman, S., Fernandez-Juricic, E., 2010. Birds in Urban Ecosystems: Population

Dynamics, Community Structure, Biodiversity, and Conservation, in: Aitkenhead-Peterson,

J., Volder, A. (Eds.), Urban Ecosystem Ecology. American Society of Agronomy, pp. 75–

86.

Stebbins, R.C., Cohen, N.W., 1995. A Natural History of Amphibians. Princeton University

Press.

Stuart, S.N., Chanson, J.S., Cox, N. a, Young, B.E., Rodrigues, A.S.L., Fischman, D.L., Waller,

R.W., 2004. Status and trends of amphibian declines and extinctions worldwide. Science

306, 1783–6. doi:10.1126/science.1103538

Tarr, T., Babbitt, K., 2002. Effects of habitat complexity and predator identity on predation of

Rana clamitans larvae. Amphibia-Reptilia 23, 13–20. doi:10.1163/156853802320877591

Team, R.C.D., 2008. R Version 3.0.1.

Tilman, D., May, R.M., Lehman, C.L., Nowak, M.A., 1994. Tilman et al. 1993 Extinction

debt.pdf 65–66.

Trenham, P.C., Shaffer, H.B., 2005. Amphibian upland habitat use and its consequences for

population viability. Ecol. Appl. 15, 1158–1168.

Van Buskirk, J., 2012. Permeability of the landscape matrix between amphibian breeding sites.

Ecol. Evol. 2, 3160–7. doi:10.1002/ece3.424

Weir, L.A., Mossman, M., 2005. North American Amphibian Monitoring Progam (NAAMP)., in:

Lannoo, M. (Ed.), Amphibian Declines: The Conservation Status of United States

Species. University of California Press, Berkeley, California, pp. 307–313.

Chapter Three

The results of Chapter One suggest that urban pollutants, especially nitrate, may be driving

patterns of wildlife use in urban wetlands. This field study contains numerous confounding

factors that are difficult to parse out. Lab studies examining nitrate effects on amphibians are

numerous, but controlled, semi-natural studies are lacking. I conducted one such study to isolate

nitrate as a possible stressor for developing amphibians in outdoor mesocosms.

Effects of nitrate on Pacific chorus frog tadpole survival and growth in outdoor mesocosms

Katie A. Holzer1*, Sharon P. Lawler1

College of Agriculture and Environmental Science, University of California, Davis, One Shields

Ave, Davis, CA 95616

*corresponding author: holzer[email protected]

Keywords: nutrient pollution, eutrophication, anuran, urban, agriculture

Abstract

Pollutants are thought to be one major cause of worldwide amphibian decline. Nitrate is a

common aquatic pollutant, and laboratory studies show that nitrate can harm amphibians, but

often only at levels higher than the maximum contaminant level for drinking water (10 mg/L).

Recent observational field studies revealed negative associations of nitrate with amphibians at

much lower levels. We conducted an outdoor mesocosm study to examine the effects of nitrate

on the survival and growth of the Pacific chorus frog (Pseudacris regilla). We used five

replicated 110 L tanks for each of seven nitrate level treatments: 0, 0.5, 1, 2, 5, 10, and 20 mg/L.

In each tank we placed 20 chorus frog eggs and allowed them to develop to metamorphosis in the

presence of algal growth, natural UV radiation, physical refugia, and naturally colonizing

invertebrate predators and competitors. Results showed that at these levels nitrate did not

decrease survival or growth of tadpoles in our outdoor mesocosms; in fact trends were toward

higher survival and growth at higher nitrate levels. Nitrate appeared to promote greater algal

growth, which likely improved food resources for tadpoles. Our results indicate that levels up to

20 mg/L of nitrate are likely not detrimental to Pacific chorus frogs. Other correlated and/or

interacting factors may drive observed negative associations between these levels of nitrates and

local amphibian declines.

1. Introduction

Amphibians are in decline worldwide due to a variety of causes (Stuart et al. 2004; Wake

& Vredenburg 2008), and research on the specific effects of drivers is needed to guide

management for their conservation. Pollution negatively affects many amphibians, and may be a

major driver in their population declines (Beebee & Griffiths 2005). Numerous negative

associations are documented between amphibians and pollutants in the field (Hamer &

McDonnell 2008; Holzer 2014), and a number of contaminants have been found to have negative

effects on some anurans in laboratory or mesocosm tests of toxicity (Egea-Serrano et al. 2012).

Field observations are limited in their ability to isolate drivers through controlling confounding

factors, while laboratory studies lack many of the realistic elements of natural environments

(Diamond 1986; Hairston 1989; Morin 1998). In contrast, outdoor experimental mesocosms

allow for control of some confounding factors and replication while also allowing for some of

the elements of the natural environments such as predators, competitors, algal growth, and

ultraviolet (UV) radiation (Egea-Serrano et al. 2012). More mesocosm tests are needed (Hamer

& McDonnell 2008; EgeaSerrano et al. 2012), especially because there is wide variance in the

susceptibility of anurans to toxicants (Kerby et al. 2010). We conducted an outdoor mesocosm

study to assess effects of a widespread aquatic pollutant, nitrate, on survival and growth of a

common frog of western North

America.

Nitrate is a naturally occurring nutrient that can stimulate primary productivity at low

concentrations, but can cause detrimental effects to aquatic ecosystems at higher levels or in

synergy with other pollutants (Rouse et al. 1999; Camargo et al. 2005; Boone et al. 2005). It is

one of the most common pollutants of aquatic systems, and levels can be particularly high in

agricultural and urban areas (Camargo et al. 2005). Major sources of nitrate in aquatic

ecosystems include fertilizers, industrial effluents, waste-water treatment discharge, animal

waste, and deposition from motor vehicles and industrial exhaust (Camargo et al. 2005). Nutrient

loads in ponds have been shown to be negatively associated with amphibian richness in some

observational field studies (Ensabella et al. 2003; Houlahan & Findlay 2003; Holzer 2014).

Several studies have been conducted examining the effects of nitrate on amphibians in controlled

laboratory settings (Egea-Serrano et al. 2012), and a few in field mesocosms (Boone et al. 2005;

Boone & Bridges-Britton 2006; Smith et al. 2011, 2013). For the most part, laboratory studies

have shown that nitrate can have detrimental effects on activity, growth, and survival, but often

only at levels above the Environmental Protection Agency’s maximum contaminant level (EPA’s

MCL) of 10 mg/L, and sometimes not until levels more than an order of magnitude higher (e.g.

Schuytema & Nebeker 1999; Boone et al. 2005; Miaud et al. 2011). Mesocosm studies have

shown that nitrate levels below 10 mg/L can be beneficial to ranid frog survival and growth

(Smith et al. 2011, 2013), can increase survival at 10 mg/L while decreasing survival at 20 mg/L,

and have mixed results on growth at both 10 and 20 mg/L (Boone et al. 2005, 2007; Boone &

Bridges-Britton 2006). The conflicting results may come from species differences as well as

confounding factors, and demonstrate the importance of studying this pollutant in replicated

semi-natural settings to better understand the effects in the natural environment.

Prior field surveys of ours in Portland, Oregon (Holzer, 2014) indicated that all water

bodies in the city had nitrate levels below 20 mg/L, but that tadpole/larvae density across several

taxa (Pseudacris regilla, Rana aurora, Ambystoma macrodactylum, Ambystoma gracile) was

negatively associated with levels above 1 mg/L (Fig 3.1). This indicates that additional stressors

or interactions in the field may cause nitrate to be more detrimental than in the laboratory (Marco

et al. 1999; Schuytema & Nebeker 1999b; Romansic et al. 2006). One outdoor mesocosm study

examined the effects of nitrate on Pseudacris regilla tadpoles, in combination with UV radiation

(Hatch & Blaustein 2003). The study found that nitrate at relatively low levels (10 and 20 mg/L)

produced detrimental effects on survival or growth of two amphibian species when combined

with UV radiation (Hatch & Blaustein 2003). However, while the study allowed for UV

radiation, it did not possess many other characteristics of field situations, such as predators,

competitors, refuge, and substantial algal growth.

The aim of this study was to assess survival and growth of a common frog (Pacific chorus

frog—Pseudacris regilla) with varying levels of nitrate in outdoor mesocosms in the presence of

several elements reflective of natural ponds. To do this we reared frog eggs in wading pools with

nitrate levels 0-20mg/L, tracking their survival and growth to metamorphosis while allowing for

UV radiation, refuge, algal growth, and natural colonization by invertebrate predators and

competitors. We tested the hypothesis that low levels of nitrate would reduce survival and growth

of the frog under these natural biotic conditions.

Fig 3.1 Relationship between amphibian larval density and NO3- levels. Each dot is one amphibian species in one

pond.

2. Methods

We conducted this experiment in 110 L wading pools in a research field in Davis,

California (38.543°N, 121.763°W). We used 35 pools as the experimental units: five replicates of

seven nitrate level treatments: 0, 0.5, 1, 2, 5, 10, and 20 mg/L. Each row constituted one block

with treatments randomized within a row (Fig 3.2). Pools were filled on 11 April, 2011 with 100

L of deionized water with ¼ teaspoon of Instant Ocean® salts (Spectrum Brands, Inc., 3001

Commerce St. Blacksburg, VA) and ~1 kg of weathered hay to provide cover. Each pool received

50 mL of pond water and ~1/3 of a plankton tow from a nearby pond at the Center for

Land-Based Learning (38.52°N, 121.91°W). We added 0.35 g of rabbit chow (Purina Mills, PO

Box 66812 St. Louis, MI) to provide a nutrient base for the food web. We added nitrate as

sodium nitrate (Carolina Biological Supply Company, 2700 York Road, Burlington, NC) starting

on 18 April, 2011 and on a weekly basis to adjust for uptake and transformation. The decrease

throughout the week followed by return to treatment levels mimics runoff situations throughout a

season. We maintained nitrate levels near the treatment concentrations throughout the experiment

by testing the nitrate level of each pool weekly with using a Nitrate combination ionselective

electrode (Fisher Scientific Accument® #13-620-534) and adding additional nitrate as needed.

We collected freshly laid Pacific chorus frog eggs from the wild (38.76°N, 120.54°W) on 19

April, 2011 and added them directly to the pools with 20 eggs per pool. Eggs were collected

under California Fish & Wildlife permit #12-032 and animals were used in the experiment under

the University of California Institutional Animal Care & Use Committee permit #16360. We left

the tanks open to the air with no covers from 11 April to 11 June, 2011 to allow for UV exposure

and colonization by insects.

Fig 3.2 Experimental tanks.

We checked tadpoles and water levels daily, maintaining water near the top of the pools.

When the first metamorphosed individual was observed we covered tanks with removable fabric

lids to prevent escape. We then removed fully metamorphosed individuals for measurement

twice daily, until all surviving individuals had metamorphosed. We measured the snout-vent

length of all metamorphs as well as their moist mass after being blotted with a moist paper towel.

Body condition was calculated as mass divided by length. After measuring, metamorphs were

euthanized with a buffered MS-222 solution. We also measured UV transmission through water

samples at 300nm using a spectrophotometer.

We considered each tadpole to be a subsample of the pool, and pool-wide averages were

taken for length, mass, days to metamorphosis, and body condition. In order to determine the

effects of nitrate level treatments on tadpole survival and growth we used multivariate linear

regression models. Significance level was determined for multiple comparisons via sequential

Bonferroni correction; for five tests at an alpha of 0.05, the lowest ‘significant’ P should be

below 0.01, the next lowest below 0.0125, the third lowest below 0.017, the next lowest below

0.025, and the highest not more than 0.05 (Holm 1979). We also conducted regression analyses

for UV transmission.

3. Results

Tadpole growth increased with higher nitrate levels, reaching an apparent asymptote at 10

mg/L (Fig 3.3). The growth metrics (snout-vent length, mass, and body condition) all increased

with nitrate level (snout-vent length: Radj=0.45, P<0.001; mass: Radj=0.60, P<0.001; body

condition: Radj=0.58, P<0.001). Time to metamorphosis decreased with nitrate level (Radj=0.25,

P=0.0209).

Fig 3.3 Measurements of tadpole survival and growth to metamorphosis in pond mesocosms subjected to various

nitrate levels. Mean +/- standard deviations.

Out of the 700 total eggs at the beginning of the experiment, 241 (34%) survived to

metamorphosis. This is within the bounds of natural survival of this species (Stebbins & Cohen

1995). Survival was very similar among most levels, but increased at the highest nitrate

treatment (Fig 3.3). Survival to metamorphosis was 25-35% in most treatments, but in 20 mg/L

60% survived and survival was more variable. Survival was positively related to nitrate level

(regression: R,dj=0.34, P<0.001).

Tanks were generally colonized by the following invertebrates, which were not

quantified: Chironomidae, Syrphidae, Culicidae, Ephydridae, and Dytiscidae. Planktonic algal

growth appeared to be higher in higher nitrate treatments (Fig 3.4). UV transmission at 300nm

was negatively correlated with nitrate level (Radj=0.34, P=0.004).

Nitrate levels tended to decrease throughout the week, whereupon nitrate was added

(Table 3.1). Higher treatments (5, 10, and 20 mg/L) were on average one half to two thirds of the

acute levels. Lower treatments (0, 0.5, 1, and 2 mg/L) tended to all average near 1 mg/L.

Treatment

Average standing

nitrate level (mg/L)

St. dev. in standing

nitrate level (mg/L)

0 0.99 0.19

0.5 0.91 0.15

1 0.89 0.09

2 1.18 0.07

5 3.11 0.30

10 6.73 1.02

20 14.16 2.15

Table 3.1 Average and standard deviation nitrate levels throughout the experiment for each treatment.

4. Discussion

Nitrate levels of 0.5-20 mg/L did not negatively impact Pacific chorus frog tadpole

survival and growth in outdoor mesocosms, but instead increased survival and growth to

metamorphosis. Nitrate was found to have mixed results on survival and growth of tadpoles in

other species

(Hyla versicolor, Lithobates clamitans, Rana sylvatica, and Bufo americanus) at these levels

(Boone et al. 2005; Boone & Bridges-Britton 2006; Smith et al. 2011, 2013). In a laboratory

experiment, these levels of nitrate were not found to be harmful to developing tadpoles of Pacific

chorus frogs (Schuytema & Nebeker 1999a); however, another outdoor mesocosm study with

this same species demonstrated reduced survival and growth at the 10 and 20 mg/L level (Hatch

& Blaustein 2003). Differences in methods between the current study and Hatch & Blaustein

2003 may have lead to the seemingly contradictory results.

In the Hatch & Blaustein 2003 study, tadpoles were fed ad libitum immediately after

hatching, and then at a rate consistent across treatments for the remainder of the experiment. In

the current study the main source of food for tadpoles was algae and periphyton growth in the

tanks. Planktonic and attached algae were more abundant in the higher nitrate treatments, and the

elevated algae levels in high nitrate tanks likely benefited tadpoles by providing increased food.

Positive effects of nitrate and another contaminant, the insecticide carbyl, in other mesocosm

studies were thought to be caused through the food web by increases in periphyton (Boone et al.

2005; Boone & Bridges-Britton 2006; Smith et al. 2013). However, positive effects of nitrate on

Rana sylvatica were not accompanied by increases in periphyton, so other mechanisms may be

present (Smith et al. 2011)

Additionally, tanks with higher nitrate had lower UV penetration, likely due to

absorbance and interference from algae in the water column. There was “very little algal growth

in the water column” in the other experiments (Hatch & Blaustein 2003 pg. 1088), whereas

growth was abundant in our experiment (Fig 3.4). This may have lowered the UV exposure of

the developing tadpoles, thereby decreasing harmful effects. Exposure to UV may also have been

reduced through the availability of hay as refuge. The tanks in the Hatch & Blaustein 2003

experiments did not offer any refuge for tadpoles to escape direct UV radiation, and they were

therefore exposed to full incoming UV levels, whereas the tadpoles in the current experiment

were able to move under the hay and therefore could behaviorally avoid full incoming UV levels.

Fig 3.4 Water samples showing planktonic algae as more abundant at higher nitrate levels. Treatments are across the

top and replicates are on the left.

Ponds in the wild usually contain planktonic algae and some sort of refuge for tadpoles,

and it is therefore possible that many tadpoles in the wild are not exposed to full incoming levels

of UV radiation. This demonstrates that natural ponds may not be subject to the detrimental

effects of synergistic nitrate and UV radiation to the extent observed in Hatch & Blaustein 2003.

However, any ponds lacking refuges, such as rock basin ponds at high elevation, or some

stormwater detainment ponds, may be at particularly high risk for these negative effects. Other

species of amphibians may also be more sensitive to these effects than the common species used

in this experiment, and additional studies are needed to determine the effects of pollutants on

other species in semi-natural mesocosms.

Observational studies of natural systems have shown negative associations between

amphibian presence/abundance and nitrate levels at the levels tested in this study (Ensabella et al.

2003; Houlahan & Findlay 2003; Holzer 2014). Due to the complex system that was simplified

greatly in the mesocosms of this study, it is likely that there are correlated pollutants that are

driving the trend alone or in synergy (Boone et al. 2005; Boone & Bridges-Britton 2006). Nitrate

was the only pollutant measured in Holzer 2014, and it was often found to be elevated in

stormwater ponds, especially those at the inlet of a facility (Holzer, personal observation). Those

ponds with increased nitrate levels are likely to also have increased levels of other pollutants

from runoff and drainage systems, many of which have been shown to negatively impact

amphibians such as salts and heavy metals (Paul & Meyer 2001; Simon et al. 2008; Willson &

Dorcas 2014). Pollutants can have detrimental synergistic effects on amphibians (Howe et al.

1998; Boone et al. 2005), and therefore nitrate may cause harm in concert with correlated factors

not measured in this study.

Throughout the world, pollutants are likely a leading cause of amphibian decline (Beebee &

Griffiths 2005). However, the results of this study indicate that nitrate at levels up to

Environmental Protection Agency’s maximum contaminant levels do not appear to be

detrimental to Pacific chorus frogs in a semi-natural setting.

Students also viewed