~red

Research You Can Use

Strategies to MaintainAmphibian Populationson Golf CoursesExploring the roles of golf courses in the environment.BY PETER W. C. PATON AND ROBERT S. EGAN

B"ologists are increasingly concernedwith amphibian populationsbecause of documented declines

on local, regional, and even globalscales. A variety of factors have beenimplicated in these declines (e.g., intro-duced predators, fertilizers, pollutants,and UV- B radiation in sunlight), andone of the leading factors is the impactof habitat fragmentation on pond-breeding amphibians (4).

This article focuses on pond-breedingamphibians because the majority ofamphibian species in the Northeastbreed in ponds (six species of sala-manders and 10 species of frogs), whilefewer species breed in streams oruplands (5). In this article, strategies arediscussed to maintain populations ofpond-breeding amphibians on golfcourses in New England based on avariety of studies conducted since 1997at the University of Rhode Island.

Amphibians can be exceptionallysensitive to changes in microclimateand microhabitat because they havepermeable skin that makes themsusceptible to desiccation. Thus, habitatecotones (mixed vegetation communi-ties formed by overlapping habitats),such as the transition between forestsand turf fairways, may represent poten-tial dispersal barriers to amphibiansmoving across the landscape. Frag-mented landscapes, such as golf courses,can impact amphibian populations.Amphibians that breed in ponds have

An adult gray treefrog is a commonspecies found in New England ponds.This species overwinters in trees andbreeds during the month of May.Because they prefer trees, golf coursefairways can be a dispersal barrier tothis species.

complex life cycles that make themparticularly vulnerable to fragmentationand loss of habitat.

Ponds are often used by adults onlyfor mating and depositing eggs, and bylarvae during development until meta-morphosis (i.e., the transformation intoterrestrial organisms). Adults are usuallyhighly site faithful to their breedingpond, returning to the same pond yearafter year, whereas metamorphs (youngof the year) tend to disperse across thelandscape and often breed in newponds. For most of the year, adults and

juveniles of most pond-breeding speciesreside in forested uplands and forestedwetlands near breeding ponds, withmany individuals traveling considerabledistances to reach their non-breedingterritories (e.g., salamanders of thegenus Ambystoma travel 180 yards andfarther).

Pond-breeding amphibians migratetwice a year, once from their non-breeding habitat to the breeding pond,and then back to their non-breedingterritory at the completion of thebreeding season. Therefore, managingthe landscape to maintain populationsof pond-breeding amphibians is a chal-lenge for golf course designers andsuperintendents because it requires adetailed understanding of the physicaland habitat characteristics of breedingponds, an understanding of habitatrequirements during the non-breedingseason, and knowledge of the inter-vening habitats used during migrationto and from ponds and non-breedinghabitat. What makes it even more diffi-cult is that biologists are just beginningto untangle the complex habitat require-ments of pond-breeding amphibians,particularly during migration and thenon-breeding season.

As part of a Wildlife Links projectfunded by the USGA, we conducteda number of short- and long-termexperiments and observational studiesto assess the impact of turf and golfcourses on pond-breeding amphibians

J U L Y - AU GUS T 2003 27

Figure IPercent of 59 ponds sampled on golf courses in southern

New England (Rhode Island, Connecticut, and Massachusetts)with various species of pond-breeding amphibians

in southern Rhode Island. Our goal inthis article is to give readers a sense ofwhat we believe are the key manage-ment issues that people working in thegolf turf profession need to understand.

70

Figure 2Relationship between pond-breeding amphibian species richness (mean number ofspecies per pond, + standard error) and hydroperiod in 137 ponds in Rhode Island.Ponds with a short hydroperiod are defined as those drying in June or July,medium-

hydroperiod ponds dry inAugust or September, long-hydroperiod ponds dry inOctober or November, and permanent ponds never dry during the year.

Short

HYDROPERIOD OFBREEDING PONDSTo assess pond-breeding amphibian useof ponds on golf courses, we used dip-nets to sample 59 ponds at 32 golfcourses in Rhode Island, Connecticut,and Massachusetts during the springand early summer of 1999. Most pondson golf courses had either green frogs(Rana clamitans) or American bullfrogs(R. catesbeiana),with few other speciesdetected (Figure 1).This was primarilybecause most of the ponds we sampledon golf courses were permanent. Inaddition, many ponds on golf courseswe sampled had fish.

During 2000 and 2001, we used dip-nets to sample amphibian communitystructure at 137 randomly selectedponds across the urbanization gradientin Rhode Island. We found that hydro-period (i.e., the number of days withstanding water in the pond basin) wasone of the most important variablesdetermining amphibian communitystructure. Ponds with a long hydro-period (drying in October orNovember) tended to have the mostspecies (Figure 2), while ponds with ashort or medium hydroperiod (dryingannually from June through September)tended to have unique species notfound in permanent ponds.

For example, wood frogs (Ranasylvatica) and marbled salamanders(Ambystoma opacum) were usually de-tected only in ponds that dried beforeSeptember. Tadpoles of both species areamong the first to complete meta-morphosis, typically emigrating fromponds by early July (5). In contrast, tad-poles of American bullfrogs were foundonly in permanent ponds, and greenfrogs were more likely to be found inlong or permanent hydroperiod ponds.Both these species have tadpoles thattake much longer to complete meta-

SpottedSalamander

Permanent

AmericanToad

PickerelFrog

Medium LongHydroperiod

Green FrogBullfrog

o

o

4.5

4.0

3.5VlQ)

3.0'uQ)a..Vl- 2.50l...Q)E 2.0::::lc:c: 1.5~Q)

:L1.0

0.5

20

10

60

50

Vl-g 40oa..-o~ 30

28 G R E ENS E C TI 0 N R E COR D

This natural pool is anexample of the areasused by pond-breedingamphibians in theNortheast. The pondusually dries everySeptember and has fivefrog species and threesalamander species thatuse it as a breeding site.

morphosis (two years for bullfrogs andone year for green frogs), thus requiringponds with longer hydroperiods forsuccessful reproduction.

The take-home message from thisresearch is that if you want to maintainthe entire amphibian community onyour golf course, you have to maintainponds with a variety of hydroperiodson or adjacent to the course. It is criticalto have ponds that dry annually becausesome species only use seasonallyflooded ponds (10).

In addition, ponds should not bestocked with fish. Fish are majorpredators of amphibian eggs and larvae,which is why many species of amphib-ians tend to avoid ponds with fish.Finally, we have found that the vege-tation in ponds can be important tocertain species. For example, wood frogstend to have larger populations inponds with extensive coverage ofbuttonbush (Cephalanthus occidentalis),whereas spring peepers tend to thrivein ponds with no canopy closure.

EFFECT OF GRASS HEIGHTAND HABITAT ONMOVEMENTSTo assess whether grass height affectsmovement behavior of amphibians,

during the 1998 field season we con-structed two square pens (50 ft. on eachside) on a four-hectare section of bent-grass, which is used by the TurfgrassGroup at the University of RhodeIsland for a variety of experiments. Theperimeter of our experimental pens wasencircled with O.5m-tall silt fence. Thepens were subdivided into four quarters(25 ft. per side). Each quarter (randomlyselected) was mowed to various grassheights (.25", .5", 1", and> 1"-2"-5").

All experiments were conducted onrainy nights, when amphibians werelikely to move. During the experiment,an individual amphibian (wood frog,American toad, green frog, bullfrog, orpickerel frog) was placed in the centerof the array, and its movements weremonitored for a three-minute period.We also constructed another set ofexperimental pens at ecotones betweena forest and mowed lawn < .5" and aforest and lawn.

During grass height experiments, wefound no evidence that frogs preferredany grass height during the three-minute trials, during which their move-ments were random with respect tograss height. This suggests that grassheight, at least in the height range wequantified, that is typical of current golf

courses in North America does nothinder or enhance amphibian move-ments. This is true for the species wesampled, but we did not have the oppor-tunity to investigate any salamanders orsome frogs (spring peepers, gray treefrogs, and wood frogs), whose move-ments could be affected by grass height.However, we did find that amphibians(frogs, in this case) preferred to moveinto forested habitats rather than eitherturf or barren areas. In both cases, theevidence shows that wooded habitatswere preferred over turf or barrenground. This suggests that amphibianspreferred forested habitat as movementcorridors over open habitats such asfairways.

EFFECT OF TURF ONDISPERSAL OF AMPHIBIANSFROM A SERIESOF PONDSWe also conducted an observationalstudy to assess the influence of habitaton movement behavior of amphibians.From 1998-2000, we monitored theimmigration and emigration of adultsand emigration of metamorphs across awooded landscape fragmented by turffields. We documented considerablevariation within and among species intheir initial departure direction from

J U L Y - AU GUS T 2003 29

breeding ponds, which suggests thathabitat near breeding ponds has littleinfluence on movement patterns.

Farther from breeding ponds, adultsof species that reside in forested habi-tats during the non-breeding seasonoccurred less often at an ecotone be-tween a turf field and woodland (e.g.,wood frog, spotted salamander, springpeeper, gray treefrog, and red-spottednewt). In contrast, species that winterin aquatic habitats readily cross theturf-woodland edge (e.g., green frog,

species were affected by small-scalevegetation removal.

Overall, these results suggest thathabitat associations of pond-breedingamphibian species during migration aresimilar to those during the non-breed-ing season. Species that reside duringthe non-breeding season and winter inforest habitats (e.g., wood frog, marbledand spotted salamander, red-spottednewt, spring peeper, gray treefrog) tendto migrate through forested habitats andavoid open expanses, such as fairways.

prefer permanent ponds for successfulreproduction. In addition, both speciesreadily cross open habitats, such as fair-ways, to reach breeding ponds/winter-ing sites.

Other researchers have documentedpatterns similar to those we found inRhode Island. For example, deMaynadierand Hunter (1,2), working in the forestsof Maine, classified wood frogs as"management sensitive" because theyavoided traveling across clear cuts. Adultspotted salamanders also generally avoid

Wood frog eggmasses attach to

button bush shrubsin the center of a

small pond inwestern Rhode

Island.An estimatedI,500 egg masses

were in this pond,covering a five-footdiameter area. Both

wood frogs andspotted salamanders

usually attach eggmasses to woody

vegetation.

American bullfrog, pickerel frog). Meta-morphs of most species tended to behabitat generalists during migration,whereas adults tended to exhibit morehabitat selection.

To further test the influence ofhabitat on migration, we removed theoverstory and understory in five smallpatches (10m by 40m) in a woodlandwhere we had been monitoring move-ments for the previous two years. Basedon this experiment, we found thatmovement patterns of at least four

30 GREEN SECTION RECORD

This is particularly true for adultamphibians that avoid open habitatsmore than young of the year.

In contrast, species that winter inaquatic habitats such as streams orponds (e.g.,American bullfrog, greenfrog, and pickerel frog) are less likely tobe impacted by forest fragmentationbecause they are willing to cross openhabitats. This explains why ponds ongolf courses tended to be dominated bythis latter group of species. As mentionedearlier, both bullfrogs and green frogs

openings in woodlands, although otherresearchers (3) suggested that migratorymovements by spotted salamanderswere unaffected by vegetation ortopographic structure.

So what does this mean for golfcourse designers and superintendentsof existing courses? Available evidencesuggests the habitat characteristics of agolf course can impact movementbehavior of some species of pond-breeding amphibians. In New England,species that winter in forested habitats

A young red-spotted newt,often referred to as an eft, isvery small when it emerges

from the breeding pond area.Efts remain on land for three

to seven years, often wanderinggreat distances before returning

to the ponds to breed.

An adult male wood froglounges near the water'ssurface. Males arrive atbreeding ponds in earlyMarch and actively call toattract females to the pond.This species spends most ofthe year in forested habitatwithin 200 yards of breedingponds. In experimentsconducted in Rhode Island,this species was reluctanttoo cross open expanses ofturf, such as fairways.

University of Rhode Islandscientists are studyingamphibian movements ongolf courses in theNortheast to gain a betterunderstanding of howwater feature design canhelp improve habitat forthese wood frogs and otheramphibians.

appear to be the most affected byhabitat fragmentation. Thus, designersshould maximize the amount of forestcover on a course while simultaneouslycreating forest travel corridors betweenbreeding ponds and non-breedinghabitat.

The species most sensitive to habitatfragmentation all primarily breed inponds that dry annually. These pondsare best identified during surveys con-ducted in March and April when theyare most likely to be flooded. If aseasonally flooded pond is found, stepsshould be taken to maintain a forestedbuffer around it. No definitive guide-lines are available on how wide thisforest buffer should be, but Semlitsch(8) estimated that approximately 95%of the population of mole salamandersusually occurs within 196 yards of thepond.

Maintaining such a wide forest bufferaround all seasonally flooded ponds ona course may be impractical.Yet, alter-

native management steps could includemaximizing the forest/shrub bufferaround ponds. This includes creatingforested travel corridors that allowmovement from seasonally floodedponds and their associated buffer tolarge patches of potential non-breedinghabitat.

REFERENCES1. deMaynadier, P. G., and M. L. Hunter. 1998.

Effects of silvicultural edges on the distribu-tion and abundance of amphibians in Maine.Conservation Biology, 12:340-352.

2. deMaynadier, P. G., and M. L. Hunter. 1999.Forest canopy closure and juvenile emigrationby pool-breeding amphibians in Maine.Journalif Wildlife Management, 63:441-450.

3. Douglas, M. E., and B. L. Monroe,Jr. 1981.Acomparative study of topographic orientationin Ambystoma (Amphibia: Caudata). Copeia,1981:160-463.

4. Lehtinen, R. M., S. M. Galatowitsch, andJ. R. Tester. 1999. Consequences of habitat lossand fragmentation for wetland amphibianassemblages. Wetlands, 19:1-12.

5. Paton, P.Wc., andW B. Crouch III. 2002.Using phenology of pond-breeding amphib-ians to develop conservation strategies.Conservation Biology, 18: 194-204.

6. Sernlitsch, R. D. 1981. Terrestrial activity andsummer home range of the mole salamander(Ambystoma taloideum). Canadian Journal ifZoology, 59:315-322.

7. Sernlitsch, R. D. 1998. Biological delineationof terrestrial buffer zones for pond-breedingsalamanders. Conservation Biology, 12:1113-1119.

8. Sernlitsch, R. D. 2000. Principles for manage-ment of aquatic-breeding amphibians. Journalif Wildlife Management, 64:615-630.

9. Snodgrass, J.W, M. J. Komoroski, A. L. BryanJr., and J. Burger. 2000. Relationships amongisolated wetland size, hydroperiod, andamphibian species richness: implications forwetlands regulations. Conservation Biology,14:414-419.

PETER PATON, PH.D., is associateprcifessorin the Department ifNatural ResourcesScience at the University ifRhode Island,and ROBERT EGAN is a wetland andwildlife ecologistat ENSR in Ulesiford,Massachusetts.

JULY-AUGUST 2003 31