Effect of a Long-Term Fire Retardant(Fire Trol 934®) on the Germination of NineMediterranean-Type Shrub Species

Alberto Cruz, Marian Serrano, Esther Navarro, Belen Luna, Jose M. Moreno

Departamento de Ciencias Ambientales, Facultad de Ciencias del Medio Ambiente,Universidad de Castilla-La Mancha, E-45071, Toledo, Spain

Received 8 December 2004; revised 26 May 2005; accepted 26 June 2005

ABSTRACT: Fire Trol 9341 is a long-term fire retardant commonly used in fire prevention and extinction.Our objective was to determine the effect of this chemical on seed germination of nine plant species fromMediterranean-type shrublands, where these chemicals are potentially used. Seeds were exposed to fivedifferent Fire Trol concentrations, (0 (control) to 10%, on a log scale) and monitored in a germinationchamber for nine weeks. Seeds from four Cistus species were subjected to an additional heat treatmentthat simulated thermal scarification caused by fire. Retardant exposure caused a significant decrease intotal germination in all species, and exposure to the highest Fire Trol concentration (10%) resulted in com-plete inhibition of germination. However, the sensitivity to Fire Trol varied across species and this differen-tial species sensitivity may potentially lead to different impacts in the soil seed banks depending onwhether sites are burned or unburned. Exposure to Fire Trol 934 may affect recruitment of shrubland spe-cies particularly during dry autumns, due to limited leaching of these chemicals from the soil surface. Con-sequently, its use should be avoided in sites where particularly sensitive plant species are present.# 2005 Wiley Periodicals, Inc. Environ Toxicol 20: 543–548, 2005.

Keywords: Cistus; Erica; Halimium; nitrogenous compounds; Lavandula; seed

INTRODUCTION

Long-term fire retardants are chemical products that, when

applied to the vegetation alter the combustion reaction and

make fire propagation more difficult for long periods of time

after an application. Because of their effectiveness in reduc-

ing the rate of fire spread, fire retardants are widely used for

the control and management of wildland fires. Such is the

case in the Mediterranean-type climate areas, which are

prone to fire. Despite the common use of fire retardants, rela-

tively little information is available on their potential envi-

ronmental effects, including impacts on plant germination.

Long-term fire retardants usually contain ammonium

salts that may be potentially toxic to biota (Gimenez et al.,

2004). Acute toxicity tests have shown harmful effects of

retardants on aquatic organisms, and ammonia is the com-

ponent which has most impact on these organisms

(Gaikowsky et al., 1996; McDonald et al., 1997; Buhl and

Hamilton, 1998, 2000; Little and Calfee, 2002). The

impacts of the liberation of these chemicals on terrestrial

environments are surprisingly poorly known. The presence

of nitrogenous compounds in the retardant solutions sug-

gests that moderate retardant applications on terrestrial eco-

systems would cause effects similar to those produced by

fertilizers. Indeed, an increase in forage yield (Larson and

Duncan, 1982) or a reduction in plant species richness

(Larson et al., 1999) are responses observed after retardant

application in grassland communities. Damage to leaves,

presumably caused by exposition to ammonium salts, has

also been reported (Bradstock et al., 1987). However, the

Correspondence to: A. Cruz; e-mail: Alberto.Cruz@uclm.es

Contract grant sponsor: ERAS project.

Contract grant number: EVG1-2002-00019.

Published online in Wiley InterScience (www.interscience.wiley.com).

DOI 10.1002/tox.20143

�C 2005 Wiley Periodicals, Inc.

543

few available studies do not permit to obtain enough infor-

mation about the effects of fire retardant use on the mecha-

nisms controlling plant dynamics.

In particular, no information is available on the effects

of these chemicals on seed germination of terrestrial plants.

The ammonium salts present in most retardant solutions

can alter the germinative response of many species (Baskin

and Baskin, 1998). Germination is often stimulated by addi-

tion of moderate concentrations of nitrogenous compounds

(Thanos and Rundel, 1995; Keeley and Fotheringham,

1998; Cruz et al., 2003; Perez-Fernandez and Rodrıguez-

Echeverrıa, 2003). Thus, we hypothesized that a stimula-

tory effect might be observed in seeds exposed to moderate

concentrations of retardant solutions. However, exposure to

high concentrations of nitrogenous compounds reduced ger-

mination in several species (Perez-Fernandez and Rod-

rıguez-Echeverrıa, 2003). As not all plant species from a

community will be equally responsive to retardant solu-

tions, their effects may differ depending on the concentra-

tion and sensitivity of the species involved.

The objective of this study is to document the effect of a

commonly used long-term fire retardant (Fire Trol 9341)

on seed germination of several plant species which are

abundant in shrublands from Mediterranean-type climate

areas of western Spain. Such shrublands represent a consid-

erable portion of land surfaces burned every year in the

Mediterranean Basin (Moreno et al., 1998) and thus they

are susceptible to be treated with retardants. Knowledge

about the impact of these chemicals on germination of

shrubland species is necessary to evaluate the environmen-

tal impact of the use of these chemicals for fire control.

MATERIALS AND METHODS

Seeds from nine shrub species were collected in Quintos de

Mora Station (Central-western Spain) during spring and

summer. Selected species were Cistus ladanifer L., C. pop-ulifolius L., C. laurifolius L., C. salviifolius L., Halimiumatriplicifolium (Lam.) Spach, H. umbellatum (L.) Spach,

Lavandula stoechas subsp. pedunculata (Miller) Samp. ex

Rozeira, Erica arborea L., and E. scoparia L. All were

widely represented in most shrubland communities of the

Mediterranean-type climate areas of central-western Spain.

Seeds were maintained in paper bags during several

weeks until the start of the germination experiment, in early

autumn. Seeds were set to germinate in Petri dishes (55 mm

diameter) on moistened filter paper (Whatman No. 1). The

dishes were placed in a germination chamber (IBERCEX,

ASL) under a photoperiod and temperature regime of 16 h

of light at 208C and 8 h of darkness at 148C. Seeds were

subjected to one of five retardant treatments of 0, 0.01, 0.1,

1, and 10% by adding 1.2 ml of the appropriate Fire Trol

9341 solution to the Petri dish. The Fire Trol solutions

were prepared by diluting the original product, as served by

the manufacturer, with distilled water. The liquid concen-

trate of Fire Trol 934 consists mainly of ammonium-poly-

phosphate complexes which act as the active agent; other

components such as sodium ferrocyanide, which serves as a

corrosion inhibitor, and Fe, Mg, Si, Na, and SO4 are present

in traces. In contrast to other fire retardants, Fire Trol 934

does not contain color additives to mark drop sites and gum

or clay thickeners to increase the viscosity of the product

(Patrice Oegema, ITRATECH, Aix-en-Provence, France,

personal communication).

Cistus species are known to increase germination after

exposure to heat (Thanos and Georghiou, 1988; Trabaud

and Oustric, 1989; Valbuena et al., 1992). Consequently,

we planned the application of a heat treatment to the seeds

of the four Cistus species in order to evaluate the interac-

tion between retardant exposure and heat. The heat treat-

ment consisted of exposing the seeds to 908C for 5 min in

an oven. Ten (Cistus spp.) or eight (the other species) repli-cate dishes were used for each treatment, and 25 seeds were

placed in each dish.

Germination was monitored every two–three days by

observing radicle emergence. Monitoring lasted for nine

weeks, after that the number of newly germinated seeds

were negligible. Total germination was determined as the

percentage of seeds germinated at the end of the monitoring

period. For each species, differences in total germination

among retardant concentrations were tested by analysis of

variance (ANOVA); for Cistus species, differences in

total germination between heat application and retardant

concentration and their interaction were tested by two-

way ANOVA. The level of significance was set at �� 0.05. A posteriori Bonferroni tests were performed to

determine differences between treatment means. Prior to

analysis, the germination data were arcsin transformed in

order to improve normality. Assumptions of normality and

homocedasticity were not met in all cases. However,

according to Underwood (1997), these assumptions are not

strictly necessary for a robust analysis of variance and can

be obviated if designs include many treatments and are

balanced.

To obtain an index to compare sensitivity of germination

to Fire Trol across species, we calculated the retardant con-

centration causing a 50% reduction of control germination

(G50) from linear regression of germination (arcsin trans-

formed) on Fire Trol concentration (log).

RESULTS

Retardant application significantly affected total seed ger-

mination of all species, where increasing Fire Trol concen-

trations caused a general decrease in germination percen-

tages (Fig. 1). However, the sensitivity to retardant solu-

tions varied across species. Erica and Lavandula species

544 CRUZ ET AL.

showed high germination percentages and a relatively high

G50 values, greater than 1% (Table 1). Indeed, Fire Trol

concentrations of 1% were needed to cause a significant

reduction in germination (Fig. 1). In Cistus and Halimiumspecies, germination in the absence of heat was low (Fig. 1).

For these species, G50 values of unheated seeds varied from

greater than 1% (C. ladanifer, C. populifolius, C. salviifolius)to lower than 0.66% (Halimium, C. laurifolius) (Table 1).

Fire Trol concentrations of 1 or even 10% were needed to

cause a significant reduction in germination of Halimiumand Cistus unheated seeds (Fig. 1). The heat treatment sig-

nificantly increased germination in all Cistus species and

affected their response to retardant (Fig. 1). Heated seeds

had generally lower G50 values than unheated seeds (with

the exception of C. laurifolius, Table 1). Indeed, Fire Trol

concentrations of only 0.1% were necessary to cause a sig-

nificant reduction in germination of C. ladanifer and C. sal-viifolius heated seeds (Fig. 1).

DISCUSSION

Retardant application caused a significant reduction in

total seed germination for the group of species studied.

These results partially contradict our initial hypothesis as

seed germination was not stimulated after exposure of

moderate Fire Trol 934 concentrations, taking into account

that the retardant contains nitrogenous salts. Exposure to

moderate ammonia or nitrate concentrations increases ger-

mination in many species, including species from genus

Erica (Cruz et al., 2003) and Cistus (Perez-Fernandez and

Rodrıguez-Echeverrıa, 2003), as those included in this

study. However, this was not observed in this study.

Keeley and Fotheringham (1998) suggested that stimula-

tion of germination by nitrogenous compounds were due

to acidic pH, and that the stimulatory effect of nitrogenous

compounds on seeds failed when pH was buffered to neu-

trality. The pH of Fire Trol 934 solutions was about 6.5,

Fig. 1. Mean (6SE) of total germination (%) of nine shrub species in response to Fire Trol934 concentration. Dissimilar letters reflect significant differences among retardant levelsin a posteriori tests (P < 0.05). In Cistus solid squares and open circles represent heatedand unheated seeds, respectively. Enclosed are significances of ANOVA tests for theeffects of retardant (R), heat (H), and retardant�heat interaction (R�H).

545EFFECT OF FIRE TROL 934 ON SEED GERMINATION

and this may explain why germination was not stimulated

in any species by moderate Fire Trol 934 concentrations.

In contrast, exposure to a relatively high retardant concen-

tration clearly reduced germination in all species. This

may be the result of the high salt concentration, reducing

the osmotic potential of the surrounding solution, which

has been shown to negatively affect seed germination

(Henig-Server et al., 1996). Alternatively, some of the

chemical substances formulated in the retardant may be

toxic to seeds. Fire Trol 934 contains ammonium-poly-

phosphates; sodium ferrocyanide and traces of Fe, Mg, Si,

Na, and SO4. (Patrice Oegema, ITRATECH, Aix-en-Pro-

vence, France, personal communication). At high concen-

trations, ammonia is potentially toxic to plants. Bradstock

et al. (1987) attributed leaf damage and death after retard-

ant exposure to the effect of ammonium salts. However,

the mechanism involved in the negative effect of Fire Trol

934 cannot be inferred from this study and needs further

verification.

Despite the general effect caused by the retardant on

germination, not all species were equally sensitive to

retardant exposure. It is clear that the highest concentra-

tion of Fire Trol 934 used in this study (10%) caused com-

plete or nearly complete inhibition of germination. How-

ever, at intermediate concentrations retardant exposure

reduced germination in some species but not in others.

The shapes of the germination–Fire Trol concentration re-

sponse curves varied among species, being almost linear

in some species (most Cistaceae species) or showing an

abrupt decrease in germination after reaching a certain re-

tardant concentration (Erica, Lavandula). This differential

sensitivity may potentially lead to different impacts in soil

seed banks of areas that remain unburned after being

sprayed with Fire-Trol 934. For instance, the sensitivity of

Cistus species, an essential component of shrubland com-

munities in the Mediterranean Basin, was much higher in

heated than in unheated seeds. Cistus species are known to

have hard coats that need to be broken by high tempera-

tures or other agents (Thanos et al., 1992). Hard coated

seeds, like those of Cistus or Halimium species, are rather

impermeable to water, showing low germination rates in

the absence of scarification. These seeds, in absence of

heat or other scarification factors, would not imbibe the

retardant solution and remain dormant; thus they would be

less affected by direct exposure to the retardant. On the

other hand, seeds of Erica or Lavandula did not need heat

stimuli to germinate: the mere presence of water activated

their germination, making them more susceptible to direct

contact with the retardant. On the basis of these results, it

can be concluded that the effect of Fire Trol 934 depends

not only on the species present but on the presence or

absence of fire. When using retardants in fire fighting, we

can expect that some sites sprayed with retardants will

remain unburned but others will burn to some extent. This

may generate patches of burned and unburned areas that

contain some retardant concentrations. In unburned

patches we would expect a greater impact of retardants on

Erica and Lavandula recruitment; however, in burned

patches the relative effect would be much greater on Cis-tus recruitment.

The results obtained in this work indicate a potential

negative effect of the use of long-term fire retardants such

as Fire Trol 934 in Mediterranean-type-shrublands. It must

be noted that the recommended field-use of Fire Trol 934

for shrubland-type communities is 2.5–3 Lm�2 concen-

trated at 20%, which is usually sprayed from aerial means.

After spraying, certain amount of the retardant solutions

will be intercepted by the vegetation and litter. The concen-

trations that fire retardants may reach in the soil are difficult

to calculate, but it is presumed that shortly after application,

TABLE I. Fire Trol concentration (%, v/v) causing a 50% reduction of maximum germination (G50) of severalspecies growing in Mediterranean environments, heated or unheated. Lower and upper limits were obtainedby applying 95% confidence intervals of the intercepts from linear regressions

Species R2 from Linear Regression Fire Trol G50 (%) 95% Lower Limit 95% Upper Limit

C. ladanifer (Unheated) 0.99 1.16 0.78 1.64

C. laurifolius (Unheated) 0.93 0.66 0.00 1.90

C. populifolius (Unheated) 0.93 1.50 0.40 3.44

C. salviifolius (Unheated) 0.86 1.36 0.02 4.42

C. ladanifer (Heated) 0.93 0.76 0.01 2.07

C. laurifolius (Heated) 0.98 1.50 0.85 2.39

C. populifolius (Heated) 0.93 0.53 0.00 1.65

C. salviifolius (Heated) 0.86 0.31 0.00 1.99

E. arborea 0.96 1.68 0.73 3.14

E. scoparia 0.98 1.22 0.63 2.04

H. umbelatum 0.94 0.51 0.00 1.51

H. atriplicifolium 0.84 0.63 0.00 2.97

L. pedunculata 0.88 1.11 0.00 3.53

546 CRUZ ET AL.

the retardant solution might be high enough to cause the

effects observed in this work. Germination of Mediterra-

nean-type species usually occurs in response to autumn

rains, either in unburned (Ortega et al., 1997) or recently

burned (Quintana et al., 2004) environments. Autumn rains

will normally cause the complete leaching of water-soluble

chemical retardants, like Fire Trol 934, from the soil sur-

face. Consequently, the impact of retardants on seeds dur-

ing wet, normal autumns might to be low. However, during

dry autumns, in which some storms occur in early autumn

followed by dry periods, leaching of chemicals would be

less intense and moderately high concentrations could

remain in the soil surface layer. Taking into account the

great among-year variability in precipitation of Mediterra-

nean-type climate areas, such a situation is likely to occur.

Consequently, the impact of Fire Trol 934 on recruitment

of plants from Mediterranean shrublands may vary from

year to year depending on autumn rainfall patterns.

In conclusion, Fire Trol 934 may significantly affect ger-

mination in areas where high concentrations of these chem-

icals may persist for a long time. Thus, its overall impacts

may be relatively small and limited and localized to certain

weather conditions. However, it is not known to what

extent the viability of seeds are affected by retardant expo-

sure. Seeds exposed to concentrated Fire Trol solutions

may not germinate but remain alive and eventually germi-

nate once the retardant is leached out of the soil. Angeler

et al. (2004) showed that retardant pre-application germina-

tion success of Typha domingensis could be recovered after

the retardant was eliminated. Although extrapolation of

results from laboratory to field conditions must be made

with caution, it is clear that Fire Trol 934 may affect seed

germination, thus its effects on plant populations cannot be

discarded. However, further studies that incorporate field

exposures are needed to better assess the environmental

impact of fire retardants and to give recommendations for

management in cases where sensitive species may be

affected.

We thank Angel Velasco for his technical support. We also

thank to the personnel of Quintos de Mora Station in which seeds

were collected. Comments by two anonymous reviewers greatly

improved the final manuscript.

REFERENCES

Angeler DG, Rodrıguez M, Martın S, Moreno JM. 2004. Assess-

ment of application-rate dependent effects of a long-term fire

retardant chemical (Fire-Trol 9341) on Typha domingensisgermination. Environ Int 30:375–381.

Baskin CC, Baskin JM. 1998. Seeds: Ecology, Biogeography, and

Evolution of Dormancy and Germination. San Diego: Academic

Press.

Bradstock R, Sanders J, Tegart A. 1987. Short-term effects on the

foliage of a eucalypt forest after an aerial application of a chem-

ical fire retardant. Aust For 50:71–80.

Buhl KJ, Hamilton SJ. 1998. Acute toxicity of fire-retardant and

foam-suppressant chemicals to early life stages of chinook sal-

mon (Oncorhynchus tshawytscha). Environ Toxicol Chem 17:

1589–1599.

Buhl KJ, Hamilton SJ. 2000. Acute toxicity of fire-control chemi-

cals nitrogenous chemicals and surfactants to rainbow trout.

Trans Amer Fish Soc 129:408–418.

Cruz A, Perez B, Velasco A, Moreno JM. 2003. Variability in seed

germination at the interpopulation, intrapopulation and intrain-

dividual levels of the shrub Erica australis in response to fire-

related cues. Plant Ecol 169:93–103.

Gaikowsky MP, Hamilton SJ, Buhl KJ, McDonald SF, Summers

CH. 1996. Acute toxicity of three fire-retardant and two fire-

suppressant foam formulations to the early life stages of rain-

bow trout (Oncorhynchus mykiss). Environ Toxicol Chem 15:

1365–1374.

Gimenez A, Pastor E, Zarate L, Planas E, Arnaldos J. 2004. Long-

term forest fire retardants: a review of quality effectiveness

application and environmental considerations. Int J Wild Fire 13:

1–15.

Henig-Server N, Eshel A, Neeman G. 1996. pH and osmotic

potential of pine ash as post-fire germination inhibitors. Physiol

Plant 96:71–76.

Keeley JE, Fotheringham CJ. 1998. Mechanisms of smoke-

induced seed germination in a postfire chaparral annual. J Ecol

86:27–36.

Larson DL, Duncan DA. 1982. Annual grassland response to fire

retardant and wildfire. J Range Manag 35:700–703.

Larson DL, Newton WE, Anderson PJ, Stein SJ. 1999. Effects of

fire retardant chemical and fire suppressant foam on shrub steppe

vegetation in northern Nevada. Int J Wild Fire 9:115–127.

Little EE, Calfee RD. 2002.Environmental persistence and toxic-

ity of fire retardant chemicals Fire-Trol1 GTS-R and Phos-

Chek1 D75-R to fathead minnows. US Forest Service Techni-

cal Report.

McDonald SF, Hamilton SJ, Buhl KJ, Heisinger JF. 1997.

Acute toxicity of fire-retardant and foam-suppressant chem-

icals to Hyalella azteca. Environ Toxicol Chem 16:1370–

1376.

Moreno JM, Vazquez A, Velez R. 1998. Recent history of

forest fires in Spain. In: Moreno JM, editor. Large Forest Fires.

Leiden, The Netherlands: Backhuys Publishers. p 159–185.

Ortega M, Levassor C, Peco B. 1997. Seasonal dynamics of Medi-

terranean pasture seed banks along environmental gradients.

J Biogeog 24:177–195.

Perez-Fernandez MA, Rodrıguez-Echeverrıa S. 2003. Effect of

smoke charred wood and nitrogenous compounds on seed ger-

mination of ten species from woodland in Central-Western

Spain. J Chem Ecol 29:237–251.

Quintana JR, Cruz A, Fernandez-Gonzalez F, Moreno JM. 2004.

Time of germination and establishment success after fire of

three obligate seeders in a Mediterranean shrubland of central

Spain. J Biogeogr 31:241–249.

547EFFECT OF FIRE TROL 934 ON SEED GERMINATION

Thanos CA, Georghiou K. 1988. Ecophysiology of fire-stimu-

lated seed germination in Cistus incanus ssp. creticus (L.)

Heywood and C. salvifolius L. Plant Cell and Environ 11:841–

849.

Thanos CA, Rundel PW. 1995. Fire-followers in chaparral: nitro-

genous compounds trigger seed germination. J Ecol 83: 207–

216.

Thanos CA, Georghiou K, Kadis C, Pantazi C. 1992. Cistaceae: a

plant family with hard seeds. Isr J Bot 41:251–263.

Trabaud L, Oustric J. 1989. Heat requirements for seed germina-

tion of three Cistus species in the garrigue of southern France.

Flora 183:321–325.

Underwood AJ. 1997. Experiments in Ecology. Their Logical

Design and Interpretation Using Analysis of Variance.

Cambridge, UK: Cambridge University Press.

Valbuena L, Tarrega R, Luis E. 1992. Influence of heat on seed

germination of Cistus laurifolius and Cistus ladanifer. Int J

Wild Fire 2:15–20.

548 CRUZ ET AL.