Soil Fertility Effects on Population Dynamics
of Soybean Insect Pests
F.M. Rhoads, J. E. Funderburk, and I. D. Teare
171MAY 29 1990
;ivi'ersity of Florida
J. E. Funderburk, Entomology and Nematology Dep.; I. D. Teare, Agronomy Dep.;
and F.M. Rhoads, Soils Dep.; North Fla. Res. and Educ. Ctr. Contribution from
the Institute of Food and Agricultural Sciences, Univ. of Fla., Route 3 Box
4370, Quincy, FL 32351. Research Report NF 90-12.
The use of fertilizers can influence the injury to a crop from arthropod
pests, through alterations in crop growth or nutritional levels (Herzog and
Funderburk, 1986). Many cotton pests are affected by soil fertility levels.
Enhancement of succulent cotton growth through fertilization renders the crop
more attractive to populations of the cotton aphid, Aphis gossypii Glover
(McGarr, 1942, 1943); the cotton fleahopper, Pseudamatamoscelis seriatus
(Reuter) (Adkisson, 1957); and the cotton bollworm, Heliothis zea (Adkisson,
1958). In short-season production areas, increasing levels of soil fertility
delay the fruiting period of the crop, thereby reducing the potential for
escape from boll weevil, Anthonomus grandis Boheman, injury (Walker et al.,
Soil fertility effects upon soybean [Glycine max (L.) Merr.] yield
(Rhoads and Barnett, 1990) have been reported for the Southeast, but the
influence of soil fertility on soybean insect pests has not been documented.
Musick (1985) and Herzog and Funderburk (1986) concluded that each crop and
pest situation must be evaluated individually and pest control decisions made
for each specific geographical location. The purpose of this study was to
determine the effect of soil fertility on population dynamics of several
foliage-inhabiting soybean pests, including larval velvetbean caterpillars
(VBC), Anticarsia gemmatalis (Hubner), and nymphal and adult southern green
stink bugs (SGSB), Nezara viridula (L.). These findings should aid in
implementation of cultural control strategies for these major insect pests of
soybean in the southeastern U.S.
:MATERIALS AND METHODS
Soybean were grown on a Norfolk loamy sand (fine, loamy, siliceous,
thermal Typic Kandiudult) at Quincy, FL. Soil samples were collected from
each plot in Feb of each year. This experiment was conducted in 1986 and 1987
at the North Fla. Res. and Educ. Ctr. on land which was previously used for
fertility research. Previous soil treatments [ P at 0, 25, 50, 100 lb. AI
applied annually as triple superphosphate for 6 yr; K at 0, 190, and 380 lb.
A applied annually as KCL for 6 yr; N at 0, 50, and 100 each year as
ammonium nitrate] are described in Rhoads and Barnett (1985). Each year of
the experiment, ten cores (1 X 6 inch) were taken from each plot in a
criss-cross pattern, composite, air-dried, and ground to pass a 0.1 inch
sieve for analysis. Melich I soil extractant was used. Soil-test levels
(ppm) in relation to treatment code are shown in Table 1.
Soybean followed wheat in 1986. No P was applied to wheat or soybean
because residual levels of P were adequate as indicated by soil test (Table
Potassium was applied in 1986 as follows:
to wheat, K1 = 0, K2 = 84 lb. K AI, K, = 166 lb. K A1;
to soybean K1 = 0, K2 = 42 lb. K A1, K3 = 84 lb. K A .
Magnesium was applied in 1986 to wheat only as follows:
Mg1 = 0, Mg1 = 60 lb. Mg A Mg3 = 120 lb. Mg A.
Soybean followed snap bean and cabbage in 1987. No P or Mg was applied to
snap bean, cabbage or soybean in 1987.
Potassium was applied in 1987 to soybean only as follows:
to soybean K, = 0, K2 = 42 lb. K A1, K3 = 84 lb. K A .
A 2-row cone planter was used to plant Braxton soybean at a planting rate
of 40 lb. Al at 1 inch soil depth on 11 June 1986 and 10 June 1987. In 1986
and 1987, 1 inch water A1' was applied preplant and at intervals during the
growing season when tensiometers placed at 6 inchessoil depth reached 0.02
MPa. Insecticides were not applied at any time during the experiment.
Nymphal and adult population densities were estimated on 8 calendar
dates/Julian date in 1986 (7-1/182, 7-12/193, 7-23/204, 8-5/217, 8-14/226,
8-26/238, 9-9/252, 9-22/265) and on 7 calendar dates/Julian date in 1987
(7-8/189, 7-22/203, 8-7/219, 8-18/230, 9-2/245, 9-14/257, 9-29/272). Sampling
was begun at early vegetative stage (V4) and continued until late seed stages
(R6) in both years.
Insect sampling was carried out as described by Kogan and Pitre (1980) and
Todd and Herzog (1980). All plots were sampled on each sampling date by
beating the plants on both sides of the row into a 36 inch square ground cloth
placed between the rows. Three samples were taken per plot on each sampling
date. Also, adjacent plants were searched at their base and the soil surface
was visually examined for VBC and SGSB.
The influence of fertilizer treatment on population densities and
population cycles of VBC larvae, SGSB nymphs, and SGSB adults were evaluated
by ANOVA. Data from each growing season were analyzed separately. The design
was a split plot over time (Steel and Torrie, 1960). The main effect of
fertilizer treatment compared the influence of fertilizer treatment on
seasonal population density. Orthogonal comparisons were used to define
fertility treatment differences. The interaction of date X treatment compared
the influence of fertilizer treatment on seasonal population cycles.
RESULTS AND DISCUSSION
A description of soybean yield in relation to soil fertility levels of P,
K, and Mg in 1986 and 1987 is given in Table 2. In 1986, mean yields were
significantly greater for the P2 P3, and P4 levels than the P, level, but
significantly similar for the P2 P3, and P4 levels. Mean yields in 1987 were
significantly greater at the P3 and P4 fertility levels than at the Pi and P,
levels, with yields significantly greater at the P2 level than at the P,
level. Yields were significantly greater both years for the K2 and K levels
than the K, level, with yields similar for the K2 and K, levels. In 1986,
yields were significantly greater for the Mg3 level than the Mg1 and Mg2
levels, with yields greater for the Mg2 than the Mg1 level. Yields were not
significantly affected by Mg fertility levels in 1987.
Insect data for individual treatments are reported in terms of population
densities and cycles which, when combined over date, describe seasonal
population dynamics (Fig. 1, 2). Population densities are described in terms
of daily and seasonal variation. Population cycles are recognized in figures
in relation to insect numbers/meter of row and stage, and date or plant
Population densities of VBC were very low each year until soybean Growth
Stage R5; then, densities increased in all treatments and were greatest on the
last sample date in both years (Fig. 1). In both years, density estimates in
all treatments were greater than the economic threshold density (12 larvae per
39 inches (meter) of row [Johnson et al., 1988]) on sample dates during
soybean Growth Stage R6. Adult populations of SGSB invaded and began
reproducing during Growth Stage R4, but data for adults are not shown due to
poor precision of sample estimates resulting from a clumped dispersion pattern
typical of adult SGSB populations (Schumann and Todd, 1982). Population
densities of SGSB nymphs in 1986 were low in all treatments until soybean
Growth Stage R6, with density estimates exceeding the economic threshold
density (3 per 36 inches of row [Johnson et al., 1988]) during R6 (Fig. 2A).
Densities remained low on all sample dates in 1987, with estimates always
below the economic threshold density.
The population densities of larval VBC differed between fertility
treatments in 1986 (F=7.6; df=7,21; P<0.01) and 1987 (F=3.1; df=10,30;
P<0.01). Orthogonal comparisons were used to separate the effects of P, K,
and Mg levels on VBC population densities. Density estimates were
significantly affected by P levels. Population cycles of VBC for treatments
at different levels of P, but constant levels of K and Mg, are shown in Figure
IA to illustrate the effect of P on density estimates. Mean densities were
significantly greater in treatments at the P4 level compared with densities in
treatments at the Pi, P2, and P3 levels in 1986 (F=9.0; df=l,21; P<0.01), but
not in 1987. Mean densities also were significantly greater in treatments at
the P2 and P3 levels than in treatments at the P1 level in 1986 (F=26.0;
df=l,21; P<0.001) and 1987 (F=7.9; df=l,30; P<0.01).
Density estimates of VBC were significantly affected by K in 1987 (Fig.
2B), but not in 1986 (Fig. 2A). Population cycles of VBC for treatments at
different levels of K, but constant levels of P and Mg, are shown in Figure
1B. Treatments at the K2 levels were not sampled in 1986. Orthogonal
treatment comparisons were used to show that estimates in 1987 were
significantly greater in treatments at the K2 and K, levels than at the K
level (F=11.0; df=1,30; P<0.01), with estimates similar at the K2 and K3
levels. Orthogonal comparisons also revealed that VBC estimates were not
significantly affected by Mg levels in 1986 or 1987 (data not shown).
The treatment x date interaction was used to determine if fertilizer
treatment influenced population cycles. This interaction was significant for
VBC in 1986 (F=4.3; df=49,168; P<0.01) and 1987 (F=2.1; df=60,198; P<0.01),
because estimates were similar on sample dates when densities were low and not
similar for some treatments on dates when estimates were greater (Fig. 1A and
The density estimates of nymphal SGSB differed between fertility treat-
ments in 1986 when populations were very great (F=2.2; df=7,21; P=0.07), but
not in 1987 when populations remained very low (F=0.05; df=10,30; P=0.87).
Population cycles of nymphal SGSB at different levels of P, but constant
levels of K and Mg, are shown in Figure 2A. Orthogonal treatment comparisons
revealed that density estimates were affected by P levels. Mean densities
were significantly greater at the P and P4 levels than at the Pi and P2
levels (F=7.8; df=l,21; P<0.01), but were similar at the P3 and P4 levels.
Mean densities also were greater at the P2 than the Pi level (F=4.7; df=l,21;
Orthogonal comparisons revealed that density estimates were not
significantly affected by K or Mg in 1986 and 1987 (data not shown). The
treatment x date interaction was not significant in 1987. This interaction
was significant in 1986 (F=2.5; df=49,168; P<0.01), because estimates were
similar on sample dates when densities were low and not similar for some
treatments on dates when estimates were greater (Fig. 2A).
Fertility levels of P, therefore, greatly affected population dynamics of
VBC and SGSB. For both pests in 1986, increased fertility levels of P that
did not result in a significant yield increase did result in significant
increases in pest population densities. Although yields were statistically
similar at the P2, P3, and P4 levels, population densities of VBC and nymphal
SGSB were significantly increased from P2 and P3 to P4. In 1987, soybean
yields were highest at and statistically similar for the P3 and P4 levels,
wjile VBC population densities were significantly highest at and statistically
similar for the P2, P3, and P4 levels. Fertility levels of K in 1986
significantly affected population densities of VBC. However, these effects
were the same as effects on soybean yield. Population densities of VBC in
1987 and SGSB nymphs in 1986 and 1987 were not significantly affected by K
levels, even at levels significantly affecting soybean yield. Although Mg
levels sometimes affected soybean yield, pest population densities were never
significantly influenced by Mg levels.
The reason why soil fertility levels affected population dynamics of these
major pests is unexplained. Soil fertility level may be directly affecting
pest populations through alterations in crop growth or nutritional level. Or,
the pest populations may be indirectly affected by effects of crop growth or
nutritional level on important natural enemies of the pests, such as bigeyed
bugs (Hemiptera:Lygaeidae), 'damsel bugs (Hemiptera: Nabidae), and spiders
(Aranaea: Araneidae). Our results showing effects of soil fertility on
population dynamics of the major pests of soybean in the extreme southern U.S.
are useful for integrated pest management programs in the region. Current
recommendations for soil fertility levels necessary to obtain optimal soybean
yields should be followed closely. In addition to reducing the cost of
fertilizer, it will also reduce the likelihood of pest outbreak and the need
for additional costs associated with pest control.
Adkisson, P.L. 1957. Influence of irrigation and fertilizer on populations
of three species of mirids attacking cotton. FAO Plant Prot. Bull.
Adkisson, P.L. 1958. The influence of fertilizer applications on population
of Heliothis zea (Boddie), and certain insect predators. J. Econ.
Herzog, D.C. and J.E. Funderburk. 1986. Ecological bases for habitat manage-
ment and cultural control. In: M. Kogan (ed.) Ecological theory and
integrated pest management practice. Wiley interscience, N.Y. p. 217-250.
Johnson, F.A., J.E. Funderburk, D.C. Herzog, and R.K. Sprenkel. 1988.
Soybean insect control. Univ. of Fla, North Fla. Res. and Educ. Ctr.,
Quincy, FL, Ext. Entomol. Rep. #58:1-27.
Kogan, M., and H. N. Pitre, Jr. 1980. General sampling methods for above-
ground populations of soybean arthropods, pp. 30-60. In: M. Kogan and D.
C. Herzog (ed.) Sampling Methods in Soybean Entomology. Springer-Verlag,
Inc. New York.
McGarr, R.L. 1942. Relation'of fertilizers to the development of the cotton
aphid. J. Econ. Entomol. 35:482-483.
McGarr, R.L,. 1943. Relation of fertilizers to the development of the cotton
aphid in 1941 and 1942. J. Econ. Entomol. 36:640.
Musick, G.J. 1985. Management of arthropod pests in conservation-tillage
systems in the southeastern U.S. pp. 191-204. In: W.L. Hargrove, F.C.
Boswell, and G.W. Langdale (ed.) Proceedings of the 1985 southern region
no-till conference. July 16-17, 1985. Griffin, Ga.
Rhoads, F.M., and R.D. Barnett. 1990. Soybean yield and soil test
phosphorous, potassium, and magnesium. Univ. of Fla. Res. and Educ. Ctr..
Quincy, FL, Res. Rep. NF-90-14. p. 1-10.
Rhoads, F.M., and R.D. Barnett. 1985. Nutritional requirement of high yield
cropping systems in the Southeast. Annual Report, IFAS, Quincy, FL.
Potash and Phosphate Institute, Atlanta, GA.
Schumann, F.W., and J.W. Todd. 1982. Population dynamics of the southern
green stink bug (Heteroptera: Pentatomidae) in relation to soybean
phenology. J. Econ. Entomol. 75:748-753.
Steel, R. G. D., and J. H. Torrie. 1960. Principles and procedures of
statistics with special reference to the biological sciencGraw-Hill, Inc.,
NY, 481 pp.
Todd, J.W., and D.C. Herzog. 1980. Sampling phytophagous pentatomidae in
soybean. In: M. Kogan and D.C. Herzog (ed.) Sampling Methods in Soybean
Entomology, Springer Verlag, Inc., New York. p.438-478.
Walker, J.K., G.A. Niles, J.R. Gannaway, D.R. Bradshaw, and R.E. Glodt. 1976.
Narrow row planting of cotton genotypes and boll weevil damage. J. Econ.
Walker, J.K., J.R. Gannaway, and G.A. Niles. 1977. Age distrubition of cotton
bolls and damage from boll weevil. J. Econ. Entomol. 70:5-8.
Our thanks to E. Brown, Agricultural Technican IV; A. Brown, Agricultural
Supervisor; A. Manning, Biological Scientist II; North Fla. Res. and Educ.
Ctr., Univ. of Fla. Quincy FL 32351; for data anaylsis and illustration,
data collection, and plot preparation and management.
Table 1. Soil' test (Melich-I extractant) from fertility
plots in the soybean-fertility-pest experiment in 1986 and
1987 at Quincy FL.
Soil-test levels (ppm) across reps.
P, = 82
P, = 75
Table 2. Soybean yields (bu/A) in relation to fertility treatment and
year with no.pesticide treatment, Quincy FL.
Year Soybean yield (bu/A) for fertility treatments across reps.
P, = 25
P, = 17
180 189 203 217 228 238 252 285 275 281
V4 R1-2 R3 R4 R5 R6 R6
4o K,P. Mg
0 ----h 7 KP. Mg,
150 180 203 217 228 238 252 285 275 28s
Mean population density [number insects/39 inches (meter) of row] of larval
velvetbean caterpillars in relation to day of year (Days Julian, 1986, and
1987) and physiological stage of soybean development in treatments differing
in soil fertility levels (A) of P, but at the same level of K and Mg and (B)
of differing soil fertility levels of K but at the same level of P and Mg.
V3 V5-6 R1R2 R4 R5 R6 RS
1988 B K,P, Mg,
180 18 203 217
V4 R1-2 13 R4 RS R6 R6
20- ----------- -----
226 230 252 265 275 225 180 189 203 217 220 236 252 268 275 285
DAY OF YEAR
Mean population density [number/39 inches (meter) of row] of nymphal southern
green stink bugs in relation to day of year (Days Julian, 1986, and 1987) and
physiological stage of soybean development in treatments differing in soil
fertility levels (A) of P, but at the same level of K and Mg and (B) of
differeing soil fertility levels of K, but at the same level of P and Mg.