Can. J. Fish. Aquat. Sci. 55: 1169–1178 (1998). © 1998 NRC Canada
1169
Phosphorus loading reductions needed
to control blue-green algal blooms
in Lake Mendota
Richard C. Lathrop, Stephen R. Carpenter, Craig A. Stow,
Patricia A. Soranno, and John C. Panuska
Abstract: We evaluated the reductions in P loading needed to control blue-green algal blooms in Lake Mendota, Wisconsin. After
developing a 21-year loading data set, we used a P mass balance model expressed as a difference equation with an annual time step
indexed from mid-April. We defined and estimated a loss parameter λ as the proportion of the lake’s April P concentration lost
through sedimentation and outflow during the following year. Using the distribution of annual λ’s and input loadings, we predicted
the steady-state distribution of April P concentrations that would result from scenarios of altered inputs due to changes in
management practices. These results were then linked to the probability of summer blue-green algal blooms. For no load
reduction, the probability of a bloom (>2 mg algae?L
–1
) on any summer day is about 60%. This probability decreases to 20% with
a load reduction of 50%. Our approach illustrates how managers can consider reducing the frequency of extreme events like algal
blooms, which may correspond more to the public’s perception of lake water quality than average conditions.
Résumé : Nous avons évalué les réductions de la charge en P nécessaires pour éliminer les efflorescences d’algues bleues dans
le lac Mendota, au Wisconsin. Après avoir établi une série de données sur la charge couvrant une période de 21 ans, nous avons
utilisé un modèle du bilan massique de P exprimé sous la forme d’une équation à différences finies avec la mi-avril comme pas
de temps annuel. Nous avons défini et estimé un paramètre de perte λ correspondant à la proportion de la teneur du lac en P au
mois d’avril qui est perdue par sédimentation et par évacuation au cours de l’année suivante. En utilisant la distribution des λ
annuels et des charges d’entrée, nous avons prédit la distribution en état d’équilibre des teneurs en P d’avril qui résulteraient de
scénarios de modification de la charge suite à des changements dans les méthodes de gestion. Nous avons alors établi un lien
entre les résultats obtenus et la probabilité d’efflorescences estivales d’algues bleues. Sans réduction de charge, la probabilité
d’une efflorescence (>2 mg d’algues?L
–1
) pendant n’importe quel jour d’été se situait à environ 60%. Cette probabilité diminue
à 20% lorsque la charge est réduite de 50%. Notre approche illustre une façon pour les gestionnaires d’envisager la réduction de
la fréquence d’événements extrêmes comme les efflorescences phytoplanctoniques, qui, aux yeux du grand public, peuvent
sembler plus révélateurs de la qualité de l’eau d’un lac que les conditions moyennes.
[Traduit par la Rédaction]
Introduction
Despite decades of management and research, blooms of blue-
green algae are still a major water quality problem in lakes and
reservoirs (National Research Council 1992). Excessive
inputs of P have been identified as the major cause of these
eutrophication symptoms (Vollenweider 1968; Schindler
1977; Sas 1989). In the United States, nonpoint pollution from
agricultural and urban runoff is the main source of P (National
Research Council 1992). Because reducing nonpoint source P
inputs can be costly, estimates of the benefits derived from
various levels of P reduction are useful in decision making.
Eutrophication models frequently have been used to predict
the lake response to P input reductions. Most of the models
are deterministic and calculate seasonal or annual averages for
P, chlorophyll, or algal concentrations, or for water transpar-
ency readings, as a function of P input (Reckhow and Chapra
1983; Cooke et al. 1993). These models, derived from cross-
sectional analyses of many lakes (e.g., Dillon and Rigler
1974a; Vollenweider 1976; Canfield and Bachmann 1981;
Reckhow and Chapra 1983), have large prediction uncertain-
ties when applied to any one lake. For widely used models,
errors in predicting P concentrations in a given lake can be as
much as ±0.05 mg?L
–1
(Cooke et al. 1993), which may be
unacceptably large for decision making by managers.
More importantly, blue-green algal blooms are extreme and
highly stochastic events whose occurrence can be masked in
average conditions (Walker 1985; Paerl 1988; Havens 1994;
Soranno 1997). It is these extreme events that are associated
with episodes of toxicity, fish kills, anoxia, and generally nox-
ious shoreline conditions resulting from wind-blown accumu-
lations of algae. Thus, the public’s perception of water quality
Received July 29, 1997. Accepted November 19, 1997.
J14143
R.C. Lathrop.
1
Wisconsin Department of Natural Resources,
1350 Femrite Dr., Monona, WI 53716, U.S.A., and Center for
Limnology, 680 North Park St., University of Wisconsin,
Madison, WI 53706, U.S.A.
S.R. Carpenter, C.A. Stow,
2
and P.A. Soranno.
3
Center for
Limnology, 680 North Park St., University of Wisconsin,
Madison, WI 53706, U.S.A.
J.C. Panuska. Wisconsin Department of Natural Resources,
1350 Femrite Dr., Monona, WI 53716, U.S.A.
1
Author to whom all correspondence should be addressed.
e-mail: rlathrop@facstaff.wisc.edu
2
Present address: Nicholas School of the Environment, Duke
University, Durham, NC 27708, U.S.A.
3
Present address: Department of Fisheries and Wildlife, Mich-
igan State University, East Lansing, MI 48824, U.S.A.
Can. J. Fish. Aquat. Sci. Downloaded from www.nrcresearchpress.com by Guangzhou Jinan University on 06/05/13
For personal use only.