deletions | additions       diff --git a/subsection_Pumping_rates_are_reduced__.tex b/subsection_Pumping_rates_are_reduced__.tex  index 17d9fb0..8334dab 100644  --- a/subsection_Pumping_rates_are_reduced__.tex  +++ b/subsection_Pumping_rates_are_reduced__.tex 
...
\subsection{Pumping rates are reduced and exhibit long correlations in a dynamic environment.}  Combined, controlling food availability and prolonged measurements enable to assay assaying  pharyngeal pumping in a dynamic environment. To demonstrate this, we have oscillated the concentration of available food between its high ($D_{600}=4.0$) and low ($0.0$) values with a period of $360$ sec. Switching between the two food reservoirs was performed using a small volume rotative valve, which eliminates back-flow and cross contamination and minimizes flow disruptions. As a result, we could not detect a response to switching between two reservoirs containing the same concentration of food (data not shown). Panels E-G in Fig. \ref{fig:results} depict examples of instantaneous pumping rates of individual animals and the corresponding food levels, as well as averaged data. Our analysis could readily detect changes in pumping rates that occurred over $\leq 10$ sec and in this assay the animals experienced a plateau of  high food availability for $90$ sec during each period. We thus hypothesized that pumping rates would track the changes in food availability and that their highest and lowest mean pumping rates would be comparable to those measured at the corresponding fixed concentrations (Fig. \ref{fig:results}C). Surprisingly, mean pumping rates in the dynamic environment were approximately $2-$fold lower than the corresponding steady state rates (Fig. \ref{fig:results}F-G). Moreover, is was not uncommon for animals to maintain low or high pumping activity regardless of oscillations in the concentration of food for up to $10$ min (see arrows in Fig. \ref{fig:results}E and Fig. \ref{fig:results}I). Due to the variability of the behavioral responses, we also calculated the response-triggered average to the periodic stimuli. We identified the time points when the instantaneous pumping rate exceeded a threshold value of $2$ Hz. Food concentration data from the $180$ sec periods that immediately preceded these time points were aligned and averaged. We found that the mean concentration of food increased from $OD_{600} \simeq 0.7 $ at earlier times to $OD_{600} \simeq 2.2$ when high activity was detected (Fig. \ref{fig:results}H). This result indicates that high pumping rates are exercised preferentially in response to an increase in food availability. Combined, our results suggest that {\it C. elegans} readily detects dynamic changes in food availability and can respond to them rapidly, yet their behavior is more complex than simply tracking the external conditions.